]> code.delx.au - gnu-emacs/blob - lisp/progmodes/cc-engine.el
; Revert "Use eldoc-documentation-functions"
[gnu-emacs] / lisp / progmodes / cc-engine.el
1 ;;; cc-engine.el --- core syntax guessing engine for CC mode -*- coding: utf-8 -*-
2
3 ;; Copyright (C) 1985, 1987, 1992-2016 Free Software Foundation, Inc.
4
5 ;; Authors: 2001- Alan Mackenzie
6 ;; 1998- Martin Stjernholm
7 ;; 1992-1999 Barry A. Warsaw
8 ;; 1987 Dave Detlefs
9 ;; 1987 Stewart Clamen
10 ;; 1985 Richard M. Stallman
11 ;; Maintainer: bug-cc-mode@gnu.org
12 ;; Created: 22-Apr-1997 (split from cc-mode.el)
13 ;; Keywords: c languages
14 ;; Package: cc-mode
15
16 ;; This file is part of GNU Emacs.
17
18 ;; GNU Emacs is free software: you can redistribute it and/or modify
19 ;; it under the terms of the GNU General Public License as published by
20 ;; the Free Software Foundation, either version 3 of the License, or
21 ;; (at your option) any later version.
22
23 ;; GNU Emacs is distributed in the hope that it will be useful,
24 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
25 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
26 ;; GNU General Public License for more details.
27
28 ;; You should have received a copy of the GNU General Public License
29 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
30
31 ;;; Commentary:
32
33 ;; The functions which have docstring documentation can be considered
34 ;; part of an API which other packages can use in CC Mode buffers.
35 ;; Otoh, undocumented functions and functions with the documentation
36 ;; in comments are considered purely internal and can change semantics
37 ;; or even disappear in the future.
38 ;;
39 ;; (This policy applies to CC Mode as a whole, not just this file. It
40 ;; probably also applies to many other Emacs packages, but here it's
41 ;; clearly spelled out.)
42
43 ;; Hidden buffer changes
44 ;;
45 ;; Various functions in CC Mode use text properties for caching and
46 ;; syntactic markup purposes, and those of them that might modify such
47 ;; properties but still don't modify the buffer in a visible way are
48 ;; said to do "hidden buffer changes". They should be used within
49 ;; `c-save-buffer-state' or a similar function that saves and restores
50 ;; buffer modifiedness, disables buffer change hooks, etc.
51 ;;
52 ;; Interactive functions are assumed to not do hidden buffer changes,
53 ;; except in the specific parts of them that do real changes.
54 ;;
55 ;; Lineup functions are assumed to do hidden buffer changes. They
56 ;; must not do real changes, though.
57 ;;
58 ;; All other functions that do hidden buffer changes have that noted
59 ;; in their doc string or comment.
60 ;;
61 ;; The intention with this system is to avoid wrapping every leaf
62 ;; function that do hidden buffer changes inside
63 ;; `c-save-buffer-state'. It should be used as near the top of the
64 ;; interactive functions as possible.
65 ;;
66 ;; Functions called during font locking are allowed to do hidden
67 ;; buffer changes since the font-lock package run them in a context
68 ;; similar to `c-save-buffer-state' (in fact, that function is heavily
69 ;; inspired by `save-buffer-state' in the font-lock package).
70
71 ;; Use of text properties
72 ;;
73 ;; CC Mode uses several text properties internally to mark up various
74 ;; positions, e.g. to improve speed and to eliminate glitches in
75 ;; interactive refontification.
76 ;;
77 ;; Note: This doc is for internal use only. Other packages should not
78 ;; assume that these text properties are used as described here.
79 ;;
80 ;; 'category
81 ;; Used for "indirection". With its help, some other property can
82 ;; be cheaply and easily switched on or off everywhere it occurs.
83 ;;
84 ;; 'syntax-table
85 ;; Used to modify the syntax of some characters. It is used to
86 ;; mark the "<" and ">" of angle bracket parens with paren syntax, to
87 ;; "hide" obtrusive characters in preprocessor lines, and to mark C++
88 ;; raw strings to enable their fontification.
89 ;;
90 ;; This property is used on single characters and is therefore
91 ;; always treated as front and rear nonsticky (or start and end open
92 ;; in XEmacs vocabulary). It's therefore installed on
93 ;; `text-property-default-nonsticky' if that variable exists (Emacs
94 ;; >= 21).
95 ;;
96 ;; 'c-is-sws and 'c-in-sws
97 ;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
98 ;; speed them up. See the comment blurb before `c-put-is-sws'
99 ;; below for further details.
100 ;;
101 ;; 'c-type
102 ;; This property is used on single characters to mark positions with
103 ;; special syntactic relevance of various sorts. Its primary use is
104 ;; to avoid glitches when multiline constructs are refontified
105 ;; interactively (on font lock decoration level 3). It's cleared in
106 ;; a region before it's fontified and is then put on relevant chars
107 ;; in that region as they are encountered during the fontification.
108 ;; The value specifies the kind of position:
109 ;;
110 ;; 'c-decl-arg-start
111 ;; Put on the last char of the token preceding each declaration
112 ;; inside a declaration style arglist (typically in a function
113 ;; prototype).
114 ;;
115 ;; 'c-decl-end
116 ;; Put on the last char of the token preceding a declaration.
117 ;; This is used in cases where declaration boundaries can't be
118 ;; recognized simply by looking for a token like ";" or "}".
119 ;; `c-type-decl-end-used' must be set if this is used (see also
120 ;; `c-find-decl-spots').
121 ;;
122 ;; 'c-<>-arg-sep
123 ;; Put on the commas that separate arguments in angle bracket
124 ;; arglists like C++ template arglists.
125 ;;
126 ;; 'c-decl-id-start and 'c-decl-type-start
127 ;; Put on the last char of the token preceding each declarator
128 ;; in the declarator list of a declaration. They are also used
129 ;; between the identifiers cases like enum declarations.
130 ;; 'c-decl-type-start is used when the declarators are types,
131 ;; 'c-decl-id-start otherwise.
132 ;;
133 ;; 'c-awk-NL-prop
134 ;; Used in AWK mode to mark the various kinds of newlines. See
135 ;; cc-awk.el.
136
137 ;;; Code:
138
139 (eval-when-compile
140 (let ((load-path
141 (if (and (boundp 'byte-compile-dest-file)
142 (stringp byte-compile-dest-file))
143 (cons (file-name-directory byte-compile-dest-file) load-path)
144 load-path)))
145 (load "cc-bytecomp" nil t)))
146
147 (cc-require 'cc-defs)
148 (cc-require-when-compile 'cc-langs)
149 (cc-require 'cc-vars)
150
151 (eval-when-compile (require 'cl))
152
153 \f
154 ;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
155
156 (defmacro c-declare-lang-variables ()
157 `(progn
158 ,@(c--mapcan (lambda (init)
159 `(,(if (elt init 2)
160 `(defvar ,(car init) nil ,(elt init 2))
161 `(defvar ,(car init) nil))
162 (make-variable-buffer-local ',(car init))))
163 (cdr c-lang-variable-inits))))
164 (c-declare-lang-variables)
165
166 \f
167 ;;; Internal state variables.
168
169 ;; Internal state of hungry delete key feature
170 (defvar c-hungry-delete-key nil)
171 (make-variable-buffer-local 'c-hungry-delete-key)
172
173 ;; The electric flag (toggled by `c-toggle-electric-state').
174 ;; If t, electric actions (like automatic reindentation, and (if
175 ;; c-auto-newline is also set) auto newlining) will happen when an electric
176 ;; key like `{' is pressed (or an electric keyword like `else').
177 (defvar c-electric-flag t)
178 (make-variable-buffer-local 'c-electric-flag)
179
180 ;; Internal state of auto newline feature.
181 (defvar c-auto-newline nil)
182 (make-variable-buffer-local 'c-auto-newline)
183
184 ;; Included in the mode line to indicate the active submodes.
185 ;; (defvar c-submode-indicators nil)
186 ;; (make-variable-buffer-local 'c-submode-indicators)
187
188 (defun c-calculate-state (arg prevstate)
189 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
190 ;; arg is nil or zero, toggle the state. If arg is negative, turn
191 ;; the state off, and if arg is positive, turn the state on
192 (if (or (not arg)
193 (zerop (setq arg (prefix-numeric-value arg))))
194 (not prevstate)
195 (> arg 0)))
196
197 \f
198 ;; Basic handling of preprocessor directives.
199
200 ;; This is a dynamically bound cache used together with
201 ;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
202 ;; works as long as point doesn't cross a macro boundary.
203 (defvar c-macro-start 'unknown)
204
205 (defsubst c-query-and-set-macro-start ()
206 (if (symbolp c-macro-start)
207 (setq c-macro-start (save-excursion
208 (c-save-buffer-state ()
209 (and (c-beginning-of-macro)
210 (point)))))
211 c-macro-start))
212
213 (defsubst c-query-macro-start ()
214 (if (symbolp c-macro-start)
215 (save-excursion
216 (c-save-buffer-state ()
217 (and (c-beginning-of-macro)
218 (point))))
219 c-macro-start))
220
221 ;; One element macro cache to cope with continual movement within very large
222 ;; CPP macros.
223 (defvar c-macro-cache nil)
224 (make-variable-buffer-local 'c-macro-cache)
225 ;; Nil or cons of the bounds of the most recent CPP form probed by
226 ;; `c-beginning-of-macro', `c-end-of-macro' or `c-syntactic-end-of-macro'.
227 ;; The cdr will be nil if we know only the start of the CPP form.
228 (defvar c-macro-cache-start-pos nil)
229 (make-variable-buffer-local 'c-macro-cache-start-pos)
230 ;; The starting position from where we determined `c-macro-cache'.
231 (defvar c-macro-cache-syntactic nil)
232 (make-variable-buffer-local 'c-macro-cache-syntactic)
233 ;; Either nil, or the syntactic end of the macro currently represented by
234 ;; `c-macro-cache'.
235 (defvar c-macro-cache-no-comment nil)
236 (make-variable-buffer-local 'c-macro-cache-no-comment)
237 ;; Either nil, or the last character of the macro currently represented by
238 ;; `c-macro-cache' which isn't in a comment. */
239
240 (defun c-invalidate-macro-cache (beg end)
241 ;; Called from a before-change function. If the change region is before or
242 ;; in the macro characterized by `c-macro-cache' etc., nullify it
243 ;; appropriately. BEG and END are the standard before-change-functions
244 ;; parameters. END isn't used.
245 (cond
246 ((null c-macro-cache))
247 ((< beg (car c-macro-cache))
248 (setq c-macro-cache nil
249 c-macro-cache-start-pos nil
250 c-macro-cache-syntactic nil
251 c-macro-cache-no-comment nil))
252 ((and (cdr c-macro-cache)
253 (< beg (cdr c-macro-cache)))
254 (setcdr c-macro-cache nil)
255 (setq c-macro-cache-start-pos beg
256 c-macro-cache-syntactic nil
257 c-macro-cache-no-comment nil))))
258
259 (defun c-macro-is-genuine-p ()
260 ;; Check that the ostensible CPP construct at point is a real one. In
261 ;; particular, if point is on the first line of a narrowed buffer, make sure
262 ;; that the "#" isn't, say, the second character of a "##" operator. Return
263 ;; t when the macro is real, nil otherwise.
264 (let ((here (point)))
265 (beginning-of-line)
266 (prog1
267 (if (and (eq (point) (point-min))
268 (/= (point) 1))
269 (save-restriction
270 (widen)
271 (beginning-of-line)
272 (and (looking-at c-anchored-cpp-prefix)
273 (eq (match-beginning 1) here)))
274 t)
275 (goto-char here))))
276
277 (defun c-beginning-of-macro (&optional lim)
278 "Go to the beginning of a preprocessor directive.
279 Leave point at the beginning of the directive and return t if in one,
280 otherwise return nil and leave point unchanged.
281
282 Note that this function might do hidden buffer changes. See the
283 comment at the start of cc-engine.el for more info."
284 (let ((here (point)))
285 (when c-opt-cpp-prefix
286 (if (and (car c-macro-cache)
287 (>= (point) (car c-macro-cache))
288 (or (and (cdr c-macro-cache)
289 (<= (point) (cdr c-macro-cache)))
290 (<= (point) c-macro-cache-start-pos)))
291 (unless (< (car c-macro-cache) (or lim (point-min)))
292 (progn (goto-char (max (or lim (point-min)) (car c-macro-cache)))
293 (setq c-macro-cache-start-pos
294 (max c-macro-cache-start-pos here))
295 t))
296 (setq c-macro-cache nil
297 c-macro-cache-start-pos nil
298 c-macro-cache-syntactic nil
299 c-macro-cache-no-comment nil)
300
301 (save-restriction
302 (if lim (narrow-to-region lim (point-max)))
303 (beginning-of-line)
304 (while (eq (char-before (1- (point))) ?\\)
305 (forward-line -1))
306 (back-to-indentation)
307 (if (and (<= (point) here)
308 (save-match-data (looking-at c-opt-cpp-start))
309 (c-macro-is-genuine-p))
310 (progn
311 (setq c-macro-cache (cons (point) nil)
312 c-macro-cache-start-pos here)
313 t)
314 (goto-char here)
315 nil))))))
316
317 (defun c-end-of-macro ()
318 "Go to the end of a preprocessor directive.
319 More accurately, move the point to the end of the closest following
320 line that doesn't end with a line continuation backslash - no check is
321 done that the point is inside a cpp directive to begin with.
322
323 Note that this function might do hidden buffer changes. See the
324 comment at the start of cc-engine.el for more info."
325 (if (and (cdr c-macro-cache)
326 (<= (point) (cdr c-macro-cache))
327 (>= (point) (car c-macro-cache)))
328 (goto-char (cdr c-macro-cache))
329 (unless (and (car c-macro-cache)
330 (<= (point) c-macro-cache-start-pos)
331 (>= (point) (car c-macro-cache)))
332 (setq c-macro-cache nil
333 c-macro-cache-start-pos nil
334 c-macro-cache-syntactic nil
335 c-macro-cache-no-comment nil))
336 (while (progn
337 (end-of-line)
338 (when (and (eq (char-before) ?\\)
339 (not (eobp)))
340 (forward-char)
341 t)))
342 (when (car c-macro-cache)
343 (setcdr c-macro-cache (point)))))
344
345 (defun c-syntactic-end-of-macro ()
346 ;; Go to the end of a CPP directive, or a "safe" pos just before.
347 ;;
348 ;; This is normally the end of the next non-escaped line. A "safe"
349 ;; position is one not within a string or comment. (The EOL on a line
350 ;; comment is NOT "safe").
351 ;;
352 ;; This function must only be called from the beginning of a CPP construct.
353 ;;
354 ;; Note that this function might do hidden buffer changes. See the comment
355 ;; at the start of cc-engine.el for more info.
356 (let* ((here (point))
357 (there (progn (c-end-of-macro) (point)))
358 s)
359 (if c-macro-cache-syntactic
360 (goto-char c-macro-cache-syntactic)
361 (setq s (parse-partial-sexp here there))
362 (while (and (or (nth 3 s) ; in a string
363 (nth 4 s)) ; in a comment (maybe at end of line comment)
364 (> there here)) ; No infinite loops, please.
365 (setq there (1- (nth 8 s)))
366 (setq s (parse-partial-sexp here there)))
367 (setq c-macro-cache-syntactic (point)))
368 (point)))
369
370 (defun c-no-comment-end-of-macro ()
371 ;; Go to the end of a CPP directive, or a pos just before which isn't in a
372 ;; comment. For this purpose, open strings are ignored.
373 ;;
374 ;; This function must only be called from the beginning of a CPP construct.
375 ;;
376 ;; Note that this function might do hidden buffer changes. See the comment
377 ;; at the start of cc-engine.el for more info.
378 (let* ((here (point))
379 (there (progn (c-end-of-macro) (point)))
380 s)
381 (if c-macro-cache-no-comment
382 (goto-char c-macro-cache-no-comment)
383 (setq s (parse-partial-sexp here there))
384 (while (and (nth 3 s) ; in a string
385 (> there here)) ; No infinite loops, please.
386 (setq here (1+ (nth 8 s)))
387 (setq s (parse-partial-sexp here there)))
388 (when (nth 4 s)
389 (goto-char (1- (nth 8 s))))
390 (setq c-macro-cache-no-comment (point)))
391 (point)))
392
393 (defun c-forward-over-cpp-define-id ()
394 ;; Assuming point is at the "#" that introduces a preprocessor
395 ;; directive, it's moved forward to the end of the identifier which is
396 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
397 ;; is returned in this case, in all other cases nil is returned and
398 ;; point isn't moved.
399 ;;
400 ;; This function might do hidden buffer changes.
401 (when (and c-opt-cpp-macro-define-id
402 (looking-at c-opt-cpp-macro-define-id))
403 (goto-char (match-end 0))))
404
405 (defun c-forward-to-cpp-define-body ()
406 ;; Assuming point is at the "#" that introduces a preprocessor
407 ;; directive, it's moved forward to the start of the definition body
408 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
409 ;; specifies). Non-nil is returned in this case, in all other cases
410 ;; nil is returned and point isn't moved.
411 ;;
412 ;; This function might do hidden buffer changes.
413 (when (and c-opt-cpp-macro-define-start
414 (looking-at c-opt-cpp-macro-define-start)
415 (not (= (match-end 0) (c-point 'eol))))
416 (goto-char (match-end 0))))
417
418 \f
419 ;;; Basic utility functions.
420
421 (defun c-delq-from-dotted-list (elt dlist)
422 ;; If ELT is a member of the (possibly dotted) list DLIST, remove all
423 ;; occurrences of it (except for any in the last cdr of DLIST).
424 ;;
425 ;; Call this as (setq DLIST (c-delq-from-dotted-list ELT DLIST)), as
426 ;; sometimes the original structure is changed, sometimes it's not.
427 ;;
428 ;; This function is needed in Emacs < 24.5, and possibly XEmacs, because
429 ;; `delq' throws an error in these versions when given a dotted list.
430 (let ((tail dlist) prev)
431 (while (consp tail)
432 (if (eq (car tail) elt)
433 (if prev
434 (setcdr prev (cdr tail))
435 (setq dlist (cdr dlist)))
436 (setq prev tail))
437 (setq tail (cdr tail)))
438 dlist))
439
440 (defun c-syntactic-content (from to paren-level)
441 ;; Return the given region as a string where all syntactic
442 ;; whitespace is removed or, where necessary, replaced with a single
443 ;; space. If PAREN-LEVEL is given then all parens in the region are
444 ;; collapsed to "()", "[]" etc.
445 ;;
446 ;; This function might do hidden buffer changes.
447
448 (save-excursion
449 (save-restriction
450 (narrow-to-region from to)
451 (goto-char from)
452 (let* ((parts (list nil)) (tail parts) pos in-paren)
453
454 (while (re-search-forward c-syntactic-ws-start to t)
455 (goto-char (setq pos (match-beginning 0)))
456 (c-forward-syntactic-ws)
457 (if (= (point) pos)
458 (forward-char)
459
460 (when paren-level
461 (save-excursion
462 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
463 pos (point))))
464
465 (if (and (> pos from)
466 (< (point) to)
467 (looking-at "\\w\\|\\s_")
468 (save-excursion
469 (goto-char (1- pos))
470 (looking-at "\\w\\|\\s_")))
471 (progn
472 (setcdr tail (list (buffer-substring-no-properties from pos)
473 " "))
474 (setq tail (cddr tail)))
475 (setcdr tail (list (buffer-substring-no-properties from pos)))
476 (setq tail (cdr tail)))
477
478 (when in-paren
479 (when (= (car (parse-partial-sexp pos to -1)) -1)
480 (setcdr tail (list (buffer-substring-no-properties
481 (1- (point)) (point))))
482 (setq tail (cdr tail))))
483
484 (setq from (point))))
485
486 (setcdr tail (list (buffer-substring-no-properties from to)))
487 (apply 'concat (cdr parts))))))
488
489 (defun c-shift-line-indentation (shift-amt)
490 ;; Shift the indentation of the current line with the specified
491 ;; amount (positive inwards). The buffer is modified only if
492 ;; SHIFT-AMT isn't equal to zero.
493 (let ((pos (- (point-max) (point)))
494 (c-macro-start c-macro-start)
495 tmp-char-inserted)
496 (if (zerop shift-amt)
497 nil
498 ;; If we're on an empty line inside a macro, we take the point
499 ;; to be at the current indentation and shift it to the
500 ;; appropriate column. This way we don't treat the extra
501 ;; whitespace out to the line continuation as indentation.
502 (when (and (c-query-and-set-macro-start)
503 (looking-at "[ \t]*\\\\$")
504 (save-excursion
505 (skip-chars-backward " \t")
506 (bolp)))
507 (insert ?x)
508 (backward-char)
509 (setq tmp-char-inserted t))
510 (unwind-protect
511 (let ((col (current-indentation)))
512 (delete-region (c-point 'bol) (c-point 'boi))
513 (beginning-of-line)
514 (indent-to (+ col shift-amt)))
515 (when tmp-char-inserted
516 (delete-char 1))))
517 ;; If initial point was within line's indentation and we're not on
518 ;; a line with a line continuation in a macro, position after the
519 ;; indentation. Else stay at same point in text.
520 (if (and (< (point) (c-point 'boi))
521 (not tmp-char-inserted))
522 (back-to-indentation)
523 (if (> (- (point-max) pos) (point))
524 (goto-char (- (point-max) pos))))))
525
526 (defsubst c-keyword-sym (keyword)
527 ;; Return non-nil if the string KEYWORD is a known keyword. More
528 ;; precisely, the value is the symbol for the keyword in
529 ;; `c-keywords-obarray'.
530 (intern-soft keyword c-keywords-obarray))
531
532 (defsubst c-keyword-member (keyword-sym lang-constant)
533 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
534 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
535 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
536 ;; nil then the result is nil.
537 (get keyword-sym lang-constant))
538
539 ;; String syntax chars, suitable for skip-syntax-(forward|backward).
540 (defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
541 "\"|"
542 "\""))
543
544 ;; Regexp matching string limit syntax.
545 (defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
546 "\\s\"\\|\\s|"
547 "\\s\""))
548
549 ;; Regexp matching WS followed by string limit syntax.
550 (defconst c-ws*-string-limit-regexp
551 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
552
553 ;; Holds formatted error strings for the few cases where parse errors
554 ;; are reported.
555 (defvar c-parsing-error nil)
556 (make-variable-buffer-local 'c-parsing-error)
557
558 (defun c-echo-parsing-error (&optional quiet)
559 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
560 (c-benign-error "%s" c-parsing-error))
561 c-parsing-error)
562
563 ;; Faces given to comments and string literals. This is used in some
564 ;; situations to speed up recognition; it isn't mandatory that font
565 ;; locking is in use. This variable is extended with the face in
566 ;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
567 (defvar c-literal-faces
568 (append '(font-lock-comment-face font-lock-string-face)
569 (when (facep 'font-lock-comment-delimiter-face)
570 ;; New in Emacs 22.
571 '(font-lock-comment-delimiter-face))))
572
573 (defsubst c-put-c-type-property (pos value)
574 ;; Put a c-type property with the given value at POS.
575 (c-put-char-property pos 'c-type value))
576
577 (defun c-clear-c-type-property (from to value)
578 ;; Remove all occurrences of the c-type property that has the given
579 ;; value in the region between FROM and TO. VALUE is assumed to not
580 ;; be nil.
581 ;;
582 ;; Note: This assumes that c-type is put on single chars only; it's
583 ;; very inefficient if matching properties cover large regions.
584 (save-excursion
585 (goto-char from)
586 (while (progn
587 (when (eq (get-text-property (point) 'c-type) value)
588 (c-clear-char-property (point) 'c-type))
589 (goto-char (c-next-single-property-change (point) 'c-type nil to))
590 (< (point) to)))))
591
592 \f
593 ;; Some debug tools to visualize various special positions. This
594 ;; debug code isn't as portable as the rest of CC Mode.
595
596 (cc-bytecomp-defun overlays-in)
597 (cc-bytecomp-defun overlay-get)
598 (cc-bytecomp-defun overlay-start)
599 (cc-bytecomp-defun overlay-end)
600 (cc-bytecomp-defun delete-overlay)
601 (cc-bytecomp-defun overlay-put)
602 (cc-bytecomp-defun make-overlay)
603
604 (defun c-debug-add-face (beg end face)
605 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
606 (while overlays
607 (setq overlay (car overlays)
608 overlays (cdr overlays))
609 (when (eq (overlay-get overlay 'face) face)
610 (setq beg (min beg (overlay-start overlay))
611 end (max end (overlay-end overlay)))
612 (delete-overlay overlay)))
613 (overlay-put (make-overlay beg end) 'face face)))
614
615 (defun c-debug-remove-face (beg end face)
616 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
617 (ol-beg beg) (ol-end end))
618 (while overlays
619 (setq overlay (car overlays)
620 overlays (cdr overlays))
621 (when (eq (overlay-get overlay 'face) face)
622 (setq ol-beg (min ol-beg (overlay-start overlay))
623 ol-end (max ol-end (overlay-end overlay)))
624 (delete-overlay overlay)))
625 (when (< ol-beg beg)
626 (overlay-put (make-overlay ol-beg beg) 'face face))
627 (when (> ol-end end)
628 (overlay-put (make-overlay end ol-end) 'face face))))
629
630 \f
631 ;; `c-beginning-of-statement-1' and accompanying stuff.
632
633 ;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
634 ;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
635 ;; better way should be implemented, but this will at least shut up
636 ;; the byte compiler.
637 (defvar c-maybe-labelp)
638
639 ;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
640
641 ;; Macros used internally in c-beginning-of-statement-1 for the
642 ;; automaton actions.
643 (defmacro c-bos-push-state ()
644 '(setq stack (cons (cons state saved-pos)
645 stack)))
646 (defmacro c-bos-pop-state (&optional do-if-done)
647 `(if (setq state (car (car stack))
648 saved-pos (cdr (car stack))
649 stack (cdr stack))
650 t
651 ,do-if-done
652 (throw 'loop nil)))
653 (defmacro c-bos-pop-state-and-retry ()
654 '(throw 'loop (setq state (car (car stack))
655 saved-pos (cdr (car stack))
656 ;; Throw nil if stack is empty, else throw non-nil.
657 stack (cdr stack))))
658 (defmacro c-bos-save-pos ()
659 '(setq saved-pos (vector pos tok ptok pptok)))
660 (defmacro c-bos-restore-pos ()
661 '(unless (eq (elt saved-pos 0) start)
662 (setq pos (elt saved-pos 0)
663 tok (elt saved-pos 1)
664 ptok (elt saved-pos 2)
665 pptok (elt saved-pos 3))
666 (goto-char pos)
667 (setq sym nil)))
668 (defmacro c-bos-save-error-info (missing got)
669 `(setq saved-pos (vector pos ,missing ,got)))
670 (defmacro c-bos-report-error ()
671 '(unless noerror
672 (setq c-parsing-error
673 (format-message
674 "No matching `%s' found for `%s' on line %d"
675 (elt saved-pos 1)
676 (elt saved-pos 2)
677 (1+ (count-lines (point-min)
678 (c-point 'bol (elt saved-pos 0))))))))
679
680 (defun c-beginning-of-statement-1 (&optional lim ignore-labels
681 noerror comma-delim)
682 "Move to the start of the current statement or declaration, or to
683 the previous one if already at the beginning of one. Only
684 statements/declarations on the same level are considered, i.e. don't
685 move into or out of sexps (not even normal expression parentheses).
686
687 If point is already at the earliest statement within braces or parens,
688 this function doesn't move back into any whitespace preceding it; it
689 returns `same' in this case.
690
691 Stop at statement continuation tokens like \"else\", \"catch\",
692 \"finally\" and the \"while\" in \"do ... while\" if the start point
693 is within the continuation. If starting at such a token, move to the
694 corresponding statement start. If at the beginning of a statement,
695 move to the closest containing statement if there is any. This might
696 also stop at a continuation clause.
697
698 Labels are treated as part of the following statements if
699 IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
700 statement start keyword.) Otherwise, each label is treated as a
701 separate statement.
702
703 Macros are ignored \(i.e. skipped over) unless point is within one, in
704 which case the content of the macro is treated as normal code. Aside
705 from any normal statement starts found in it, stop at the first token
706 of the content in the macro, i.e. the expression of an \"#if\" or the
707 start of the definition in a \"#define\". Also stop at start of
708 macros before leaving them.
709
710 Return:
711 `label' if stopped at a label or \"case...:\" or \"default:\";
712 `same' if stopped at the beginning of the current statement;
713 `up' if stepped to a containing statement;
714 `previous' if stepped to a preceding statement;
715 `beginning' if stepped from a statement continuation clause to
716 its start clause; or
717 `macro' if stepped to a macro start.
718 Note that `same' and not `label' is returned if stopped at the same
719 label without crossing the colon character.
720
721 LIM may be given to limit the search. If the search hits the limit,
722 point will be left at the closest following token, or at the start
723 position if that is less (`same' is returned in this case).
724
725 NOERROR turns off error logging to `c-parsing-error'.
726
727 Normally only `;' and virtual semicolons are considered to delimit
728 statements, but if COMMA-DELIM is non-nil then `,' is treated
729 as a delimiter too.
730
731 Note that this function might do hidden buffer changes. See the
732 comment at the start of cc-engine.el for more info."
733
734 ;; The bulk of this function is a pushdown automaton that looks at statement
735 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
736 ;; purpose is to keep track of nested statements, ensuring that such
737 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
738 ;; does with nested braces/brackets/parentheses).
739 ;;
740 ;; Note: The position of a boundary is the following token.
741 ;;
742 ;; Beginning with the current token (the one following point), move back one
743 ;; sexp at a time (where a sexp is, more or less, either a token or the
744 ;; entire contents of a brace/bracket/paren pair). Each time a statement
745 ;; boundary is crossed or a "while"-like token is found, update the state of
746 ;; the PDA. Stop at the beginning of a statement when the stack (holding
747 ;; nested statement info) is empty and the position has been moved.
748 ;;
749 ;; The following variables constitute the PDA:
750 ;;
751 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
752 ;; scanned back over, 'boundary if we've just gone back over a
753 ;; statement boundary, or nil otherwise.
754 ;; state: takes one of the values (nil else else-boundary while
755 ;; while-boundary catch catch-boundary).
756 ;; nil means "no "while"-like token yet scanned".
757 ;; 'else, for example, means "just gone back over an else".
758 ;; 'else-boundary means "just gone back over a statement boundary
759 ;; immediately after having gone back over an else".
760 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
761 ;; of error reporting information.
762 ;; stack: The stack onto which the PDA pushes its state. Each entry
763 ;; consists of a saved value of state and saved-pos. An entry is
764 ;; pushed when we move back over a "continuation" token (e.g. else)
765 ;; and popped when we encounter the corresponding opening token
766 ;; (e.g. if).
767 ;;
768 ;;
769 ;; The following diagram briefly outlines the PDA.
770 ;;
771 ;; Common state:
772 ;; "else": Push state, goto state `else'.
773 ;; "while": Push state, goto state `while'.
774 ;; "catch" or "finally": Push state, goto state `catch'.
775 ;; boundary: Pop state.
776 ;; other: Do nothing special.
777 ;;
778 ;; State `else':
779 ;; boundary: Goto state `else-boundary'.
780 ;; other: Error, pop state, retry token.
781 ;;
782 ;; State `else-boundary':
783 ;; "if": Pop state.
784 ;; boundary: Error, pop state.
785 ;; other: See common state.
786 ;;
787 ;; State `while':
788 ;; boundary: Save position, goto state `while-boundary'.
789 ;; other: Pop state, retry token.
790 ;;
791 ;; State `while-boundary':
792 ;; "do": Pop state.
793 ;; boundary: Restore position if it's not at start, pop state. [*see below]
794 ;; other: See common state.
795 ;;
796 ;; State `catch':
797 ;; boundary: Goto state `catch-boundary'.
798 ;; other: Error, pop state, retry token.
799 ;;
800 ;; State `catch-boundary':
801 ;; "try": Pop state.
802 ;; "catch": Goto state `catch'.
803 ;; boundary: Error, pop state.
804 ;; other: See common state.
805 ;;
806 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
807 ;; searching for a "do" which would have opened a do-while. If we didn't
808 ;; find it, we discard the analysis done since the "while", go back to this
809 ;; token in the buffer and restart the scanning there, this time WITHOUT
810 ;; pushing the 'while state onto the stack.
811 ;;
812 ;; In addition to the above there is some special handling of labels
813 ;; and macros.
814
815 (let ((case-fold-search nil)
816 (start (point))
817 macro-start
818 (delims (if comma-delim '(?\; ?,) '(?\;)))
819 (c-stmt-delim-chars (if comma-delim
820 c-stmt-delim-chars-with-comma
821 c-stmt-delim-chars))
822 c-in-literal-cache c-maybe-labelp after-case:-pos saved
823 ;; Current position.
824 pos
825 ;; Position of last stmt boundary character (e.g. ;).
826 boundary-pos
827 ;; The position of the last sexp or bound that follows the
828 ;; first found colon, i.e. the start of the nonlabel part of
829 ;; the statement. It's `start' if a colon is found just after
830 ;; the start.
831 after-labels-pos
832 ;; Like `after-labels-pos', but the first such position inside
833 ;; a label, i.e. the start of the last label before the start
834 ;; of the nonlabel part of the statement.
835 last-label-pos
836 ;; The last position where a label is possible provided the
837 ;; statement started there. It's nil as long as no invalid
838 ;; label content has been found (according to
839 ;; `c-nonlabel-token-key'). It's `start' if no valid label
840 ;; content was found in the label. Note that we might still
841 ;; regard it a label if it starts with `c-label-kwds'.
842 label-good-pos
843 ;; Putative positions of the components of a bitfield declaration,
844 ;; e.g. "int foo : NUM_FOO_BITS ;"
845 bitfield-type-pos bitfield-id-pos bitfield-size-pos
846 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
847 ;; See above.
848 sym
849 ;; Current state in the automaton. See above.
850 state
851 ;; Current saved positions. See above.
852 saved-pos
853 ;; Stack of conses (state . saved-pos).
854 stack
855 ;; Regexp which matches "for", "if", etc.
856 (cond-key (or c-opt-block-stmt-key
857 "\\<\\>")) ; Matches nothing.
858 ;; Return value.
859 (ret 'same)
860 ;; Positions of the last three sexps or bounds we've stopped at.
861 tok ptok pptok)
862
863 (save-restriction
864 (if lim (narrow-to-region lim (point-max)))
865
866 (if (save-excursion
867 (and (c-beginning-of-macro)
868 (/= (point) start)))
869 (setq macro-start (point)))
870
871 ;; Try to skip back over unary operator characters, to register
872 ;; that we've moved.
873 (while (progn
874 (setq pos (point))
875 (c-backward-syntactic-ws)
876 ;; Protect post-++/-- operators just before a virtual semicolon.
877 (and (not (c-at-vsemi-p))
878 (/= (skip-chars-backward "-+!*&~@`#") 0))))
879
880 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
881 ;; done. Later on we ignore the boundaries for statements that don't
882 ;; contain any sexp. The only thing that is affected is that the error
883 ;; checking is a little less strict, and we really don't bother.
884 (if (and (memq (char-before) delims)
885 (progn (forward-char -1)
886 (setq saved (point))
887 (c-backward-syntactic-ws)
888 (or (memq (char-before) delims)
889 (memq (char-before) '(?: nil))
890 (eq (char-syntax (char-before)) ?\()
891 (c-at-vsemi-p))))
892 (setq ret 'previous
893 pos saved)
894
895 ;; Begin at start and not pos to detect macros if we stand
896 ;; directly after the #.
897 (goto-char start)
898 (if (looking-at "\\<\\|\\W")
899 ;; Record this as the first token if not starting inside it.
900 (setq tok start))
901
902 ;; The following while loop goes back one sexp (balanced parens,
903 ;; etc. with contents, or symbol or suchlike) each iteration. This
904 ;; movement is accomplished with a call to c-backward-sexp approx 170
905 ;; lines below.
906 ;;
907 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
908 ;; 1. On reaching the start of a macro;
909 ;; 2. On having passed a stmt boundary with the PDA stack empty;
910 ;; 3. On reaching the start of an Objective C method def;
911 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
912 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
913 (while
914 (catch 'loop ;; Throw nil to break, non-nil to continue.
915 (cond
916 ;; Are we in a macro, just after the opening #?
917 ((save-excursion
918 (and macro-start ; Always NIL for AWK.
919 (progn (skip-chars-backward " \t")
920 (eq (char-before) ?#))
921 (progn (setq saved (1- (point)))
922 (beginning-of-line)
923 (not (eq (char-before (1- (point))) ?\\)))
924 (looking-at c-opt-cpp-start)
925 (progn (skip-chars-forward " \t")
926 (eq (point) saved))))
927 (goto-char saved)
928 (if (and (c-forward-to-cpp-define-body)
929 (progn (c-forward-syntactic-ws start)
930 (< (point) start)))
931 ;; Stop at the first token in the content of the macro.
932 (setq pos (point)
933 ignore-labels t) ; Avoid the label check on exit.
934 (setq pos saved
935 ret 'macro
936 ignore-labels t))
937 (throw 'loop nil)) ; 1. Start of macro.
938
939 ;; Do a round through the automaton if we've just passed a
940 ;; statement boundary or passed a "while"-like token.
941 ((or sym
942 (and (looking-at cond-key)
943 (setq sym (intern (match-string 1)))))
944
945 (when (and (< pos start) (null stack))
946 (throw 'loop nil)) ; 2. Statement boundary.
947
948 ;; The PDA state handling.
949 ;;
950 ;; Refer to the description of the PDA in the opening
951 ;; comments. In the following OR form, the first leaf
952 ;; attempts to handles one of the specific actions detailed
953 ;; (e.g., finding token "if" whilst in state `else-boundary').
954 ;; We drop through to the second leaf (which handles common
955 ;; state) if no specific handler is found in the first cond.
956 ;; If a parsing error is detected (e.g. an "else" with no
957 ;; preceding "if"), we throw to the enclosing catch.
958 ;;
959 ;; Note that the (eq state 'else) means
960 ;; "we've just passed an else", NOT "we're looking for an
961 ;; else".
962 (or (cond
963 ((eq state 'else)
964 (if (eq sym 'boundary)
965 (setq state 'else-boundary)
966 (c-bos-report-error)
967 (c-bos-pop-state-and-retry)))
968
969 ((eq state 'else-boundary)
970 (cond ((eq sym 'if)
971 (c-bos-pop-state (setq ret 'beginning)))
972 ((eq sym 'boundary)
973 (c-bos-report-error)
974 (c-bos-pop-state))))
975
976 ((eq state 'while)
977 (if (and (eq sym 'boundary)
978 ;; Since this can cause backtracking we do a
979 ;; little more careful analysis to avoid it:
980 ;; If there's a label in front of the while
981 ;; it can't be part of a do-while.
982 (not after-labels-pos))
983 (progn (c-bos-save-pos)
984 (setq state 'while-boundary))
985 (c-bos-pop-state-and-retry))) ; Can't be a do-while
986
987 ((eq state 'while-boundary)
988 (cond ((eq sym 'do)
989 (c-bos-pop-state (setq ret 'beginning)))
990 ((eq sym 'boundary) ; isn't a do-while
991 (c-bos-restore-pos) ; the position of the while
992 (c-bos-pop-state)))) ; no longer searching for do.
993
994 ((eq state 'catch)
995 (if (eq sym 'boundary)
996 (setq state 'catch-boundary)
997 (c-bos-report-error)
998 (c-bos-pop-state-and-retry)))
999
1000 ((eq state 'catch-boundary)
1001 (cond
1002 ((eq sym 'try)
1003 (c-bos-pop-state (setq ret 'beginning)))
1004 ((eq sym 'catch)
1005 (setq state 'catch))
1006 ((eq sym 'boundary)
1007 (c-bos-report-error)
1008 (c-bos-pop-state)))))
1009
1010 ;; This is state common. We get here when the previous
1011 ;; cond statement found no particular state handler.
1012 (cond ((eq sym 'boundary)
1013 ;; If we have a boundary at the start
1014 ;; position we push a frame to go to the
1015 ;; previous statement.
1016 (if (>= pos start)
1017 (c-bos-push-state)
1018 (c-bos-pop-state)))
1019 ((eq sym 'else)
1020 (c-bos-push-state)
1021 (c-bos-save-error-info 'if 'else)
1022 (setq state 'else))
1023 ((eq sym 'while)
1024 ;; Is this a real while, or a do-while?
1025 ;; The next `when' triggers unless we are SURE that
1026 ;; the `while' is not the tail end of a `do-while'.
1027 (when (or (not pptok)
1028 (memq (char-after pptok) delims)
1029 ;; The following kludge is to prevent
1030 ;; infinite recursion when called from
1031 ;; c-awk-after-if-for-while-condition-p,
1032 ;; or the like.
1033 (and (eq (point) start)
1034 (c-vsemi-status-unknown-p))
1035 (c-at-vsemi-p pptok))
1036 ;; Since this can cause backtracking we do a
1037 ;; little more careful analysis to avoid it: If
1038 ;; the while isn't followed by a (possibly
1039 ;; virtual) semicolon it can't be a do-while.
1040 (c-bos-push-state)
1041 (setq state 'while)))
1042 ((memq sym '(catch finally))
1043 (c-bos-push-state)
1044 (c-bos-save-error-info 'try sym)
1045 (setq state 'catch))))
1046
1047 (when c-maybe-labelp
1048 ;; We're either past a statement boundary or at the
1049 ;; start of a statement, so throw away any label data
1050 ;; for the previous one.
1051 (setq after-labels-pos nil
1052 last-label-pos nil
1053 c-maybe-labelp nil))))
1054
1055 ;; Step to the previous sexp, but not if we crossed a
1056 ;; boundary, since that doesn't consume an sexp.
1057 (if (eq sym 'boundary)
1058 (setq ret 'previous)
1059
1060 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
1061 ;; BACKWARDS THROUGH THE SOURCE.
1062
1063 (c-backward-syntactic-ws)
1064 (let ((before-sws-pos (point))
1065 ;; The end position of the area to search for statement
1066 ;; barriers in this round.
1067 (maybe-after-boundary-pos pos))
1068
1069 ;; Go back over exactly one logical sexp, taking proper
1070 ;; account of macros and escaped EOLs.
1071 (while
1072 (progn
1073 (unless (c-safe (c-backward-sexp) t)
1074 ;; Give up if we hit an unbalanced block. Since the
1075 ;; stack won't be empty the code below will report a
1076 ;; suitable error.
1077 (throw 'loop nil))
1078 (cond
1079 ;; Have we moved into a macro?
1080 ((and (not macro-start)
1081 (c-beginning-of-macro))
1082 ;; Have we crossed a statement boundary? If not,
1083 ;; keep going back until we find one or a "real" sexp.
1084 (and
1085 (save-excursion
1086 (c-end-of-macro)
1087 (not (c-crosses-statement-barrier-p
1088 (point) maybe-after-boundary-pos)))
1089 (setq maybe-after-boundary-pos (point))))
1090 ;; Have we just gone back over an escaped NL? This
1091 ;; doesn't count as a sexp.
1092 ((looking-at "\\\\$")))))
1093
1094 ;; Have we crossed a statement boundary?
1095 (setq boundary-pos
1096 (cond
1097 ;; Are we at a macro beginning?
1098 ((and (not macro-start)
1099 c-opt-cpp-prefix
1100 (looking-at c-opt-cpp-prefix))
1101 (save-excursion
1102 (c-end-of-macro)
1103 (c-crosses-statement-barrier-p
1104 (point) maybe-after-boundary-pos)))
1105 ;; Just gone back over a brace block?
1106 ((and
1107 (eq (char-after) ?{)
1108 (not (c-looking-at-inexpr-block lim nil t))
1109 (save-excursion
1110 (c-backward-token-2 1 t nil)
1111 (not (looking-at "=\\([^=]\\|$\\)"))))
1112 (save-excursion
1113 (c-forward-sexp) (point)))
1114 ;; Just gone back over some paren block?
1115 ((looking-at "\\s(")
1116 (save-excursion
1117 (goto-char (1+ (c-down-list-backward
1118 before-sws-pos)))
1119 (c-crosses-statement-barrier-p
1120 (point) maybe-after-boundary-pos)))
1121 ;; Just gone back over an ordinary symbol of some sort?
1122 (t (c-crosses-statement-barrier-p
1123 (point) maybe-after-boundary-pos))))
1124
1125 (when boundary-pos
1126 (setq pptok ptok
1127 ptok tok
1128 tok boundary-pos
1129 sym 'boundary)
1130 ;; Like a C "continue". Analyze the next sexp.
1131 (throw 'loop t))))
1132
1133 ;; ObjC method def?
1134 (when (and c-opt-method-key
1135 (setq saved (c-in-method-def-p)))
1136 (setq pos saved
1137 ignore-labels t) ; Avoid the label check on exit.
1138 (throw 'loop nil)) ; 3. ObjC method def.
1139
1140 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1141 (if c-has-bitfields
1142 (cond
1143 ;; The : <size> and <id> fields?
1144 ((and (numberp c-maybe-labelp)
1145 (not bitfield-size-pos)
1146 (save-excursion
1147 (goto-char (or tok start))
1148 (not (looking-at c-keywords-regexp)))
1149 (not (looking-at c-keywords-regexp))
1150 (not (c-punctuation-in (point) c-maybe-labelp)))
1151 (setq bitfield-size-pos (or tok start)
1152 bitfield-id-pos (point)))
1153 ;; The <type> field?
1154 ((and bitfield-id-pos
1155 (not bitfield-type-pos))
1156 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1157 (not (looking-at c-not-primitive-type-keywords-regexp))
1158 (not (c-punctuation-in (point) tok)))
1159 (setq bitfield-type-pos (point))
1160 (setq bitfield-size-pos nil
1161 bitfield-id-pos nil)))))
1162
1163 ;; Handle labels.
1164 (unless (eq ignore-labels t)
1165 (when (numberp c-maybe-labelp)
1166 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1167 ;; might be in a label now. Have we got a real label
1168 ;; (including a case label) or something like C++'s "public:"?
1169 ;; A case label might use an expression rather than a token.
1170 (setq after-case:-pos (or tok start))
1171 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1172 ;; Catch C++'s inheritance construct "class foo : bar".
1173 (save-excursion
1174 (and
1175 (c-safe (c-backward-sexp) t)
1176 (looking-at c-nonlabel-token-2-key))))
1177 (setq c-maybe-labelp nil)
1178 (if after-labels-pos ; Have we already encountered a label?
1179 (if (not last-label-pos)
1180 (setq last-label-pos (or tok start)))
1181 (setq after-labels-pos (or tok start)))
1182 (setq c-maybe-labelp t
1183 label-good-pos nil))) ; bogus "label"
1184
1185 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1186 ; been found.
1187 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
1188 ;; We're in a potential label and it's the first
1189 ;; time we've found something that isn't allowed in
1190 ;; one.
1191 (setq label-good-pos (or tok start))))
1192
1193 ;; We've moved back by a sexp, so update the token positions.
1194 (setq sym nil
1195 pptok ptok
1196 ptok tok
1197 tok (point)
1198 pos tok) ; always non-nil
1199 ) ; end of (catch loop ....)
1200 ) ; end of sexp-at-a-time (while ....)
1201
1202 ;; If the stack isn't empty there might be errors to report.
1203 (while stack
1204 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1205 (c-bos-report-error))
1206 (setq saved-pos (cdr (car stack))
1207 stack (cdr stack)))
1208
1209 (when (and (eq ret 'same)
1210 (not (memq sym '(boundary ignore nil))))
1211 ;; Need to investigate closer whether we've crossed
1212 ;; between a substatement and its containing statement.
1213 (if (setq saved
1214 (cond ((and (looking-at c-block-stmt-1-2-key)
1215 (eq (char-after ptok) ?\())
1216 pptok)
1217 ((looking-at c-block-stmt-1-key)
1218 ptok)
1219 (t pptok)))
1220 (cond ((> start saved) (setq pos saved))
1221 ((= start saved) (setq ret 'up)))))
1222
1223 (when (and (not ignore-labels)
1224 (eq c-maybe-labelp t)
1225 (not (eq ret 'beginning))
1226 after-labels-pos
1227 (not bitfield-type-pos) ; Bitfields take precedence over labels.
1228 (or (not label-good-pos)
1229 (<= label-good-pos pos)
1230 (progn
1231 (goto-char (if (and last-label-pos
1232 (< last-label-pos start))
1233 last-label-pos
1234 pos))
1235 (looking-at c-label-kwds-regexp))))
1236 ;; We're in a label. Maybe we should step to the statement
1237 ;; after it.
1238 (if (< after-labels-pos start)
1239 (setq pos after-labels-pos)
1240 (setq ret 'label)
1241 (if (and last-label-pos (< last-label-pos start))
1242 ;; Might have jumped over several labels. Go to the last one.
1243 (setq pos last-label-pos)))))
1244
1245 ;; Have we got "case <expression>:"?
1246 (goto-char pos)
1247 (when (and after-case:-pos
1248 (not (eq ret 'beginning))
1249 (looking-at c-case-kwds-regexp))
1250 (if (< after-case:-pos start)
1251 (setq pos after-case:-pos))
1252 (if (eq ret 'same)
1253 (setq ret 'label)))
1254
1255 ;; Skip over the unary operators that can start the statement.
1256 (while (progn
1257 (c-backward-syntactic-ws)
1258 ;; protect AWK post-inc/decrement operators, etc.
1259 (and (not (c-at-vsemi-p (point)))
1260 (/= (skip-chars-backward "-+!*&~@`#") 0)))
1261 (setq pos (point)))
1262 (goto-char pos)
1263 ret)))
1264
1265 (defun c-punctuation-in (from to)
1266 "Return non-nil if there is a non-comment non-macro punctuation character
1267 between FROM and TO. FROM must not be in a string or comment. The returned
1268 value is the position of the first such character."
1269 (save-excursion
1270 (goto-char from)
1271 (let ((pos (point)))
1272 (while (progn (skip-chars-forward c-symbol-chars to)
1273 (c-forward-syntactic-ws to)
1274 (> (point) pos))
1275 (setq pos (point))))
1276 (and (< (point) to) (point))))
1277
1278 (defun c-crosses-statement-barrier-p (from to)
1279 "Return non-nil if buffer positions FROM to TO cross one or more
1280 statement or declaration boundaries. The returned value is actually
1281 the position of the earliest boundary char. FROM must not be within
1282 a string or comment.
1283
1284 The variable `c-maybe-labelp' is set to the position of the first `:' that
1285 might start a label (i.e. not part of `::' and not preceded by `?'). If a
1286 single `?' is found, then `c-maybe-labelp' is cleared.
1287
1288 For AWK, a statement which is terminated by an EOL (not a ; or a }) is
1289 regarded as having a \"virtual semicolon\" immediately after the last token on
1290 the line. If this virtual semicolon is _at_ from, the function recognizes it.
1291
1292 Note that this function might do hidden buffer changes. See the
1293 comment at the start of cc-engine.el for more info."
1294 (let* ((skip-chars
1295 ;; If the current language has CPP macros, insert # into skip-chars.
1296 (if c-opt-cpp-symbol
1297 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1298 c-opt-cpp-symbol ; usually "#"
1299 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1300 c-stmt-delim-chars))
1301 (non-skip-list
1302 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1303 lit-range lit-start vsemi-pos)
1304 (save-restriction
1305 (widen)
1306 (save-excursion
1307 (catch 'done
1308 (goto-char from)
1309 (while (progn (skip-chars-forward
1310 skip-chars
1311 (min to (c-point 'bonl)))
1312 (< (point) to))
1313 (cond
1314 ;; Virtual semicolon?
1315 ((and (bolp)
1316 (save-excursion
1317 (progn
1318 (if (setq lit-start (c-literal-start from)) ; Have we landed in a string/comment?
1319 (goto-char lit-start))
1320 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1321 (setq vsemi-pos (point))
1322 (c-at-vsemi-p))))
1323 (throw 'done vsemi-pos))
1324 ;; In a string/comment?
1325 ((setq lit-range (c-literal-limits from))
1326 (goto-char (cdr lit-range)))
1327 ((eq (char-after) ?:)
1328 (forward-char)
1329 (if (and (eq (char-after) ?:)
1330 (< (point) to))
1331 ;; Ignore scope operators.
1332 (forward-char)
1333 (setq c-maybe-labelp (1- (point)))))
1334 ((eq (char-after) ??)
1335 ;; A question mark. Can't be a label, so stop
1336 ;; looking for more : and ?.
1337 (setq c-maybe-labelp nil
1338 skip-chars (substring c-stmt-delim-chars 0 -2)))
1339 ;; At a CPP construct or a "#" or "##" operator?
1340 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol))
1341 (if (save-excursion
1342 (skip-chars-backward " \t")
1343 (and (bolp)
1344 (or (bobp)
1345 (not (eq (char-before (1- (point))) ?\\)))))
1346 (c-end-of-macro)
1347 (skip-chars-forward c-opt-cpp-symbol)))
1348 ((memq (char-after) non-skip-list)
1349 (throw 'done (point)))))
1350 ;; In trailing space after an as yet undetected virtual semicolon?
1351 (c-backward-syntactic-ws from)
1352 (when (and (bolp) (not (bobp))) ; Can happen in AWK Mode with an
1353 ; unterminated string/regexp.
1354 (backward-char))
1355 (if (and (< (point) to)
1356 (c-at-vsemi-p))
1357 (point)
1358 nil))))))
1359
1360 (defun c-at-statement-start-p ()
1361 "Return non-nil if the point is at the first token in a statement
1362 or somewhere in the syntactic whitespace before it.
1363
1364 A \"statement\" here is not restricted to those inside code blocks.
1365 Any kind of declaration-like construct that occur outside function
1366 bodies is also considered a \"statement\".
1367
1368 Note that this function might do hidden buffer changes. See the
1369 comment at the start of cc-engine.el for more info."
1370
1371 (save-excursion
1372 (let ((end (point))
1373 c-maybe-labelp)
1374 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1375 (or (bobp)
1376 (eq (char-before) ?})
1377 (and (eq (char-before) ?{)
1378 (not (and c-special-brace-lists
1379 (progn (backward-char)
1380 (c-looking-at-special-brace-list)))))
1381 (c-crosses-statement-barrier-p (point) end)))))
1382
1383 (defun c-at-expression-start-p ()
1384 "Return non-nil if the point is at the first token in an expression or
1385 statement, or somewhere in the syntactic whitespace before it.
1386
1387 An \"expression\" here is a bit different from the normal language
1388 grammar sense: It's any sequence of expression tokens except commas,
1389 unless they are enclosed inside parentheses of some kind. Also, an
1390 expression never continues past an enclosing parenthesis, but it might
1391 contain parenthesis pairs of any sort except braces.
1392
1393 Since expressions never cross statement boundaries, this function also
1394 recognizes statement beginnings, just like `c-at-statement-start-p'.
1395
1396 Note that this function might do hidden buffer changes. See the
1397 comment at the start of cc-engine.el for more info."
1398
1399 (save-excursion
1400 (let ((end (point))
1401 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1402 c-maybe-labelp)
1403 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1404 (or (bobp)
1405 (memq (char-before) '(?{ ?}))
1406 (save-excursion (backward-char)
1407 (looking-at "\\s("))
1408 (c-crosses-statement-barrier-p (point) end)))))
1409
1410 \f
1411 ;; A set of functions that covers various idiosyncrasies in
1412 ;; implementations of `forward-comment'.
1413
1414 ;; Note: Some emacsen considers incorrectly that any line comment
1415 ;; ending with a backslash continues to the next line. I can't think
1416 ;; of any way to work around that in a reliable way without changing
1417 ;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1418 ;; changing the syntax for backslash doesn't work since we must treat
1419 ;; escapes in string literals correctly.)
1420
1421 (defun c-forward-single-comment ()
1422 "Move forward past whitespace and the closest following comment, if any.
1423 Return t if a comment was found, nil otherwise. In either case, the
1424 point is moved past the following whitespace. Line continuations,
1425 i.e. a backslashes followed by line breaks, are treated as whitespace.
1426 The line breaks that end line comments are considered to be the
1427 comment enders, so the point will be put on the beginning of the next
1428 line if it moved past a line comment.
1429
1430 This function does not do any hidden buffer changes."
1431
1432 (let ((start (point)))
1433 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1434 (goto-char (match-end 0)))
1435
1436 (when (forward-comment 1)
1437 (if (eobp)
1438 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1439 ;; forwards at eob.
1440 nil
1441
1442 ;; Emacs includes the ending newline in a b-style (c++)
1443 ;; comment, but XEmacs doesn't. We depend on the Emacs
1444 ;; behavior (which also is symmetric).
1445 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1446 (condition-case nil (forward-char 1)))
1447
1448 t))))
1449
1450 (defsubst c-forward-comments ()
1451 "Move forward past all following whitespace and comments.
1452 Line continuations, i.e. a backslashes followed by line breaks, are
1453 treated as whitespace.
1454
1455 Note that this function might do hidden buffer changes. See the
1456 comment at the start of cc-engine.el for more info."
1457
1458 (while (or
1459 ;; If forward-comment in at least XEmacs 21 is given a large
1460 ;; positive value, it'll loop all the way through if it hits
1461 ;; eob.
1462 (and (forward-comment 5)
1463 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1464 ;; forwards at eob.
1465 (not (eobp)))
1466
1467 (when (looking-at "\\\\[\n\r]")
1468 (forward-char 2)
1469 t))))
1470
1471 (defun c-backward-single-comment ()
1472 "Move backward past whitespace and the closest preceding comment, if any.
1473 Return t if a comment was found, nil otherwise. In either case, the
1474 point is moved past the preceding whitespace. Line continuations,
1475 i.e. a backslashes followed by line breaks, are treated as whitespace.
1476 The line breaks that end line comments are considered to be the
1477 comment enders, so the point cannot be at the end of the same line to
1478 move over a line comment.
1479
1480 This function does not do any hidden buffer changes."
1481
1482 (let ((start (point)))
1483 ;; When we got newline terminated comments, forward-comment in all
1484 ;; supported emacsen so far will stop at eol of each line not
1485 ;; ending with a comment when moving backwards. This corrects for
1486 ;; that, and at the same time handles line continuations.
1487 (while (progn
1488 (skip-chars-backward " \t\n\r\f\v")
1489 (and (looking-at "[\n\r]")
1490 (eq (char-before) ?\\)))
1491 (backward-char))
1492
1493 (if (bobp)
1494 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1495 ;; backwards at bob.
1496 nil
1497
1498 ;; Leave point after the closest following newline if we've
1499 ;; backed up over any above, since forward-comment won't move
1500 ;; backward over a line comment if point is at the end of the
1501 ;; same line.
1502 (re-search-forward "\\=\\s *[\n\r]" start t)
1503
1504 (if (if (forward-comment -1)
1505 (if (eolp)
1506 ;; If forward-comment above succeeded and we're at eol
1507 ;; then the newline we moved over above didn't end a
1508 ;; line comment, so we give it another go.
1509 (forward-comment -1)
1510 t))
1511
1512 ;; Emacs <= 20 and XEmacs move back over the closer of a
1513 ;; block comment that lacks an opener.
1514 (if (looking-at "\\*/")
1515 (progn (forward-char 2) nil)
1516 t)))))
1517
1518 (defsubst c-backward-comments ()
1519 "Move backward past all preceding whitespace and comments.
1520 Line continuations, i.e. a backslashes followed by line breaks, are
1521 treated as whitespace. The line breaks that end line comments are
1522 considered to be the comment enders, so the point cannot be at the end
1523 of the same line to move over a line comment. Unlike
1524 c-backward-syntactic-ws, this function doesn't move back over
1525 preprocessor directives.
1526
1527 Note that this function might do hidden buffer changes. See the
1528 comment at the start of cc-engine.el for more info."
1529
1530 (let ((start (point)))
1531 (while (and
1532 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
1533 ;; return t when moving backwards at bob.
1534 (not (bobp))
1535
1536 (if (let (moved-comment)
1537 (while
1538 (and (not (setq moved-comment (forward-comment -1)))
1539 ;; Cope specifically with ^M^J here -
1540 ;; forward-comment sometimes gets stuck after ^Ms,
1541 ;; sometimes after ^M^J.
1542 (or
1543 (when (eq (char-before) ?\r)
1544 (backward-char)
1545 t)
1546 (when (and (eq (char-before) ?\n)
1547 (eq (char-before (1- (point))) ?\r))
1548 (backward-char 2)
1549 t))))
1550 moved-comment)
1551 (if (looking-at "\\*/")
1552 ;; Emacs <= 20 and XEmacs move back over the
1553 ;; closer of a block comment that lacks an opener.
1554 (progn (forward-char 2) nil)
1555 t)
1556
1557 ;; XEmacs treats line continuations as whitespace but
1558 ;; only in the backward direction, which seems a bit
1559 ;; odd. Anyway, this is necessary for Emacs.
1560 (when (and (looking-at "[\n\r]")
1561 (eq (char-before) ?\\)
1562 (< (point) start))
1563 (backward-char)
1564 t))))))
1565
1566 \f
1567 ;; Tools for skipping over syntactic whitespace.
1568
1569 ;; The following functions use text properties to cache searches over
1570 ;; large regions of syntactic whitespace. It works as follows:
1571 ;;
1572 ;; o If a syntactic whitespace region contains anything but simple
1573 ;; whitespace (i.e. space, tab and line breaks), the text property
1574 ;; `c-in-sws' is put over it. At places where we have stopped
1575 ;; within that region there's also a `c-is-sws' text property.
1576 ;; That since there typically are nested whitespace inside that
1577 ;; must be handled separately, e.g. whitespace inside a comment or
1578 ;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1579 ;; to jump to another point with that property within the same
1580 ;; `c-in-sws' region. It can be likened to a ladder where
1581 ;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1582 ;;
1583 ;; o The `c-is-sws' property is put on the simple whitespace chars at
1584 ;; a "rung position" and also maybe on the first following char.
1585 ;; As many characters as can be conveniently found in this range
1586 ;; are marked, but no assumption can be made that the whole range
1587 ;; is marked (it could be clobbered by later changes, for
1588 ;; instance).
1589 ;;
1590 ;; Note that some part of the beginning of a sequence of simple
1591 ;; whitespace might be part of the end of a preceding line comment
1592 ;; or cpp directive and must not be considered part of the "rung".
1593 ;; Such whitespace is some amount of horizontal whitespace followed
1594 ;; by a newline. In the case of cpp directives it could also be
1595 ;; two newlines with horizontal whitespace between them.
1596 ;;
1597 ;; The reason to include the first following char is to cope with
1598 ;; "rung positions" that don't have any ordinary whitespace. If
1599 ;; `c-is-sws' is put on a token character it does not have
1600 ;; `c-in-sws' set simultaneously. That's the only case when that
1601 ;; can occur, and the reason for not extending the `c-in-sws'
1602 ;; region to cover it is that the `c-in-sws' region could then be
1603 ;; accidentally merged with a following one if the token is only
1604 ;; one character long.
1605 ;;
1606 ;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1607 ;; removed in the changed region. If the change was inside
1608 ;; syntactic whitespace that means that the "ladder" is broken, but
1609 ;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1610 ;; parts on either side and use an ordinary search only to "repair"
1611 ;; the gap.
1612 ;;
1613 ;; Special care needs to be taken if a region is removed: If there
1614 ;; are `c-in-sws' on both sides of it which do not connect inside
1615 ;; the region then they can't be joined. If e.g. a marked macro is
1616 ;; broken, syntactic whitespace inside the new text might be
1617 ;; marked. If those marks would become connected with the old
1618 ;; `c-in-sws' range around the macro then we could get a ladder
1619 ;; with one end outside the macro and the other at some whitespace
1620 ;; within it.
1621 ;;
1622 ;; The main motivation for this system is to increase the speed in
1623 ;; skipping over the large whitespace regions that can occur at the
1624 ;; top level in e.g. header files that contain a lot of comments and
1625 ;; cpp directives. For small comments inside code it's probably
1626 ;; slower than using `forward-comment' straightforwardly, but speed is
1627 ;; not a significant factor there anyway.
1628
1629 ; (defface c-debug-is-sws-face
1630 ; '((t (:background "GreenYellow")))
1631 ; "Debug face to mark the `c-is-sws' property.")
1632 ; (defface c-debug-in-sws-face
1633 ; '((t (:underline t)))
1634 ; "Debug face to mark the `c-in-sws' property.")
1635
1636 ; (defun c-debug-put-sws-faces ()
1637 ; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1638 ; ;; properties in the buffer.
1639 ; (interactive)
1640 ; (save-excursion
1641 ; (c-save-buffer-state (in-face)
1642 ; (goto-char (point-min))
1643 ; (setq in-face (if (get-text-property (point) 'c-is-sws)
1644 ; (point)))
1645 ; (while (progn
1646 ; (goto-char (next-single-property-change
1647 ; (point) 'c-is-sws nil (point-max)))
1648 ; (if in-face
1649 ; (progn
1650 ; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1651 ; (setq in-face nil))
1652 ; (setq in-face (point)))
1653 ; (not (eobp))))
1654 ; (goto-char (point-min))
1655 ; (setq in-face (if (get-text-property (point) 'c-in-sws)
1656 ; (point)))
1657 ; (while (progn
1658 ; (goto-char (next-single-property-change
1659 ; (point) 'c-in-sws nil (point-max)))
1660 ; (if in-face
1661 ; (progn
1662 ; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1663 ; (setq in-face nil))
1664 ; (setq in-face (point)))
1665 ; (not (eobp)))))))
1666
1667 (defmacro c-debug-sws-msg (&rest args)
1668 ;;`(message ,@args)
1669 )
1670
1671 (defmacro c-put-is-sws (beg end)
1672 ;; This macro does a hidden buffer change.
1673 `(let ((beg ,beg) (end ,end))
1674 (put-text-property beg end 'c-is-sws t)
1675 ,@(when (facep 'c-debug-is-sws-face)
1676 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1677
1678 (defmacro c-put-in-sws (beg end)
1679 ;; This macro does a hidden buffer change.
1680 `(let ((beg ,beg) (end ,end))
1681 (put-text-property beg end 'c-in-sws t)
1682 ,@(when (facep 'c-debug-is-sws-face)
1683 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1684
1685 (defmacro c-remove-is-sws (beg end)
1686 ;; This macro does a hidden buffer change.
1687 `(let ((beg ,beg) (end ,end))
1688 (remove-text-properties beg end '(c-is-sws nil))
1689 ,@(when (facep 'c-debug-is-sws-face)
1690 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1691
1692 (defmacro c-remove-in-sws (beg end)
1693 ;; This macro does a hidden buffer change.
1694 `(let ((beg ,beg) (end ,end))
1695 (remove-text-properties beg end '(c-in-sws nil))
1696 ,@(when (facep 'c-debug-is-sws-face)
1697 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1698
1699 (defmacro c-remove-is-and-in-sws (beg end)
1700 ;; This macro does a hidden buffer change.
1701 `(let ((beg ,beg) (end ,end))
1702 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1703 ,@(when (facep 'c-debug-is-sws-face)
1704 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1705 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1706
1707 (defsubst c-invalidate-sws-region-after (beg end)
1708 ;; Called from `after-change-functions'. Note that if
1709 ;; `c-forward-sws' or `c-backward-sws' are used outside
1710 ;; `c-save-buffer-state' or similar then this will remove the cache
1711 ;; properties right after they're added.
1712 ;;
1713 ;; This function does hidden buffer changes.
1714
1715 (save-excursion
1716 ;; Adjust the end to remove the properties in any following simple
1717 ;; ws up to and including the next line break, if there is any
1718 ;; after the changed region. This is necessary e.g. when a rung
1719 ;; marked empty line is converted to a line comment by inserting
1720 ;; "//" before the line break. In that case the line break would
1721 ;; keep the rung mark which could make a later `c-backward-sws'
1722 ;; move into the line comment instead of over it.
1723 (goto-char end)
1724 (skip-chars-forward " \t\f\v")
1725 (when (and (eolp) (not (eobp)))
1726 (setq end (1+ (point)))))
1727
1728 (when (and (= beg end)
1729 (get-text-property beg 'c-in-sws)
1730 (> beg (point-min))
1731 (get-text-property (1- beg) 'c-in-sws))
1732 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1733 ;; safe to keep a range that was continuous before the change. E.g:
1734 ;;
1735 ;; #define foo
1736 ;; \
1737 ;; bar
1738 ;;
1739 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1740 ;; after "foo" is removed then "bar" will become part of the cpp
1741 ;; directive instead of a syntactically relevant token. In that
1742 ;; case there's no longer syntactic ws from "#" to "b".
1743 (setq beg (1- beg)))
1744
1745 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1746 (c-remove-is-and-in-sws beg end))
1747
1748 (defun c-forward-sws ()
1749 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
1750 ;;
1751 ;; This function might do hidden buffer changes.
1752
1753 (let (;; `rung-pos' is set to a position as early as possible in the
1754 ;; unmarked part of the simple ws region.
1755 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1756 rung-is-marked next-rung-is-marked simple-ws-end
1757 ;; `safe-start' is set when it's safe to cache the start position.
1758 ;; It's not set if we've initially skipped over comments and line
1759 ;; continuations since we might have gone out through the end of a
1760 ;; macro then. This provision makes `c-forward-sws' not populate the
1761 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1762 ;; more common.
1763 safe-start)
1764
1765 ;; Skip simple ws and do a quick check on the following character to see
1766 ;; if it's anything that can't start syntactic ws, so we can bail out
1767 ;; early in the majority of cases when there just are a few ws chars.
1768 (skip-chars-forward " \t\n\r\f\v")
1769 (when (or (looking-at c-syntactic-ws-start)
1770 (and c-opt-cpp-prefix
1771 (looking-at c-noise-macro-name-re)))
1772
1773 (setq rung-end-pos (min (1+ (point)) (point-max)))
1774 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1775 'c-is-sws t))
1776 ;; Find the last rung position to avoid setting properties in all
1777 ;; the cases when the marked rung is complete.
1778 ;; (`next-single-property-change' is certain to move at least one
1779 ;; step forward.)
1780 (setq rung-pos (1- (c-next-single-property-change
1781 rung-is-marked 'c-is-sws nil rung-end-pos)))
1782 ;; Got no marked rung here. Since the simple ws might have started
1783 ;; inside a line comment or cpp directive we must set `rung-pos' as
1784 ;; high as possible.
1785 (setq rung-pos (point)))
1786
1787 (with-silent-modifications
1788 (while
1789 (progn
1790 ;; In the following while form, we move over a "ladder" and
1791 ;; following simple WS each time round the loop, appending the WS
1792 ;; onto the ladder, joining adjacent ladders, and terminating when
1793 ;; there is no more WS or we reach EOB.
1794 (while
1795 (when (and rung-is-marked
1796 (get-text-property (point) 'c-in-sws))
1797
1798 ;; The following search is the main reason that `c-in-sws'
1799 ;; and `c-is-sws' aren't combined to one property.
1800 (goto-char (c-next-single-property-change
1801 (point) 'c-in-sws nil (point-max)))
1802 (unless (get-text-property (point) 'c-is-sws)
1803 ;; If the `c-in-sws' region extended past the last
1804 ;; `c-is-sws' char we have to go back a bit.
1805 (or (get-text-property (1- (point)) 'c-is-sws)
1806 (goto-char (previous-single-property-change
1807 (point) 'c-is-sws)))
1808 (backward-char))
1809
1810 (c-debug-sws-msg
1811 "c-forward-sws cached move %s -> %s (max %s)"
1812 rung-pos (point) (point-max))
1813
1814 (setq rung-pos (point))
1815 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1816 (not (eobp))))
1817
1818 ;; We'll loop here if there is simple ws after the last rung.
1819 ;; That means that there's been some change in it and it's
1820 ;; possible that we've stepped into another ladder, so extend
1821 ;; the previous one to join with it if there is one, and try to
1822 ;; use the cache again.
1823 (c-debug-sws-msg
1824 "c-forward-sws extending rung with [%s..%s] (max %s)"
1825 (1+ rung-pos) (1+ (point)) (point-max))
1826 (unless (get-text-property (point) 'c-is-sws)
1827 ;; Remove any `c-in-sws' property from the last char of
1828 ;; the rung before we mark it with `c-is-sws', so that we
1829 ;; won't connect with the remains of a broken "ladder".
1830 (c-remove-in-sws (point) (1+ (point))))
1831 (c-put-is-sws (1+ rung-pos)
1832 (1+ (point)))
1833 (c-put-in-sws rung-pos
1834 (setq rung-pos (point)
1835 last-put-in-sws-pos rung-pos)))
1836
1837 ;; Now move over any comments (x)or a CPP construct.
1838 (setq simple-ws-end (point))
1839 (c-forward-comments)
1840
1841 (cond
1842 ((/= (point) simple-ws-end)
1843 ;; Skipped over comments. Don't cache at eob in case the buffer
1844 ;; is narrowed.
1845 (not (eobp)))
1846
1847 ((save-excursion
1848 (and c-opt-cpp-prefix
1849 (looking-at c-opt-cpp-start)
1850 (progn (skip-chars-backward " \t")
1851 (bolp))
1852 (or (bobp)
1853 (progn (backward-char)
1854 (not (eq (char-before) ?\\))))))
1855 ;; Skip a preprocessor directive.
1856 (end-of-line)
1857 (while (and (eq (char-before) ?\\)
1858 (= (forward-line 1) 0))
1859 (end-of-line))
1860 (forward-line 1)
1861 (setq safe-start t)
1862 ;; Don't cache at eob in case the buffer is narrowed.
1863 (not (eobp)))
1864
1865 ((and c-opt-cpp-prefix
1866 (looking-at c-noise-macro-name-re))
1867 ;; Skip over a noise macro.
1868 (goto-char (match-end 1))
1869 (setq safe-start t)
1870 (not (eobp)))))
1871
1872 ;; We've searched over a piece of non-white syntactic ws. See if this
1873 ;; can be cached.
1874 (setq next-rung-pos (point))
1875 (skip-chars-forward " \t\n\r\f\v")
1876 (setq rung-end-pos (min (1+ (point)) (point-max)))
1877
1878 (if (or
1879 ;; Cache if we haven't skipped comments only, and if we started
1880 ;; either from a marked rung or from a completely uncached
1881 ;; position.
1882 (and safe-start
1883 (or rung-is-marked
1884 (not (get-text-property simple-ws-end 'c-in-sws))))
1885
1886 ;; See if there's a marked rung in the encountered simple ws. If
1887 ;; so then we can cache, unless `safe-start' is nil. Even then
1888 ;; we need to do this to check if the cache can be used for the
1889 ;; next step.
1890 (and (setq next-rung-is-marked
1891 (text-property-any next-rung-pos rung-end-pos
1892 'c-is-sws t))
1893 safe-start))
1894
1895 (progn
1896 (c-debug-sws-msg
1897 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1898 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1899 (point-max))
1900
1901 ;; Remove the properties for any nested ws that might be cached.
1902 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1903 ;; anyway.
1904 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1905 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1906 (c-put-is-sws rung-pos
1907 (1+ simple-ws-end))
1908 (setq rung-is-marked t))
1909 (c-put-in-sws rung-pos
1910 (setq rung-pos (point)
1911 last-put-in-sws-pos rung-pos))
1912 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1913 ;; Remove any `c-in-sws' property from the last char of
1914 ;; the rung before we mark it with `c-is-sws', so that we
1915 ;; won't connect with the remains of a broken "ladder".
1916 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1917 (c-put-is-sws next-rung-pos
1918 rung-end-pos))
1919
1920 (c-debug-sws-msg
1921 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1922 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1923 (point-max))
1924
1925 ;; Set `rung-pos' for the next rung. It's the same thing here as
1926 ;; initially, except that the rung position is set as early as
1927 ;; possible since we can't be in the ending ws of a line comment or
1928 ;; cpp directive now.
1929 (if (setq rung-is-marked next-rung-is-marked)
1930 (setq rung-pos (1- (c-next-single-property-change
1931 rung-is-marked 'c-is-sws nil rung-end-pos)))
1932 (setq rung-pos next-rung-pos))
1933 (setq safe-start t)))
1934
1935 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1936 ;; another one after the point (which might occur when editing inside a
1937 ;; comment or macro).
1938 (when (eq last-put-in-sws-pos (point))
1939 (cond ((< last-put-in-sws-pos (point-max))
1940 (c-debug-sws-msg
1941 "c-forward-sws clearing at %s for cache separation"
1942 last-put-in-sws-pos)
1943 (c-remove-in-sws last-put-in-sws-pos
1944 (1+ last-put-in-sws-pos)))
1945 (t
1946 ;; If at eob we have to clear the last character before the end
1947 ;; instead since the buffer might be narrowed and there might
1948 ;; be a `c-in-sws' after (point-max). In this case it's
1949 ;; necessary to clear both properties.
1950 (c-debug-sws-msg
1951 "c-forward-sws clearing thoroughly at %s for cache separation"
1952 (1- last-put-in-sws-pos))
1953 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1954 last-put-in-sws-pos))))
1955 ))))
1956
1957 (defun c-backward-sws ()
1958 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
1959 ;;
1960 ;; This function might do hidden buffer changes.
1961
1962 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1963 ;; part of the simple ws region.
1964 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1965 rung-is-marked simple-ws-beg cmt-skip-pos)
1966
1967 ;; Skip simple horizontal ws and do a quick check on the preceding
1968 ;; character to see if it's anything that can't end syntactic ws, so we can
1969 ;; bail out early in the majority of cases when there just are a few ws
1970 ;; chars. Newlines are complicated in the backward direction, so we can't
1971 ;; skip over them.
1972 (skip-chars-backward " \t\f")
1973 (when (and (not (bobp))
1974 (save-excursion
1975 (backward-char)
1976 (or (looking-at c-syntactic-ws-end)
1977 (and c-opt-cpp-prefix
1978 (looking-at c-symbol-char-key)
1979 (progn (c-beginning-of-current-token)
1980 (looking-at c-noise-macro-name-re))))))
1981 ;; Try to find a rung position in the simple ws preceding point, so that
1982 ;; we can get a cache hit even if the last bit of the simple ws has
1983 ;; changed recently.
1984 (setq simple-ws-beg (point))
1985 (skip-chars-backward " \t\n\r\f\v")
1986 (if (setq rung-is-marked (text-property-any
1987 (point) (min (1+ rung-pos) (point-max))
1988 'c-is-sws t))
1989 ;; `rung-pos' will be the earliest marked position, which means that
1990 ;; there might be later unmarked parts in the simple ws region.
1991 ;; It's not worth the effort to fix that; the last part of the
1992 ;; simple ws is also typically edited often, so it could be wasted.
1993 (goto-char (setq rung-pos rung-is-marked))
1994 (goto-char simple-ws-beg))
1995
1996 (with-silent-modifications
1997 (while
1998 (progn
1999 ;; Each time round the next while form, we move back over a ladder
2000 ;; and append any simple WS preceding it, if possible joining with
2001 ;; the previous ladder.
2002 (while
2003 (when (and rung-is-marked
2004 (not (bobp))
2005 (get-text-property (1- (point)) 'c-in-sws))
2006
2007 ;; The following search is the main reason that `c-in-sws'
2008 ;; and `c-is-sws' aren't combined to one property.
2009 (goto-char (previous-single-property-change
2010 (point) 'c-in-sws nil (point-min)))
2011 (unless (get-text-property (point) 'c-is-sws)
2012 ;; If the `c-in-sws' region extended past the first
2013 ;; `c-is-sws' char we have to go forward a bit.
2014 (goto-char (c-next-single-property-change
2015 (point) 'c-is-sws)))
2016
2017 (c-debug-sws-msg
2018 "c-backward-sws cached move %s <- %s (min %s)"
2019 (point) rung-pos (point-min))
2020
2021 (setq rung-pos (point))
2022 (if (and (< (min (skip-chars-backward " \t\f\v")
2023 (progn
2024 (setq simple-ws-beg (point))
2025 (skip-chars-backward " \t\n\r\f\v")))
2026 0)
2027 (setq rung-is-marked
2028 (text-property-any (point) rung-pos
2029 'c-is-sws t)))
2030 t
2031 (goto-char simple-ws-beg)
2032 nil))
2033
2034 ;; We'll loop here if there is simple ws before the first rung.
2035 ;; That means that there's been some change in it and it's
2036 ;; possible that we've stepped into another ladder, so extend
2037 ;; the previous one to join with it if there is one, and try to
2038 ;; use the cache again.
2039 (c-debug-sws-msg
2040 "c-backward-sws extending rung with [%s..%s] (min %s)"
2041 rung-is-marked rung-pos (point-min))
2042 (unless (get-text-property (1- rung-pos) 'c-is-sws)
2043 ;; Remove any `c-in-sws' property from the last char of
2044 ;; the rung before we mark it with `c-is-sws', so that we
2045 ;; won't connect with the remains of a broken "ladder".
2046 (c-remove-in-sws (1- rung-pos) rung-pos))
2047 (c-put-is-sws rung-is-marked
2048 rung-pos)
2049 (c-put-in-sws rung-is-marked
2050 (1- rung-pos))
2051 (setq rung-pos rung-is-marked
2052 last-put-in-sws-pos rung-pos))
2053
2054 (c-backward-comments)
2055 (setq cmt-skip-pos (point))
2056
2057 (cond
2058 ((and c-opt-cpp-prefix
2059 (/= cmt-skip-pos simple-ws-beg)
2060 (c-beginning-of-macro))
2061 ;; Inside a cpp directive. See if it should be skipped over.
2062 (let ((cpp-beg (point)))
2063
2064 ;; Move back over all line continuations in the region skipped
2065 ;; over by `c-backward-comments'. If we go past it then we
2066 ;; started inside the cpp directive.
2067 (goto-char simple-ws-beg)
2068 (beginning-of-line)
2069 (while (and (> (point) cmt-skip-pos)
2070 (progn (backward-char)
2071 (eq (char-before) ?\\)))
2072 (beginning-of-line))
2073
2074 (if (< (point) cmt-skip-pos)
2075 ;; Don't move past the cpp directive if we began inside
2076 ;; it. Note that the position at the end of the last line
2077 ;; of the macro is also considered to be within it.
2078 (progn (goto-char cmt-skip-pos)
2079 nil)
2080
2081 ;; It's worthwhile to spend a little bit of effort on finding
2082 ;; the end of the macro, to get a good `simple-ws-beg'
2083 ;; position for the cache. Note that `c-backward-comments'
2084 ;; could have stepped over some comments before going into
2085 ;; the macro, and then `simple-ws-beg' must be kept on the
2086 ;; same side of those comments.
2087 (goto-char simple-ws-beg)
2088 (skip-chars-backward " \t\n\r\f\v")
2089 (if (eq (char-before) ?\\)
2090 (forward-char))
2091 (forward-line 1)
2092 (if (< (point) simple-ws-beg)
2093 ;; Might happen if comments after the macro were skipped
2094 ;; over.
2095 (setq simple-ws-beg (point)))
2096
2097 (goto-char cpp-beg)
2098 t)))
2099
2100 ((/= (save-excursion
2101 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
2102 (setq next-rung-pos (point)))
2103 simple-ws-beg)
2104 ;; Skipped over comments. Must put point at the end of
2105 ;; the simple ws at point since we might be after a line
2106 ;; comment or cpp directive that's been partially
2107 ;; narrowed out, and we can't risk marking the simple ws
2108 ;; at the end of it.
2109 (goto-char next-rung-pos)
2110 t)
2111
2112 ((and c-opt-cpp-prefix
2113 (save-excursion
2114 (and (< (skip-syntax-backward "w_") 0)
2115 (progn (setq next-rung-pos (point))
2116 (looking-at c-noise-macro-name-re)))))
2117 ;; Skipped over a noise macro
2118 (goto-char next-rung-pos)
2119 t)))
2120
2121 ;; We've searched over a piece of non-white syntactic ws. See if this
2122 ;; can be cached.
2123 (setq next-rung-pos (point))
2124 (skip-chars-backward " \t\f\v")
2125
2126 (if (or
2127 ;; Cache if we started either from a marked rung or from a
2128 ;; completely uncached position.
2129 rung-is-marked
2130 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
2131
2132 ;; Cache if there's a marked rung in the encountered simple ws.
2133 (save-excursion
2134 (skip-chars-backward " \t\n\r\f\v")
2135 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
2136 'c-is-sws t)))
2137
2138 (progn
2139 (c-debug-sws-msg
2140 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
2141 (point) (1+ next-rung-pos)
2142 simple-ws-beg (min (1+ rung-pos) (point-max))
2143 (point-min))
2144
2145 ;; Remove the properties for any nested ws that might be cached.
2146 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
2147 ;; anyway.
2148 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
2149 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
2150 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
2151 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
2152 ;; Remove any `c-in-sws' property from the last char of
2153 ;; the rung before we mark it with `c-is-sws', so that we
2154 ;; won't connect with the remains of a broken "ladder".
2155 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
2156 (c-put-is-sws simple-ws-beg
2157 rung-end-pos)
2158 (setq rung-is-marked t)))
2159 (c-put-in-sws (setq simple-ws-beg (point)
2160 last-put-in-sws-pos simple-ws-beg)
2161 rung-pos)
2162 (c-put-is-sws (setq rung-pos simple-ws-beg)
2163 (1+ next-rung-pos)))
2164
2165 (c-debug-sws-msg
2166 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
2167 (point) (1+ next-rung-pos)
2168 simple-ws-beg (min (1+ rung-pos) (point-max))
2169 (point-min))
2170 (setq rung-pos next-rung-pos
2171 simple-ws-beg (point))
2172 ))
2173
2174 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
2175 ;; another one before the point (which might occur when editing inside a
2176 ;; comment or macro).
2177 (when (eq last-put-in-sws-pos (point))
2178 (cond ((< (point-min) last-put-in-sws-pos)
2179 (c-debug-sws-msg
2180 "c-backward-sws clearing at %s for cache separation"
2181 (1- last-put-in-sws-pos))
2182 (c-remove-in-sws (1- last-put-in-sws-pos)
2183 last-put-in-sws-pos))
2184 ((> (point-min) 1)
2185 ;; If at bob and the buffer is narrowed, we have to clear the
2186 ;; character we're standing on instead since there might be a
2187 ;; `c-in-sws' before (point-min). In this case it's necessary
2188 ;; to clear both properties.
2189 (c-debug-sws-msg
2190 "c-backward-sws clearing thoroughly at %s for cache separation"
2191 last-put-in-sws-pos)
2192 (c-remove-is-and-in-sws last-put-in-sws-pos
2193 (1+ last-put-in-sws-pos)))))
2194 ))))
2195
2196 \f
2197 ;; Other whitespace tools
2198 (defun c-partial-ws-p (beg end)
2199 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2200 ;; region? This is a "heuristic" function. .....
2201 ;;
2202 ;; The motivation for the second bit is to check whether removing this
2203 ;; region would coalesce two symbols.
2204 ;;
2205 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2206 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2207 (save-excursion
2208 (let ((end+1 (min (1+ end) (point-max))))
2209 (or (progn (goto-char (max (point-min) (1- beg)))
2210 (c-skip-ws-forward end)
2211 (eq (point) end))
2212 (progn (goto-char beg)
2213 (c-skip-ws-forward end+1)
2214 (eq (point) end+1))))))
2215 \f
2216 ;; A system for finding noteworthy parens before the point.
2217
2218 (defconst c-state-cache-too-far 5000)
2219 ;; A maximum comfortable scanning distance, e.g. between
2220 ;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
2221 ;; this distance is exceeded, we take "emergency measures", e.g. by clearing
2222 ;; the cache and starting again from point-min or a beginning of defun. This
2223 ;; value can be tuned for efficiency or set to a lower value for testing.
2224
2225 (defvar c-state-cache nil)
2226 (make-variable-buffer-local 'c-state-cache)
2227 ;; The state cache used by `c-parse-state' to cut down the amount of
2228 ;; searching. It's the result from some earlier `c-parse-state' call. See
2229 ;; `c-parse-state''s doc string for details of its structure.
2230 ;;
2231 ;; The use of the cached info is more effective if the next
2232 ;; `c-parse-state' call is on a line close by the one the cached state
2233 ;; was made at; the cache can actually slow down a little if the
2234 ;; cached state was made very far back in the buffer. The cache is
2235 ;; most effective if `c-parse-state' is used on each line while moving
2236 ;; forward.
2237
2238 (defvar c-state-cache-good-pos 1)
2239 (make-variable-buffer-local 'c-state-cache-good-pos)
2240 ;; This is a position where `c-state-cache' is known to be correct, or
2241 ;; nil (see below). It's a position inside one of the recorded unclosed
2242 ;; parens or the top level, but not further nested inside any literal or
2243 ;; subparen that is closed before the last recorded position.
2244 ;;
2245 ;; The exact position is chosen to try to be close to yet earlier than
2246 ;; the position where `c-state-cache' will be called next. Right now
2247 ;; the heuristic is to set it to the position after the last found
2248 ;; closing paren (of any type) before the line on which
2249 ;; `c-parse-state' was called. That is chosen primarily to work well
2250 ;; with refontification of the current line.
2251 ;;
2252 ;; 2009-07-28: When `c-state-point-min' and the last position where
2253 ;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2254 ;; both in the same literal, there is no such "good position", and
2255 ;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2256 ;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2257 ;;
2258 ;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2259 ;; the middle of the desert, as long as it is not within a brace pair
2260 ;; recorded in `c-state-cache' or a paren/bracket pair.
2261
2262 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2263 ;; We maintain a simple cache of positions which aren't in a literal, so as to
2264 ;; speed up testing for non-literality.
2265 (defconst c-state-nonlit-pos-interval 3000)
2266 ;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2267
2268 (defvar c-state-nonlit-pos-cache nil)
2269 (make-variable-buffer-local 'c-state-nonlit-pos-cache)
2270 ;; A list of buffer positions which are known not to be in a literal or a cpp
2271 ;; construct. This is ordered with higher positions at the front of the list.
2272 ;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
2273
2274 (defvar c-state-nonlit-pos-cache-limit 1)
2275 (make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2276 ;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2277 ;; reduced by buffer changes, and increased by invocations of
2278 ;; `c-state-literal-at'.
2279
2280 (defvar c-state-semi-nonlit-pos-cache nil)
2281 (make-variable-buffer-local 'c-state-semi-nonlit-pos-cache)
2282 ;; A list of elements which are either buffer positions (when such positions
2283 ;; are not in literals) or lists of the form (POS TYPE START), where POS is
2284 ;; a buffer position inside a literal, TYPE is the type of the literal
2285 ;; ('string, 'c, or 'c++) and START is the start of the literal.
2286
2287 (defvar c-state-semi-nonlit-pos-cache-limit 1)
2288 (make-variable-buffer-local 'c-state-semi-nonlit-pos-cache-limit)
2289 ;; An upper limit on valid entries in `c-state-semi-nonlit-pos-cache'. This
2290 ;; is reduced by buffer changes, and increased by invocations of
2291 ;; `c-parse-ps-state-below'.
2292
2293 (defsubst c-truncate-semi-nonlit-pos-cache (pos)
2294 ;; Truncate the upper bound of the cache `c-state-semi-nonlit-pos-cache' to
2295 ;; POS, if it is higher than that position.
2296 (setq c-state-semi-nonlit-pos-cache-limit
2297 (min c-state-semi-nonlit-pos-cache-limit pos)))
2298
2299 (defun c-state-semi-pp-to-literal (here &optional not-in-delimiter)
2300 ;; Do a parse-partial-sexp from a position in the buffer before HERE which
2301 ;; isn't in a literal, and return information about HERE, either:
2302 ;; (STATE TYPE BEG) if HERE is in a literal; or
2303 ;; (STATE) otherwise,
2304 ;; where STATE is the parsing state at HERE, TYPE is the type of the literal
2305 ;; enclosing HERE, (one of 'string, 'c, 'c++) and BEG is the starting
2306 ;; position of that literal (including the delimiter).
2307 ;;
2308 ;; Unless NOT-IN-DELIMITER is non-nil, when TO is inside a two-character
2309 ;; comment opener, this is recognized as being in a comment literal.
2310 ;;
2311 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote), 7
2312 ;; (comment type), and 8 (start of comment/string), and possibly 10 (in
2313 ;; newer Emacsen only, the syntax of a position after a potential first char
2314 ;; of a two char construct) of STATE are valid.
2315 (save-excursion
2316 (save-match-data
2317 (let* ((base-and-state (c-parse-ps-state-below here))
2318 (base (car base-and-state))
2319 (s (cdr base-and-state))
2320 (s (parse-partial-sexp base here nil nil s))
2321 ty)
2322 (cond
2323 ((or (nth 3 s) (nth 4 s)) ; in a string or comment
2324 (setq ty (cond
2325 ((nth 3 s) 'string)
2326 ((nth 7 s) 'c++)
2327 (t 'c)))
2328 (list s ty (nth 8 s)))
2329
2330 ((and (not not-in-delimiter) ; inside a comment starter
2331 (not (bobp))
2332 (progn (backward-char)
2333 (and (not (and (memq 'category-properties c-emacs-features)
2334 (looking-at "\\s!")))
2335 (looking-at c-comment-start-regexp))))
2336 (setq ty (if (looking-at c-block-comment-start-regexp) 'c 'c++))
2337 (list s ty (point)))
2338
2339 (t (list s)))))))
2340
2341 (defun c-state-full-pp-to-literal (here &optional not-in-delimiter)
2342 ;; This function will supersede c-state-pp-to-literal.
2343 ;;
2344 ;; Do a parse-partial-sexp from a position in the buffer before HERE which
2345 ;; isn't in a literal, and return information about HERE, either:
2346 ;; (STATE TYPE (BEG . END)) if HERE is in a literal; or
2347 ;; (STATE) otherwise,
2348 ;; where STATE is the parsing state at HERE, TYPE is the type of the literal
2349 ;; enclosing HERE, (one of 'string, 'c, 'c++) and (BEG . END) is the
2350 ;; boundaries of that literal (including the delimiters).
2351 ;;
2352 ;; Unless NOT-IN-DELIMITER is non-nil, when TO is inside a two-character
2353 ;; comment opener, this is recognized as being in a comment literal.
2354 ;;
2355 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote), 7
2356 ;; (comment type), and 8 (start of comment/string), and possibly 10 (in
2357 ;; newer Emacsen only, the syntax of a position after a potential first char
2358 ;; of a two char construct) of STATE are valid.
2359 (save-excursion
2360 (save-match-data
2361 (let* ((base-and-state (c-parse-ps-state-below here))
2362 (base (car base-and-state))
2363 (s (cdr base-and-state))
2364 (s (parse-partial-sexp base here nil nil s))
2365 ty start)
2366 (cond
2367 ((or (nth 3 s) (nth 4 s)) ; in a string or comment
2368 (setq ty (cond
2369 ((nth 3 s) 'string)
2370 ((nth 7 s) 'c++)
2371 (t 'c)))
2372 (setq start (nth 8 s))
2373 (parse-partial-sexp here (point-max)
2374 nil ; TARGETDEPTH
2375 nil ; STOPBEFORE
2376 s ; OLDSTATE
2377 'syntax-table) ; stop at end of literal
2378 (list s ty (cons start (point))))
2379
2380 ((and (not not-in-delimiter) ; inside a comment starter
2381 (not (bobp))
2382 (progn (backward-char)
2383 (and (not (and (memq 'category-properties c-emacs-features)
2384 (looking-at "\\s!")))
2385 (looking-at c-comment-start-regexp))))
2386 (setq ty (if (looking-at c-block-comment-start-regexp) 'c 'c++)
2387 start (point))
2388 (forward-comment 1)
2389 (list s ty (cons start (point))))
2390
2391 (t (list s)))))))
2392
2393 (defsubst c-state-pp-to-literal (from to &optional not-in-delimiter)
2394 ;; Do a parse-partial-sexp from FROM to TO, returning either
2395 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2396 ;; (STATE) otherwise,
2397 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2398 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal,
2399 ;; including the delimiters.
2400 ;;
2401 ;; Unless NOT-IN-DELIMITER is non-nil, when TO is inside a two-character
2402 ;; comment opener, this is recognized as being in a comment literal.
2403 ;;
2404 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2405 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2406 ;; STATE are valid.
2407 (save-excursion
2408 (save-match-data
2409 (let ((s (parse-partial-sexp from to))
2410 ty co-st)
2411 (cond
2412 ((or (nth 3 s) (nth 4 s)) ; in a string or comment
2413 (setq ty (cond
2414 ((nth 3 s) 'string)
2415 ((nth 7 s) 'c++)
2416 (t 'c)))
2417 (parse-partial-sexp (point) (point-max)
2418 nil ; TARGETDEPTH
2419 nil ; STOPBEFORE
2420 s ; OLDSTATE
2421 'syntax-table) ; stop at end of literal
2422 `(,s ,ty (,(nth 8 s) . ,(point))))
2423
2424 ((and (not not-in-delimiter) ; inside a comment starter
2425 (not (bobp))
2426 (progn (backward-char)
2427 (and (not (looking-at "\\s!"))
2428 (looking-at c-comment-start-regexp))))
2429 (setq ty (if (looking-at c-block-comment-start-regexp) 'c 'c++)
2430 co-st (point))
2431 (forward-comment 1)
2432 `(,s ,ty (,co-st . ,(point))))
2433
2434 (t `(,s)))))))
2435
2436 (defun c-cache-to-parse-ps-state (elt)
2437 ;; Create a list suitable to use as the old-state parameter to
2438 ;; `parse-partial-sexp', out of ELT. ELT is either just a number, a buffer
2439 ;; position, or it is a list (POS TYPE STARTING-POS). Here POS is the
2440 ;; buffer position the other elements are pertinent for, TYPE is either 'c
2441 ;; or 'c++ (for a comment) or a character (for a string delimiter) or t
2442 ;; (meaning a string fence opened the string), STARTING-POS is the starting
2443 ;; position of the comment or string.
2444 (if (consp elt)
2445 (let ((depth 0) (containing nil) (last nil)
2446 in-string in-comment (after-quote nil)
2447 (min-depth 0) com-style com-str-start (intermediate nil)
2448 (between-syntax nil)
2449 (type (cadr elt)))
2450 (setq com-str-start (car (cddr elt)))
2451 (cond
2452 ((or (numberp type) (eq type t)) ; A string
2453 (setq in-string type))
2454 ((memq type '(c c++)) ; A comment
2455 (setq in-comment t
2456 com-style (if (eq type 'c++) 1 nil)))
2457 (t (c-benign-error "Invalid type %s in c-cache-to-parse-ps-state"
2458 elt)))
2459 (list depth containing last
2460 in-string in-comment after-quote
2461 min-depth com-style com-str-start
2462 intermediate nil))
2463 (copy-tree '(0 nil nil nil nil nil 0 nil nil nil nil))))
2464
2465 (defun c-parse-ps-state-to-cache (state)
2466 ;; Convert STATE, a `parse-partial-sexp' state valid at POINT, to an element
2467 ;; for the `c-state-semi-nonlit-pos-cache' cache. This is either POINT
2468 ;; (when point is not in a literal) or a list (POINT TYPE STARTING-POS),
2469 ;; where TYPE is the type of the literal, either 'string, 'c, or 'c++, and
2470 ;; STARTING-POS is the starting position of the comment or string.
2471 (cond
2472 ((nth 3 state) ; A string
2473 (list (point) (nth 3 state) (nth 8 state)))
2474 ((nth 4 state) ; A comment
2475 (list (point)
2476 (if (eq (nth 7 state) 1) 'c++ 'c)
2477 (nth 8 state)))
2478 (t ; Neither string nor comment.
2479 (point))))
2480
2481 (defsubst c-ps-state-cache-pos (elt)
2482 ;; Get the buffer position from ELT, an element from the cache
2483 ;; `c-state-semi-nonlit-pos-cache'.
2484 (if (atom elt)
2485 elt
2486 (car elt)))
2487
2488 (defun c-parse-ps-state-below (here)
2489 ;; Given a buffer position HERE, Return a cons (CACHE-POS . STATE), where
2490 ;; CACHE-POS is a position not very far before HERE for which the
2491 ;; parse-partial-sexp STATE is valid. Note that the only valid elements of
2492 ;; STATE are those concerning comments and strings; STATE is the state of a
2493 ;; null `parse-partial-sexp' scan when CACHE-POS is not in a comment or
2494 ;; string.
2495 (save-restriction
2496 (widen)
2497 (save-excursion
2498 (let ((c c-state-semi-nonlit-pos-cache)
2499 elt state pos npos high-elt)
2500 ;; Trim the cache to take account of buffer changes.
2501 (while (and c (> (c-ps-state-cache-pos (car c))
2502 c-state-semi-nonlit-pos-cache-limit))
2503 (setq c (cdr c)))
2504 (setq c-state-semi-nonlit-pos-cache c)
2505
2506 (while (and c (> (c-ps-state-cache-pos (car c)) here))
2507 (setq high-elt (car c))
2508 (setq c (cdr c)))
2509 (setq pos (or (and c (c-ps-state-cache-pos (car c)))
2510 (point-min)))
2511
2512 (if high-elt
2513 (setq state (c-cache-to-parse-ps-state (car c)))
2514 (setq elt (if c (car c) (point-min)))
2515 (setq state
2516 (if c
2517 (c-cache-to-parse-ps-state (car c))
2518 (copy-tree '(0 nil nil nil nil nil 0 nil nil nil nil))))
2519 (while
2520 ;; Add an element to `c-state-semi-nonlit-pos-cache' each iteration.
2521 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2522 (setq state (parse-partial-sexp pos npos nil nil state))
2523 (setq elt (c-parse-ps-state-to-cache state))
2524 (setq c-state-semi-nonlit-pos-cache
2525 (cons elt c-state-semi-nonlit-pos-cache))
2526 (setq pos npos)))
2527
2528 (if (> pos c-state-semi-nonlit-pos-cache-limit)
2529 (setq c-state-semi-nonlit-pos-cache-limit pos))
2530
2531 (cons pos state)))))
2532
2533 (defun c-state-safe-place (here)
2534 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2535 ;; string, comment, or macro.
2536 ;;
2537 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2538 ;; MAY NOT contain any positions within macros, since macros are frequently
2539 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2540 ;; We cannot rely on this mechanism whilst determining a cache pos since
2541 ;; this function is also called from outwith `c-parse-state'.
2542 (save-restriction
2543 (widen)
2544 (save-excursion
2545 (let ((c c-state-nonlit-pos-cache)
2546 pos npos high-pos lit macro-beg macro-end)
2547 ;; Trim the cache to take account of buffer changes.
2548 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2549 (setq c (cdr c)))
2550 (setq c-state-nonlit-pos-cache c)
2551
2552 (while (and c (> (car c) here))
2553 (setq high-pos (car c))
2554 (setq c (cdr c)))
2555 (setq pos (or (car c) (point-min)))
2556
2557 (unless high-pos
2558 (while
2559 ;; Add an element to `c-state-nonlit-pos-cache' each iteration.
2560 (and
2561 (setq npos
2562 (when (<= (+ pos c-state-nonlit-pos-interval) here)
2563 (+ pos c-state-nonlit-pos-interval)))
2564
2565 ;; Test for being in a literal. If so, go to after it.
2566 (progn
2567 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2568 (or (null lit)
2569 (prog1 (<= (cdr lit) here)
2570 (setq npos (cdr lit)))))
2571
2572 ;; Test for being in a macro. If so, go to after it.
2573 (progn
2574 (goto-char npos)
2575 (setq macro-beg
2576 (and (c-beginning-of-macro) (/= (point) npos) (point)))
2577 (when macro-beg
2578 (c-syntactic-end-of-macro)
2579 (or (eobp) (forward-char))
2580 (setq macro-end (point)))
2581 (or (null macro-beg)
2582 (prog1 (<= macro-end here)
2583 (setq npos macro-end)))))
2584
2585 (setq pos npos)
2586 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2587 ;; Add one extra element above HERE so as to to avoid the previous
2588 ;; expensive calculation when the next call is close to the current
2589 ;; one. This is especially useful when inside a large macro.
2590 (when npos
2591 (setq c-state-nonlit-pos-cache
2592 (cons npos c-state-nonlit-pos-cache))))
2593
2594 (if (> pos c-state-nonlit-pos-cache-limit)
2595 (setq c-state-nonlit-pos-cache-limit pos))
2596 pos))))
2597
2598 (defun c-state-literal-at (here)
2599 ;; If position HERE is inside a literal, return (START . END), the
2600 ;; boundaries of the literal (which may be outside the accessible bit of the
2601 ;; buffer). Otherwise, return nil.
2602 ;;
2603 ;; This function is almost the same as `c-literal-limits'. Previously, it
2604 ;; differed in that it was a lower level function, and that it rigorously
2605 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2606 ;; virtually identical to this function.
2607 (save-restriction
2608 (widen)
2609 (save-excursion
2610 (let ((pos (c-state-safe-place here)))
2611 (car (cddr (c-state-pp-to-literal pos here)))))))
2612
2613 (defsubst c-state-lit-beg (pos)
2614 ;; Return the start of the literal containing POS, or POS itself.
2615 (or (car (c-state-literal-at pos))
2616 pos))
2617
2618 (defsubst c-state-cache-non-literal-place (pos state)
2619 ;; Return a position outside of a string/comment/macro at or before POS.
2620 ;; STATE is the parse-partial-sexp state at POS.
2621 (let ((res (if (or (nth 3 state) ; in a string?
2622 (nth 4 state)) ; in a comment?
2623 (nth 8 state)
2624 pos)))
2625 (save-excursion
2626 (goto-char res)
2627 (if (c-beginning-of-macro)
2628 (point)
2629 res))))
2630
2631 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2632 ;; Stuff to do with point-min, and coping with any literal there.
2633 (defvar c-state-point-min 1)
2634 (make-variable-buffer-local 'c-state-point-min)
2635 ;; This is (point-min) when `c-state-cache' was last calculated. A change of
2636 ;; narrowing is likely to affect the parens that are visible before the point.
2637
2638 (defvar c-state-point-min-lit-type nil)
2639 (make-variable-buffer-local 'c-state-point-min-lit-type)
2640 (defvar c-state-point-min-lit-start nil)
2641 (make-variable-buffer-local 'c-state-point-min-lit-start)
2642 ;; These two variables define the literal, if any, containing point-min.
2643 ;; Their values are, respectively, 'string, c, or c++, and the start of the
2644 ;; literal. If there's no literal there, they're both nil.
2645
2646 (defvar c-state-min-scan-pos 1)
2647 (make-variable-buffer-local 'c-state-min-scan-pos)
2648 ;; This is the earliest buffer-pos from which scanning can be done. It is
2649 ;; either the end of the literal containing point-min, or point-min itself.
2650 ;; It becomes nil if the buffer is changed earlier than this point.
2651 (defun c-state-get-min-scan-pos ()
2652 ;; Return the lowest valid scanning pos. This will be the end of the
2653 ;; literal enclosing point-min, or point-min itself.
2654 (or c-state-min-scan-pos
2655 (save-restriction
2656 (save-excursion
2657 (widen)
2658 (goto-char c-state-point-min-lit-start)
2659 (if (eq c-state-point-min-lit-type 'string)
2660 (forward-sexp)
2661 (forward-comment 1))
2662 (setq c-state-min-scan-pos (point))))))
2663
2664 (defun c-state-mark-point-min-literal ()
2665 ;; Determine the properties of any literal containing POINT-MIN, setting the
2666 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2667 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2668 (let ((p-min (point-min))
2669 lit)
2670 (save-restriction
2671 (widen)
2672 (setq lit (c-state-literal-at p-min))
2673 (if lit
2674 (setq c-state-point-min-lit-type
2675 (save-excursion
2676 (goto-char (car lit))
2677 (cond
2678 ((looking-at c-block-comment-start-regexp) 'c)
2679 ((looking-at c-line-comment-starter) 'c++)
2680 (t 'string)))
2681 c-state-point-min-lit-start (car lit)
2682 c-state-min-scan-pos (cdr lit))
2683 (setq c-state-point-min-lit-type nil
2684 c-state-point-min-lit-start nil
2685 c-state-min-scan-pos p-min)))))
2686
2687
2688 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2689 ;; A variable which signals a brace dessert - helpful for reducing the number
2690 ;; of fruitless backward scans.
2691 (defvar c-state-brace-pair-desert nil)
2692 (make-variable-buffer-local 'c-state-brace-pair-desert)
2693 ;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
2694 ;; that defun has searched backwards for a brace pair and not found one. Its
2695 ;; value is either nil or a cons (PA . FROM), where PA is the position of the
2696 ;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2697 ;; nil when at top level) and FROM is where the backward search started. It
2698 ;; is reset to nil in `c-invalidate-state-cache'.
2699
2700
2701 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2702 ;; Lowish level functions/macros which work directly on `c-state-cache', or a
2703 ;; list of like structure.
2704 (defmacro c-state-cache-top-lparen (&optional cache)
2705 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2706 ;; (default `c-state-cache') (or nil).
2707 (let ((cash (or cache 'c-state-cache)))
2708 `(if (consp (car ,cash))
2709 (caar ,cash)
2710 (car ,cash))))
2711
2712 (defmacro c-state-cache-top-paren (&optional cache)
2713 ;; Return the address of the latest brace/bracket/paren (whether left or
2714 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2715 (let ((cash (or cache 'c-state-cache)))
2716 `(if (consp (car ,cash))
2717 (cdar ,cash)
2718 (car ,cash))))
2719
2720 (defmacro c-state-cache-after-top-paren (&optional cache)
2721 ;; Return the position just after the latest brace/bracket/paren (whether
2722 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2723 (let ((cash (or cache 'c-state-cache)))
2724 `(if (consp (car ,cash))
2725 (cdar ,cash)
2726 (and (car ,cash)
2727 (1+ (car ,cash))))))
2728
2729 (defun c-get-cache-scan-pos (here)
2730 ;; From the state-cache, determine the buffer position from which we might
2731 ;; scan forward to HERE to update this cache. This position will be just
2732 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2733 ;; return the earliest position in the accessible region which isn't within
2734 ;; a literal. If the visible portion of the buffer is entirely within a
2735 ;; literal, return NIL.
2736 (let ((c c-state-cache) elt)
2737 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2738 (while (and c
2739 (>= (c-state-cache-top-lparen c) here))
2740 (setq c (cdr c)))
2741
2742 (setq elt (car c))
2743 (cond
2744 ((consp elt)
2745 (if (> (cdr elt) here)
2746 (1+ (car elt))
2747 (cdr elt)))
2748 (elt (1+ elt))
2749 ((<= (c-state-get-min-scan-pos) here)
2750 (c-state-get-min-scan-pos))
2751 (t nil))))
2752
2753 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2754 ;; Variables which keep track of preprocessor constructs.
2755 (defvar c-state-old-cpp-beg-marker nil)
2756 (make-variable-buffer-local 'c-state-old-cpp-beg-marker)
2757 (defvar c-state-old-cpp-beg nil)
2758 (make-variable-buffer-local 'c-state-old-cpp-beg)
2759 (defvar c-state-old-cpp-end-marker nil)
2760 (make-variable-buffer-local 'c-state-old-cpp-end-marker)
2761 (defvar c-state-old-cpp-end nil)
2762 (make-variable-buffer-local 'c-state-old-cpp-end)
2763 ;; These are the limits of the macro containing point at the previous call of
2764 ;; `c-parse-state', or nil.
2765
2766 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2767 ;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
2768 (defun c-get-fallback-scan-pos (here)
2769 ;; Return a start position for building `c-state-cache' from
2770 ;; scratch. This will be at the top level, 2 defuns back.
2771 (save-excursion
2772 ;; Go back 2 bods, but ignore any bogus positions returned by
2773 ;; beginning-of-defun (i.e. open paren in column zero).
2774 (goto-char here)
2775 (let ((cnt 2))
2776 (while (not (or (bobp) (zerop cnt)))
2777 (c-beginning-of-defun-1) ; Pure elisp BOD.
2778 (if (eq (char-after) ?\{)
2779 (setq cnt (1- cnt)))))
2780 (point)))
2781
2782 (defun c-state-balance-parens-backwards (here- here+ top)
2783 ;; Return the position of the opening paren/brace/bracket before HERE- which
2784 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2785 ;; there's a macro, HERE- and HERE+ are the same. Like this:
2786 ;;
2787 ;; ............................................
2788 ;; | |
2789 ;; ( [ ( .........#macro.. ) ( ) ] )
2790 ;; ^ ^ ^ ^
2791 ;; | | | |
2792 ;; return HERE- HERE+ TOP
2793 ;;
2794 ;; If there aren't enough opening paren/brace/brackets, return the position
2795 ;; of the outermost one found, or HERE- if there are none. If there are no
2796 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
2797 ;; must not be inside literals. Only the accessible portion of the buffer
2798 ;; will be scanned.
2799
2800 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2801 ;; `here'. Go round the next loop each time we pass over such a ")". These
2802 ;; probably match "("s before `here-'.
2803 (let (pos pa ren+1 lonely-rens)
2804 (save-excursion
2805 (save-restriction
2806 (narrow-to-region (point-min) top) ; This can move point, sometimes.
2807 (setq pos here+)
2808 (c-safe
2809 (while
2810 (setq ren+1 (c-sc-scan-lists pos 1 1)) ; might signal
2811 (setq lonely-rens (cons ren+1 lonely-rens)
2812 pos ren+1)))))
2813
2814 ;; PART 2: Scan back before `here-' searching for the "("s
2815 ;; matching/mismatching the ")"s found above. We only need to direct the
2816 ;; caller to scan when we've encountered unmatched right parens.
2817 (setq pos here-)
2818 (when lonely-rens
2819 (c-safe
2820 (while
2821 (and lonely-rens ; actual values aren't used.
2822 (setq pa (c-sc-scan-lists pos -1 1)))
2823 (setq pos pa)
2824 (setq lonely-rens (cdr lonely-rens)))))
2825 pos))
2826
2827 (defun c-parse-state-get-strategy (here good-pos)
2828 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
2829 ;; to minimize the amount of scanning. HERE is the pertinent position in
2830 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2831 ;; its head trimmed) is known to be good, or nil if there is no such
2832 ;; position.
2833 ;;
2834 ;; The return value is a list, one of the following:
2835 ;;
2836 ;; o - ('forward START-POINT) - scan forward from START-POINT,
2837 ;; which is not less than the highest position in `c-state-cache' below HERE,
2838 ;; which is after GOOD-POS.
2839 ;; o - ('backward nil) - scan backwards (from HERE).
2840 ;; o - ('back-and-forward START-POINT) - like 'forward, but when HERE is earlier
2841 ;; than GOOD-POS.
2842 ;; o - ('BOD START-POINT) - scan forwards from START-POINT, which is at the
2843 ;; top level.
2844 ;; o - ('IN-LIT nil) - point is inside the literal containing point-min.
2845 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2846 BOD-pos ; position of 2nd BOD before HERE.
2847 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2848 start-point
2849 how-far) ; putative scanning distance.
2850 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2851 (cond
2852 ((< here (c-state-get-min-scan-pos))
2853 (setq strategy 'IN-LIT
2854 start-point nil
2855 cache-pos nil
2856 how-far 0))
2857 ((<= good-pos here)
2858 (setq strategy 'forward
2859 start-point (max good-pos cache-pos)
2860 how-far (- here start-point)))
2861 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2862 (setq strategy 'backward
2863 how-far (- good-pos here)))
2864 (t
2865 (setq strategy 'back-and-forward
2866 start-point cache-pos
2867 how-far (- here start-point))))
2868
2869 ;; Might we be better off starting from the top level, two defuns back,
2870 ;; instead? This heuristic no longer works well in C++, where
2871 ;; declarations inside namespace brace blocks are frequently placed at
2872 ;; column zero. (2015-11-10): Remove the condition on C++ Mode.
2873 (when (and (or (not (memq 'col-0-paren c-emacs-features))
2874 open-paren-in-column-0-is-defun-start)
2875 ;; (not (c-major-mode-is 'c++-mode))
2876 (> how-far c-state-cache-too-far))
2877 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2878 (if (< (- here BOD-pos) how-far)
2879 (setq strategy 'BOD
2880 start-point BOD-pos)))
2881
2882 (list strategy start-point)))
2883
2884
2885 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2886 ;; Routines which change `c-state-cache' and associated values.
2887 (defun c-renarrow-state-cache ()
2888 ;; The region (more precisely, point-min) has changed since we
2889 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2890 (if (< (point-min) c-state-point-min)
2891 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2892 ;; It would be possible to do a better job here and recalculate the top
2893 ;; only.
2894 (progn
2895 (c-state-mark-point-min-literal)
2896 (setq c-state-cache nil
2897 c-state-cache-good-pos c-state-min-scan-pos
2898 c-state-brace-pair-desert nil))
2899
2900 ;; point-min has MOVED FORWARD.
2901
2902 ;; Is the new point-min inside a (different) literal?
2903 (unless (and c-state-point-min-lit-start ; at prev. point-min
2904 (< (point-min) (c-state-get-min-scan-pos)))
2905 (c-state-mark-point-min-literal))
2906
2907 ;; Cut off a bit of the tail from `c-state-cache'.
2908 (let ((ptr (cons nil c-state-cache))
2909 pa)
2910 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2911 (>= pa (point-min)))
2912 (setq ptr (cdr ptr)))
2913
2914 (when (consp ptr)
2915 (if (or (eq (cdr ptr) c-state-cache)
2916 (and (consp (cadr ptr))
2917 (> (cdr (cadr ptr)) (point-min)))) ; Our new point-min is
2918 ; inside a recorded
2919 ; brace pair.
2920 (setq c-state-cache nil
2921 c-state-cache-good-pos c-state-min-scan-pos)
2922 (setcdr ptr nil)
2923 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2924 )))
2925
2926 (setq c-state-point-min (point-min)))
2927
2928 (defun c-append-lower-brace-pair-to-state-cache (from here &optional upper-lim)
2929 ;; If there is a brace pair preceding FROM in the buffer, at the same level
2930 ;; of nesting (not necessarily immediately preceding), push a cons onto
2931 ;; `c-state-cache' to represent it. FROM must not be inside a literal. If
2932 ;; UPPER-LIM is non-nil, we append the highest brace pair whose "}" is below
2933 ;; UPPER-LIM.
2934 ;;
2935 ;; Return non-nil when this has been done.
2936 ;;
2937 ;; The situation it copes with is this transformation:
2938 ;;
2939 ;; OLD: { (.) {...........}
2940 ;; ^ ^
2941 ;; FROM HERE
2942 ;;
2943 ;; NEW: { {....} (.) {.........
2944 ;; ^ ^ ^
2945 ;; LOWER BRACE PAIR HERE or HERE
2946 ;;
2947 ;; This routine should be fast. Since it can get called a LOT, we maintain
2948 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2949 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2950 (save-excursion
2951 (save-restriction
2952 (let* (new-cons
2953 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2954 (macro-start-or-from
2955 (progn (goto-char from)
2956 (c-beginning-of-macro)
2957 (point)))
2958 (bra ; Position of "{".
2959 ;; Don't start scanning in the middle of a CPP construct unless
2960 ;; it contains HERE - these constructs, in Emacs, are "commented
2961 ;; out" with category properties.
2962 (if (eq (c-get-char-property macro-start-or-from 'category)
2963 'c-cpp-delimiter)
2964 macro-start-or-from
2965 from))
2966 ce) ; Position of "}"
2967 (or upper-lim (setq upper-lim from))
2968
2969 ;; If we're essentially repeating a fruitless search, just give up.
2970 (unless (and c-state-brace-pair-desert
2971 (eq cache-pos (car c-state-brace-pair-desert))
2972 (or (null (car c-state-brace-pair-desert))
2973 (> from (car c-state-brace-pair-desert)))
2974 (<= from (cdr c-state-brace-pair-desert)))
2975 ;; DESERT-LIM. Avoid repeated searching through the cached desert.
2976 (let ((desert-lim
2977 (and c-state-brace-pair-desert
2978 (eq cache-pos (car c-state-brace-pair-desert))
2979 (>= from (cdr c-state-brace-pair-desert))
2980 (cdr c-state-brace-pair-desert)))
2981 ;; CACHE-LIM. This limit will be necessary when an opening
2982 ;; paren at `cache-pos' has just had its matching close paren
2983 ;; inserted into the buffer. `cache-pos' continues to be a
2984 ;; search bound, even though the algorithm below would skip
2985 ;; over the new paren pair.
2986 (cache-lim (and cache-pos (< cache-pos from) cache-pos)))
2987 (narrow-to-region
2988 (cond
2989 ((and desert-lim cache-lim)
2990 (max desert-lim cache-lim))
2991 (desert-lim)
2992 (cache-lim)
2993 ((point-min)))
2994 ;; The top limit is EOB to ensure that `bra' is inside the
2995 ;; accessible part of the buffer at the next scan operation.
2996 (1+ (buffer-size))))
2997
2998 ;; In the next pair of nested loops, the inner one moves back past a
2999 ;; pair of (mis-)matching parens or brackets; the outer one moves
3000 ;; back over a sequence of unmatched close brace/paren/bracket each
3001 ;; time round.
3002 (while
3003 (progn
3004 (c-safe
3005 (while
3006 (and (setq ce (c-sc-scan-lists bra -1 -1)) ; back past )/]/}; might signal
3007 (setq bra (c-sc-scan-lists ce -1 1)) ; back past (/[/{; might signal
3008 (or (> bra here) ;(> ce here)
3009 (and
3010 (< ce here)
3011 (or (not (eq (char-after bra) ?\{))
3012 (and (goto-char bra)
3013 (c-beginning-of-macro)
3014 (< (point) macro-start-or-from))))))))
3015 (and ce (< ce bra)))
3016 (setq bra ce)) ; If we just backed over an unbalanced closing
3017 ; brace, ignore it.
3018
3019 (if (and ce (< ce here) (< bra ce) (eq (char-after bra) ?\{))
3020 ;; We've found the desired brace-pair.
3021 (progn
3022 (setq new-cons (cons bra (1+ ce)))
3023 (cond
3024 ((consp (car c-state-cache))
3025 (setcar c-state-cache new-cons))
3026 ((and (numberp (car c-state-cache)) ; probably never happens
3027 (< ce (car c-state-cache)))
3028 (setcdr c-state-cache
3029 (cons new-cons (cdr c-state-cache))))
3030 (t (setq c-state-cache (cons new-cons c-state-cache)))))
3031
3032 ;; We haven't found a brace pair. Record this in the cache.
3033 (setq c-state-brace-pair-desert
3034 (cons (if (and ce (< bra ce) (> ce here)) ; {..} straddling HERE?
3035 bra
3036 (point-min))
3037 (min here from)))))))))
3038
3039 (defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
3040 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
3041 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
3042 ;; "push" "a" brace pair onto `c-state-cache'.
3043 ;;
3044 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
3045 ;; otherwise push it normally.
3046 ;;
3047 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
3048 ;; latter is inside a macro, not being a macro containing
3049 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
3050 ;; base pair. This latter case is assumed to be rare.
3051 ;;
3052 ;; Note: POINT is not preserved in this routine.
3053 (if bra+1
3054 (if (or (> bra+1 macro-start-or-here)
3055 (progn (goto-char bra+1)
3056 (not (c-beginning-of-macro))))
3057 (setq c-state-cache
3058 (cons (cons (1- bra+1)
3059 (c-sc-scan-lists bra+1 1 1))
3060 (if (consp (car c-state-cache))
3061 (cdr c-state-cache)
3062 c-state-cache)))
3063 ;; N.B. This defsubst codes one method for the simple, normal case,
3064 ;; and a more sophisticated, slower way for the general case. Don't
3065 ;; eliminate this defsubst - it's a speed optimization.
3066 (c-append-lower-brace-pair-to-state-cache (1- bra+1) (point-max)))))
3067
3068 (defun c-append-to-state-cache (from here)
3069 ;; Scan the buffer from FROM to HERE, adding elements into `c-state-cache'
3070 ;; for braces etc. Return a candidate for `c-state-cache-good-pos'.
3071 ;;
3072 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
3073 ;; any. Typically, it is immediately after it. It must not be inside a
3074 ;; literal.
3075 (let ((here-bol (c-point 'bol here))
3076 (macro-start-or-here
3077 (save-excursion (goto-char here)
3078 (if (c-beginning-of-macro)
3079 (point)
3080 here)))
3081 pa+1 ; pos just after an opening PAren (or brace).
3082 (ren+1 from) ; usually a pos just after an closing paREN etc.
3083 ; Is actually the pos. to scan for a (/{/[ from,
3084 ; which sometimes is after a silly )/}/].
3085 paren+1 ; Pos after some opening or closing paren.
3086 paren+1s ; A list of `paren+1's; used to determine a
3087 ; good-pos.
3088 bra+1 ; just after L bra-ce.
3089 bra+1s ; list of OLD values of bra+1.
3090 mstart) ; start of a macro.
3091
3092 (save-excursion
3093 (save-restriction
3094 (narrow-to-region (point-min) here)
3095 ;; Each time round the following loop, we enter a successively deeper
3096 ;; level of brace/paren nesting. (Except sometimes we "continue at
3097 ;; the existing level".) `pa+1' is a pos inside an opening
3098 ;; brace/paren/bracket, usually just after it.
3099 (while
3100 (progn
3101 ;; Each time round the next loop moves forward over an opening then
3102 ;; a closing brace/bracket/paren. This loop is white hot, so it
3103 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
3104 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
3105 ;; call of `scan-lists' signals an error, which happens when there
3106 ;; are no more b/b/p's to scan.
3107 (c-safe
3108 (while t
3109 (setq pa+1 (c-sc-scan-lists ren+1 1 -1) ; Into (/{/[; might signal
3110 paren+1s (cons pa+1 paren+1s))
3111 (setq ren+1 (c-sc-scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
3112 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
3113 (setq bra+1 pa+1))
3114 (setcar paren+1s ren+1)))
3115
3116 (if (and pa+1 (> pa+1 ren+1))
3117 ;; We've just entered a deeper nesting level.
3118 (progn
3119 ;; Insert the brace pair (if present) and the single open
3120 ;; paren/brace/bracket into `c-state-cache' It cannot be
3121 ;; inside a macro, except one around point, because of what
3122 ;; `c-neutralize-syntax-in-CPP' has done.
3123 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
3124 ;; Insert the opening brace/bracket/paren position.
3125 (setq c-state-cache (cons (1- pa+1) c-state-cache))
3126 ;; Clear admin stuff for the next more nested part of the scan.
3127 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
3128 t) ; Carry on the loop
3129
3130 ;; All open p/b/b's at this nesting level, if any, have probably
3131 ;; been closed by matching/mismatching ones. We're probably
3132 ;; finished - we just need to check for having found an
3133 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
3134 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
3135 (c-safe (setq ren+1 (c-sc-scan-lists ren+1 1 1)))))) ; acts as loop control.
3136
3137 ;; Record the final, innermost, brace-pair if there is one.
3138 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
3139
3140 ;; Determine a good pos
3141 (while (and (setq paren+1 (car paren+1s))
3142 (> (if (> paren+1 macro-start-or-here)
3143 paren+1
3144 (goto-char paren+1)
3145 (setq mstart (and (c-beginning-of-macro)
3146 (point)))
3147 (or mstart paren+1))
3148 here-bol))
3149 (setq paren+1s (cdr paren+1s)))
3150 (cond
3151 ((and paren+1 mstart)
3152 (min paren+1 mstart))
3153 (paren+1)
3154 (t from))))))
3155
3156 (defun c-remove-stale-state-cache (start-point here pps-point)
3157 ;; Remove stale entries from the `c-cache-state', i.e. those which will
3158 ;; not be in it when it is amended for position HERE. This may involve
3159 ;; replacing a CONS element for a brace pair containing HERE with its car.
3160 ;; Additionally, the "outermost" open-brace entry before HERE will be
3161 ;; converted to a cons if the matching close-brace is below HERE.
3162 ;;
3163 ;; START-POINT is a "maximal" "safe position" - there must be no open
3164 ;; parens/braces/brackets between START-POINT and HERE.
3165 ;;
3166 ;; As a second thing, calculate the result of parse-partial-sexp at
3167 ;; PPS-POINT, w.r.t. START-POINT. The motivation here is that
3168 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
3169 ;; adjust it to get outside a string/comment. (Sorry about this! The code
3170 ;; needs to be FAST).
3171 ;;
3172 ;; Return a list (GOOD-POS SCAN-BACK-POS CONS-SEPARATED PPS-STATE), where
3173 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
3174 ;; to be good (we aim for this to be as high as possible);
3175 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
3176 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
3177 ;; position to scan backwards from. It is the position of the "{" of the
3178 ;; last element to be removed from `c-state-cache', when that elt is a
3179 ;; cons, otherwise nil.
3180 ;; o - CONS-SEPARATED is t when a cons element in `c-state-cache' has been
3181 ;; replaced by its car because HERE lies inside the brace pair represented
3182 ;; by the cons.
3183 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
3184 (save-excursion
3185 (save-restriction
3186 (narrow-to-region 1 (point-max))
3187 (let* ((in-macro-start ; start of macro containing HERE or nil.
3188 (save-excursion
3189 (goto-char here)
3190 (and (c-beginning-of-macro)
3191 (point))))
3192 (start-point-actual-macro-start ; Start of macro containing
3193 ; start-point or nil
3194 (and (< start-point here)
3195 (save-excursion
3196 (goto-char start-point)
3197 (and (c-beginning-of-macro)
3198 (point)))))
3199 (start-point-actual-macro-end ; End of this macro, (maybe
3200 ; HERE), or nil.
3201 (and start-point-actual-macro-start
3202 (save-excursion
3203 (goto-char start-point-actual-macro-start)
3204 (c-end-of-macro)
3205 (point))))
3206 pps-state ; Will be 9 or 10 elements long.
3207 pos
3208 upper-lim ; ,beyond which `c-state-cache' entries are removed
3209 scan-back-pos
3210 cons-separated
3211 pair-beg pps-point-state target-depth)
3212
3213 ;; Remove entries beyond HERE. Also remove any entries inside
3214 ;; a macro, unless HERE is in the same macro.
3215 (setq upper-lim
3216 (if (or (null c-state-old-cpp-beg)
3217 (and (> here c-state-old-cpp-beg)
3218 (< here c-state-old-cpp-end)))
3219 here
3220 (min here c-state-old-cpp-beg)))
3221 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
3222 (setq scan-back-pos (car-safe (car c-state-cache)))
3223 (setq c-state-cache (cdr c-state-cache)))
3224
3225 ;; If `upper-lim' is inside the last recorded brace pair, remove its
3226 ;; RBrace and indicate we'll need to search backwards for a previous
3227 ;; brace pair.
3228 (when (and c-state-cache
3229 (consp (car c-state-cache))
3230 (> (cdar c-state-cache) upper-lim))
3231 (setcar c-state-cache (caar c-state-cache))
3232 (setq scan-back-pos (car c-state-cache)
3233 cons-separated t))
3234
3235 ;; The next loop jumps forward out of a nested level of parens each
3236 ;; time round; the corresponding elements in `c-state-cache' are
3237 ;; removed. `pos' is just after the brace-pair or the open paren at
3238 ;; (car c-state-cache). There can be no open parens/braces/brackets
3239 ;; between `start-point'/`start-point-actual-macro-start' and HERE,
3240 ;; due to the interface spec to this function.
3241 (setq pos (if (and start-point-actual-macro-end
3242 (not (eq start-point-actual-macro-start
3243 in-macro-start)))
3244 (1+ start-point-actual-macro-end) ; get outside the macro as
3245 ; marked by a `category' text property.
3246 start-point))
3247 (goto-char pos)
3248 (while (and c-state-cache
3249 (or (numberp (car c-state-cache)) ; Have we a { at all?
3250 (cdr c-state-cache))
3251 (< (point) here))
3252 (cond
3253 ((null pps-state) ; first time through
3254 (setq target-depth -1))
3255 ((eq (car pps-state) target-depth) ; found closing ),},]
3256 (setq target-depth (1- (car pps-state))))
3257 ;; Do nothing when we've merely reached pps-point.
3258 )
3259
3260 ;; Scan!
3261 (setq pps-state
3262 (c-sc-parse-partial-sexp
3263 (point) (if (< (point) pps-point) pps-point here)
3264 target-depth
3265 nil pps-state))
3266
3267 (if (= (point) pps-point)
3268 (setq pps-point-state pps-state))
3269
3270 (when (eq (car pps-state) target-depth)
3271 (setq pos (point)) ; POS is now just after an R-paren/brace.
3272 (cond
3273 ((and (consp (car c-state-cache))
3274 (eq (point) (cdar c-state-cache)))
3275 ;; We've just moved out of the paren pair containing the brace-pair
3276 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
3277 ;; and is potentially where the open brace of a cons in
3278 ;; c-state-cache will be.
3279 (setq pair-beg (car-safe (cdr c-state-cache))
3280 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
3281 ((numberp (car c-state-cache))
3282 (setq pair-beg (car c-state-cache)
3283 c-state-cache (cdr c-state-cache))) ; remove this
3284 ; containing Lparen
3285 ((numberp (cadr c-state-cache))
3286 (setq pair-beg (cadr c-state-cache)
3287 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
3288 ; together with enclosed brace pair.
3289 ;; (t nil) ; Ignore an unmated Rparen.
3290 )))
3291
3292 (if (< (point) pps-point)
3293 (setq pps-state (c-sc-parse-partial-sexp
3294 (point) pps-point
3295 nil nil ; TARGETDEPTH, STOPBEFORE
3296 pps-state)))
3297
3298 ;; If the last paren pair we moved out of was actually a brace pair,
3299 ;; insert it into `c-state-cache'.
3300 (when (and pair-beg (eq (char-after pair-beg) ?{))
3301 (if (consp (car-safe c-state-cache))
3302 (setq c-state-cache (cdr c-state-cache)))
3303 (setq c-state-cache (cons (cons pair-beg pos)
3304 c-state-cache)))
3305
3306 (list pos scan-back-pos cons-separated pps-state)))))
3307
3308 (defun c-remove-stale-state-cache-backwards (here)
3309 ;; Strip stale elements of `c-state-cache' by moving backwards through the
3310 ;; buffer, and inform the caller of the scenario detected.
3311 ;;
3312 ;; HERE is the position we're setting `c-state-cache' for.
3313 ;; CACHE-POS (a locally bound variable) is just after the latest recorded
3314 ;; position in `c-state-cache' before HERE, or a position at or near
3315 ;; point-min which isn't in a literal.
3316 ;;
3317 ;; This function must only be called only when (> `c-state-cache-good-pos'
3318 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
3319 ;; optimized to eliminate (or minimize) scanning between these two
3320 ;; positions.
3321 ;;
3322 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
3323 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
3324 ;; could become so after missing elements are inserted into
3325 ;; `c-state-cache'. This is JUST AFTER an opening or closing
3326 ;; brace/paren/bracket which is already in `c-state-cache' or just before
3327 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
3328 ;; before `here''s line, or the start of the literal containing it.
3329 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
3330 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
3331 ;; to scan backwards from.
3332 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
3333 ;; POS and HERE which aren't recorded in `c-state-cache'.
3334 ;;
3335 ;; The comments in this defun use "paren" to mean parenthesis or square
3336 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
3337 ;;
3338 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
3339 ;; | | | | | |
3340 ;; CP E here D C good
3341 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
3342 (pos c-state-cache-good-pos)
3343 pa ren ; positions of "(" and ")"
3344 dropped-cons ; whether the last element dropped from `c-state-cache'
3345 ; was a cons (representing a brace-pair)
3346 good-pos ; see above.
3347 lit ; (START . END) of a literal containing some point.
3348 here-lit-start here-lit-end ; bounds of literal containing `here'
3349 ; or `here' itself.
3350 here- here+ ; start/end of macro around HERE, or HERE
3351 (here-bol (c-point 'bol here))
3352 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
3353
3354 ;; Remove completely irrelevant entries from `c-state-cache'.
3355 (while (and c-state-cache
3356 (>= (setq pa (c-state-cache-top-lparen)) here))
3357 (setq dropped-cons (consp (car c-state-cache)))
3358 (setq c-state-cache (cdr c-state-cache))
3359 (setq pos pa))
3360 ;; At this stage, (>= pos here);
3361 ;; (< (c-state-cache-top-lparen) here) (or is nil).
3362
3363 (cond
3364 ((and (consp (car c-state-cache))
3365 (> (cdar c-state-cache) here))
3366 ;; CASE 1: The top of the cache is a brace pair which now encloses
3367 ;; `here'. As good-pos, return the address. of the "{". Since we've no
3368 ;; knowledge of what's inside these braces, we have no alternative but
3369 ;; to direct the caller to scan the buffer from the opening brace.
3370 (setq pos (caar c-state-cache))
3371 (setcar c-state-cache pos)
3372 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
3373 ; entry into a { entry, so the caller needs to
3374 ; search for a brace pair before the {.
3375
3376 ;; `here' might be inside a literal. Check for this.
3377 ((progn
3378 (setq lit (c-state-literal-at here)
3379 here-lit-start (or (car lit) here)
3380 here-lit-end (or (cdr lit) here))
3381 ;; Has `here' just "newly entered" a macro?
3382 (save-excursion
3383 (goto-char here-lit-start)
3384 (if (and (c-beginning-of-macro)
3385 (or (null c-state-old-cpp-beg)
3386 (not (= (point) c-state-old-cpp-beg))))
3387 (progn
3388 (setq here- (point))
3389 (c-end-of-macro)
3390 (setq here+ (point)))
3391 (setq here- here-lit-start
3392 here+ here-lit-end)))
3393
3394 ;; `here' might be nested inside any depth of parens (or brackets but
3395 ;; not braces). Scan backwards to find the outermost such opening
3396 ;; paren, if there is one. This will be the scan position to return.
3397 (save-restriction
3398 (narrow-to-region cache-pos (point-max))
3399 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
3400 nil)) ; for the cond
3401
3402 ((< pos here-lit-start)
3403 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
3404 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
3405 ;; a brace pair preceding this, it will already be in `c-state-cache',
3406 ;; unless there was a brace pair after it, i.e. there'll only be one to
3407 ;; scan for if we've just deleted one.
3408 (list pos (and dropped-cons pos) t)) ; Return value.
3409
3410 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
3411 ;; Further forward scanning isn't needed, but we still need to find a
3412 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
3413 ((progn
3414 (save-restriction
3415 (narrow-to-region here-bol (point-max))
3416 (setq pos here-lit-start)
3417 (c-safe (while (setq pa (c-sc-scan-lists pos -1 1))
3418 (setq pos pa)))) ; might signal
3419 nil)) ; for the cond
3420
3421 ((save-restriction
3422 (narrow-to-region too-far-back (point-max))
3423 (setq ren (c-safe (c-sc-scan-lists pos -1 -1))))
3424 ;; CASE 3: After a }/)/] before `here''s BOL.
3425 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
3426
3427 ((progn (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
3428 (>= cache-pos good-pos))
3429 ;; CASE 3.5: Just after an existing entry in `c-state-cache' on `here''s
3430 ;; line or the previous line.
3431 (list cache-pos nil nil))
3432
3433 (t
3434 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
3435 ;; literal containing it.
3436 (list good-pos (and dropped-cons good-pos) nil)))))
3437
3438
3439 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3440 ;; Externally visible routines.
3441
3442 (defun c-state-cache-init ()
3443 (setq c-state-cache nil
3444 c-state-cache-good-pos 1
3445 c-state-nonlit-pos-cache nil
3446 c-state-nonlit-pos-cache-limit 1
3447 c-state-semi-nonlit-pos-cache nil
3448 c-state-semi-nonlit-pos-cache-limit 1
3449 c-state-brace-pair-desert nil
3450 c-state-point-min 1
3451 c-state-point-min-lit-type nil
3452 c-state-point-min-lit-start nil
3453 c-state-min-scan-pos 1
3454 c-state-old-cpp-beg nil
3455 c-state-old-cpp-end nil)
3456 (c-state-mark-point-min-literal))
3457
3458 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3459 ;; Debugging routines to dump `c-state-cache' in a "replayable" form.
3460 ;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
3461 ;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
3462 ;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
3463 ;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
3464 ;; (defun c-state-dump ()
3465 ;; ;; For debugging.
3466 ;; ;(message
3467 ;; (concat
3468 ;; (c-sc-qde c-state-cache)
3469 ;; (c-sc-de c-state-cache-good-pos)
3470 ;; (c-sc-qde c-state-nonlit-pos-cache)
3471 ;; (c-sc-de c-state-nonlit-pos-cache-limit)
3472 ;; (c-sc-qde c-state-brace-pair-desert)
3473 ;; (c-sc-de c-state-point-min)
3474 ;; (c-sc-de c-state-point-min-lit-type)
3475 ;; (c-sc-de c-state-point-min-lit-start)
3476 ;; (c-sc-de c-state-min-scan-pos)
3477 ;; (c-sc-de c-state-old-cpp-beg)
3478 ;; (c-sc-de c-state-old-cpp-end)))
3479 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3480
3481 (defun c-invalidate-state-cache-1 (here)
3482 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
3483 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
3484 ;; left in a consistent state.
3485 ;;
3486 ;; This is much like `c-whack-state-after', but it never changes a paren
3487 ;; pair element into an open paren element. Doing that would mean that the
3488 ;; new open paren wouldn't have the required preceding paren pair element.
3489 ;;
3490 ;; This function is called from c-before-change.
3491
3492 ;; The caches of non-literals:
3493 ;; Note that we use "<=" for the possibility of the second char of a two-char
3494 ;; comment opener being typed; this would invalidate any cache position at
3495 ;; HERE.
3496 (if (<= here c-state-nonlit-pos-cache-limit)
3497 (setq c-state-nonlit-pos-cache-limit (1- here)))
3498 (c-truncate-semi-nonlit-pos-cache here)
3499
3500 ;; `c-state-cache':
3501 ;; Case 1: if `here' is in a literal containing point-min, everything
3502 ;; becomes (or is already) nil.
3503 (if (or (null c-state-cache-good-pos)
3504 (< here (c-state-get-min-scan-pos)))
3505 (setq c-state-cache nil
3506 c-state-cache-good-pos nil
3507 c-state-min-scan-pos nil)
3508
3509 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
3510 ;; below `here'. To maintain its consistency, we may need to insert a new
3511 ;; brace pair.
3512 (let ((here-bol (c-point 'bol here))
3513 too-high-pa ; recorded {/(/[ next above or just below here, or nil.
3514 dropped-cons ; was the last removed element a brace pair?
3515 pa)
3516 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
3517 (while (and c-state-cache
3518 (>= (setq pa (c-state-cache-top-paren)) here))
3519 (setq dropped-cons (consp (car c-state-cache))
3520 too-high-pa (c-state-cache-top-lparen)
3521 c-state-cache (cdr c-state-cache)))
3522
3523 ;; Do we need to add in an earlier brace pair, having lopped one off?
3524 (if (and dropped-cons
3525 (<= too-high-pa here))
3526 (c-append-lower-brace-pair-to-state-cache too-high-pa here here-bol))
3527 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3528 (c-state-get-min-scan-pos)))))
3529
3530 ;; The brace-pair desert marker:
3531 (when (car c-state-brace-pair-desert)
3532 (if (< here (car c-state-brace-pair-desert))
3533 (setq c-state-brace-pair-desert nil)
3534 (if (< here (cdr c-state-brace-pair-desert))
3535 (setcdr c-state-brace-pair-desert here)))))
3536
3537 (defun c-parse-state-1 ()
3538 ;; Find and record all noteworthy parens between some good point earlier in
3539 ;; the file and point. That good point is at least the beginning of the
3540 ;; top-level construct we are in, or the beginning of the preceding
3541 ;; top-level construct if we aren't in one.
3542 ;;
3543 ;; The returned value is a list of the noteworthy parens with the last one
3544 ;; first. If an element in the list is an integer, it's the position of an
3545 ;; open paren (of any type) which has not been closed before the point. If
3546 ;; an element is a cons, it gives the position of a closed BRACE paren
3547 ;; pair[*]; the car is the start brace position and the cdr is the position
3548 ;; following the closing brace. Only the last closed brace paren pair
3549 ;; before each open paren and before the point is recorded, and thus the
3550 ;; state never contains two cons elements in succession. When a close brace
3551 ;; has no matching open brace (e.g., the matching brace is outside the
3552 ;; visible region), it is not represented in the returned value.
3553 ;;
3554 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3555 ;; This defun explicitly treats mismatching parens/braces/brackets as
3556 ;; matching. It is the open brace which makes it a "brace" pair.
3557 ;;
3558 ;; If POINT is within a macro, open parens and brace pairs within
3559 ;; THIS macro MIGHT be recorded. This depends on whether their
3560 ;; syntactic properties have been suppressed by
3561 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
3562 ;;
3563 ;; Currently no characters which are given paren syntax with the
3564 ;; syntax-table property are recorded, i.e. angle bracket arglist
3565 ;; parens are never present here. Note that this might change.
3566 ;;
3567 ;; BUG: This function doesn't cope entirely well with unbalanced
3568 ;; parens in macros. (2008-12-11: this has probably been resolved
3569 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3570 ;; following case the brace before the macro isn't balanced with the
3571 ;; one after it:
3572 ;;
3573 ;; {
3574 ;; #define X {
3575 ;; }
3576 ;;
3577 ;; Note to maintainers: this function DOES get called with point
3578 ;; within comments and strings, so don't assume it doesn't!
3579 ;;
3580 ;; This function might do hidden buffer changes.
3581 (let* ((here (point))
3582 (here-bopl (c-point 'bopl))
3583 strategy ; 'forward, 'backward etc..
3584 ;; Candidate positions to start scanning from:
3585 cache-pos ; highest position below HERE already existing in
3586 ; cache (or 1).
3587 good-pos
3588 start-point ; (when scanning forward) a place below HERE where there
3589 ; are no open parens/braces between it and HERE.
3590 bopl-state
3591 res
3592 cons-separated
3593 scan-backward-pos scan-forward-p) ; used for 'backward.
3594 ;; If POINT-MIN has changed, adjust the cache
3595 (unless (= (point-min) c-state-point-min)
3596 (c-renarrow-state-cache))
3597
3598 ;; Strategy?
3599 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3600 strategy (car res)
3601 start-point (cadr res))
3602
3603 (when (eq strategy 'BOD)
3604 (setq c-state-cache nil
3605 c-state-cache-good-pos start-point))
3606
3607 ;; SCAN!
3608 (cond
3609 ((memq strategy '(forward back-and-forward BOD))
3610 (setq res (c-remove-stale-state-cache start-point here here-bopl))
3611 (setq cache-pos (car res)
3612 scan-backward-pos (cadr res)
3613 cons-separated (car (cddr res))
3614 bopl-state (cadr (cddr res))) ; will be nil if (< here-bopl
3615 ; start-point)
3616 (if (and scan-backward-pos
3617 (or cons-separated (eq strategy 'forward))) ;scan-backward-pos
3618 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3619 (setq good-pos
3620 (c-append-to-state-cache cache-pos here))
3621 (setq c-state-cache-good-pos
3622 (if (and bopl-state
3623 (< good-pos (- here c-state-cache-too-far)))
3624 (c-state-cache-non-literal-place here-bopl bopl-state)
3625 good-pos)))
3626
3627 ((eq strategy 'backward)
3628 (setq res (c-remove-stale-state-cache-backwards here)
3629 good-pos (car res)
3630 scan-backward-pos (cadr res)
3631 scan-forward-p (car (cddr res)))
3632 (if scan-backward-pos
3633 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3634 (setq c-state-cache-good-pos
3635 (if scan-forward-p
3636 (c-append-to-state-cache good-pos here)
3637 good-pos)))
3638
3639 (t ; (eq strategy 'IN-LIT)
3640 (setq c-state-cache nil
3641 c-state-cache-good-pos nil))))
3642
3643 c-state-cache)
3644
3645 (defun c-invalidate-state-cache (here)
3646 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3647 ;;
3648 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3649 ;; of all parens in preprocessor constructs, except for any such construct
3650 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3651 ;; worrying further about macros and template delimiters.
3652 (if (eval-when-compile (memq 'category-properties c-emacs-features))
3653 ;; Emacs
3654 (c-with-<->-as-parens-suppressed
3655 (if (and c-state-old-cpp-beg
3656 (< c-state-old-cpp-beg here))
3657 (c-with-all-but-one-cpps-commented-out
3658 c-state-old-cpp-beg
3659 c-state-old-cpp-end
3660 (c-invalidate-state-cache-1 here))
3661 (c-with-cpps-commented-out
3662 (c-invalidate-state-cache-1 here))))
3663 ;; XEmacs
3664 (c-invalidate-state-cache-1 here)))
3665
3666 (defmacro c-state-maybe-marker (place marker)
3667 ;; If PLACE is non-nil, return a marker marking it, otherwise nil.
3668 ;; We (re)use MARKER.
3669 `(and ,place
3670 (or ,marker (setq ,marker (make-marker)))
3671 (set-marker ,marker ,place)))
3672
3673 (defun c-parse-state ()
3674 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3675 ;; description of the functionality and return value.
3676 ;;
3677 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3678 ;; of all parens in preprocessor constructs, except for any such construct
3679 ;; containing point. We can then call `c-parse-state-1' without worrying
3680 ;; further about macros and template delimiters.
3681 (let (here-cpp-beg here-cpp-end)
3682 (save-excursion
3683 (when (c-beginning-of-macro)
3684 (setq here-cpp-beg (point))
3685 (unless
3686 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3687 here-cpp-beg)
3688 (setq here-cpp-beg nil here-cpp-end nil))))
3689 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3690 ;; subsystem.
3691 (prog1
3692 (if (eval-when-compile (memq 'category-properties c-emacs-features))
3693 ;; Emacs
3694 (c-with-<->-as-parens-suppressed
3695 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3696 (c-with-all-but-one-cpps-commented-out
3697 here-cpp-beg here-cpp-end
3698 (c-parse-state-1))
3699 (c-with-cpps-commented-out
3700 (c-parse-state-1))))
3701 ;; XEmacs
3702 (c-parse-state-1))
3703 (setq c-state-old-cpp-beg
3704 (c-state-maybe-marker here-cpp-beg c-state-old-cpp-beg-marker)
3705 c-state-old-cpp-end
3706 (c-state-maybe-marker here-cpp-end c-state-old-cpp-end-marker)))))
3707
3708 ;; Debug tool to catch cache inconsistencies. This is called from
3709 ;; 000tests.el.
3710 (defvar c-debug-parse-state nil)
3711 (unless (fboundp 'c-real-parse-state)
3712 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3713 (cc-bytecomp-defun c-real-parse-state)
3714
3715 (defvar c-parse-state-point nil)
3716 (defvar c-parse-state-state nil)
3717 (make-variable-buffer-local 'c-parse-state-state)
3718 (defun c-record-parse-state-state ()
3719 (setq c-parse-state-point (point))
3720 (when (markerp (cdr (assq 'c-state-old-cpp-beg c-parse-state-state)))
3721 (move-marker (cdr (assq 'c-state-old-cpp-beg c-parse-state-state)) nil)
3722 (move-marker (cdr (assq 'c-state-old-cpp-end c-parse-state-state)) nil))
3723 (setq c-parse-state-state
3724 (mapcar
3725 (lambda (arg)
3726 (let ((val (symbol-value arg)))
3727 (cons arg
3728 (cond ((consp val) (copy-tree val))
3729 ((markerp val) (copy-marker val))
3730 (t val)))))
3731 '(c-state-cache
3732 c-state-cache-good-pos
3733 c-state-nonlit-pos-cache
3734 c-state-nonlit-pos-cache-limit
3735 c-state-semi-nonlit-pos-cache
3736 c-state-semi-nonlit-pos-cache-limit
3737 c-state-brace-pair-desert
3738 c-state-point-min
3739 c-state-point-min-lit-type
3740 c-state-point-min-lit-start
3741 c-state-min-scan-pos
3742 c-state-old-cpp-beg
3743 c-state-old-cpp-end
3744 c-parse-state-point))))
3745 (defun c-replay-parse-state-state ()
3746 (message "%s"
3747 (concat "(setq "
3748 (mapconcat
3749 (lambda (arg)
3750 (format "%s %s%s" (car arg)
3751 (if (atom (cdr arg)) "" "'")
3752 (if (markerp (cdr arg))
3753 (format "(copy-marker %s)" (marker-position (cdr arg)))
3754 (cdr arg))))
3755 c-parse-state-state " ")
3756 ")")))
3757
3758 (defun c-debug-parse-state-double-cons (state)
3759 (let (state-car conses-not-ok)
3760 (while state
3761 (setq state-car (car state)
3762 state (cdr state))
3763 (if (and (consp state-car)
3764 (consp (car state)))
3765 (setq conses-not-ok t)))
3766 conses-not-ok))
3767
3768 (defun c-debug-parse-state ()
3769 (let ((here (point)) (min-point (point-min)) (res1 (c-real-parse-state)) res2)
3770 (let ((c-state-cache nil)
3771 (c-state-cache-good-pos 1)
3772 (c-state-nonlit-pos-cache nil)
3773 (c-state-nonlit-pos-cache-limit 1)
3774 (c-state-brace-pair-desert nil)
3775 (c-state-point-min 1)
3776 (c-state-point-min-lit-type nil)
3777 (c-state-point-min-lit-start nil)
3778 (c-state-min-scan-pos 1)
3779 (c-state-old-cpp-beg nil)
3780 (c-state-old-cpp-end nil))
3781 (setq res2 (c-real-parse-state)))
3782 (unless (equal res1 res2)
3783 ;; The cache can actually go further back due to the ad-hoc way
3784 ;; the first paren is found, so try to whack off a bit of its
3785 ;; start before complaining.
3786 ;; (save-excursion
3787 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3788 ;; (c-beginning-of-defun-1)
3789 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3790 ;; (c-beginning-of-defun-1))
3791 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3792 ;; (message (concat "c-parse-state inconsistency at %s: "
3793 ;; "using cache: %s, from scratch: %s")
3794 ;; here res1 res2)))
3795 (message (concat "c-parse-state inconsistency at %s: "
3796 "using cache: %s, from scratch: %s. POINT-MIN: %s")
3797 here res1 res2 min-point)
3798 (message "Old state:")
3799 (c-replay-parse-state-state))
3800
3801 (when (c-debug-parse-state-double-cons res1)
3802 (message "c-parse-state INVALIDITY at %s: %s"
3803 here res1)
3804 (message "Old state:")
3805 (c-replay-parse-state-state))
3806
3807 (c-record-parse-state-state)
3808 res2 ; res1 correct a cascading series of errors ASAP
3809 ))
3810
3811 (defun c-toggle-parse-state-debug (&optional arg)
3812 (interactive "P")
3813 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3814 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3815 'c-debug-parse-state
3816 'c-real-parse-state)))
3817 (c-keep-region-active)
3818 (message "c-debug-parse-state %sabled"
3819 (if c-debug-parse-state "en" "dis")))
3820 (when c-debug-parse-state
3821 (c-toggle-parse-state-debug 1))
3822
3823 \f
3824 (defun c-whack-state-before (bufpos paren-state)
3825 ;; Whack off any state information from PAREN-STATE which lies
3826 ;; before BUFPOS. Not destructive on PAREN-STATE.
3827 (let* ((newstate (list nil))
3828 (ptr newstate)
3829 car)
3830 (while paren-state
3831 (setq car (car paren-state)
3832 paren-state (cdr paren-state))
3833 (if (< (if (consp car) (car car) car) bufpos)
3834 (setq paren-state nil)
3835 (setcdr ptr (list car))
3836 (setq ptr (cdr ptr))))
3837 (cdr newstate)))
3838
3839 (defun c-whack-state-after (bufpos paren-state)
3840 ;; Whack off any state information from PAREN-STATE which lies at or
3841 ;; after BUFPOS. Not destructive on PAREN-STATE.
3842 (catch 'done
3843 (while paren-state
3844 (let ((car (car paren-state)))
3845 (if (consp car)
3846 ;; just check the car, because in a balanced brace
3847 ;; expression, it must be impossible for the corresponding
3848 ;; close brace to be before point, but the open brace to
3849 ;; be after.
3850 (if (<= bufpos (car car))
3851 nil ; whack it off
3852 (if (< bufpos (cdr car))
3853 ;; its possible that the open brace is before
3854 ;; bufpos, but the close brace is after. In that
3855 ;; case, convert this to a non-cons element. The
3856 ;; rest of the state is before bufpos, so we're
3857 ;; done.
3858 (throw 'done (cons (car car) (cdr paren-state)))
3859 ;; we know that both the open and close braces are
3860 ;; before bufpos, so we also know that everything else
3861 ;; on state is before bufpos.
3862 (throw 'done paren-state)))
3863 (if (<= bufpos car)
3864 nil ; whack it off
3865 ;; it's before bufpos, so everything else should too.
3866 (throw 'done paren-state)))
3867 (setq paren-state (cdr paren-state)))
3868 nil)))
3869
3870 (defun c-most-enclosing-brace (paren-state &optional bufpos)
3871 ;; Return the bufpos of the innermost enclosing open paren before
3872 ;; bufpos, or nil if none was found.
3873 (let (enclosingp)
3874 (or bufpos (setq bufpos 134217727))
3875 (while paren-state
3876 (setq enclosingp (car paren-state)
3877 paren-state (cdr paren-state))
3878 (if (or (consp enclosingp)
3879 (>= enclosingp bufpos))
3880 (setq enclosingp nil)
3881 (setq paren-state nil)))
3882 enclosingp))
3883
3884 (defun c-least-enclosing-brace (paren-state)
3885 ;; Return the bufpos of the outermost enclosing open paren, or nil
3886 ;; if none was found.
3887 (let (pos elem)
3888 (while paren-state
3889 (setq elem (car paren-state)
3890 paren-state (cdr paren-state))
3891 (if (integerp elem)
3892 (setq pos elem)))
3893 pos))
3894
3895 (defun c-safe-position (bufpos paren-state)
3896 ;; Return the closest "safe" position recorded on PAREN-STATE that
3897 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3898 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3899 ;; find the closest limit before a given limit that might be nil.
3900 ;;
3901 ;; A "safe" position is a position at or after a recorded open
3902 ;; paren, or after a recorded close paren. The returned position is
3903 ;; thus either the first position after a close brace, or the first
3904 ;; position after an enclosing paren, or at the enclosing paren in
3905 ;; case BUFPOS is immediately after it.
3906 (when bufpos
3907 (let (elem)
3908 (catch 'done
3909 (while paren-state
3910 (setq elem (car paren-state))
3911 (if (consp elem)
3912 (cond ((< (cdr elem) bufpos)
3913 (throw 'done (cdr elem)))
3914 ((< (car elem) bufpos)
3915 ;; See below.
3916 (throw 'done (min (1+ (car elem)) bufpos))))
3917 (if (< elem bufpos)
3918 ;; elem is the position at and not after the opening paren, so
3919 ;; we can go forward one more step unless it's equal to
3920 ;; bufpos. This is useful in some cases avoid an extra paren
3921 ;; level between the safe position and bufpos.
3922 (throw 'done (min (1+ elem) bufpos))))
3923 (setq paren-state (cdr paren-state)))))))
3924
3925 (defun c-beginning-of-syntax ()
3926 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3927 ;; goes to the closest previous point that is known to be outside
3928 ;; any string literal or comment. `c-state-cache' is used if it has
3929 ;; a position in the vicinity.
3930 (let* ((paren-state c-state-cache)
3931 elem
3932
3933 (pos (catch 'done
3934 ;; Note: Similar code in `c-safe-position'. The
3935 ;; difference is that we accept a safe position at
3936 ;; the point and don't bother to go forward past open
3937 ;; parens.
3938 (while paren-state
3939 (setq elem (car paren-state))
3940 (if (consp elem)
3941 (cond ((<= (cdr elem) (point))
3942 (throw 'done (cdr elem)))
3943 ((<= (car elem) (point))
3944 (throw 'done (car elem))))
3945 (if (<= elem (point))
3946 (throw 'done elem)))
3947 (setq paren-state (cdr paren-state)))
3948 (point-min))))
3949
3950 (if (> pos (- (point) 4000))
3951 (goto-char pos)
3952 ;; The position is far back. Try `c-beginning-of-defun-1'
3953 ;; (although we can't be entirely sure it will go to a position
3954 ;; outside a comment or string in current emacsen). FIXME:
3955 ;; Consult `syntax-ppss' here.
3956 (c-beginning-of-defun-1)
3957 (if (< (point) pos)
3958 (goto-char pos)))))
3959
3960 \f
3961 ;; Tools for scanning identifiers and other tokens.
3962
3963 (defun c-on-identifier ()
3964 "Return non-nil if the point is on or directly after an identifier.
3965 Keywords are recognized and not considered identifiers. If an
3966 identifier is detected, the returned value is its starting position.
3967 If an identifier ends at the point and another begins at it \(can only
3968 happen in Pike) then the point for the preceding one is returned.
3969
3970 Note that this function might do hidden buffer changes. See the
3971 comment at the start of cc-engine.el for more info."
3972
3973 ;; FIXME: Shouldn't this function handle "operator" in C++?
3974
3975 (save-excursion
3976 (skip-syntax-backward "w_")
3977
3978 (or
3979
3980 ;; Check for a normal (non-keyword) identifier.
3981 (and (looking-at c-symbol-start)
3982 (not (looking-at c-keywords-regexp))
3983 (point))
3984
3985 (when (c-major-mode-is 'pike-mode)
3986 ;; Handle the `<operator> syntax in Pike.
3987 (let ((pos (point)))
3988 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3989 (and (if (< (skip-chars-backward "`") 0)
3990 t
3991 (goto-char pos)
3992 (eq (char-after) ?\`))
3993 (looking-at c-symbol-key)
3994 (>= (match-end 0) pos)
3995 (point))))
3996
3997 ;; Handle the "operator +" syntax in C++.
3998 (when (and c-overloadable-operators-regexp
3999 (= (c-backward-token-2 0) 0))
4000
4001 (cond ((and (looking-at c-overloadable-operators-regexp)
4002 (or (not c-opt-op-identifier-prefix)
4003 (and (= (c-backward-token-2 1) 0)
4004 (looking-at c-opt-op-identifier-prefix))))
4005 (point))
4006
4007 ((save-excursion
4008 (and c-opt-op-identifier-prefix
4009 (looking-at c-opt-op-identifier-prefix)
4010 (= (c-forward-token-2 1) 0)
4011 (looking-at c-overloadable-operators-regexp)))
4012 (point))))
4013
4014 )))
4015
4016 (defsubst c-simple-skip-symbol-backward ()
4017 ;; If the point is at the end of a symbol then skip backward to the
4018 ;; beginning of it. Don't move otherwise. Return non-nil if point
4019 ;; moved.
4020 ;;
4021 ;; This function might do hidden buffer changes.
4022 (or (< (skip-syntax-backward "w_") 0)
4023 (and (c-major-mode-is 'pike-mode)
4024 ;; Handle the `<operator> syntax in Pike.
4025 (let ((pos (point)))
4026 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
4027 (< (skip-chars-backward "`") 0)
4028 (looking-at c-symbol-key)
4029 (>= (match-end 0) pos))
4030 t
4031 (goto-char pos)
4032 nil)))))
4033
4034 (defun c-beginning-of-current-token (&optional back-limit)
4035 ;; Move to the beginning of the current token. Do not move if not
4036 ;; in the middle of one. BACK-LIMIT may be used to bound the
4037 ;; backward search; if given it's assumed to be at the boundary
4038 ;; between two tokens. Return non-nil if the point is moved, nil
4039 ;; otherwise.
4040 ;;
4041 ;; This function might do hidden buffer changes.
4042 (let ((start (point)))
4043 (if (looking-at "\\w\\|\\s_")
4044 (skip-syntax-backward "w_" back-limit)
4045 (when (< (skip-syntax-backward ".()" back-limit) 0)
4046 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
4047 (match-end 0))
4048 ;; `c-nonsymbol-token-regexp' should always match
4049 ;; since we've skipped backward over punctuation
4050 ;; or paren syntax, but consume one char in case
4051 ;; it doesn't so that we don't leave point before
4052 ;; some earlier incorrect token.
4053 (1+ (point)))))
4054 (if (<= pos start)
4055 (goto-char pos))))))
4056 (< (point) start)))
4057
4058 (defun c-end-of-current-token (&optional back-limit)
4059 ;; Move to the end of the current token. Do not move if not in the
4060 ;; middle of one. BACK-LIMIT may be used to bound the backward
4061 ;; search; if given it's assumed to be at the boundary between two
4062 ;; tokens. Return non-nil if the point is moved, nil otherwise.
4063 ;;
4064 ;; This function might do hidden buffer changes.
4065 (let ((start (point)))
4066 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
4067 (skip-syntax-forward "w_"))
4068 ((< (skip-syntax-backward ".()" back-limit) 0)
4069 (while (progn
4070 (if (looking-at c-nonsymbol-token-regexp)
4071 (goto-char (match-end 0))
4072 ;; `c-nonsymbol-token-regexp' should always match since
4073 ;; we've skipped backward over punctuation or paren
4074 ;; syntax, but move forward in case it doesn't so that
4075 ;; we don't leave point earlier than we started with.
4076 (forward-char))
4077 (< (point) start)))))
4078 (> (point) start)))
4079
4080 (defconst c-jump-syntax-balanced
4081 (if (memq 'gen-string-delim c-emacs-features)
4082 "\\w\\|\\s_\\|\\s(\\|\\s)\\|\\s\"\\|\\s|"
4083 "\\w\\|\\s_\\|\\s(\\|\\s)\\|\\s\""))
4084
4085 (defconst c-jump-syntax-unbalanced
4086 (if (memq 'gen-string-delim c-emacs-features)
4087 "\\w\\|\\s_\\|\\s\"\\|\\s|"
4088 "\\w\\|\\s_\\|\\s\""))
4089
4090 (defun c-forward-token-2 (&optional count balanced limit)
4091 "Move forward by tokens.
4092 A token is defined as all symbols and identifiers which aren't
4093 syntactic whitespace \(note that multicharacter tokens like \"==\" are
4094 treated properly). Point is always either left at the beginning of a
4095 token or not moved at all. COUNT specifies the number of tokens to
4096 move; a negative COUNT moves in the opposite direction. A COUNT of 0
4097 moves to the next token beginning only if not already at one. If
4098 BALANCED is true, move over balanced parens, otherwise move into them.
4099 Also, if BALANCED is true, never move out of an enclosing paren.
4100
4101 LIMIT sets the limit for the movement and defaults to the point limit.
4102 The case when LIMIT is set in the middle of a token, comment or macro
4103 is handled correctly, i.e. the point won't be left there.
4104
4105 Return the number of tokens left to move \(positive or negative). If
4106 BALANCED is true, a move over a balanced paren counts as one. Note
4107 that if COUNT is 0 and no appropriate token beginning is found, 1 will
4108 be returned. Thus, a return value of 0 guarantees that point is at
4109 the requested position and a return value less \(without signs) than
4110 COUNT guarantees that point is at the beginning of some token.
4111
4112 Note that this function might do hidden buffer changes. See the
4113 comment at the start of cc-engine.el for more info."
4114
4115 (or count (setq count 1))
4116 (if (< count 0)
4117 (- (c-backward-token-2 (- count) balanced limit))
4118
4119 (let ((jump-syntax (if balanced
4120 c-jump-syntax-balanced
4121 c-jump-syntax-unbalanced))
4122 (last (point))
4123 (prev (point)))
4124
4125 (if (zerop count)
4126 ;; If count is zero we should jump if in the middle of a token.
4127 (c-end-of-current-token))
4128
4129 (save-restriction
4130 (if limit (narrow-to-region (point-min) limit))
4131 (if (/= (point)
4132 (progn (c-forward-syntactic-ws) (point)))
4133 ;; Skip whitespace. Count this as a move if we did in
4134 ;; fact move.
4135 (setq count (max (1- count) 0)))
4136
4137 (if (eobp)
4138 ;; Moved out of bounds. Make sure the returned count isn't zero.
4139 (progn
4140 (if (zerop count) (setq count 1))
4141 (goto-char last))
4142
4143 ;; Use `condition-case' to avoid having the limit tests
4144 ;; inside the loop.
4145 (condition-case nil
4146 (while (and
4147 (> count 0)
4148 (progn
4149 (setq last (point))
4150 (cond ((looking-at jump-syntax)
4151 (goto-char (scan-sexps (point) 1))
4152 t)
4153 ((looking-at c-nonsymbol-token-regexp)
4154 (goto-char (match-end 0))
4155 t)
4156 ;; `c-nonsymbol-token-regexp' above should always
4157 ;; match if there are correct tokens. Try to
4158 ;; widen to see if the limit was set in the
4159 ;; middle of one, else fall back to treating
4160 ;; the offending thing as a one character token.
4161 ((and limit
4162 (save-restriction
4163 (widen)
4164 (looking-at c-nonsymbol-token-regexp)))
4165 nil)
4166 (t
4167 (forward-char)
4168 t))))
4169 (c-forward-syntactic-ws)
4170 (setq prev last
4171 count (1- count)))
4172 (error (goto-char last)))
4173
4174 (when (eobp)
4175 (goto-char prev)
4176 (setq count (1+ count)))))
4177
4178 count)))
4179
4180 (defun c-backward-token-2 (&optional count balanced limit)
4181 "Move backward by tokens.
4182 See `c-forward-token-2' for details."
4183
4184 (or count (setq count 1))
4185 (if (< count 0)
4186 (- (c-forward-token-2 (- count) balanced limit))
4187
4188 (or limit (setq limit (point-min)))
4189 (let ((jump-syntax (if balanced
4190 c-jump-syntax-balanced
4191 c-jump-syntax-unbalanced))
4192 (last (point)))
4193
4194 (if (zerop count)
4195 ;; The count is zero so try to skip to the beginning of the
4196 ;; current token.
4197 (if (> (point)
4198 (progn (c-beginning-of-current-token) (point)))
4199 (if (< (point) limit)
4200 ;; The limit is inside the same token, so return 1.
4201 (setq count 1))
4202
4203 ;; We're not in the middle of a token. If there's
4204 ;; whitespace after the point then we must move backward,
4205 ;; so set count to 1 in that case.
4206 (and (looking-at c-syntactic-ws-start)
4207 ;; If we're looking at a '#' that might start a cpp
4208 ;; directive then we have to do a more elaborate check.
4209 (or (/= (char-after) ?#)
4210 (not c-opt-cpp-prefix)
4211 (save-excursion
4212 (and (= (point)
4213 (progn (beginning-of-line)
4214 (looking-at "[ \t]*")
4215 (match-end 0)))
4216 (or (bobp)
4217 (progn (backward-char)
4218 (not (eq (char-before) ?\\)))))))
4219 (setq count 1))))
4220
4221 ;; Use `condition-case' to avoid having to check for buffer
4222 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
4223 (condition-case nil
4224 (while (and
4225 (> count 0)
4226 (progn
4227 (c-backward-syntactic-ws)
4228 (backward-char)
4229 (if (looking-at jump-syntax)
4230 (goto-char (scan-sexps (1+ (point)) -1))
4231 ;; This can be very inefficient if there's a long
4232 ;; sequence of operator tokens without any separation.
4233 ;; That doesn't happen in practice, anyway.
4234 (c-beginning-of-current-token))
4235 (>= (point) limit)))
4236 (setq last (point)
4237 count (1- count)))
4238 (error (goto-char last)))
4239
4240 (if (< (point) limit)
4241 (goto-char last))
4242
4243 count)))
4244
4245 (defun c-forward-token-1 (&optional count balanced limit)
4246 "Like `c-forward-token-2' but doesn't treat multicharacter operator
4247 tokens like \"==\" as single tokens, i.e. all sequences of symbol
4248 characters are jumped over character by character. This function is
4249 for compatibility only; it's only a wrapper over `c-forward-token-2'."
4250 (let ((c-nonsymbol-token-regexp "\\s."))
4251 (c-forward-token-2 count balanced limit)))
4252
4253 (defun c-backward-token-1 (&optional count balanced limit)
4254 "Like `c-backward-token-2' but doesn't treat multicharacter operator
4255 tokens like \"==\" as single tokens, i.e. all sequences of symbol
4256 characters are jumped over character by character. This function is
4257 for compatibility only; it's only a wrapper over `c-backward-token-2'."
4258 (let ((c-nonsymbol-token-regexp "\\s."))
4259 (c-backward-token-2 count balanced limit)))
4260
4261 \f
4262 ;; Tools for doing searches restricted to syntactically relevant text.
4263
4264 (defun c-syntactic-re-search-forward (regexp &optional bound noerror
4265 paren-level not-inside-token
4266 lookbehind-submatch)
4267 "Like `re-search-forward', but only report matches that are found
4268 in syntactically significant text. I.e. matches in comments, macros
4269 or string literals are ignored. The start point is assumed to be
4270 outside any comment, macro or string literal, or else the content of
4271 that region is taken as syntactically significant text.
4272
4273 If PAREN-LEVEL is non-nil, an additional restriction is added to
4274 ignore matches in nested paren sexps. The search will also not go
4275 outside the current list sexp, which has the effect that if the point
4276 should be moved to BOUND when no match is found \(i.e. NOERROR is
4277 neither nil nor t), then it will be at the closing paren if the end of
4278 the current list sexp is encountered first.
4279
4280 If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
4281 ignored. Things like multicharacter operators and special symbols
4282 \(e.g. \"`()\" in Pike) are handled but currently not floating point
4283 constants.
4284
4285 If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
4286 subexpression in REGEXP. The end of that submatch is used as the
4287 position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
4288 isn't used or if that subexpression didn't match then the start
4289 position of the whole match is used instead. The \"look behind\"
4290 subexpression is never tested before the starting position, so it
4291 might be a good idea to include \\=\\= as a match alternative in it.
4292
4293 Optimization note: Matches might be missed if the \"look behind\"
4294 subexpression can match the end of nonwhite syntactic whitespace,
4295 i.e. the end of comments or cpp directives. This since the function
4296 skips over such things before resuming the search. It's on the other
4297 hand not safe to assume that the \"look behind\" subexpression never
4298 matches syntactic whitespace.
4299
4300 Bug: Unbalanced parens inside cpp directives are currently not handled
4301 correctly \(i.e. they don't get ignored as they should) when
4302 PAREN-LEVEL is set.
4303
4304 Note that this function might do hidden buffer changes. See the
4305 comment at the start of cc-engine.el for more info."
4306
4307 (or bound (setq bound (point-max)))
4308 (if paren-level (setq paren-level -1))
4309
4310 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
4311
4312 (let ((start (point))
4313 tmp
4314 ;; Start position for the last search.
4315 search-pos
4316 ;; The `parse-partial-sexp' state between the start position
4317 ;; and the point.
4318 state
4319 ;; The current position after the last state update. The next
4320 ;; `parse-partial-sexp' continues from here.
4321 (state-pos (point))
4322 ;; The position at which to check the state and the state
4323 ;; there. This is separate from `state-pos' since we might
4324 ;; need to back up before doing the next search round.
4325 check-pos check-state
4326 ;; Last position known to end a token.
4327 (last-token-end-pos (point-min))
4328 ;; Set when a valid match is found.
4329 found)
4330
4331 (condition-case err
4332 (while
4333 (and
4334 (progn
4335 (setq search-pos (point))
4336 (re-search-forward regexp bound noerror))
4337
4338 (progn
4339 (setq state (parse-partial-sexp
4340 state-pos (match-beginning 0) paren-level nil state)
4341 state-pos (point))
4342 (if (setq check-pos (and lookbehind-submatch
4343 (or (not paren-level)
4344 (>= (car state) 0))
4345 (match-end lookbehind-submatch)))
4346 (setq check-state (parse-partial-sexp
4347 state-pos check-pos paren-level nil state))
4348 (setq check-pos state-pos
4349 check-state state))
4350
4351 ;; NOTE: If we got a look behind subexpression and get
4352 ;; an insignificant match in something that isn't
4353 ;; syntactic whitespace (i.e. strings or in nested
4354 ;; parentheses), then we can never skip more than a
4355 ;; single character from the match start position
4356 ;; (i.e. `state-pos' here) before continuing the
4357 ;; search. That since the look behind subexpression
4358 ;; might match the end of the insignificant region in
4359 ;; the next search.
4360
4361 (cond
4362 ((elt check-state 7)
4363 ;; Match inside a line comment. Skip to eol. Use
4364 ;; `re-search-forward' instead of `skip-chars-forward' to get
4365 ;; the right bound behavior.
4366 (re-search-forward "[\n\r]" bound noerror))
4367
4368 ((elt check-state 4)
4369 ;; Match inside a block comment. Skip to the '*/'.
4370 (search-forward "*/" bound noerror))
4371
4372 ((and (not (elt check-state 5))
4373 (eq (char-before check-pos) ?/)
4374 (not (c-get-char-property (1- check-pos) 'syntax-table))
4375 (memq (char-after check-pos) '(?/ ?*)))
4376 ;; Match in the middle of the opener of a block or line
4377 ;; comment.
4378 (if (= (char-after check-pos) ?/)
4379 (re-search-forward "[\n\r]" bound noerror)
4380 (search-forward "*/" bound noerror)))
4381
4382 ;; The last `parse-partial-sexp' above might have
4383 ;; stopped short of the real check position if the end
4384 ;; of the current sexp was encountered in paren-level
4385 ;; mode. The checks above are always false in that
4386 ;; case, and since they can do better skipping in
4387 ;; lookbehind-submatch mode, we do them before
4388 ;; checking the paren level.
4389
4390 ((and paren-level
4391 (/= (setq tmp (car check-state)) 0))
4392 ;; Check the paren level first since we're short of the
4393 ;; syntactic checking position if the end of the
4394 ;; current sexp was encountered by `parse-partial-sexp'.
4395 (if (> tmp 0)
4396
4397 ;; Inside a nested paren sexp.
4398 (if lookbehind-submatch
4399 ;; See the NOTE above.
4400 (progn (goto-char state-pos) t)
4401 ;; Skip out of the paren quickly.
4402 (setq state (parse-partial-sexp state-pos bound 0 nil state)
4403 state-pos (point)))
4404
4405 ;; Have exited the current paren sexp.
4406 (if noerror
4407 (progn
4408 ;; The last `parse-partial-sexp' call above
4409 ;; has left us just after the closing paren
4410 ;; in this case, so we can modify the bound
4411 ;; to leave the point at the right position
4412 ;; upon return.
4413 (setq bound (1- (point)))
4414 nil)
4415 (signal 'search-failed (list regexp)))))
4416
4417 ((setq tmp (elt check-state 3))
4418 ;; Match inside a string.
4419 (if (or lookbehind-submatch
4420 (not (integerp tmp)))
4421 ;; See the NOTE above.
4422 (progn (goto-char state-pos) t)
4423 ;; Skip to the end of the string before continuing.
4424 (let ((ender (make-string 1 tmp)) (continue t))
4425 (while (if (search-forward ender bound noerror)
4426 (progn
4427 (setq state (parse-partial-sexp
4428 state-pos (point) nil nil state)
4429 state-pos (point))
4430 (elt state 3))
4431 (setq continue nil)))
4432 continue)))
4433
4434 ((save-excursion
4435 (save-match-data
4436 (c-beginning-of-macro start)))
4437 ;; Match inside a macro. Skip to the end of it.
4438 (c-end-of-macro)
4439 (cond ((<= (point) bound) t)
4440 (noerror nil)
4441 (t (signal 'search-failed (list regexp)))))
4442
4443 ((and not-inside-token
4444 (or (< check-pos last-token-end-pos)
4445 (< check-pos
4446 (save-excursion
4447 (goto-char check-pos)
4448 (save-match-data
4449 (c-end-of-current-token last-token-end-pos))
4450 (setq last-token-end-pos (point))))))
4451 ;; Inside a token.
4452 (if lookbehind-submatch
4453 ;; See the NOTE above.
4454 (goto-char state-pos)
4455 (goto-char (min last-token-end-pos bound))))
4456
4457 (t
4458 ;; A real match.
4459 (setq found t)
4460 nil)))
4461
4462 ;; Should loop to search again, but take care to avoid
4463 ;; looping on the same spot.
4464 (or (/= search-pos (point))
4465 (if (= (point) bound)
4466 (if noerror
4467 nil
4468 (signal 'search-failed (list regexp)))
4469 (forward-char)
4470 t))))
4471
4472 (error
4473 (goto-char start)
4474 (signal (car err) (cdr err))))
4475
4476 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
4477
4478 (if found
4479 (progn
4480 (goto-char (match-end 0))
4481 (match-end 0))
4482
4483 ;; Search failed. Set point as appropriate.
4484 (if (eq noerror t)
4485 (goto-char start)
4486 (goto-char bound))
4487 nil)))
4488
4489 (defvar safe-pos-list) ; bound in c-syntactic-skip-backward
4490
4491 (defsubst c-ssb-lit-begin ()
4492 ;; Return the start of the literal point is in, or nil.
4493 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
4494 ;; bound in the caller.
4495
4496 ;; Use `parse-partial-sexp' from a safe position down to the point to check
4497 ;; if it's outside comments and strings.
4498 (save-excursion
4499 (let ((pos (point)) safe-pos state)
4500 ;; Pick a safe position as close to the point as possible.
4501 ;;
4502 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
4503 ;; position.
4504
4505 (while (and safe-pos-list
4506 (> (car safe-pos-list) (point)))
4507 (setq safe-pos-list (cdr safe-pos-list)))
4508 (unless (setq safe-pos (car-safe safe-pos-list))
4509 (setq safe-pos (max (or (c-safe-position
4510 (point) (c-parse-state))
4511 0)
4512 (point-min))
4513 safe-pos-list (list safe-pos)))
4514
4515 ;; Cache positions along the way to use if we have to back up more. We
4516 ;; cache every closing paren on the same level. If the paren cache is
4517 ;; relevant in this region then we're typically already on the same
4518 ;; level as the target position. Note that we might cache positions
4519 ;; after opening parens in case safe-pos is in a nested list. That's
4520 ;; both uncommon and harmless.
4521 (while (progn
4522 (setq state (parse-partial-sexp
4523 safe-pos pos 0))
4524 (< (point) pos))
4525 (setq safe-pos (point)
4526 safe-pos-list (cons safe-pos safe-pos-list)))
4527
4528 ;; If the state contains the start of the containing sexp we cache that
4529 ;; position too, so that parse-partial-sexp in the next run has a bigger
4530 ;; chance of starting at the same level as the target position and thus
4531 ;; will get more good safe positions into the list.
4532 (if (elt state 1)
4533 (setq safe-pos (1+ (elt state 1))
4534 safe-pos-list (cons safe-pos safe-pos-list)))
4535
4536 (if (or (elt state 3) (elt state 4))
4537 ;; Inside string or comment. Continue search at the
4538 ;; beginning of it.
4539 (elt state 8)))))
4540
4541 (defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4542 "Like `skip-chars-backward' but only look at syntactically relevant chars,
4543 i.e. don't stop at positions inside syntactic whitespace or string
4544 literals. Preprocessor directives are also ignored, with the exception
4545 of the one that the point starts within, if any. If LIMIT is given,
4546 it's assumed to be at a syntactically relevant position.
4547
4548 If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4549 sexps, and the search will also not go outside the current paren sexp.
4550 However, if LIMIT or the buffer limit is reached inside a nested paren
4551 then the point will be left at the limit.
4552
4553 Non-nil is returned if the point moved, nil otherwise.
4554
4555 Note that this function might do hidden buffer changes. See the
4556 comment at the start of cc-engine.el for more info."
4557
4558 (c-self-bind-state-cache
4559 (let ((start (point))
4560 state-2
4561 ;; A list of syntactically relevant positions in descending
4562 ;; order. It's used to avoid scanning repeatedly over
4563 ;; potentially large regions with `parse-partial-sexp' to verify
4564 ;; each position. Used in `c-ssb-lit-begin'
4565 safe-pos-list
4566 ;; The result from `c-beginning-of-macro' at the start position or the
4567 ;; start position itself if it isn't within a macro. Evaluated on
4568 ;; demand.
4569 start-macro-beg
4570 ;; The earliest position after the current one with the same paren
4571 ;; level. Used only when `paren-level' is set.
4572 lit-beg
4573 (paren-level-pos (point)))
4574
4575 (while
4576 (progn
4577 ;; The next loop "tries" to find the end point each time round,
4578 ;; loops when it hasn't succeeded.
4579 (while
4580 (and
4581 (let ((pos (point)))
4582 (while (and
4583 (< (skip-chars-backward skip-chars limit) 0)
4584 ;; Don't stop inside a literal.
4585 (when (setq lit-beg (c-ssb-lit-begin))
4586 (goto-char lit-beg)
4587 t)))
4588 (< (point) pos))
4589
4590 (let ((pos (point)) state-2 pps-end-pos)
4591
4592 (cond
4593 ((and paren-level
4594 (save-excursion
4595 (setq state-2 (parse-partial-sexp
4596 pos paren-level-pos -1)
4597 pps-end-pos (point))
4598 (/= (car state-2) 0)))
4599 ;; Not at the right level.
4600
4601 (if (and (< (car state-2) 0)
4602 ;; We stop above if we go out of a paren.
4603 ;; Now check whether it precedes or is
4604 ;; nested in the starting sexp.
4605 (save-excursion
4606 (setq state-2
4607 (parse-partial-sexp
4608 pps-end-pos paren-level-pos
4609 nil nil state-2))
4610 (< (car state-2) 0)))
4611
4612 ;; We've stopped short of the starting position
4613 ;; so the hit was inside a nested list. Go up
4614 ;; until we are at the right level.
4615 (condition-case nil
4616 (progn
4617 (goto-char (scan-lists pos -1
4618 (- (car state-2))))
4619 (setq paren-level-pos (point))
4620 (if (and limit (>= limit paren-level-pos))
4621 (progn
4622 (goto-char limit)
4623 nil)
4624 t))
4625 (error
4626 (goto-char (or limit (point-min)))
4627 nil))
4628
4629 ;; The hit was outside the list at the start
4630 ;; position. Go to the start of the list and exit.
4631 (goto-char (1+ (elt state-2 1)))
4632 nil))
4633
4634 ((c-beginning-of-macro limit)
4635 ;; Inside a macro.
4636 (if (< (point)
4637 (or start-macro-beg
4638 (setq start-macro-beg
4639 (save-excursion
4640 (goto-char start)
4641 (c-beginning-of-macro limit)
4642 (point)))))
4643 t
4644
4645 ;; It's inside the same macro we started in so it's
4646 ;; a relevant match.
4647 (goto-char pos)
4648 nil))))))
4649
4650 (> (point)
4651 (progn
4652 ;; Skip syntactic ws afterwards so that we don't stop at the
4653 ;; end of a comment if `skip-chars' is something like "^/".
4654 (c-backward-syntactic-ws)
4655 (point)))))
4656
4657 ;; We might want to extend this with more useful return values in
4658 ;; the future.
4659 (/= (point) start))))
4660
4661 ;; The following is an alternative implementation of
4662 ;; `c-syntactic-skip-backward' that uses backward movement to keep
4663 ;; track of the syntactic context. It turned out to be generally
4664 ;; slower than the one above which uses forward checks from earlier
4665 ;; safe positions.
4666 ;;
4667 ;;(defconst c-ssb-stop-re
4668 ;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4669 ;; ;; stop at to avoid going into comments and literals.
4670 ;; (concat
4671 ;; ;; Match comment end syntax and string literal syntax. Also match
4672 ;; ;; '/' for block comment endings (not covered by comment end
4673 ;; ;; syntax).
4674 ;; "\\s>\\|/\\|\\s\""
4675 ;; (if (memq 'gen-string-delim c-emacs-features)
4676 ;; "\\|\\s|"
4677 ;; "")
4678 ;; (if (memq 'gen-comment-delim c-emacs-features)
4679 ;; "\\|\\s!"
4680 ;; "")))
4681 ;;
4682 ;;(defconst c-ssb-stop-paren-re
4683 ;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4684 ;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4685 ;;
4686 ;;(defconst c-ssb-sexp-end-re
4687 ;; ;; Regexp matching the ending syntax of a complex sexp.
4688 ;; (concat c-string-limit-regexp "\\|\\s)"))
4689 ;;
4690 ;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4691 ;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4692 ;;i.e. don't stop at positions inside syntactic whitespace or string
4693 ;;literals. Preprocessor directives are also ignored. However, if the
4694 ;;point is within a comment, string literal or preprocessor directory to
4695 ;;begin with, its contents is treated as syntactically relevant chars.
4696 ;;If LIMIT is given, it limits the backward search and the point will be
4697 ;;left there if no earlier position is found.
4698 ;;
4699 ;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4700 ;;sexps, and the search will also not go outside the current paren sexp.
4701 ;;However, if LIMIT or the buffer limit is reached inside a nested paren
4702 ;;then the point will be left at the limit.
4703 ;;
4704 ;;Non-nil is returned if the point moved, nil otherwise.
4705 ;;
4706 ;;Note that this function might do hidden buffer changes. See the
4707 ;;comment at the start of cc-engine.el for more info."
4708 ;;
4709 ;; (save-restriction
4710 ;; (when limit
4711 ;; (narrow-to-region limit (point-max)))
4712 ;;
4713 ;; (let ((start (point)))
4714 ;; (catch 'done
4715 ;; (while (let ((last-pos (point))
4716 ;; (stop-pos (progn
4717 ;; (skip-chars-backward skip-chars)
4718 ;; (point))))
4719 ;;
4720 ;; ;; Skip back over the same region as
4721 ;; ;; `skip-chars-backward' above, but keep to
4722 ;; ;; syntactically relevant positions.
4723 ;; (goto-char last-pos)
4724 ;; (while (and
4725 ;; ;; `re-search-backward' with a single char regexp
4726 ;; ;; should be fast.
4727 ;; (re-search-backward
4728 ;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4729 ;; stop-pos 'move)
4730 ;;
4731 ;; (progn
4732 ;; (cond
4733 ;; ((looking-at "\\s(")
4734 ;; ;; `paren-level' is set and we've found the
4735 ;; ;; start of the containing paren.
4736 ;; (forward-char)
4737 ;; (throw 'done t))
4738 ;;
4739 ;; ((looking-at c-ssb-sexp-end-re)
4740 ;; ;; We're at the end of a string literal or paren
4741 ;; ;; sexp (if `paren-level' is set).
4742 ;; (forward-char)
4743 ;; (condition-case nil
4744 ;; (c-backward-sexp)
4745 ;; (error
4746 ;; (goto-char limit)
4747 ;; (throw 'done t))))
4748 ;;
4749 ;; (t
4750 ;; (forward-char)
4751 ;; ;; At the end of some syntactic ws or possibly
4752 ;; ;; after a plain '/' operator.
4753 ;; (let ((pos (point)))
4754 ;; (c-backward-syntactic-ws)
4755 ;; (if (= pos (point))
4756 ;; ;; Was a plain '/' operator. Go past it.
4757 ;; (backward-char)))))
4758 ;;
4759 ;; (> (point) stop-pos))))
4760 ;;
4761 ;; ;; Now the point is either at `stop-pos' or at some
4762 ;; ;; position further back if `stop-pos' was at a
4763 ;; ;; syntactically irrelevant place.
4764 ;;
4765 ;; ;; Skip additional syntactic ws so that we don't stop
4766 ;; ;; at the end of a comment if `skip-chars' is
4767 ;; ;; something like "^/".
4768 ;; (c-backward-syntactic-ws)
4769 ;;
4770 ;; (< (point) stop-pos))))
4771 ;;
4772 ;; ;; We might want to extend this with more useful return values
4773 ;; ;; in the future.
4774 ;; (/= (point) start))))
4775
4776 \f
4777 ;; Tools for handling comments and string literals.
4778
4779 (defun c-in-literal (&optional lim detect-cpp)
4780 "Return the type of literal point is in, if any.
4781 The return value is `c' if in a C-style comment, `c++' if in a C++
4782 style comment, `string' if in a string literal, `pound' if DETECT-CPP
4783 is non-nil and in a preprocessor line, or nil if somewhere else.
4784 Optional LIM is used as the backward limit of the search. If omitted,
4785 or nil, `c-beginning-of-defun' is used.
4786
4787 The last point calculated is cached if the cache is enabled, i.e. if
4788 `c-in-literal-cache' is bound to a two element vector.
4789
4790 Note that this function might do hidden buffer changes. See the
4791 comment at the start of cc-engine.el for more info."
4792 (save-restriction
4793 (widen)
4794 (let ((lit (c-state-semi-pp-to-literal (point))))
4795 (or (cadr lit)
4796 (and detect-cpp
4797 (save-excursion (c-beginning-of-macro))
4798 'pound)))))
4799
4800 (defun c-literal-limits (&optional lim near not-in-delimiter)
4801 "Return a cons of the beginning and end positions of the comment or
4802 string surrounding point (including both delimiters), or nil if point
4803 isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4804 to start parsing from. If NEAR is non-nil, then the limits of any
4805 literal next to point is returned. \"Next to\" means there's only
4806 spaces and tabs between point and the literal. The search for such a
4807 literal is done first in forward direction. If NOT-IN-DELIMITER is
4808 non-nil, the case when point is inside a starting delimiter won't be
4809 recognized. This only has effect for comments which have starting
4810 delimiters with more than one character.
4811
4812 Note that this function might do hidden buffer changes. See the
4813 comment at the start of cc-engine.el for more info."
4814
4815 (save-excursion
4816 (let*
4817 ((pos (point))
4818 (lit-limits
4819 (if lim
4820 (let ((s (parse-partial-sexp lim (point))))
4821 (when (or (nth 3 s) (nth 4 s))
4822 (cons (nth 8 s)
4823 (progn (parse-partial-sexp (point) (point-max)
4824 nil nil
4825 s
4826 'syntax-table)
4827 (point)))))
4828 (let ((pp-to-lit (c-state-full-pp-to-literal pos not-in-delimiter)))
4829 (car (cddr pp-to-lit))))))
4830 (cond
4831 (lit-limits)
4832
4833 (near
4834 (goto-char pos)
4835 ;; Search forward for a literal.
4836 (skip-chars-forward " \t")
4837 (cond
4838 ((looking-at c-string-limit-regexp) ; String.
4839 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4840 (point-max))))
4841
4842 ((looking-at c-comment-start-regexp) ; Line or block comment.
4843 (cons (point) (progn (c-forward-single-comment) (point))))
4844
4845 (t
4846 ;; Search backward.
4847 (skip-chars-backward " \t")
4848
4849 (let ((end (point)) beg)
4850 (cond
4851 ((save-excursion
4852 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4853 (setq beg (c-safe (c-backward-sexp 1) (point))))
4854
4855 ((and (c-safe (forward-char -2) t)
4856 (looking-at "*/"))
4857 ;; Block comment. Due to the nature of line
4858 ;; comments, they will always be covered by the
4859 ;; normal case above.
4860 (goto-char end)
4861 (c-backward-single-comment)
4862 ;; If LIM is bogus, beg will be bogus.
4863 (setq beg (point))))
4864
4865 (if beg (cons beg end))))))
4866 ))))
4867
4868 (defun c-literal-start (&optional safe-pos)
4869 "Return the start of the string or comment surrounding point, or nil if
4870 point isn't in one. SAFE-POS, if non-nil, is a position before point which is
4871 a known \"safe position\", i.e. outside of any string or comment."
4872 (if safe-pos
4873 (let ((s (parse-partial-sexp safe-pos (point))))
4874 (and (or (nth 3 s) (nth 4 s))
4875 (nth 8 s)))
4876 (car (cddr (c-state-semi-pp-to-literal (point))))))
4877
4878 ;; In case external callers use this; it did have a docstring.
4879 (defalias 'c-literal-limits-fast 'c-literal-limits)
4880
4881 (defun c-collect-line-comments (range)
4882 "If the argument is a cons of two buffer positions (such as returned by
4883 `c-literal-limits'), and that range contains a C++ style line comment,
4884 then an extended range is returned that contains all adjacent line
4885 comments (i.e. all comments that starts in the same column with no
4886 empty lines or non-whitespace characters between them). Otherwise the
4887 argument is returned.
4888
4889 Note that this function might do hidden buffer changes. See the
4890 comment at the start of cc-engine.el for more info."
4891
4892 (save-excursion
4893 (condition-case nil
4894 (if (and (consp range) (progn
4895 (goto-char (car range))
4896 (looking-at c-line-comment-starter)))
4897 (let ((col (current-column))
4898 (beg (point))
4899 (bopl (c-point 'bopl))
4900 (end (cdr range)))
4901 ;; Got to take care in the backward direction to handle
4902 ;; comments which are preceded by code.
4903 (while (and (c-backward-single-comment)
4904 (>= (point) bopl)
4905 (looking-at c-line-comment-starter)
4906 (= col (current-column)))
4907 (setq beg (point)
4908 bopl (c-point 'bopl)))
4909 (goto-char end)
4910 (while (and (progn (skip-chars-forward " \t")
4911 (looking-at c-line-comment-starter))
4912 (= col (current-column))
4913 (prog1 (zerop (forward-line 1))
4914 (setq end (point)))))
4915 (cons beg end))
4916 range)
4917 (error range))))
4918
4919 (defun c-literal-type (range)
4920 "Convenience function that given the result of `c-literal-limits',
4921 returns nil or the type of literal that the range surrounds, one
4922 of the symbols `c', `c++' or `string'. It's much faster than using
4923 `c-in-literal' and is intended to be used when you need both the
4924 type of a literal and its limits.
4925
4926 Note that this function might do hidden buffer changes. See the
4927 comment at the start of cc-engine.el for more info."
4928
4929 (if (consp range)
4930 (save-excursion
4931 (goto-char (car range))
4932 (cond ((looking-at c-string-limit-regexp) 'string)
4933 ((or (looking-at "//") ; c++ line comment
4934 (and (looking-at "\\s<") ; comment starter
4935 (looking-at "#"))) ; awk comment.
4936 'c++)
4937 (t 'c))) ; Assuming the range is valid.
4938 range))
4939
4940 (defsubst c-determine-limit-get-base (start try-size)
4941 ;; Get a "safe place" approximately TRY-SIZE characters before START.
4942 ;; This defsubst doesn't preserve point.
4943 (let* ((pos (max (- start try-size) (point-min)))
4944 (s (c-state-semi-pp-to-literal pos)))
4945 (or (car (cddr s)) pos)))
4946
4947 (defun c-determine-limit (how-far-back &optional start try-size)
4948 ;; Return a buffer position HOW-FAR-BACK non-literal characters from START
4949 ;; (default point). This is done by going back further in the buffer then
4950 ;; searching forward for literals. The position found won't be in a
4951 ;; literal. We start searching for the sought position TRY-SIZE (default
4952 ;; twice HOW-FAR-BACK) bytes back from START. This function must be fast.
4953 ;; :-)
4954 (save-excursion
4955 (let* ((start (or start (point)))
4956 (try-size (or try-size (* 2 how-far-back)))
4957 (base (c-determine-limit-get-base start try-size))
4958 (pos base)
4959
4960 (s (parse-partial-sexp pos pos)) ; null state.
4961 stack elt size
4962 (count 0))
4963 (while (< pos start)
4964 ;; Move forward one literal each time round this loop.
4965 ;; Move forward to the start of a comment or string.
4966 (setq s (parse-partial-sexp
4967 pos
4968 start
4969 nil ; target-depth
4970 nil ; stop-before
4971 s ; state
4972 'syntax-table)) ; stop-comment
4973
4974 ;; Gather details of the non-literal-bit - starting pos and size.
4975 (setq size (- (if (or (nth 4 s) (nth 3 s))
4976 (nth 8 s)
4977 (point))
4978 pos))
4979 (if (> size 0)
4980 (setq stack (cons (cons pos size) stack)))
4981
4982 ;; Move forward to the end of the comment/string.
4983 (if (or (nth 4 s) (nth 3 s))
4984 (setq s (parse-partial-sexp
4985 (point)
4986 start
4987 nil ; target-depth
4988 nil ; stop-before
4989 s ; state
4990 'syntax-table))) ; stop-comment
4991 (setq pos (point)))
4992
4993 ;; Now try and find enough non-literal characters recorded on the stack.
4994 ;; Go back one recorded literal each time round this loop.
4995 (while (and (< count how-far-back)
4996 stack)
4997 (setq elt (car stack)
4998 stack (cdr stack))
4999 (setq count (+ count (cdr elt))))
5000
5001 ;; Have we found enough yet?
5002 (cond
5003 ((>= count how-far-back)
5004 (+ (car elt) (- count how-far-back)))
5005 ((eq base (point-min))
5006 (point-min))
5007 (t
5008 (c-determine-limit (- how-far-back count) base try-size))))))
5009
5010 (defun c-determine-+ve-limit (how-far &optional start-pos)
5011 ;; Return a buffer position about HOW-FAR non-literal characters forward
5012 ;; from START-POS (default point), which must not be inside a literal.
5013 (save-excursion
5014 (let ((pos (or start-pos (point)))
5015 (count how-far)
5016 (s (parse-partial-sexp (point) (point)))) ; null state
5017 (while (and (not (eobp))
5018 (> count 0))
5019 ;; Scan over counted characters.
5020 (setq s (parse-partial-sexp
5021 pos
5022 (min (+ pos count) (point-max))
5023 nil ; target-depth
5024 nil ; stop-before
5025 s ; state
5026 'syntax-table)) ; stop-comment
5027 (setq count (- count (- (point) pos) 1)
5028 pos (point))
5029 ;; Scan over literal characters.
5030 (if (nth 8 s)
5031 (setq s (parse-partial-sexp
5032 pos
5033 (point-max)
5034 nil ; target-depth
5035 nil ; stop-before
5036 s ; state
5037 'syntax-table) ; stop-comment
5038 pos (point))))
5039 (point))))
5040
5041 \f
5042 ;; `c-find-decl-spots' and accompanying stuff.
5043
5044 ;; Variables used in `c-find-decl-spots' to cache the search done for
5045 ;; the first declaration in the last call. When that function starts,
5046 ;; it needs to back up over syntactic whitespace to look at the last
5047 ;; token before the region being searched. That can sometimes cause
5048 ;; moves back and forth over a quite large region of comments and
5049 ;; macros, which would be repeated for each changed character when
5050 ;; we're called during fontification, since font-lock refontifies the
5051 ;; current line for each change. Thus it's worthwhile to cache the
5052 ;; first match.
5053 ;;
5054 ;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
5055 ;; the syntactic whitespace less or equal to some start position.
5056 ;; There's no cached value if it's nil.
5057 ;;
5058 ;; `c-find-decl-match-pos' is the match position if
5059 ;; `c-find-decl-prefix-search' matched before the syntactic whitespace
5060 ;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
5061 (defvar c-find-decl-syntactic-pos nil)
5062 (make-variable-buffer-local 'c-find-decl-syntactic-pos)
5063 (defvar c-find-decl-match-pos nil)
5064 (make-variable-buffer-local 'c-find-decl-match-pos)
5065
5066 (defsubst c-invalidate-find-decl-cache (change-min-pos)
5067 (and c-find-decl-syntactic-pos
5068 (< change-min-pos c-find-decl-syntactic-pos)
5069 (setq c-find-decl-syntactic-pos nil)))
5070
5071 ; (defface c-debug-decl-spot-face
5072 ; '((t (:background "Turquoise")))
5073 ; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
5074 ; (defface c-debug-decl-sws-face
5075 ; '((t (:background "Khaki")))
5076 ; "Debug face to mark the syntactic whitespace between the declaration
5077 ; spots and the preceding token end.")
5078
5079 (defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
5080 (when (facep 'c-debug-decl-spot-face)
5081 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
5082 (c-debug-add-face (max match-pos (point-min)) decl-pos
5083 'c-debug-decl-sws-face)
5084 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
5085 'c-debug-decl-spot-face))))
5086 (defmacro c-debug-remove-decl-spot-faces (beg end)
5087 (when (facep 'c-debug-decl-spot-face)
5088 `(c-save-buffer-state ()
5089 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
5090 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
5091
5092 (defmacro c-find-decl-prefix-search ()
5093 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
5094 ;; but it contains lots of free variables that refer to things
5095 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
5096 ;; if there is a match, otherwise at `cfd-limit'.
5097 ;;
5098 ;; The macro moves point forward to the next putative start of a declaration
5099 ;; or cfd-limit. This decl start is the next token after a "declaration
5100 ;; prefix". The declaration prefix is the earlier of `cfd-prop-match' and
5101 ;; `cfd-re-match'. `cfd-match-pos' is set to the decl prefix.
5102 ;;
5103 ;; This macro might do hidden buffer changes.
5104
5105 '(progn
5106 ;; Find the next property match position if we haven't got one already.
5107 (unless cfd-prop-match
5108 (save-excursion
5109 (while (progn
5110 (goto-char (c-next-single-property-change
5111 (point) 'c-type nil cfd-limit))
5112 (and (< (point) cfd-limit)
5113 (not (eq (c-get-char-property (1- (point)) 'c-type)
5114 'c-decl-end)))))
5115 (setq cfd-prop-match (point))))
5116
5117 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
5118 ;; got one already.
5119 (unless cfd-re-match
5120
5121 (if (> cfd-re-match-end (point))
5122 (goto-char cfd-re-match-end))
5123
5124 ;; Each time round, the next `while' moves forward over a pseudo match
5125 ;; of `c-decl-prefix-or-start-re' which is either inside a literal, or
5126 ;; is a ":" not preceded by "public", etc.. `cfd-re-match' and
5127 ;; `cfd-re-match-end' get set.
5128 (while
5129 (progn
5130 (setq cfd-re-match-end (re-search-forward c-decl-prefix-or-start-re
5131 cfd-limit 'move))
5132 (cond
5133 ((null cfd-re-match-end)
5134 ;; No match. Finish up and exit the loop.
5135 (setq cfd-re-match cfd-limit)
5136 nil)
5137 ((c-got-face-at
5138 (if (setq cfd-re-match (match-end 1))
5139 ;; Matched the end of a token preceding a decl spot.
5140 (progn
5141 (goto-char cfd-re-match)
5142 (1- cfd-re-match))
5143 ;; Matched a token that start a decl spot.
5144 (goto-char (match-beginning 0))
5145 (point))
5146 c-literal-faces)
5147 ;; Pseudo match inside a comment or string literal. Skip out
5148 ;; of comments and string literals.
5149 (while (progn
5150 (goto-char (c-next-single-property-change
5151 (point) 'face nil cfd-limit))
5152 (and (< (point) cfd-limit)
5153 (c-got-face-at (point) c-literal-faces))))
5154 t) ; Continue the loop over pseudo matches.
5155 ((and (match-string 1)
5156 (string= (match-string 1) ":")
5157 (save-excursion
5158 (or (/= (c-backward-token-2 2) 0) ; no search limit. :-(
5159 (not (looking-at c-decl-start-colon-kwd-re)))))
5160 ;; Found a ":" which isn't part of "public:", etc.
5161 t)
5162 (t nil)))) ;; Found a real match. Exit the pseudo-match loop.
5163
5164 ;; If our match was at the decl start, we have to back up over the
5165 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
5166 ;; any decl spots in the syntactic ws.
5167 (unless cfd-re-match
5168 (c-backward-syntactic-ws)
5169 (setq cfd-re-match (point))))
5170
5171 ;; Choose whichever match is closer to the start.
5172 (if (< cfd-re-match cfd-prop-match)
5173 (setq cfd-match-pos cfd-re-match
5174 cfd-re-match nil)
5175 (setq cfd-match-pos cfd-prop-match
5176 cfd-prop-match nil))
5177
5178 (goto-char cfd-match-pos)
5179
5180 (when (< cfd-match-pos cfd-limit)
5181 ;; Skip forward past comments only so we don't skip macros.
5182 (c-forward-comments)
5183 ;; Set the position to continue at. We can avoid going over
5184 ;; the comments skipped above a second time, but it's possible
5185 ;; that the comment skipping has taken us past `cfd-prop-match'
5186 ;; since the property might be used inside comments.
5187 (setq cfd-continue-pos (if cfd-prop-match
5188 (min cfd-prop-match (point))
5189 (point))))))
5190
5191 (defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
5192 ;; Call CFD-FUN for each possible spot for a declaration, cast or
5193 ;; label from the point to CFD-LIMIT.
5194 ;;
5195 ;; CFD-FUN is called with point at the start of the spot. It's passed two
5196 ;; arguments: The first is the end position of the token preceding the spot,
5197 ;; or 0 for the implicit match at bob. The second is a flag that is t when
5198 ;; the match is inside a macro. Point should be moved forward by at least
5199 ;; one token.
5200 ;;
5201 ;; If CFD-FUN adds `c-decl-end' properties somewhere below the current spot,
5202 ;; it should return non-nil to ensure that the next search will find them.
5203 ;;
5204 ;; Such a spot is:
5205 ;; o The first token after bob.
5206 ;; o The first token after the end of submatch 1 in
5207 ;; `c-decl-prefix-or-start-re' when that submatch matches. This
5208 ;; submatch is typically a (L or R) brace or paren, a ;, or a ,.
5209 ;; o The start of each `c-decl-prefix-or-start-re' match when
5210 ;; submatch 1 doesn't match. This is, for example, the keyword
5211 ;; "class" in Pike.
5212 ;; o The start of a previously recognized declaration; "recognized"
5213 ;; means that the last char of the previous token has a `c-type'
5214 ;; text property with the value `c-decl-end'; this only holds
5215 ;; when `c-type-decl-end-used' is set.
5216 ;;
5217 ;; Only a spot that match CFD-DECL-RE and whose face is in the
5218 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
5219 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
5220 ;;
5221 ;; If the match is inside a macro then the buffer is narrowed to the
5222 ;; end of it, so that CFD-FUN can investigate the following tokens
5223 ;; without matching something that begins inside a macro and ends
5224 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
5225 ;; CFD-FACE-CHECKLIST checks exist.
5226 ;;
5227 ;; The spots are visited approximately in order from top to bottom.
5228 ;; It's however the positions where `c-decl-prefix-or-start-re'
5229 ;; matches and where `c-decl-end' properties are found that are in
5230 ;; order. Since the spots often are at the following token, they
5231 ;; might be visited out of order insofar as more spots are reported
5232 ;; later on within the syntactic whitespace between the match
5233 ;; positions and their spots.
5234 ;;
5235 ;; It's assumed that comments and strings are fontified in the
5236 ;; searched range.
5237 ;;
5238 ;; This is mainly used in fontification, and so has an elaborate
5239 ;; cache to handle repeated calls from the same start position; see
5240 ;; the variables above.
5241 ;;
5242 ;; All variables in this function begin with `cfd-' to avoid name
5243 ;; collision with the (dynamically bound) variables used in CFD-FUN.
5244 ;;
5245 ;; This function might do hidden buffer changes.
5246
5247 (let ((cfd-start-pos (point)) ; never changed
5248 (cfd-buffer-end (point-max))
5249 ;; The end of the token preceding the decl spot last found
5250 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
5251 ;; no match.
5252 cfd-re-match
5253 ;; The end position of the last `c-decl-prefix-or-start-re'
5254 ;; match. If this is greater than `cfd-continue-pos', the
5255 ;; next regexp search is started here instead.
5256 (cfd-re-match-end (point-min))
5257 ;; The end of the last `c-decl-end' found by
5258 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
5259 ;; match. If searching for the property isn't needed then we
5260 ;; disable it by setting it to `cfd-limit' directly.
5261 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
5262 ;; The end of the token preceding the decl spot last found by
5263 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
5264 ;; bob. `cfd-limit' if there's no match. In other words,
5265 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
5266 (cfd-match-pos cfd-limit)
5267 ;; The position to continue searching at.
5268 cfd-continue-pos
5269 ;; The position of the last "real" token we've stopped at.
5270 ;; This can be greater than `cfd-continue-pos' when we get
5271 ;; hits inside macros or at `c-decl-end' positions inside
5272 ;; comments.
5273 (cfd-token-pos 0)
5274 ;; The end position of the last entered macro.
5275 (cfd-macro-end 0))
5276
5277 ;; Initialize by finding a syntactically relevant start position
5278 ;; before the point, and do the first `c-decl-prefix-or-start-re'
5279 ;; search unless we're at bob.
5280
5281 (let (start-in-literal start-in-macro syntactic-pos)
5282 ;; Must back up a bit since we look for the end of the previous
5283 ;; statement or declaration, which is earlier than the first
5284 ;; returned match.
5285
5286 ;; This `cond' moves back over any literals or macros. It has special
5287 ;; handling for when the region being searched is entirely within a
5288 ;; macro. It sets `cfd-continue-pos' (unless we've reached
5289 ;; `cfd-limit').
5290 (cond
5291 ;; First we need to move to a syntactically relevant position.
5292 ;; Begin by backing out of comment or string literals.
5293 ;;
5294 ;; This arm of the cond actually triggers if we're in a literal,
5295 ;; and cfd-limit is at most at BONL.
5296 ((and
5297 ;; This arm of the `and' moves backwards out of a literal when
5298 ;; the face at point is a literal face. In this case, its value
5299 ;; is always non-nil.
5300 (when (c-got-face-at (point) c-literal-faces)
5301 ;; Try to use the faces to back up to the start of the
5302 ;; literal. FIXME: What if the point is on a declaration
5303 ;; inside a comment?
5304 (while (and (not (bobp))
5305 (c-got-face-at (1- (point)) c-literal-faces))
5306 (goto-char (previous-single-property-change
5307 (point) 'face nil (point-min))))
5308
5309 ;; XEmacs doesn't fontify the quotes surrounding string
5310 ;; literals.
5311 (and (featurep 'xemacs)
5312 (eq (get-text-property (point) 'face)
5313 'font-lock-string-face)
5314 (not (bobp))
5315 (progn (backward-char)
5316 (not (looking-at c-string-limit-regexp)))
5317 (forward-char))
5318
5319 ;; Don't trust the literal to contain only literal faces
5320 ;; (the font lock package might not have fontified the
5321 ;; start of it at all, for instance) so check that we have
5322 ;; arrived at something that looks like a start or else
5323 ;; resort to `c-literal-limits'.
5324 (unless (looking-at c-literal-start-regexp)
5325 (let ((lit-start (c-literal-start)))
5326 (if lit-start (goto-char lit-start)))
5327 )
5328
5329 (setq start-in-literal (point))) ; end of `and' arm.
5330
5331 ;; The start is in a literal. If the limit is in the same
5332 ;; one we don't have to find a syntactic position etc. We
5333 ;; only check that if the limit is at or before bonl to save
5334 ;; time; it covers the by far most common case when font-lock
5335 ;; refontifies the current line only.
5336 (<= cfd-limit (c-point 'bonl cfd-start-pos))
5337 (save-excursion
5338 (goto-char cfd-start-pos)
5339 (while (progn
5340 (goto-char (c-next-single-property-change
5341 (point) 'face nil cfd-limit))
5342 (and (< (point) cfd-limit)
5343 (c-got-face-at (point) c-literal-faces))))
5344 (= (point) cfd-limit))) ; end of `cond' arm condition
5345
5346 ;; Completely inside a literal. Set up variables to trig the
5347 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
5348 ;; find a suitable start position.
5349 (setq cfd-continue-pos start-in-literal)) ; end of `cond' arm
5350
5351 ;; Check if the region might be completely inside a macro, to
5352 ;; optimize that like the completely-inside-literal above.
5353 ((save-excursion
5354 (and (= (forward-line 1) 0)
5355 (bolp) ; forward-line has funny behavior at eob.
5356 (>= (point) cfd-limit)
5357 (progn (backward-char)
5358 (eq (char-before) ?\\))))
5359 ;; (Maybe) completely inside a macro. Only need to trig the
5360 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
5361 ;; set things up.
5362 (setq cfd-continue-pos (1- cfd-start-pos)
5363 start-in-macro t))
5364
5365 ;; The default arm of the `cond' moves back over any macro we're in
5366 ;; and over any syntactic WS. It sets `c-find-decl-syntactic-pos'.
5367 (t
5368 ;; Back out of any macro so we don't miss any declaration
5369 ;; that could follow after it.
5370 (when (c-beginning-of-macro)
5371 (setq start-in-macro t))
5372
5373 ;; Now we're at a proper syntactically relevant position so we
5374 ;; can use the cache. But first clear it if it applied
5375 ;; further down.
5376 (c-invalidate-find-decl-cache cfd-start-pos)
5377
5378 (setq syntactic-pos (point))
5379 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
5380 ;; Don't have to do this if the cache is relevant here,
5381 ;; typically if the same line is refontified again. If
5382 ;; we're just some syntactic whitespace further down we can
5383 ;; still use the cache to limit the skipping.
5384 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
5385
5386 ;; If we hit `c-find-decl-syntactic-pos' and
5387 ;; `c-find-decl-match-pos' is set then we install the cached
5388 ;; values. If we hit `c-find-decl-syntactic-pos' and
5389 ;; `c-find-decl-match-pos' is nil then we know there's no decl
5390 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
5391 ;; and so we can continue the search from this point. If we
5392 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
5393 ;; the right spot to begin searching anyway.
5394 (if (and (eq (point) c-find-decl-syntactic-pos)
5395 c-find-decl-match-pos)
5396 (setq cfd-match-pos c-find-decl-match-pos
5397 cfd-continue-pos syntactic-pos)
5398
5399 (setq c-find-decl-syntactic-pos syntactic-pos)
5400
5401 (when (if (bobp)
5402 ;; Always consider bob a match to get the first
5403 ;; declaration in the file. Do this separately instead of
5404 ;; letting `c-decl-prefix-or-start-re' match bob, so that
5405 ;; regexp always can consume at least one character to
5406 ;; ensure that we won't get stuck in an infinite loop.
5407 (setq cfd-re-match 0)
5408 (backward-char)
5409 (c-beginning-of-current-token)
5410 (< (point) cfd-limit))
5411 ;; Do an initial search now. In the bob case above it's
5412 ;; only done to search for a `c-decl-end' spot.
5413 (c-find-decl-prefix-search)) ; sets cfd-continue-pos
5414
5415 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
5416 cfd-match-pos))))) ; end of `cond'
5417
5418 ;; Advance `cfd-continue-pos' if it's before the start position.
5419 ;; The closest continue position that might have effect at or
5420 ;; after the start depends on what we started in. This also
5421 ;; finds a suitable start position in the special cases when the
5422 ;; region is completely within a literal or macro.
5423 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
5424
5425 (cond
5426 (start-in-macro
5427 ;; If we're in a macro then it's the closest preceding token
5428 ;; in the macro. Check this before `start-in-literal',
5429 ;; since if we're inside a literal in a macro, the preceding
5430 ;; token is earlier than any `c-decl-end' spot inside the
5431 ;; literal (comment).
5432 (goto-char (or start-in-literal cfd-start-pos))
5433 ;; The only syntactic ws in macros are comments.
5434 (c-backward-comments)
5435 (backward-char)
5436 (c-beginning-of-current-token))
5437
5438 (start-in-literal
5439 ;; If we're in a comment it can only be the closest
5440 ;; preceding `c-decl-end' position within that comment, if
5441 ;; any. Go back to the beginning of such a property so that
5442 ;; `c-find-decl-prefix-search' will find the end of it.
5443 ;; (Can't stop at the end and install it directly on
5444 ;; `cfd-prop-match' since that variable might be cleared
5445 ;; after `cfd-fun' below.)
5446 ;;
5447 ;; Note that if the literal is a string then the property
5448 ;; search will simply skip to the beginning of it right
5449 ;; away.
5450 (if (not c-type-decl-end-used)
5451 (goto-char start-in-literal)
5452 (goto-char cfd-start-pos)
5453 (while (progn
5454 (goto-char (previous-single-property-change
5455 (point) 'c-type nil start-in-literal))
5456 (and (> (point) start-in-literal)
5457 (not (eq (c-get-char-property (point) 'c-type)
5458 'c-decl-end))))))
5459
5460 (when (= (point) start-in-literal)
5461 ;; Didn't find any property inside the comment, so we can
5462 ;; skip it entirely. (This won't skip past a string, but
5463 ;; that'll be handled quickly by the next
5464 ;; `c-find-decl-prefix-search' anyway.)
5465 (c-forward-single-comment)
5466 (if (> (point) cfd-limit)
5467 (goto-char cfd-limit))))
5468
5469 (t
5470 ;; If we started in normal code, the only match that might
5471 ;; apply before the start is what we already got in
5472 ;; `cfd-match-pos' so we can continue at the start position.
5473 ;; (Note that we don't get here if the first match is below
5474 ;; it.)
5475 (goto-char cfd-start-pos))) ; end of `cond'
5476
5477 ;; Delete found matches if they are before our new continue
5478 ;; position, so that `c-find-decl-prefix-search' won't back up
5479 ;; to them later on.
5480 (setq cfd-continue-pos (point))
5481 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
5482 (setq cfd-re-match nil))
5483 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
5484 (setq cfd-prop-match nil))) ; end of `when'
5485
5486 (if syntactic-pos
5487 ;; This is the normal case and we got a proper syntactic
5488 ;; position. If there's a match then it's always outside
5489 ;; macros and comments, so advance to the next token and set
5490 ;; `cfd-token-pos'. The loop below will later go back using
5491 ;; `cfd-continue-pos' to fix declarations inside the
5492 ;; syntactic ws.
5493 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
5494 (goto-char syntactic-pos)
5495 (c-forward-syntactic-ws)
5496 (and cfd-continue-pos
5497 (< cfd-continue-pos (point))
5498 (setq cfd-token-pos (point))))
5499
5500 ;; Have one of the special cases when the region is completely
5501 ;; within a literal or macro. `cfd-continue-pos' is set to a
5502 ;; good start position for the search, so do it.
5503 (c-find-decl-prefix-search)))
5504
5505 ;; Now loop, one decl spot per iteration. We already have the first
5506 ;; match in `cfd-match-pos'.
5507 (while (progn
5508 ;; Go forward over "false matches", one per iteration.
5509 (while (and
5510 (< cfd-match-pos cfd-limit)
5511
5512 (or
5513 ;; Kludge to filter out matches on the "<" that
5514 ;; aren't open parens, for the sake of languages
5515 ;; that got `c-recognize-<>-arglists' set.
5516 (and (eq (char-before cfd-match-pos) ?<)
5517 (not (c-get-char-property (1- cfd-match-pos)
5518 'syntax-table)))
5519
5520 ;; If `cfd-continue-pos' is less or equal to
5521 ;; `cfd-token-pos', we've got a hit inside a macro
5522 ;; that's in the syntactic whitespace before the last
5523 ;; "real" declaration we've checked. If they're equal
5524 ;; we've arrived at the declaration a second time, so
5525 ;; there's nothing to do.
5526 (= cfd-continue-pos cfd-token-pos)
5527
5528 (progn
5529 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
5530 ;; we're still searching for declarations embedded in
5531 ;; the syntactic whitespace. In that case we need
5532 ;; only to skip comments and not macros, since they
5533 ;; can't be nested, and that's already been done in
5534 ;; `c-find-decl-prefix-search'.
5535 (when (> cfd-continue-pos cfd-token-pos)
5536 (c-forward-syntactic-ws)
5537 (setq cfd-token-pos (point)))
5538
5539 ;; Continue if the following token fails the
5540 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
5541 (when (or (>= (point) cfd-limit)
5542 (not (looking-at cfd-decl-re))
5543 (and cfd-face-checklist
5544 (not (c-got-face-at
5545 (point) cfd-face-checklist))))
5546 (goto-char cfd-continue-pos)
5547 t)))
5548
5549 (< (point) cfd-limit)) ; end of "false matches" condition
5550 (c-find-decl-prefix-search)) ; end of "false matches" loop
5551
5552 (< (point) cfd-limit)) ; end of condition for "decl-spot" while
5553
5554 (when (and
5555 (>= (point) cfd-start-pos)
5556
5557 (progn
5558 ;; Narrow to the end of the macro if we got a hit inside
5559 ;; one, to avoid recognizing things that start inside the
5560 ;; macro and end outside it.
5561 (when (> cfd-match-pos cfd-macro-end)
5562 ;; Not in the same macro as in the previous round.
5563 (save-excursion
5564 (goto-char cfd-match-pos)
5565 (setq cfd-macro-end
5566 (if (save-excursion (and (c-beginning-of-macro)
5567 (< (point) cfd-match-pos)))
5568 (progn (c-end-of-macro)
5569 (point))
5570 0))))
5571
5572 (if (zerop cfd-macro-end)
5573 t
5574 (if (> cfd-macro-end (point))
5575 (progn (narrow-to-region (point-min) cfd-macro-end)
5576 t)
5577 ;; The matched token was the last thing in the macro,
5578 ;; so the whole match is bogus.
5579 (setq cfd-macro-end 0)
5580 nil)))) ; end of when condition
5581
5582 (c-debug-put-decl-spot-faces cfd-match-pos (point))
5583 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
5584 (setq cfd-prop-match nil))
5585
5586 (when (/= cfd-macro-end 0)
5587 ;; Restore limits if we did macro narrowing above.
5588 (narrow-to-region (point-min) cfd-buffer-end)))
5589
5590 (goto-char cfd-continue-pos)
5591 (if (= cfd-continue-pos cfd-limit)
5592 (setq cfd-match-pos cfd-limit)
5593 (c-find-decl-prefix-search))))) ; Moves point, sets cfd-continue-pos,
5594 ; cfd-match-pos, etc.
5595
5596 \f
5597 ;; A cache for found types.
5598
5599 ;; Buffer local variable that contains an obarray with the types we've
5600 ;; found. If a declaration is recognized somewhere we record the
5601 ;; fully qualified identifier in it to recognize it as a type
5602 ;; elsewhere in the file too. This is not accurate since we do not
5603 ;; bother with the scoping rules of the languages, but in practice the
5604 ;; same name is seldom used as both a type and something else in a
5605 ;; file, and we only use this as a last resort in ambiguous cases (see
5606 ;; `c-forward-decl-or-cast-1').
5607 ;;
5608 ;; Not every type need be in this cache. However, things which have
5609 ;; ceased to be types must be removed from it.
5610 ;;
5611 ;; Template types in C++ are added here too but with the template
5612 ;; arglist replaced with "<>" in references or "<" for the one in the
5613 ;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
5614 ;; "Foo<>::Bar<". This avoids storing very long strings (since C++
5615 ;; template specs can be fairly sized programs in themselves) and
5616 ;; improves the hit ratio (it's a type regardless of the template
5617 ;; args; it's just not the same type, but we're only interested in
5618 ;; recognizing types, not telling distinct types apart). Note that
5619 ;; template types in references are added here too; from the example
5620 ;; above there will also be an entry "Foo<".
5621 (defvar c-found-types nil)
5622 (make-variable-buffer-local 'c-found-types)
5623
5624 (defsubst c-clear-found-types ()
5625 ;; Clears `c-found-types'.
5626 (setq c-found-types (make-vector 53 0)))
5627
5628 (defun c-add-type (from to)
5629 ;; Add the given region as a type in `c-found-types'. If the region
5630 ;; doesn't match an existing type but there is a type which is equal
5631 ;; to the given one except that the last character is missing, then
5632 ;; the shorter type is removed. That's done to avoid adding all
5633 ;; prefixes of a type as it's being entered and font locked. This
5634 ;; doesn't cover cases like when characters are removed from a type
5635 ;; or added in the middle. We'd need the position of point when the
5636 ;; font locking is invoked to solve this well.
5637 ;;
5638 ;; This function might do hidden buffer changes.
5639 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
5640 (unless (intern-soft type c-found-types)
5641 (unintern (substring type 0 -1) c-found-types)
5642 (intern type c-found-types))))
5643
5644 (defun c-unfind-type (name)
5645 ;; Remove the "NAME" from c-found-types, if present.
5646 (unintern name c-found-types))
5647
5648 (defsubst c-check-type (from to)
5649 ;; Return non-nil if the given region contains a type in
5650 ;; `c-found-types'.
5651 ;;
5652 ;; This function might do hidden buffer changes.
5653 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
5654 c-found-types))
5655
5656 (defun c-list-found-types ()
5657 ;; Return all the types in `c-found-types' as a sorted list of
5658 ;; strings.
5659 (let (type-list)
5660 (mapatoms (lambda (type)
5661 (setq type-list (cons (symbol-name type)
5662 type-list)))
5663 c-found-types)
5664 (sort type-list 'string-lessp)))
5665
5666 ;; Shut up the byte compiler.
5667 (defvar c-maybe-stale-found-type)
5668
5669 (defun c-trim-found-types (beg end old-len)
5670 ;; An after change function which, in conjunction with the info in
5671 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
5672 ;; from `c-found-types', should this type have become stale. For
5673 ;; example, this happens to "foo" when "foo \n bar();" becomes
5674 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
5675 ;; the fontification.
5676 ;;
5677 ;; Have we, perhaps, added non-ws characters to the front/back of a found
5678 ;; type?
5679 (when (> end beg)
5680 (save-excursion
5681 (when (< end (point-max))
5682 (goto-char end)
5683 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5684 (progn (goto-char end)
5685 (c-end-of-current-token)))
5686 (c-unfind-type (buffer-substring-no-properties
5687 end (point)))))
5688 (when (> beg (point-min))
5689 (goto-char beg)
5690 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5691 (progn (goto-char beg)
5692 (c-beginning-of-current-token)))
5693 (c-unfind-type (buffer-substring-no-properties
5694 (point) beg))))))
5695
5696 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5697 (cond
5698 ;; Changing the amount of (already existing) whitespace - don't do anything.
5699 ((and (c-partial-ws-p beg end)
5700 (or (= beg end) ; removal of WS
5701 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5702
5703 ;; The syntactic relationship which defined a "found type" has been
5704 ;; destroyed.
5705 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5706 (c-unfind-type (cadr c-maybe-stale-found-type)))
5707 ;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5708 )))
5709
5710 \f
5711 ;; Setting and removing syntax properties on < and > in languages (C++
5712 ;; and Java) where they can be template/generic delimiters as well as
5713 ;; their normal meaning of "less/greater than".
5714
5715 ;; Normally, < and > have syntax 'punctuation'. When they are found to
5716 ;; be delimiters, they are marked as such with the category properties
5717 ;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5718
5719 ;; STRATEGY:
5720 ;;
5721 ;; It is impossible to determine with certainty whether a <..> pair in
5722 ;; C++ is two comparison operators or is template delimiters, unless
5723 ;; one duplicates a lot of a C++ compiler. For example, the following
5724 ;; code fragment:
5725 ;;
5726 ;; foo (a < b, c > d) ;
5727 ;;
5728 ;; could be a function call with two integer parameters (each a
5729 ;; relational expression), or it could be a constructor for class foo
5730 ;; taking one parameter d of templated type "a < b, c >". They are
5731 ;; somewhat easier to distinguish in Java.
5732 ;;
5733 ;; The strategy now (2010-01) adopted is to mark and unmark < and
5734 ;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5735 ;; individually when their context so indicated. This gave rise to
5736 ;; intractable problems when one of a matching pair was deleted, or
5737 ;; pulled into a literal.]
5738 ;;
5739 ;; At each buffer change, the syntax-table properties are removed in a
5740 ;; before-change function and reapplied, when needed, in an
5741 ;; after-change function. It is far more important that the
5742 ;; properties get removed when they they are spurious than that they
5743 ;; be present when wanted.
5744 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5745 (defun c-clear-<-pair-props (&optional pos)
5746 ;; POS (default point) is at a < character. If it is marked with
5747 ;; open paren syntax-table text property, remove the property,
5748 ;; together with the close paren property on the matching > (if
5749 ;; any).
5750 (save-excursion
5751 (if pos
5752 (goto-char pos)
5753 (setq pos (point)))
5754 (when (equal (c-get-char-property (point) 'syntax-table)
5755 c-<-as-paren-syntax)
5756 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5757 (c-go-list-forward))
5758 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5759 c->-as-paren-syntax) ; should always be true.
5760 (c-unmark-<->-as-paren (1- (point))))
5761 (c-unmark-<->-as-paren pos))))
5762
5763 (defun c-clear->-pair-props (&optional pos)
5764 ;; POS (default point) is at a > character. If it is marked with
5765 ;; close paren syntax-table property, remove the property, together
5766 ;; with the open paren property on the matching < (if any).
5767 (save-excursion
5768 (if pos
5769 (goto-char pos)
5770 (setq pos (point)))
5771 (when (equal (c-get-char-property (point) 'syntax-table)
5772 c->-as-paren-syntax)
5773 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5774 (c-go-up-list-backward))
5775 (when (equal (c-get-char-property (point) 'syntax-table)
5776 c-<-as-paren-syntax) ; should always be true.
5777 (c-unmark-<->-as-paren (point)))
5778 (c-unmark-<->-as-paren pos))))
5779
5780 (defun c-clear-<>-pair-props (&optional pos)
5781 ;; POS (default point) is at a < or > character. If it has an
5782 ;; open/close paren syntax-table property, remove this property both
5783 ;; from the current character and its partner (which will also be
5784 ;; thusly marked).
5785 (cond
5786 ((eq (char-after) ?\<)
5787 (c-clear-<-pair-props pos))
5788 ((eq (char-after) ?\>)
5789 (c-clear->-pair-props pos))
5790 (t (c-benign-error
5791 "c-clear-<>-pair-props called from wrong position"))))
5792
5793 (defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5794 ;; POS (default point) is at a < character. If it is both marked
5795 ;; with open/close paren syntax-table property, and has a matching >
5796 ;; (also marked) which is after LIM, remove the property both from
5797 ;; the current > and its partner. Return t when this happens, nil
5798 ;; when it doesn't.
5799 (save-excursion
5800 (if pos
5801 (goto-char pos)
5802 (setq pos (point)))
5803 (when (equal (c-get-char-property (point) 'syntax-table)
5804 c-<-as-paren-syntax)
5805 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5806 (c-go-list-forward))
5807 (when (and (>= (point) lim)
5808 (equal (c-get-char-property (1- (point)) 'syntax-table)
5809 c->-as-paren-syntax)) ; should always be true.
5810 (c-unmark-<->-as-paren (1- (point)))
5811 (c-unmark-<->-as-paren pos))
5812 t)))
5813
5814 (defun c-clear->-pair-props-if-match-before (lim &optional pos)
5815 ;; POS (default point) is at a > character. If it is both marked
5816 ;; with open/close paren syntax-table property, and has a matching <
5817 ;; (also marked) which is before LIM, remove the property both from
5818 ;; the current < and its partner. Return t when this happens, nil
5819 ;; when it doesn't.
5820 (save-excursion
5821 (if pos
5822 (goto-char pos)
5823 (setq pos (point)))
5824 (when (equal (c-get-char-property (point) 'syntax-table)
5825 c->-as-paren-syntax)
5826 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5827 (c-go-up-list-backward))
5828 (when (and (<= (point) lim)
5829 (equal (c-get-char-property (point) 'syntax-table)
5830 c-<-as-paren-syntax)) ; should always be true.
5831 (c-unmark-<->-as-paren (point))
5832 (c-unmark-<->-as-paren pos))
5833 t)))
5834
5835 ;; Set by c-common-init in cc-mode.el.
5836 (defvar c-new-BEG)
5837 (defvar c-new-END)
5838 ;; Set by c-after-change in cc-mode.el.
5839 (defvar c-old-BEG)
5840 (defvar c-old-END)
5841
5842 (defun c-before-change-check-<>-operators (beg end)
5843 ;; Unmark certain pairs of "< .... >" which are currently marked as
5844 ;; template/generic delimiters. (This marking is via syntax-table text
5845 ;; properties), and expand the (c-new-BEG c-new-END) region to include all
5846 ;; unmarked < and > operators within the certain bounds (see below).
5847 ;;
5848 ;; These pairs are those which are in the current "statement" (i.e.,
5849 ;; the region between the {, }, or ; before BEG and the one after
5850 ;; END), and which enclose any part of the interval (BEG END).
5851 ;;
5852 ;; Note that in C++ (?and Java), template/generic parens cannot
5853 ;; enclose a brace or semicolon, so we use these as bounds on the
5854 ;; region we must work on.
5855 ;;
5856 ;; This function is called from before-change-functions (via
5857 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5858 ;; and point is undefined, both at entry and exit.
5859 ;;
5860 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5861 ;; 2010-01-29.
5862 (save-excursion
5863 (c-save-buffer-state
5864 ((beg-lit-start (progn (goto-char beg) (c-literal-start)))
5865 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5866 new-beg new-end beg-limit end-limit)
5867 ;; Locate the earliest < after the barrier before the changed region,
5868 ;; which isn't already marked as a paren.
5869 (goto-char (or beg-lit-start beg))
5870 (setq beg-limit (c-determine-limit 512))
5871
5872 ;; Remove the syntax-table/category properties from each pertinent <...>
5873 ;; pair. Firstly, the ones with the < before beg and > after beg....
5874 (while (progn (c-syntactic-skip-backward "^;{}<" beg-limit)
5875 (eq (char-before) ?<))
5876 (c-backward-token-2)
5877 (when (eq (char-after) ?<)
5878 (c-clear-<-pair-props-if-match-after beg)))
5879 (c-forward-syntactic-ws)
5880 (setq new-beg (point))
5881
5882 ;; ...Then the ones with < before end and > after end.
5883 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5884 (setq end-limit (c-determine-+ve-limit 512))
5885 (while (and (c-syntactic-re-search-forward "[;{}>]" end-limit 'end)
5886 (eq (char-before) ?>))
5887 (c-end-of-current-token)
5888 (when (eq (char-before) ?>)
5889 (c-clear->-pair-props-if-match-before end (1- (point)))))
5890 (c-backward-syntactic-ws)
5891 (setq new-end (point))
5892
5893 ;; Extend the fontification region, if needed.
5894 (and new-beg
5895 (< new-beg c-new-BEG)
5896 (setq c-new-BEG new-beg))
5897 (and new-end
5898 (> new-end c-new-END)
5899 (setq c-new-END new-end)))))
5900
5901 (defun c-after-change-check-<>-operators (beg end)
5902 ;; This is called from `after-change-functions' when
5903 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5904 ;; chars with paren syntax become part of another operator like "<<"
5905 ;; or ">=".
5906 ;;
5907 ;; This function might do hidden buffer changes.
5908
5909 (save-excursion
5910 (goto-char beg)
5911 (when (or (looking-at "[<>]")
5912 (< (skip-chars-backward "<>") 0))
5913
5914 (goto-char beg)
5915 (c-beginning-of-current-token)
5916 (when (and (< (point) beg)
5917 (looking-at c-<>-multichar-token-regexp)
5918 (< beg (setq beg (match-end 0))))
5919 (while (progn (skip-chars-forward "^<>" beg)
5920 (< (point) beg))
5921 (c-clear-<>-pair-props)
5922 (forward-char))))
5923
5924 (when (< beg end)
5925 (goto-char end)
5926 (when (or (looking-at "[<>]")
5927 (< (skip-chars-backward "<>") 0))
5928
5929 (goto-char end)
5930 (c-beginning-of-current-token)
5931 (when (and (< (point) end)
5932 (looking-at c-<>-multichar-token-regexp)
5933 (< end (setq end (match-end 0))))
5934 (while (progn (skip-chars-forward "^<>" end)
5935 (< (point) end))
5936 (c-clear-<>-pair-props)
5937 (forward-char)))))))
5938
5939 (defun c-restore-<>-properties (_beg _end _old-len)
5940 ;; This function is called as an after-change function. It restores the
5941 ;; category/syntax-table properties on template/generic <..> pairs between
5942 ;; c-new-BEG and c-new-END. It may do hidden buffer changes.
5943 (c-save-buffer-state ((c-parse-and-markup-<>-arglists t)
5944 c-restricted-<>-arglists lit-limits)
5945 (goto-char c-new-BEG)
5946 (if (setq lit-limits (c-literal-limits))
5947 (goto-char (cdr lit-limits)))
5948 (while (and (< (point) c-new-END)
5949 (c-syntactic-re-search-forward "<" c-new-END 'bound))
5950 (backward-char)
5951 (save-excursion
5952 (c-backward-token-2)
5953 (setq c-restricted-<>-arglists
5954 (and (not (looking-at c-opt-<>-sexp-key))
5955 (progn (c-backward-syntactic-ws) ; to ( or ,
5956 (and (memq (char-before) '(?\( ?,)) ; what about <?
5957 (not (eq (c-get-char-property (point) 'c-type)
5958 'c-decl-arg-start)))))))
5959 (or (c-forward-<>-arglist nil)
5960 (forward-char)))))
5961
5962 \f
5963 ;; Functions to handle C++ raw strings.
5964 ;;
5965 ;; A valid C++ raw string looks like
5966 ;; R"<id>(<contents>)<id>"
5967 ;; , where <id> is an identifier from 0 to 16 characters long, not containing
5968 ;; spaces, control characters, double quote or left/right paren. <contents>
5969 ;; can include anything which isn't the terminating )<id>", including new
5970 ;; lines, "s, parentheses, etc.
5971 ;;
5972 ;; CC Mode handles C++ raw strings by the use of `syntax-table' text
5973 ;; properties as follows:
5974 ;;
5975 ;; (i) On a validly terminated raw string, no `syntax-table' text properties
5976 ;; are applied to the opening and closing delimiters, but any " in the
5977 ;; contents is given the property value "punctuation" (`(1)') to prevent it
5978 ;; interacting with the "s in the delimiters.
5979 ;;
5980 ;; The font locking routine `c-font-lock-c++-raw-strings' (in cc-fonts.el)
5981 ;; recognizes valid raw strings, and fontifies the delimiters (apart from
5982 ;; the parentheses) with the default face and the parentheses and the
5983 ;; <contents> with font-lock-string-face.
5984 ;;
5985 ;; (ii) A valid, but unterminated, raw string opening delimiter gets the
5986 ;; "punctuation" value (`(1)') of the `syntax-table' text property, and the
5987 ;; open parenthesis gets the "string fence" value (`(15)').
5988 ;;
5989 ;; `c-font-lock-c++-raw-strings' puts c-font-lock-warning-face on the entire
5990 ;; unmatched opening delimiter (from the R up to the open paren), and allows
5991 ;; the rest of the buffer to get font-lock-string-face, caused by the
5992 ;; unmatched "string fence" `syntax-table' text property value.
5993 ;;
5994 ;; (iii) Inside a macro, a valid raw string is handled as in (i). An
5995 ;; unmatched opening delimiter is handled slightly differently. In addition
5996 ;; to the "punctuation" and "string fence" properties on the delimiter,
5997 ;; another "string fence" `syntax-table' property is applied to the last
5998 ;; possible character of the macro before the terminating linefeed (if there
5999 ;; is such a character after the "("). This "last possible" character is
6000 ;; never a backslash escaping the end of line. If the character preceding
6001 ;; this "last possible" character is itself a backslash, this preceding
6002 ;; character gets a "punctuation" `syntax-table' value. If the "(" is
6003 ;; already at the end of the macro, it gets the "punctuation" value, and no
6004 ;; "string fence"s are used.
6005 ;;
6006 ;; The effect on the fontification of either of these tactics is that rest of
6007 ;; the macro (if any) after the "(" gets font-lock-string-face, but the rest
6008 ;; of the file is fontified normally.
6009
6010
6011 (defun c-raw-string-pos ()
6012 ;; Get POINT's relationship to any containing raw string.
6013 ;; If point isn't in a raw string, return nil.
6014 ;; Otherwise, return the following list:
6015 ;;
6016 ;; (POS B\" B\( E\) E\")
6017 ;;
6018 ;; , where POS is the symbol `open-delim' if point is in the opening
6019 ;; delimiter, the symbol `close-delim' if it's in the closing delimiter, and
6020 ;; nil if it's in the string body. B\", B\(, E\), E\" are the positions of
6021 ;; the opening and closing quotes and parentheses of a correctly terminated
6022 ;; raw string. (N.B.: E\) and E\" are NOT on the "outside" of these
6023 ;; characters.) If the raw string is not terminated, E\) and E\" are set to
6024 ;; nil.
6025 ;;
6026 ;; Note: this routine is dependant upon the correct syntax-table text
6027 ;; properties being set.
6028 (let ((state (c-state-semi-pp-to-literal (point)))
6029 open-quote-pos open-paren-pos close-paren-pos close-quote-pos id)
6030 (save-excursion
6031 (when
6032 (and
6033 (cond
6034 ((null (cadr state))
6035 (or (eq (char-after) ?\")
6036 (search-backward "\"" (max (- (point) 17) (point-min)) t)))
6037 ((and (eq (cadr state) 'string)
6038 (goto-char (nth 2 state))
6039 (or (eq (char-after) ?\")
6040 (search-backward "\"" (max (- (point) 17) (point-min)) t))
6041 (not (bobp)))))
6042 (eq (char-before) ?R)
6043 (looking-at "\"\\([^ ()\\\n\r\t]\\{0,16\\}\\)("))
6044 (setq open-quote-pos (point)
6045 open-paren-pos (match-end 1)
6046 id (match-string-no-properties 1))
6047 (goto-char (1+ open-paren-pos))
6048 (when (and (not (c-get-char-property open-paren-pos 'syntax-table))
6049 (search-forward (concat ")" id "\"") nil t))
6050 (setq close-paren-pos (match-beginning 0)
6051 close-quote-pos (1- (point))))))
6052 (and open-quote-pos
6053 (list
6054 (cond
6055 ((<= (point) open-paren-pos)
6056 'open-delim)
6057 ((and close-paren-pos
6058 (> (point) close-paren-pos))
6059 'close-delim)
6060 (t nil))
6061 open-quote-pos open-paren-pos close-paren-pos close-quote-pos))))
6062
6063 (defun c-depropertize-raw-string (id open-quote open-paren bound)
6064 ;; Point is immediately after a raw string opening delimiter. Remove any
6065 ;; `syntax-table' text properties associated with the delimiter (if it's
6066 ;; unmatched) or the raw string.
6067 ;;
6068 ;; ID, a string, is the delimiter's identifier. OPEN-QUOTE and OPEN-PAREN
6069 ;; are the buffer positions of the delimiter's components. BOUND is the
6070 ;; bound for searching for a matching closing delimiter; it is usually nil,
6071 ;; but if we're inside a macro, it's the end of the macro.
6072 ;;
6073 ;; Point is moved to after the (terminated) raw string, or left after the
6074 ;; unmatched opening delimiter, as the case may be. The return value is of
6075 ;; no significance.
6076 (let ((open-paren-prop (c-get-char-property open-paren 'syntax-table)))
6077 (cond
6078 ((null open-paren-prop)
6079 ;; A terminated raw string
6080 (when (search-forward (concat ")" id "\"") nil t)
6081 (let* ((closing-paren (match-beginning 0))
6082 (first-punctuation
6083 (save-match-data
6084 (goto-char (1+ open-paren))
6085 (and (c-search-forward-char-property 'syntax-table '(1)
6086 closing-paren)
6087 (1- (point)))))
6088 )
6089 (when first-punctuation
6090 (c-clear-char-property-with-value
6091 first-punctuation (match-beginning 0) 'syntax-table '(1))
6092 (c-truncate-semi-nonlit-pos-cache first-punctuation)
6093 ))))
6094 ((or (and (equal open-paren-prop '(15)) (null bound))
6095 (equal open-paren-prop '(1)))
6096 ;; An unterminated raw string either not in a macro, or in a macro with
6097 ;; the open parenthesis right up against the end of macro
6098 (c-clear-char-property open-quote 'syntax-table)
6099 (c-truncate-semi-nonlit-pos-cache open-quote)
6100 (c-clear-char-property open-paren 'syntax-table))
6101 (t
6102 ;; An unterminated string in a macro, with at least one char after the
6103 ;; open paren
6104 (c-clear-char-property open-quote 'syntax-table)
6105 (c-truncate-semi-nonlit-pos-cache open-quote)
6106 (c-clear-char-property open-paren 'syntax-table)
6107 (let ((after-string-fence-pos
6108 (save-excursion
6109 (goto-char (1+ open-paren))
6110 (c-search-forward-char-property 'syntax-table '(15) bound))))
6111 (when after-string-fence-pos
6112 (c-clear-char-property (1- after-string-fence-pos) 'syntax-table)))
6113 ))))
6114
6115 (defun c-depropertize-raw-strings-in-region (start finish)
6116 ;; Remove any `syntax-table' text properties associated with C++ raw strings
6117 ;; contained in the region (START FINISH). Point is undefined at entry and
6118 ;; exit, and the return value has no significance.
6119 (goto-char start)
6120 (while (and (< (point) finish)
6121 (re-search-forward
6122 (concat "\\(" ; 1
6123 c-anchored-cpp-prefix ; 2
6124 "\\)\\|\\(" ; 3
6125 "R\"\\([^ ()\\\n\r\t]\\{0,16\\}\\)(" ; 4
6126 "\\)")
6127 finish t))
6128 (when (save-excursion
6129 (goto-char (match-beginning 0)) (not (c-in-literal)))
6130 (if (match-beginning 4) ; the id
6131 ;; We've found a raw string
6132 (c-depropertize-raw-string
6133 (match-string-no-properties 4) ; id
6134 (1+ (match-beginning 3)) ; open quote
6135 (match-end 4) ; open paren
6136 nil) ; bound
6137 ;; We've found a CPP construct. Search for raw strings within it.
6138 (goto-char (match-beginning 2)) ; the "#"
6139 (c-end-of-macro)
6140 (let ((eom (point)))
6141 (goto-char (match-end 2)) ; after the "#".
6142 (while (and (< (point) eom)
6143 (c-syntactic-re-search-forward
6144 "R\"\\([^ ()\\\n\r\t]\\{0,16\\}\\)(" eom t))
6145 (c-depropertize-raw-string
6146 (match-string-no-properties 1) ; id
6147 (1+ (match-beginning 0)) ; open quote
6148 (match-end 1) ; open paren
6149 eom))))))) ; bound.
6150
6151 (defun c-before-change-check-raw-strings (beg end)
6152 ;; This function clears `syntax-table' text properties from C++ raw strings
6153 ;; in the region (c-new-BEG c-new-END). BEG and END are the standard
6154 ;; arguments supplied to any before-change function.
6155 ;;
6156 ;; Point is undefined on both entry and exit, and the return value has no
6157 ;; significance.
6158 ;;
6159 ;; This function is called as a before-change function solely due to its
6160 ;; membership of the C++ value of `c-get-state-before-change-functions'.
6161 (c-save-buffer-state
6162 ((beg-rs (progn (goto-char beg) (c-raw-string-pos)))
6163 (beg-plus (if (null beg-rs)
6164 beg
6165 (max beg
6166 (1+ (or (nth 4 beg-rs) (nth 2 beg-rs))))))
6167 (end-rs (progn (goto-char end) (c-raw-string-pos))) ; FIXME!!!
6168 ; Optimize this so that we don't call
6169 ; `c-raw-string-pos' twice when once
6170 ; will do. (2016-06-02).
6171 (end-minus (if (null end-rs)
6172 end
6173 (min end (cadr end-rs))))
6174 )
6175 (when beg-rs
6176 (setq c-new-BEG (min c-new-BEG (1- (cadr beg-rs)))))
6177 (c-depropertize-raw-strings-in-region c-new-BEG beg-plus)
6178
6179 (when end-rs
6180 (setq c-new-END (max c-new-END
6181 (1+ (or (nth 4 end-rs)
6182 (nth 2 end-rs))))))
6183 (c-depropertize-raw-strings-in-region end-minus c-new-END)))
6184
6185 (defun c-propertize-raw-string-opener (id open-quote open-paren bound)
6186 ;; Point is immediately after a raw string opening delimiter. Apply any
6187 ;; pertinent `syntax-table' text properties to the delimiter and also the
6188 ;; raw string, should there be a valid matching closing delimiter.
6189 ;;
6190 ;; ID, a string, is the delimiter's identifier. OPEN-QUOTE and OPEN-PAREN
6191 ;; are the buffer positions of the delimiter's components. BOUND is the
6192 ;; bound for searching for a matching closing delimiter; it is usually nil,
6193 ;; but if we're inside a macro, it's the end of the macro.
6194 ;;
6195 ;; Point is moved to after the (terminated) raw string, or left after the
6196 ;; unmatched opening delimiter, as the case may be. The return value is of
6197 ;; no significance.
6198 (if (search-forward (concat ")" id "\"") bound t)
6199 (let ((end-string (match-beginning 0))
6200 (after-quote (match-end 0)))
6201 (goto-char open-paren)
6202 (while (progn (skip-syntax-forward "^\"" end-string)
6203 (< (point) end-string))
6204 (c-put-char-property (point) 'syntax-table '(1)) ; punctuation
6205 (c-truncate-semi-nonlit-pos-cache (point))
6206 (forward-char))
6207 (goto-char after-quote))
6208 (c-put-char-property open-quote 'syntax-table '(1)) ; punctuation
6209 (c-truncate-semi-nonlit-pos-cache open-quote)
6210 (c-put-char-property open-paren 'syntax-table '(15)) ; generic string
6211 (when bound
6212 ;; In a CPP construct, we try to apply a generic-string `syntax-table'
6213 ;; text property to the last possible character in the string, so that
6214 ;; only characters within the macro get "stringed out".
6215 (goto-char bound)
6216 (if (save-restriction
6217 (narrow-to-region (1+ open-paren) (point-max))
6218 (re-search-backward
6219 (eval-when-compile
6220 ;; This regular expression matches either an escape pair (which
6221 ;; isn't an escaped NL) (submatch 5) or a non-escaped character
6222 ;; (which isn't itself a backslash) (submatch 10). The long
6223 ;; preambles to these (respectively submatches 2-4 and 6-9)
6224 ;; ensure that we have the correct parity for sequences of
6225 ;; backslashes, etc..
6226 (concat "\\(" ; 1
6227 "\\(\\`[^\\]?\\|[^\\][^\\]\\)\\(\\\\\\(.\\|\n\\)\\)*" ; 2-4
6228 "\\(\\\\.\\)" ; 5
6229 "\\|"
6230 "\\(\\`\\|[^\\]\\|\\(\\`[^\\]?\\|[^\\][^\\]\\)\\(\\\\\\(.\\|\n\\)\\)+\\)" ; 6-9
6231 "\\([^\\]\\)" ; 10
6232 "\\)"
6233 "\\(\\\\\n\\)*\\=")) ; 11
6234 (1+ open-paren) t))
6235 (if (match-beginning 10)
6236 (progn
6237 (c-put-char-property (match-beginning 10) 'syntax-table '(15))
6238 (c-truncate-semi-nonlit-pos-cache (match-beginning 10)))
6239 (c-put-char-property (match-beginning 5) 'syntax-table '(1))
6240 (c-put-char-property (1+ (match-beginning 5)) 'syntax-table '(15))
6241 (c-truncate-semi-nonlit-pos-cache (1+ (match-beginning 5))))
6242 (c-put-char-property open-paren 'syntax-table '(1)))
6243 (goto-char bound))))
6244
6245 (defun c-after-change-re-mark-raw-strings (beg end old-len)
6246 ;; This function applies `syntax-table' text properties to C++ raw strings
6247 ;; beginning in the region (c-new-BEG c-new-END). BEG, END, and OLD-LEN are
6248 ;; the standard arguments supplied to any after-change function.
6249 ;;
6250 ;; Point is undefined on both entry and exit, and the return value has no
6251 ;; significance.
6252 ;;
6253 ;; This function is called as an after-change function solely due to its
6254 ;; membership of the C++ value of `c-before-font-lock-functions'.
6255 (c-save-buffer-state ()
6256 ;; If the region (c-new-BEG c-new-END) has expanded, remove
6257 ;; `syntax-table' text-properties from the new piece(s).
6258 (when (< c-new-BEG c-old-BEG)
6259 (let ((beg-rs (progn (goto-char c-old-BEG) (c-raw-string-pos))))
6260 (c-depropertize-raw-strings-in-region
6261 c-new-BEG
6262 (if beg-rs
6263 (1+ (or (nth 4 beg-rs) (nth 2 beg-rs)))
6264 c-old-BEG))))
6265 (when (> c-new-END c-old-END)
6266 (let ((end-rs (progn (goto-char c-old-END) (c-raw-string-pos))))
6267 (c-depropertize-raw-strings-in-region
6268 (if end-rs
6269 (cadr end-rs)
6270 c-old-END)
6271 c-new-END)))
6272
6273 (goto-char c-new-BEG)
6274 (while (and (< (point) c-new-END)
6275 (re-search-forward
6276 (concat "\\(" ; 1
6277 c-anchored-cpp-prefix ; 2
6278 "\\)\\|\\(" ; 3
6279 "R\"\\([^ ()\\\n\r\t]\\{0,16\\}\\)(" ; 4
6280 "\\)")
6281 c-new-END t))
6282 (when (save-excursion
6283 (goto-char (match-beginning 0)) (not (c-in-literal)))
6284 (if (match-beginning 4) ; the id
6285 ;; We've found a raw string.
6286 (c-propertize-raw-string-opener
6287 (match-string-no-properties 4) ; id
6288 (1+ (match-beginning 3)) ; open quote
6289 (match-end 4) ; open paren
6290 nil) ; bound
6291 ;; We've found a CPP construct. Search for raw strings within it.
6292 (goto-char (match-beginning 2)) ; the "#"
6293 (c-end-of-macro)
6294 (let ((eom (point)))
6295 (goto-char (match-end 2)) ; after the "#".
6296 (while (and (< (point) eom)
6297 (c-syntactic-re-search-forward
6298 "R\"\\([^ ()\\\n\r\t]\\{0,16\\}\\)(" eom t))
6299 (c-propertize-raw-string-opener
6300 (match-string-no-properties 1) ; id
6301 (1+ (match-beginning 0)) ; open quote
6302 (match-end 1) ; open paren
6303 eom)))))))) ; bound
6304
6305 \f
6306 ;; Handling of small scale constructs like types and names.
6307
6308 ;; Dynamically bound variable that instructs `c-forward-type' to also
6309 ;; treat possible types (i.e. those that it normally returns 'maybe or
6310 ;; 'found for) as actual types (and always return 'found for them).
6311 ;; This means that it records them in `c-record-type-identifiers' if
6312 ;; that is set, and that it adds them to `c-found-types'.
6313 (defvar c-promote-possible-types nil)
6314
6315 ;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
6316 ;; mark up successfully parsed arglists with paren syntax properties on
6317 ;; the surrounding angle brackets and with `c-<>-arg-sep' in the
6318 ;; `c-type' property of each argument separating comma.
6319 ;;
6320 ;; Setting this variable also makes `c-forward-<>-arglist' recurse into
6321 ;; all arglists for side effects (i.e. recording types), otherwise it
6322 ;; exploits any existing paren syntax properties to quickly jump to the
6323 ;; end of already parsed arglists.
6324 ;;
6325 ;; Marking up the arglists is not the default since doing that correctly
6326 ;; depends on a proper value for `c-restricted-<>-arglists'.
6327 (defvar c-parse-and-markup-<>-arglists nil)
6328
6329 ;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
6330 ;; not accept arglists that contain binary operators.
6331 ;;
6332 ;; This is primarily used to handle C++ template arglists. C++
6333 ;; disambiguates them by checking whether the preceding name is a
6334 ;; template or not. We can't do that, so we assume it is a template
6335 ;; if it can be parsed as one. That usually works well since
6336 ;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
6337 ;; in almost all cases would be pointless.
6338 ;;
6339 ;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
6340 ;; should let the comma separate the function arguments instead. And
6341 ;; in a context where the value of the expression is taken, e.g. in
6342 ;; "if (a < b || c > d)", it's probably not a template.
6343 (defvar c-restricted-<>-arglists nil)
6344
6345 ;; Dynamically bound variables that instructs
6346 ;; `c-forward-keyword-clause', `c-forward-<>-arglist',
6347 ;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
6348 ;; `c-forward-label' to record the ranges of all the type and
6349 ;; reference identifiers they encounter. They will build lists on
6350 ;; these variables where each element is a cons of the buffer
6351 ;; positions surrounding each identifier. This recording is only
6352 ;; activated when `c-record-type-identifiers' is non-nil.
6353 ;;
6354 ;; All known types that can't be identifiers are recorded, and also
6355 ;; other possible types if `c-promote-possible-types' is set.
6356 ;; Recording is however disabled inside angle bracket arglists that
6357 ;; are encountered inside names and other angle bracket arglists.
6358 ;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
6359 ;; instead.
6360 ;;
6361 ;; Only the names in C++ template style references (e.g. "tmpl" in
6362 ;; "tmpl<a,b>::foo") are recorded as references, other references
6363 ;; aren't handled here.
6364 ;;
6365 ;; `c-forward-label' records the label identifier(s) on
6366 ;; `c-record-ref-identifiers'.
6367 (defvar c-record-type-identifiers nil)
6368 (defvar c-record-ref-identifiers nil)
6369
6370 ;; This variable will receive a cons cell of the range of the last
6371 ;; single identifier symbol stepped over by `c-forward-name' if it's
6372 ;; successful. This is the range that should be put on one of the
6373 ;; record lists above by the caller. It's assigned nil if there's no
6374 ;; such symbol in the name.
6375 (defvar c-last-identifier-range nil)
6376
6377 (defmacro c-record-type-id (range)
6378 (if (eq (car-safe range) 'cons)
6379 ;; Always true.
6380 `(setq c-record-type-identifiers
6381 (cons ,range c-record-type-identifiers))
6382 `(let ((range ,range))
6383 (if range
6384 (setq c-record-type-identifiers
6385 (cons range c-record-type-identifiers))))))
6386
6387 (defmacro c-record-ref-id (range)
6388 (if (eq (car-safe range) 'cons)
6389 ;; Always true.
6390 `(setq c-record-ref-identifiers
6391 (cons ,range c-record-ref-identifiers))
6392 `(let ((range ,range))
6393 (if range
6394 (setq c-record-ref-identifiers
6395 (cons range c-record-ref-identifiers))))))
6396
6397 ;; Dynamically bound variable that instructs `c-forward-type' to
6398 ;; record the ranges of types that only are found. Behaves otherwise
6399 ;; like `c-record-type-identifiers'.
6400 (defvar c-record-found-types nil)
6401
6402 (defmacro c-forward-keyword-prefixed-id (type)
6403 ;; Used internally in `c-forward-keyword-clause' to move forward
6404 ;; over a type (if TYPE is 'type) or a name (otherwise) which
6405 ;; possibly is prefixed by keywords and their associated clauses.
6406 ;; Try with a type/name first to not trip up on those that begin
6407 ;; with a keyword. Return t if a known or found type is moved
6408 ;; over. The point is clobbered if nil is returned. If range
6409 ;; recording is enabled, the identifier is recorded on as a type
6410 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
6411 ;;
6412 ;; This macro might do hidden buffer changes.
6413 `(let (res)
6414 (setq c-last-identifier-range nil)
6415 (while (if (setq res ,(if (eq type 'type)
6416 `(c-forward-type)
6417 `(c-forward-name)))
6418 nil
6419 (cond ((looking-at c-keywords-regexp)
6420 (c-forward-keyword-clause 1))
6421 ((and c-opt-cpp-prefix
6422 (looking-at c-noise-macro-with-parens-name-re))
6423 (c-forward-noise-clause)))))
6424 (when (memq res '(t known found prefix maybe))
6425 (when c-record-type-identifiers
6426 ,(if (eq type 'type)
6427 `(c-record-type-id c-last-identifier-range)
6428 `(c-record-ref-id c-last-identifier-range)))
6429 t)))
6430
6431 (defmacro c-forward-id-comma-list (type update-safe-pos)
6432 ;; Used internally in `c-forward-keyword-clause' to move forward
6433 ;; over a comma separated list of types or names using
6434 ;; `c-forward-keyword-prefixed-id'.
6435 ;;
6436 ;; This macro might do hidden buffer changes.
6437 `(while (and (progn
6438 ,(when update-safe-pos
6439 `(setq safe-pos (point)))
6440 (eq (char-after) ?,))
6441 (progn
6442 (forward-char)
6443 (c-forward-syntactic-ws)
6444 (c-forward-keyword-prefixed-id ,type)))))
6445
6446 (defun c-forward-noise-clause ()
6447 ;; Point is at a c-noise-macro-with-parens-names macro identifier. Go
6448 ;; forward over this name, any parenthesis expression which follows it, and
6449 ;; any syntactic WS, ending up at the next token. If there is an unbalanced
6450 ;; paren expression, leave point at it. Always Return t.
6451 (c-forward-token-2)
6452 (if (and (eq (char-after) ?\()
6453 (c-go-list-forward))
6454 (c-forward-syntactic-ws))
6455 t)
6456
6457 (defun c-forward-keyword-clause (match)
6458 ;; Submatch MATCH in the current match data is assumed to surround a
6459 ;; token. If it's a keyword, move over it and any immediately
6460 ;; following clauses associated with it, stopping at the start of
6461 ;; the next token. t is returned in that case, otherwise the point
6462 ;; stays and nil is returned. The kind of clauses that are
6463 ;; recognized are those specified by `c-type-list-kwds',
6464 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
6465 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
6466 ;; and `c-<>-arglist-kwds'.
6467 ;;
6468 ;; This function records identifier ranges on
6469 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6470 ;; `c-record-type-identifiers' is non-nil.
6471 ;;
6472 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
6473 ;; apply directly after the keyword, the type list is moved over
6474 ;; only when there is no unaccounted token before it (i.e. a token
6475 ;; that isn't moved over due to some other keyword list). The
6476 ;; identifier ranges in the list are still recorded if that should
6477 ;; be done, though.
6478 ;;
6479 ;; This function might do hidden buffer changes.
6480
6481 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
6482 ;; The call to `c-forward-<>-arglist' below is made after
6483 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
6484 ;; are angle bracket arglists and `c-restricted-<>-arglists'
6485 ;; should therefore be nil.
6486 (c-parse-and-markup-<>-arglists t)
6487 c-restricted-<>-arglists)
6488
6489 (when kwd-sym
6490 (goto-char (match-end match))
6491 (c-forward-syntactic-ws)
6492 (setq safe-pos (point))
6493
6494 (cond
6495 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
6496 (c-forward-keyword-prefixed-id type))
6497 ;; There's a type directly after a keyword in `c-type-list-kwds'.
6498 (c-forward-id-comma-list type t))
6499
6500 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
6501 (c-forward-keyword-prefixed-id ref))
6502 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
6503 (c-forward-id-comma-list ref t))
6504
6505 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
6506 (eq (char-after) ?\())
6507 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
6508
6509 (forward-char)
6510 (when (and (setq pos (c-up-list-forward))
6511 (eq (char-before pos) ?\)))
6512 (when (and c-record-type-identifiers
6513 (c-keyword-member kwd-sym 'c-paren-type-kwds))
6514 ;; Use `c-forward-type' on every identifier we can find
6515 ;; inside the paren, to record the types.
6516 (while (c-syntactic-re-search-forward c-symbol-start pos t)
6517 (goto-char (match-beginning 0))
6518 (unless (c-forward-type)
6519 (looking-at c-symbol-key) ; Always matches.
6520 (goto-char (match-end 0)))))
6521
6522 (goto-char pos)
6523 (c-forward-syntactic-ws)
6524 (setq safe-pos (point))))
6525
6526 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
6527 (eq (char-after) ?<)
6528 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
6529 (c-forward-syntactic-ws)
6530 (setq safe-pos (point)))
6531
6532 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
6533 (not (looking-at c-symbol-start))
6534 (c-safe (c-forward-sexp) t))
6535 (c-forward-syntactic-ws)
6536 (setq safe-pos (point))))
6537
6538 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
6539 (if (eq (char-after) ?:)
6540 ;; If we are at the colon already, we move over the type
6541 ;; list after it.
6542 (progn
6543 (forward-char)
6544 (c-forward-syntactic-ws)
6545 (when (c-forward-keyword-prefixed-id type)
6546 (c-forward-id-comma-list type t)))
6547 ;; Not at the colon, so stop here. But the identifier
6548 ;; ranges in the type list later on should still be
6549 ;; recorded.
6550 (and c-record-type-identifiers
6551 (progn
6552 ;; If a keyword matched both one of the types above and
6553 ;; this one, we match `c-colon-type-list-re' after the
6554 ;; clause matched above.
6555 (goto-char safe-pos)
6556 (looking-at c-colon-type-list-re))
6557 (progn
6558 (goto-char (match-end 0))
6559 (c-forward-syntactic-ws)
6560 (c-forward-keyword-prefixed-id type))
6561 ;; There's a type after the `c-colon-type-list-re' match
6562 ;; after a keyword in `c-colon-type-list-kwds'.
6563 (c-forward-id-comma-list type nil))))
6564
6565 (goto-char safe-pos)
6566 t)))
6567
6568 ;; cc-mode requires cc-fonts.
6569 (declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
6570
6571 (defun c-forward-<>-arglist (all-types)
6572 ;; The point is assumed to be at a "<". Try to treat it as the open
6573 ;; paren of an angle bracket arglist and move forward to the
6574 ;; corresponding ">". If successful, the point is left after the
6575 ;; ">" and t is returned, otherwise the point isn't moved and nil is
6576 ;; returned. If ALL-TYPES is t then all encountered arguments in
6577 ;; the arglist that might be types are treated as found types.
6578 ;;
6579 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
6580 ;; function handles text properties on the angle brackets and argument
6581 ;; separating commas.
6582 ;;
6583 ;; `c-restricted-<>-arglists' controls how lenient the template
6584 ;; arglist recognition should be.
6585 ;;
6586 ;; This function records identifier ranges on
6587 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6588 ;; `c-record-type-identifiers' is non-nil.
6589 ;;
6590 ;; This function might do hidden buffer changes.
6591
6592 (let ((start (point))
6593 ;; If `c-record-type-identifiers' is set then activate
6594 ;; recording of any found types that constitute an argument in
6595 ;; the arglist.
6596 (c-record-found-types (if c-record-type-identifiers t)))
6597 (if (catch 'angle-bracket-arglist-escape
6598 (setq c-record-found-types
6599 (c-forward-<>-arglist-recur all-types)))
6600 (progn
6601 (when (consp c-record-found-types)
6602 (setq c-record-type-identifiers
6603 ;; `nconc' doesn't mind that the tail of
6604 ;; `c-record-found-types' is t.
6605 (nconc c-record-found-types c-record-type-identifiers)))
6606 t)
6607
6608 (goto-char start)
6609 nil)))
6610
6611 (defun c-forward-<>-arglist-recur (all-types)
6612 ;; Recursive part of `c-forward-<>-arglist'.
6613 ;;
6614 ;; This function might do hidden buffer changes.
6615 (let ((start (point)) res pos
6616 ;; Cover this so that any recorded found type ranges are
6617 ;; automatically lost if it turns out to not be an angle
6618 ;; bracket arglist. It's propagated through the return value
6619 ;; on successful completion.
6620 (c-record-found-types c-record-found-types)
6621 ;; List that collects the positions after the argument
6622 ;; separating ',' in the arglist.
6623 arg-start-pos)
6624 ;; If the '<' has paren open syntax then we've marked it as an angle
6625 ;; bracket arglist before, so skip to the end.
6626 (if (and (not c-parse-and-markup-<>-arglists)
6627 (c-get-char-property (point) 'syntax-table))
6628
6629 (progn
6630 (forward-char)
6631 (if (and (c-go-up-list-forward)
6632 (eq (char-before) ?>))
6633 t
6634 ;; Got unmatched paren angle brackets. We don't clear the paren
6635 ;; syntax properties and retry, on the basis that it's very
6636 ;; unlikely that paren angle brackets become operators by code
6637 ;; manipulation. It's far more likely that it doesn't match due
6638 ;; to narrowing or some temporary change.
6639 (goto-char start)
6640 nil))
6641
6642 (forward-char) ; Forward over the opening '<'.
6643
6644 (unless (looking-at c-<-op-cont-regexp)
6645 ;; go forward one non-alphanumeric character (group) per iteration of
6646 ;; this loop.
6647 (while (and
6648 (progn
6649 (c-forward-syntactic-ws)
6650 (when (or (and c-record-type-identifiers all-types)
6651 (not (equal c-inside-<>-type-key "\\(\\<\\>\\)")))
6652 (c-forward-syntactic-ws)
6653 (cond
6654 ((eq (char-after) ??)
6655 (forward-char))
6656 ((and (looking-at c-identifier-start)
6657 (not (looking-at c-keywords-regexp)))
6658 (if (or (and all-types c-record-type-identifiers)
6659 (c-major-mode-is 'java-mode))
6660 ;; All encountered identifiers are types, so set the
6661 ;; promote flag and parse the type.
6662 (let ((c-promote-possible-types t)
6663 (c-record-found-types t))
6664 (c-forward-type))
6665 (c-forward-token-2))))
6666
6667 (c-forward-syntactic-ws)
6668
6669 (when (looking-at c-inside-<>-type-key)
6670 (goto-char (match-end 1))
6671 (c-forward-syntactic-ws)
6672 (let ((c-promote-possible-types t)
6673 (c-record-found-types t))
6674 (c-forward-type))
6675 (c-forward-syntactic-ws)))
6676
6677 (setq pos (point)) ; e.g. first token inside the '<'
6678
6679 ;; Note: These regexps exploit the match order in \| so
6680 ;; that "<>" is matched by "<" rather than "[^>:-]>".
6681 (c-syntactic-re-search-forward
6682 ;; Stop on ',', '|', '&', '+' and '-' to catch
6683 ;; common binary operators that could be between
6684 ;; two comparison expressions "a<b" and "c>d".
6685 ;; 2016-02-11: C++11 templates can now contain arithmetic
6686 ;; expressions, so template detection in C++ is now less
6687 ;; robust than it was.
6688 c-<>-notable-chars-re
6689 nil t t))
6690
6691 (cond
6692 ((eq (char-before) ?>)
6693 ;; Either an operator starting with '>' or the end of
6694 ;; the angle bracket arglist.
6695
6696 (if (save-excursion
6697 (c-backward-token-2)
6698 (looking-at c-multichar->-op-not->>-regexp))
6699 (progn
6700 (goto-char (match-end 0))
6701 t) ; Continue the loop.
6702
6703 ;; The angle bracket arglist is finished.
6704 (when c-parse-and-markup-<>-arglists
6705 (while arg-start-pos
6706 (c-put-c-type-property (1- (car arg-start-pos))
6707 'c-<>-arg-sep)
6708 (setq arg-start-pos (cdr arg-start-pos)))
6709 (c-mark-<-as-paren start)
6710 (c-mark->-as-paren (1- (point))))
6711 (setq res t)
6712 nil)) ; Exit the loop.
6713
6714 ((eq (char-before) ?<)
6715 ;; Either an operator starting with '<' or a nested arglist.
6716 (setq pos (point))
6717 (let (id-start id-end subres keyword-match)
6718 (cond
6719 ;; The '<' begins a multi-char operator.
6720 ((looking-at c-<-op-cont-regexp)
6721 (goto-char (match-end 0)))
6722 ;; We're at a nested <.....>
6723 ((progn
6724 (backward-char) ; to the '<'
6725 (and
6726 (save-excursion
6727 ;; There's always an identifier before an angle
6728 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
6729 ;; or `c-<>-arglist-kwds'.
6730 (c-backward-syntactic-ws)
6731 (setq id-end (point))
6732 (c-simple-skip-symbol-backward)
6733 (when (or (setq keyword-match
6734 (looking-at c-opt-<>-sexp-key))
6735 (not (looking-at c-keywords-regexp)))
6736 (setq id-start (point))))
6737 (setq subres
6738 (let ((c-promote-possible-types t)
6739 (c-record-found-types t))
6740 (c-forward-<>-arglist-recur
6741 (and keyword-match
6742 (c-keyword-member
6743 (c-keyword-sym (match-string 1))
6744 'c-<>-type-kwds))))))
6745 (or subres (goto-char pos))
6746 subres)
6747 ;; It was an angle bracket arglist.
6748 (setq c-record-found-types subres)
6749
6750 ;; Record the identifier before the template as a type
6751 ;; or reference depending on whether the arglist is last
6752 ;; in a qualified identifier.
6753 (when (and c-record-type-identifiers
6754 (not keyword-match))
6755 (if (and c-opt-identifier-concat-key
6756 (progn
6757 (c-forward-syntactic-ws)
6758 (looking-at c-opt-identifier-concat-key)))
6759 (c-record-ref-id (cons id-start id-end))
6760 (c-record-type-id (cons id-start id-end)))))
6761
6762 ;; At a "less than" operator.
6763 (t
6764 ;; (forward-char) ; NO! We've already gone over the <.
6765 )))
6766 t) ; carry on looping.
6767
6768 ((and
6769 (eq (char-before) ?\()
6770 (c-go-up-list-forward)
6771 (eq (char-before) ?\))))
6772
6773 ((and (not c-restricted-<>-arglists)
6774 (or (and (eq (char-before) ?&)
6775 (not (eq (char-after) ?&)))
6776 (eq (char-before) ?,)))
6777 ;; Just another argument. Record the position. The
6778 ;; type check stuff that made us stop at it is at
6779 ;; the top of the loop.
6780 (setq arg-start-pos (cons (point) arg-start-pos)))
6781
6782 (t
6783 ;; Got a character that can't be in an angle bracket
6784 ;; arglist argument. Abort using `throw', since
6785 ;; it's useless to try to find a surrounding arglist
6786 ;; if we're nested.
6787 (throw 'angle-bracket-arglist-escape nil))))))
6788 (if res
6789 (or c-record-found-types t)))))
6790
6791 (defun c-backward-<>-arglist (all-types &optional limit)
6792 ;; The point is assumed to be directly after a ">". Try to treat it
6793 ;; as the close paren of an angle bracket arglist and move back to
6794 ;; the corresponding "<". If successful, the point is left at
6795 ;; the "<" and t is returned, otherwise the point isn't moved and
6796 ;; nil is returned. ALL-TYPES is passed on to
6797 ;; `c-forward-<>-arglist'.
6798 ;;
6799 ;; If the optional LIMIT is given, it bounds the backward search.
6800 ;; It's then assumed to be at a syntactically relevant position.
6801 ;;
6802 ;; This is a wrapper around `c-forward-<>-arglist'. See that
6803 ;; function for more details.
6804
6805 (let ((start (point)))
6806 (backward-char)
6807 (if (and (not c-parse-and-markup-<>-arglists)
6808 (c-get-char-property (point) 'syntax-table))
6809
6810 (if (and (c-go-up-list-backward)
6811 (eq (char-after) ?<))
6812 t
6813 ;; See corresponding note in `c-forward-<>-arglist'.
6814 (goto-char start)
6815 nil)
6816
6817 (while (progn
6818 (c-syntactic-skip-backward "^<;{}" limit t)
6819
6820 (and
6821 (if (eq (char-before) ?<)
6822 t
6823 ;; Stopped at bob or a char that isn't allowed in an
6824 ;; arglist, so we've failed.
6825 (goto-char start)
6826 nil)
6827
6828 (if (> (point)
6829 (progn (c-beginning-of-current-token)
6830 (point)))
6831 ;; If we moved then the "<" was part of some
6832 ;; multicharacter token.
6833 t
6834
6835 (backward-char)
6836 (let ((beg-pos (point)))
6837 (if (c-forward-<>-arglist all-types)
6838 (cond ((= (point) start)
6839 ;; Matched the arglist. Break the while.
6840 (goto-char beg-pos)
6841 nil)
6842 ((> (point) start)
6843 ;; We started from a non-paren ">" inside an
6844 ;; arglist.
6845 (goto-char start)
6846 nil)
6847 (t
6848 ;; Matched a shorter arglist. Can be a nested
6849 ;; one so continue looking.
6850 (goto-char beg-pos)
6851 t))
6852 t))))))
6853
6854 (/= (point) start))))
6855
6856 (defun c-forward-name ()
6857 ;; Move forward over a complete name if at the beginning of one,
6858 ;; stopping at the next following token. A keyword, as such,
6859 ;; doesn't count as a name. If the point is not at something that
6860 ;; is recognized as a name then it stays put.
6861 ;;
6862 ;; A name could be something as simple as "foo" in C or something as
6863 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
6864 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
6865 ;; int>::*volatile const" in C++ (this function is actually little
6866 ;; more than a `looking-at' call in all modes except those that,
6867 ;; like C++, have `c-recognize-<>-arglists' set).
6868 ;;
6869 ;; Return
6870 ;; o - nil if no name is found;
6871 ;; o - 'template if it's an identifier ending with an angle bracket
6872 ;; arglist;
6873 ;; o - 'operator of it's an operator identifier;
6874 ;; o - t if it's some other kind of name.
6875 ;;
6876 ;; This function records identifier ranges on
6877 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6878 ;; `c-record-type-identifiers' is non-nil.
6879 ;;
6880 ;; This function might do hidden buffer changes.
6881
6882 (let ((pos (point)) (start (point)) res id-start id-end
6883 ;; Turn off `c-promote-possible-types' here since we might
6884 ;; call `c-forward-<>-arglist' and we don't want it to promote
6885 ;; every suspect thing in the arglist to a type. We're
6886 ;; typically called from `c-forward-type' in this case, and
6887 ;; the caller only wants the top level type that it finds to
6888 ;; be promoted.
6889 c-promote-possible-types)
6890 (while
6891 (and
6892 (looking-at c-identifier-key)
6893
6894 (progn
6895 ;; Check for keyword. We go to the last symbol in
6896 ;; `c-identifier-key' first.
6897 (goto-char (setq id-end (match-end 0)))
6898 (c-simple-skip-symbol-backward)
6899 (setq id-start (point))
6900
6901 (if (looking-at c-keywords-regexp)
6902 (when (and (c-major-mode-is 'c++-mode)
6903 (looking-at
6904 (cc-eval-when-compile
6905 (concat "\\(operator\\|\\(template\\)\\)"
6906 "\\(" (c-lang-const c-nonsymbol-key c++)
6907 "\\|$\\)")))
6908 (if (match-beginning 2)
6909 ;; "template" is only valid inside an
6910 ;; identifier if preceded by "::".
6911 (save-excursion
6912 (c-backward-syntactic-ws)
6913 (and (c-safe (backward-char 2) t)
6914 (looking-at "::")))
6915 t))
6916
6917 ;; Handle a C++ operator or template identifier.
6918 (goto-char id-end)
6919 (c-forward-syntactic-ws)
6920 (cond ((eq (char-before id-end) ?e)
6921 ;; Got "... ::template".
6922 (let ((subres (c-forward-name)))
6923 (when subres
6924 (setq pos (point)
6925 res subres))))
6926
6927 ((looking-at c-identifier-start)
6928 ;; Got a cast operator.
6929 (when (c-forward-type)
6930 (setq pos (point)
6931 res 'operator)
6932 ;; Now we should match a sequence of either
6933 ;; '*', '&' or a name followed by ":: *",
6934 ;; where each can be followed by a sequence
6935 ;; of `c-opt-type-modifier-key'.
6936 (while (cond ((looking-at "[*&]")
6937 (goto-char (match-end 0))
6938 t)
6939 ((looking-at c-identifier-start)
6940 (and (c-forward-name)
6941 (looking-at "::")
6942 (progn
6943 (goto-char (match-end 0))
6944 (c-forward-syntactic-ws)
6945 (eq (char-after) ?*))
6946 (progn
6947 (forward-char)
6948 t))))
6949 (while (progn
6950 (c-forward-syntactic-ws)
6951 (setq pos (point))
6952 (looking-at c-opt-type-modifier-key))
6953 (goto-char (match-end 1))))))
6954
6955 ((looking-at c-overloadable-operators-regexp)
6956 ;; Got some other operator.
6957 (setq c-last-identifier-range
6958 (cons (point) (match-end 0)))
6959 (goto-char (match-end 0))
6960 (c-forward-syntactic-ws)
6961 (setq pos (point)
6962 res 'operator)))
6963
6964 nil)
6965
6966 ;; `id-start' is equal to `id-end' if we've jumped over
6967 ;; an identifier that doesn't end with a symbol token.
6968 ;; That can occur e.g. for Java import directives on the
6969 ;; form "foo.bar.*".
6970 (when (and id-start (/= id-start id-end))
6971 (setq c-last-identifier-range
6972 (cons id-start id-end)))
6973 (goto-char id-end)
6974 (c-forward-syntactic-ws)
6975 (setq pos (point)
6976 res t)))
6977
6978 (progn
6979 (goto-char pos)
6980 (when (or c-opt-identifier-concat-key
6981 c-recognize-<>-arglists)
6982
6983 (cond
6984 ((and c-opt-identifier-concat-key
6985 (looking-at c-opt-identifier-concat-key))
6986 ;; Got a concatenated identifier. This handles the
6987 ;; cases with tricky syntactic whitespace that aren't
6988 ;; covered in `c-identifier-key'.
6989 (goto-char (match-end 0))
6990 (c-forward-syntactic-ws)
6991 t)
6992
6993 ((and c-recognize-<>-arglists
6994 (eq (char-after) ?<))
6995 ;; Maybe an angle bracket arglist.
6996 (when (let (c-last-identifier-range)
6997 (c-forward-<>-arglist nil))
6998
6999 (c-forward-syntactic-ws)
7000 (unless (eq (char-after) ?\()
7001 (setq c-last-identifier-range nil)
7002 (c-add-type start (1+ pos)))
7003 (setq pos (point))
7004
7005 (if (and c-opt-identifier-concat-key
7006 (looking-at c-opt-identifier-concat-key))
7007
7008 ;; Continue if there's an identifier concatenation
7009 ;; operator after the template argument.
7010 (progn
7011 (when (and c-record-type-identifiers id-start)
7012 (c-record-ref-id (cons id-start id-end)))
7013 (forward-char 2)
7014 (c-forward-syntactic-ws)
7015 t)
7016
7017 (when (and c-record-type-identifiers id-start
7018 (not (eq (char-after) ?\()))
7019 (c-record-type-id (cons id-start id-end)))
7020 (setq res 'template)
7021 nil)))
7022 )))))
7023
7024 (goto-char pos)
7025 res))
7026
7027 (defun c-forward-type (&optional brace-block-too)
7028 ;; Move forward over a type spec if at the beginning of one,
7029 ;; stopping at the next following token. The keyword "typedef"
7030 ;; isn't part of a type spec here.
7031 ;;
7032 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
7033 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
7034 ;; The current (2009-03-10) intention is to convert all uses of
7035 ;; `c-forward-type' to call with this parameter set, then to
7036 ;; eliminate it.
7037 ;;
7038 ;; Return
7039 ;; o - t if it's a known type that can't be a name or other
7040 ;; expression;
7041 ;; o - 'known if it's an otherwise known type (according to
7042 ;; `*-font-lock-extra-types');
7043 ;; o - 'prefix if it's a known prefix of a type;
7044 ;; o - 'found if it's a type that matches one in `c-found-types';
7045 ;; o - 'maybe if it's an identifier that might be a type;
7046 ;; o - 'decltype if it's a decltype(variable) declaration; - or
7047 ;; o - nil if it can't be a type (the point isn't moved then).
7048 ;;
7049 ;; The point is assumed to be at the beginning of a token.
7050 ;;
7051 ;; Note that this function doesn't skip past the brace definition
7052 ;; that might be considered part of the type, e.g.
7053 ;; "enum {a, b, c} foo".
7054 ;;
7055 ;; This function records identifier ranges on
7056 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7057 ;; `c-record-type-identifiers' is non-nil.
7058 ;;
7059 ;; This function might do hidden buffer changes.
7060 (when (and c-recognize-<>-arglists
7061 (looking-at "<"))
7062 (c-forward-<>-arglist t)
7063 (c-forward-syntactic-ws))
7064
7065 (let ((start (point)) pos res name-res id-start id-end id-range)
7066
7067 ;; Skip leading type modifiers. If any are found we know it's a
7068 ;; prefix of a type.
7069 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
7070 (while (looking-at c-opt-type-modifier-key)
7071 (goto-char (match-end 1))
7072 (c-forward-syntactic-ws)
7073 (setq res 'prefix)))
7074
7075 (cond
7076 ((looking-at c-typeof-key) ; e.g. C++'s "decltype".
7077 (goto-char (match-end 1))
7078 (c-forward-syntactic-ws)
7079 (setq res (and (eq (char-after) ?\()
7080 (c-safe (c-forward-sexp))
7081 'decltype))
7082 (if res
7083 (c-forward-syntactic-ws)
7084 (goto-char start)))
7085
7086 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
7087 ; "typedef".
7088 (goto-char (match-end 1))
7089 (c-forward-syntactic-ws)
7090
7091 (while (cond
7092 ((looking-at c-decl-hangon-key)
7093 (c-forward-keyword-clause 1))
7094 ((and c-opt-cpp-prefix
7095 (looking-at c-noise-macro-with-parens-name-re))
7096 (c-forward-noise-clause))))
7097
7098 (setq pos (point))
7099
7100 (setq name-res (c-forward-name))
7101 (setq res (not (null name-res)))
7102 (when (eq name-res t)
7103 ;; In many languages the name can be used without the
7104 ;; prefix, so we add it to `c-found-types'.
7105 (c-add-type pos (point))
7106 (when (and c-record-type-identifiers
7107 c-last-identifier-range)
7108 (c-record-type-id c-last-identifier-range)))
7109 (when (and brace-block-too
7110 (memq res '(t nil))
7111 (eq (char-after) ?\{)
7112 (save-excursion
7113 (c-safe
7114 (progn (c-forward-sexp)
7115 (c-forward-syntactic-ws)
7116 (setq pos (point))))))
7117 (goto-char pos)
7118 (setq res t))
7119 (unless res (goto-char start))) ; invalid syntax
7120
7121 ((progn
7122 (setq pos nil)
7123 (if (looking-at c-identifier-start)
7124 (save-excursion
7125 (setq id-start (point)
7126 name-res (c-forward-name))
7127 (when name-res
7128 (setq id-end (point)
7129 id-range c-last-identifier-range))))
7130 (and (cond ((looking-at c-primitive-type-key)
7131 (setq res t))
7132 ((c-with-syntax-table c-identifier-syntax-table
7133 (looking-at c-known-type-key))
7134 (setq res 'known)))
7135 (or (not id-end)
7136 (>= (save-excursion
7137 (save-match-data
7138 (goto-char (match-end 1))
7139 (c-forward-syntactic-ws)
7140 (setq pos (point))))
7141 id-end)
7142 (setq res nil))))
7143 ;; Looking at a primitive or known type identifier. We've
7144 ;; checked for a name first so that we don't go here if the
7145 ;; known type match only is a prefix of another name.
7146
7147 (setq id-end (match-end 1))
7148
7149 (when (and c-record-type-identifiers
7150 (or c-promote-possible-types (eq res t)))
7151 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
7152
7153 (if (and c-opt-type-component-key
7154 (save-match-data
7155 (looking-at c-opt-type-component-key)))
7156 ;; There might be more keywords for the type.
7157 (let (safe-pos)
7158 (c-forward-keyword-clause 1)
7159 (while (progn
7160 (setq safe-pos (point))
7161 (looking-at c-opt-type-component-key))
7162 (when (and c-record-type-identifiers
7163 (looking-at c-primitive-type-key))
7164 (c-record-type-id (cons (match-beginning 1)
7165 (match-end 1))))
7166 (c-forward-keyword-clause 1))
7167 (if (looking-at c-primitive-type-key)
7168 (progn
7169 (when c-record-type-identifiers
7170 (c-record-type-id (cons (match-beginning 1)
7171 (match-end 1))))
7172 (c-forward-keyword-clause 1)
7173 (setq res t))
7174 (goto-char safe-pos)
7175 (setq res 'prefix)))
7176 (unless (save-match-data (c-forward-keyword-clause 1))
7177 (if pos
7178 (goto-char pos)
7179 (goto-char (match-end 1))
7180 (c-forward-syntactic-ws)))))
7181
7182 (name-res
7183 (cond ((eq name-res t)
7184 ;; A normal identifier.
7185 (goto-char id-end)
7186 (if (or res c-promote-possible-types)
7187 (progn
7188 (c-add-type id-start id-end)
7189 (when (and c-record-type-identifiers id-range)
7190 (c-record-type-id id-range))
7191 (unless res
7192 (setq res 'found)))
7193 (setq res (if (c-check-type id-start id-end)
7194 ;; It's an identifier that has been used as
7195 ;; a type somewhere else.
7196 'found
7197 ;; It's an identifier that might be a type.
7198 'maybe))))
7199 ((eq name-res 'template)
7200 ;; A template is sometimes a type.
7201 (goto-char id-end)
7202 (c-forward-syntactic-ws)
7203 (setq res
7204 (if (eq (char-after) ?\()
7205 (if (c-check-type id-start id-end)
7206 ;; It's an identifier that has been used as
7207 ;; a type somewhere else.
7208 'found
7209 ;; It's an identifier that might be a type.
7210 'maybe)
7211 t)))
7212 (t
7213 ;; Otherwise it's an operator identifier, which is not a type.
7214 (goto-char start)
7215 (setq res nil)))))
7216
7217 (when res
7218 ;; Skip trailing type modifiers. If any are found we know it's
7219 ;; a type.
7220 (when c-opt-type-modifier-key
7221 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
7222 (goto-char (match-end 1))
7223 (c-forward-syntactic-ws)
7224 (setq res t)))
7225
7226 ;; Step over any type suffix operator. Do not let the existence
7227 ;; of these alter the classification of the found type, since
7228 ;; these operators typically are allowed in normal expressions
7229 ;; too.
7230 (when c-opt-type-suffix-key ; e.g. "..."
7231 (while (looking-at c-opt-type-suffix-key)
7232 (goto-char (match-end 1))
7233 (c-forward-syntactic-ws)))
7234
7235 (when c-opt-type-concat-key ; Only/mainly for pike.
7236 ;; Look for a trailing operator that concatenates the type
7237 ;; with a following one, and if so step past that one through
7238 ;; a recursive call. Note that we don't record concatenated
7239 ;; types in `c-found-types' - it's the component types that
7240 ;; are recorded when appropriate.
7241 (setq pos (point))
7242 (let* ((c-promote-possible-types (or (memq res '(t known))
7243 c-promote-possible-types))
7244 ;; If we can't promote then set `c-record-found-types' so that
7245 ;; we can merge in the types from the second part afterwards if
7246 ;; it turns out to be a known type there.
7247 (c-record-found-types (and c-record-type-identifiers
7248 (not c-promote-possible-types)))
7249 subres)
7250 (if (and (looking-at c-opt-type-concat-key)
7251
7252 (progn
7253 (goto-char (match-end 1))
7254 (c-forward-syntactic-ws)
7255 (setq subres (c-forward-type))))
7256
7257 (progn
7258 ;; If either operand certainly is a type then both are, but we
7259 ;; don't let the existence of the operator itself promote two
7260 ;; uncertain types to a certain one.
7261 (cond ((eq res t))
7262 ((eq subres t)
7263 (unless (eq name-res 'template)
7264 (c-add-type id-start id-end))
7265 (when (and c-record-type-identifiers id-range)
7266 (c-record-type-id id-range))
7267 (setq res t))
7268 ((eq res 'known))
7269 ((eq subres 'known)
7270 (setq res 'known))
7271 ((eq res 'found))
7272 ((eq subres 'found)
7273 (setq res 'found))
7274 (t
7275 (setq res 'maybe)))
7276
7277 (when (and (eq res t)
7278 (consp c-record-found-types))
7279 ;; Merge in the ranges of any types found by the second
7280 ;; `c-forward-type'.
7281 (setq c-record-type-identifiers
7282 ;; `nconc' doesn't mind that the tail of
7283 ;; `c-record-found-types' is t.
7284 (nconc c-record-found-types
7285 c-record-type-identifiers))))
7286
7287 (goto-char pos))))
7288
7289 (when (and c-record-found-types (memq res '(known found)) id-range)
7290 (setq c-record-found-types
7291 (cons id-range c-record-found-types))))
7292
7293 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
7294
7295 res))
7296
7297 (defun c-forward-annotation ()
7298 ;; Used for Java code only at the moment. Assumes point is on the @, moves
7299 ;; forward an annotation and returns t. Leaves point unmoved and returns
7300 ;; nil if there is no annotation at point.
7301 (let ((pos (point)))
7302 (or
7303 (and (looking-at "@")
7304 (not (looking-at c-keywords-regexp))
7305 (progn (forward-char) t)
7306 (looking-at c-symbol-key)
7307 (progn (goto-char (match-end 0))
7308 (c-forward-syntactic-ws)
7309 t)
7310 (if (looking-at "(")
7311 (c-go-list-forward)
7312 t))
7313 (progn (goto-char pos) nil))))
7314
7315 (defmacro c-pull-open-brace (ps)
7316 ;; Pull the next open brace from PS (which has the form of paren-state),
7317 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
7318 `(progn
7319 (while (consp (car ,ps))
7320 (setq ,ps (cdr ,ps)))
7321 (prog1 (car ,ps)
7322 (setq ,ps (cdr ,ps)))))
7323
7324 (defun c-back-over-compound-identifier ()
7325 ;; Point is putatively just after a "compound identifier", i.e. something
7326 ;; looking (in C++) like this "FQN::of::base::Class". Move to the start of
7327 ;; this construct and return t. If the parsing fails, return nil, leaving
7328 ;; point unchanged.
7329 (let ((here (point))
7330 end
7331 )
7332 (if (not (c-simple-skip-symbol-backward))
7333 nil
7334 (while
7335 (progn
7336 (setq end (point))
7337 (c-backward-syntactic-ws)
7338 (c-backward-token-2)
7339 (and
7340 c-opt-identifier-concat-key
7341 (looking-at c-opt-identifier-concat-key)
7342 (progn
7343 (c-backward-syntactic-ws)
7344 (c-simple-skip-symbol-backward))))
7345 (setq end (point)))
7346 (goto-char end)
7347 t)))
7348
7349 (defun c-back-over-member-initializer-braces ()
7350 ;; Point is just after a closing brace/parenthesis. Try to parse this as a
7351 ;; C++ member initializer list, going back to just after the introducing ":"
7352 ;; and returning t. Otherwise return nil, leaving point unchanged.
7353 (let ((here (point)) res)
7354 (setq res
7355 (catch 'done
7356 (when (not (c-go-list-backward))
7357 (throw 'done nil))
7358 (c-backward-syntactic-ws)
7359 (when (not (c-back-over-compound-identifier))
7360 (throw 'done nil))
7361 (c-backward-syntactic-ws)
7362
7363 (while (eq (char-before) ?,)
7364 (backward-char)
7365 (c-backward-syntactic-ws)
7366 (when (not (memq (char-before) '(?\) ?})))
7367 (throw 'done nil))
7368 (when (not (c-go-list-backward))
7369 (throw 'done nil))
7370 (c-backward-syntactic-ws)
7371 (when (not (c-back-over-compound-identifier))
7372 (throw 'done nil))
7373 (c-backward-syntactic-ws))
7374
7375 (eq (char-before) ?:)))
7376 (or res (goto-char here))
7377 res))
7378
7379 (defmacro c-back-over-list-of-member-inits ()
7380 ;; Go back over a list of elements, each looking like:
7381 ;; <symbol> (<expression>) ,
7382 ;; or <symbol> {<expression>} ,
7383 ;; when we are putatively immediately after a comma. Stop when we don't see
7384 ;; a comma. If either of <symbol> or bracketed <expression> is missing,
7385 ;; throw nil to 'level. If the terminating } or ) is unmatched, throw nil
7386 ;; to 'done. This is not a general purpose macro!
7387 `(while (eq (char-before) ?,)
7388 (backward-char)
7389 (c-backward-syntactic-ws)
7390 (when (not (memq (char-before) '(?\) ?})))
7391 (throw 'level nil))
7392 (when (not (c-go-list-backward))
7393 (throw 'done nil))
7394 (c-backward-syntactic-ws)
7395 (when (not (c-back-over-compound-identifier))
7396 (throw 'level nil))
7397 (c-backward-syntactic-ws)))
7398
7399 (defun c-back-over-member-initializers ()
7400 ;; Test whether we are in a C++ member initializer list, and if so, go back
7401 ;; to the introducing ":", returning the position of the opening paren of
7402 ;; the function's arglist. Otherwise return nil, leaving point unchanged.
7403 (let ((here (point))
7404 (paren-state (c-parse-state))
7405 pos level-plausible at-top-level res)
7406 ;; Assume tentatively that we're at the top level. Try to go back to the
7407 ;; colon we seek.
7408 (setq res
7409 (catch 'done
7410 (setq level-plausible
7411 (catch 'level
7412 (c-backward-syntactic-ws)
7413 (when (memq (char-before) '(?\) ?}))
7414 (when (not (c-go-list-backward))
7415 (throw 'done nil))
7416 (c-backward-syntactic-ws))
7417 (when (c-back-over-compound-identifier)
7418 (c-backward-syntactic-ws))
7419 (c-back-over-list-of-member-inits)
7420 (and (eq (char-before) ?:)
7421 (save-excursion
7422 (c-backward-token-2)
7423 (not (looking-at c-:$-multichar-token-regexp)))
7424 (c-just-after-func-arglist-p))))
7425
7426 (while (and (not (and level-plausible
7427 (setq at-top-level (c-at-toplevel-p))))
7428 (setq pos (c-pull-open-brace paren-state))) ; might be a paren.
7429 (setq level-plausible
7430 (catch 'level
7431 (goto-char pos)
7432 (c-backward-syntactic-ws)
7433 (when (not (c-back-over-compound-identifier))
7434 (throw 'level nil))
7435 (c-backward-syntactic-ws)
7436 (c-back-over-list-of-member-inits)
7437 (and (eq (char-before) ?:)
7438 (save-excursion
7439 (c-backward-token-2)
7440 (not (looking-at c-:$-multichar-token-regexp)))
7441 (c-just-after-func-arglist-p)))))
7442
7443 (and at-top-level level-plausible)))
7444 (or res (goto-char here))
7445 res))
7446
7447 \f
7448 ;; Handling of large scale constructs like statements and declarations.
7449
7450 ;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
7451 ;; defsubst or perhaps even a defun, but it contains lots of free
7452 ;; variables that refer to things inside `c-forward-decl-or-cast-1'.
7453 (defmacro c-fdoc-shift-type-backward (&optional short)
7454 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
7455 ;; of types when parsing a declaration, which means that it
7456 ;; sometimes consumes the identifier in the declaration as a type.
7457 ;; This is used to "backtrack" and make the last type be treated as
7458 ;; an identifier instead.
7459 `(progn
7460 ,(unless short
7461 ;; These identifiers are bound only in the inner let.
7462 '(setq identifier-type at-type
7463 identifier-start type-start
7464 got-parens nil
7465 got-identifier t
7466 got-suffix t
7467 got-suffix-after-parens id-start
7468 paren-depth 0))
7469
7470 (if (setq at-type (if (eq backup-at-type 'prefix)
7471 t
7472 backup-at-type))
7473 (setq type-start backup-type-start
7474 id-start backup-id-start)
7475 (setq type-start start-pos
7476 id-start start-pos))
7477
7478 ;; When these flags already are set we've found specifiers that
7479 ;; unconditionally signal these attributes - backtracking doesn't
7480 ;; change that. So keep them set in that case.
7481 (or at-type-decl
7482 (setq at-type-decl backup-at-type-decl))
7483 (or maybe-typeless
7484 (setq maybe-typeless backup-maybe-typeless))
7485
7486 ,(unless short
7487 ;; This identifier is bound only in the inner let.
7488 '(setq start id-start))))
7489
7490 (defun c-forward-declarator (&optional limit accept-anon)
7491 ;; Assuming point is at the start of a declarator, move forward over it,
7492 ;; leaving point at the next token after it (e.g. a ) or a ; or a ,).
7493 ;;
7494 ;; Return a list (ID-START ID-END BRACKETS-AFTER-ID GOT-INIT), where ID-START and
7495 ;; ID-END are the bounds of the declarator's identifier, and
7496 ;; BRACKETS-AFTER-ID is non-nil if a [...] pair is present after the id.
7497 ;; GOT-INIT is non-nil when the declarator is followed by "=" or "(".
7498 ;;
7499 ;; If ACCEPT-ANON is non-nil, move forward over any "anonymous declarator",
7500 ;; i.e. something like the (*) in int (*), such as might be found in a
7501 ;; declaration. In such a case ID-START and ID-END in the return value are
7502 ;; both set to nil. A "null" "anonymous declarator" gives a non-nil result.
7503 ;;
7504 ;; If no declarator is found, leave point unmoved and return nil. LIMIT is
7505 ;; an optional limit for forward searching.
7506 ;;
7507 ;; Note that the global variable `c-last-identifier-range' is written to, so
7508 ;; the caller should bind it if necessary.
7509
7510 ;; Inside the following "condition form", we move forward over the
7511 ;; declarator's identifier up as far as any opening bracket (for array
7512 ;; size) or paren (for parameters of function-type) or brace (for
7513 ;; array/struct initialization) or "=" or terminating delimiter
7514 ;; (e.g. "," or ";" or "}").
7515 (let ((here (point))
7516 id-start id-end brackets-after-id paren-depth)
7517 (or limit (setq limit (point-max)))
7518 (if (and
7519 (< (point) limit)
7520
7521 ;; The following form moves forward over the declarator's
7522 ;; identifier (and what precedes it), returning t. If there
7523 ;; wasn't one, it returns nil.
7524 (let (got-identifier)
7525 (setq paren-depth 0)
7526 ;; Skip over type decl prefix operators, one for each iteration
7527 ;; of the while. These are, e.g. "*" in "int *foo" or "(" and
7528 ;; "*" in "int (*foo) (void)" (Note similar code in
7529 ;; `c-forward-decl-or-cast-1'.)
7530 (while
7531 (cond
7532 ((looking-at c-decl-hangon-key)
7533 (c-forward-keyword-clause 1))
7534 ((and c-opt-cpp-prefix
7535 (looking-at c-noise-macro-with-parens-name-re))
7536 (c-forward-noise-clause))
7537 ((and (looking-at c-type-decl-prefix-key)
7538 (if (and (c-major-mode-is 'c++-mode)
7539 (match-beginning 3))
7540 ;; If the third submatch matches in C++ then
7541 ;; we're looking at an identifier that's a
7542 ;; prefix only if it specifies a member pointer.
7543 (progn
7544 (setq id-start (point))
7545 (c-forward-name)
7546 (if (looking-at "\\(::\\)")
7547 ;; We only check for a trailing "::" and
7548 ;; let the "*" that should follow be
7549 ;; matched in the next round.
7550 t
7551 ;; It turned out to be the real identifier,
7552 ;; so flag that and stop.
7553 (setq got-identifier t)
7554 nil))
7555 t))
7556 (if (eq (char-after) ?\()
7557 (progn
7558 (setq paren-depth (1+ paren-depth))
7559 (forward-char))
7560 (goto-char (match-end 1)))
7561 (c-forward-syntactic-ws)
7562 t)))
7563
7564 ;; If we haven't passed the identifier already, do it now.
7565 (unless got-identifier
7566 (setq id-start (point)))
7567 (cond
7568 ((or got-identifier
7569 (c-forward-name))
7570 (save-excursion
7571 (c-backward-syntactic-ws)
7572 (setq id-end (point))))
7573 (accept-anon
7574 (setq id-start nil id-end nil)
7575 t)
7576 (t (/= (point) here))))
7577
7578 ;; Skip out of the parens surrounding the identifier. If closing
7579 ;; parens are missing, this form returns nil.
7580 (or (= paren-depth 0)
7581 (c-safe (goto-char (scan-lists (point) 1 paren-depth))))
7582
7583 (<= (point) limit)
7584
7585 ;; Skip over any trailing bit, such as "__attribute__".
7586 (progn
7587 (while (cond
7588 ((looking-at c-decl-hangon-key)
7589 (c-forward-keyword-clause 1))
7590 ((and c-opt-cpp-prefix
7591 (looking-at c-noise-macro-with-parens-name-re))
7592 (c-forward-noise-clause))))
7593 (<= (point) limit))
7594
7595 ;; Search syntactically to the end of the declarator (";",
7596 ;; ",", a closing paren, eob etc) or to the beginning of an
7597 ;; initializer or function prototype ("=" or "\\s\(").
7598 ;; Note that square brackets are now not also treated as
7599 ;; initializers, since this broke when there were also
7600 ;; initializing brace lists.
7601 (let (found)
7602 (while
7603 (and (setq found (c-syntactic-re-search-forward
7604 "[;,]\\|\\s)\\|\\'\\|\\(=\\|\\s(\\)" limit t t))
7605 (eq (char-before) ?\[)
7606 (c-go-up-list-forward))
7607 (setq brackets-after-id t))
7608 (backward-char)
7609 found))
7610 (list id-start id-end brackets-after-id (match-beginning 1))
7611
7612 (goto-char here)
7613 nil)))
7614
7615 (defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
7616 ;; Move forward over a declaration or a cast if at the start of one.
7617 ;; The point is assumed to be at the start of some token. Nil is
7618 ;; returned if no declaration or cast is recognized, and the point
7619 ;; is clobbered in that case.
7620 ;;
7621 ;; If a declaration is parsed:
7622 ;;
7623 ;; The point is left at the first token after the first complete
7624 ;; declarator, if there is one. The return value is a list of 4 elements,
7625 ;; where the first is the position of the first token in the declarator.
7626 ;; (See below for the other three.)
7627 ;; Some examples:
7628 ;;
7629 ;; void foo (int a, char *b) stuff ...
7630 ;; car ^ ^ point
7631 ;; float (*a)[], b;
7632 ;; car ^ ^ point
7633 ;; unsigned int a = c_style_initializer, b;
7634 ;; car ^ ^ point
7635 ;; unsigned int a (cplusplus_style_initializer), b;
7636 ;; car ^ ^ point (might change)
7637 ;; class Foo : public Bar {}
7638 ;; car ^ ^ point
7639 ;; class PikeClass (int a, string b) stuff ...
7640 ;; car ^ ^ point
7641 ;; enum bool;
7642 ;; car ^ ^ point
7643 ;; enum bool flag;
7644 ;; car ^ ^ point
7645 ;; void cplusplus_function (int x) throw (Bad);
7646 ;; car ^ ^ point
7647 ;; Foo::Foo (int b) : Base (b) {}
7648 ;; car ^ ^ point
7649 ;;
7650 ;; auto foo = 5;
7651 ;; car ^ ^ point
7652 ;; auto cplusplus_11 (int a, char *b) -> decltype (bar):
7653 ;; car ^ ^ point
7654 ;;
7655 ;;
7656 ;;
7657 ;; The second element of the return value is non-nil when a
7658 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
7659 ;; Specifically it is a dotted pair (A . B) where B is t when a
7660 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
7661 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
7662 ;; specifier is present. I.e., (some of) the declared
7663 ;; identifier(s) are types.
7664 ;;
7665 ;; The third element of the return value is non-nil when the declaration
7666 ;; parsed might be an expression. The fourth element is the position of
7667 ;; the start of the type identifier.
7668 ;;
7669 ;; If a cast is parsed:
7670 ;;
7671 ;; The point is left at the first token after the closing paren of
7672 ;; the cast. The return value is `cast'. Note that the start
7673 ;; position must be at the first token inside the cast parenthesis
7674 ;; to recognize it.
7675 ;;
7676 ;; PRECEDING-TOKEN-END is the first position after the preceding
7677 ;; token, i.e. on the other side of the syntactic ws from the point.
7678 ;; Use a value less than or equal to (point-min) if the point is at
7679 ;; the first token in (the visible part of) the buffer.
7680 ;;
7681 ;; CONTEXT is a symbol that describes the context at the point:
7682 ;; 'decl In a comma-separated declaration context (typically
7683 ;; inside a function declaration arglist).
7684 ;; '<> In an angle bracket arglist.
7685 ;; 'arglist Some other type of arglist.
7686 ;; nil Some other context or unknown context. Includes
7687 ;; within the parens of an if, for, ... construct.
7688 ;;
7689 ;; LAST-CAST-END is the first token after the closing paren of a
7690 ;; preceding cast, or nil if none is known. If
7691 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
7692 ;; the position after the closest preceding call where a cast was
7693 ;; matched. In that case it's used to discover chains of casts like
7694 ;; "(a) (b) c".
7695 ;;
7696 ;; This function records identifier ranges on
7697 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7698 ;; `c-record-type-identifiers' is non-nil.
7699 ;;
7700 ;; This function might do hidden buffer changes.
7701
7702 (let (;; `start-pos' is used below to point to the start of the
7703 ;; first type, i.e. after any leading specifiers. It might
7704 ;; also point at the beginning of the preceding syntactic
7705 ;; whitespace.
7706 (start-pos (point))
7707 ;; Set to the result of `c-forward-type'.
7708 at-type
7709 ;; The position of the first token in what we currently
7710 ;; believe is the type in the declaration or cast, after any
7711 ;; specifiers and their associated clauses.
7712 type-start
7713 ;; The position of the first token in what we currently
7714 ;; believe is the declarator for the first identifier. Set
7715 ;; when the type is found, and moved forward over any
7716 ;; `c-decl-hangon-kwds' and their associated clauses that
7717 ;; occurs after the type.
7718 id-start
7719 ;; These store `at-type', `type-start' and `id-start' of the
7720 ;; identifier before the one in those variables. The previous
7721 ;; identifier might turn out to be the real type in a
7722 ;; declaration if the last one has to be the declarator in it.
7723 ;; If `backup-at-type' is nil then the other variables have
7724 ;; undefined values.
7725 backup-at-type backup-type-start backup-id-start
7726 ;; This stores `kwd-sym' of the symbol before the current one.
7727 ;; This is needed to distinguish the C++11 version of "auto" from
7728 ;; the pre C++11 meaning.
7729 backup-kwd-sym
7730 ;; Set if we've found a specifier (apart from "typedef") that makes
7731 ;; the defined identifier(s) types.
7732 at-type-decl
7733 ;; Set if we've a "typedef" keyword.
7734 at-typedef
7735 ;; Set if we've found a specifier that can start a declaration
7736 ;; where there's no type.
7737 maybe-typeless
7738 ;; Save the value of kwd-sym between loops of the "Check for a
7739 ;; type" loop. Needed to distinguish a C++11 "auto" from a pre
7740 ;; C++11 one.
7741 prev-kwd-sym
7742 ;; If a specifier is found that also can be a type prefix,
7743 ;; these flags are set instead of those above. If we need to
7744 ;; back up an identifier, they are copied to the real flag
7745 ;; variables. Thus they only take effect if we fail to
7746 ;; interpret it as a type.
7747 backup-at-type-decl backup-maybe-typeless
7748 ;; Whether we've found a declaration or a cast. We might know
7749 ;; this before we've found the type in it. It's 'ids if we've
7750 ;; found two consecutive identifiers (usually a sure sign, but
7751 ;; we should allow that in labels too), and t if we've found a
7752 ;; specifier keyword (a 100% sure sign).
7753 at-decl-or-cast
7754 ;; Set when we need to back up to parse this as a declaration
7755 ;; but not as a cast.
7756 backup-if-not-cast
7757 ;; For casts, the return position.
7758 cast-end
7759 ;; Have we got a new-style C++11 "auto"?
7760 new-style-auto
7761 ;; Set when the symbol before `preceding-token-end' is known to
7762 ;; terminate the previous construct, or when we're at point-min.
7763 at-decl-start
7764 ;; Save `c-record-type-identifiers' and
7765 ;; `c-record-ref-identifiers' since ranges are recorded
7766 ;; speculatively and should be thrown away if it turns out
7767 ;; that it isn't a declaration or cast.
7768 (save-rec-type-ids c-record-type-identifiers)
7769 (save-rec-ref-ids c-record-ref-identifiers)
7770 ;; Set when we parse a declaration which might also be an expression,
7771 ;; such as "a *b". See CASE 16 and CASE 17.
7772 maybe-expression)
7773
7774 (save-excursion
7775 (goto-char preceding-token-end)
7776 (setq at-decl-start
7777 (or (bobp)
7778 (let ((tok-end (point)))
7779 (c-backward-token-2)
7780 (member (buffer-substring-no-properties (point) tok-end)
7781 c-pre-start-tokens)))))
7782
7783 (while (c-forward-annotation)
7784 (c-forward-syntactic-ws))
7785
7786 ;; Check for a type. Unknown symbols are treated as possible
7787 ;; types, but they could also be specifiers disguised through
7788 ;; macros like __INLINE__, so we recognize both types and known
7789 ;; specifiers after them too.
7790 (while
7791 (let* ((start (point)) kwd-sym kwd-clause-end found-type noise-start)
7792
7793 (cond
7794 ;; Look for a specifier keyword clause.
7795 ((or (looking-at c-prefix-spec-kwds-re)
7796 (and (c-major-mode-is 'java-mode)
7797 (looking-at "@[A-Za-z0-9]+")))
7798 (save-match-data
7799 (if (looking-at c-typedef-key)
7800 (setq at-typedef t)))
7801 (setq kwd-sym (c-keyword-sym (match-string 1)))
7802 (save-excursion
7803 (c-forward-keyword-clause 1)
7804 (setq kwd-clause-end (point))))
7805 ((and c-opt-cpp-prefix
7806 (looking-at c-noise-macro-with-parens-name-re))
7807 (setq noise-start (point))
7808 (c-forward-noise-clause)
7809 (setq kwd-clause-end (point))))
7810
7811 (when (setq found-type (c-forward-type t)) ; brace-block-too
7812 ;; Found a known or possible type or a prefix of a known type.
7813 (when (and (c-major-mode-is 'c++-mode) ; C++11 style "auto"?
7814 (eq prev-kwd-sym (c-keyword-sym "auto"))
7815 (looking-at "[=(]")) ; FIXME!!! proper regexp.
7816 (setq new-style-auto t)
7817 (setq found-type nil)
7818 (goto-char start)) ; position of foo in "auto foo"
7819
7820 (when at-type
7821 ;; Got two identifiers with nothing but whitespace
7822 ;; between them. That can only happen in declarations.
7823 (setq at-decl-or-cast 'ids)
7824
7825 (when (eq at-type 'found)
7826 ;; If the previous identifier is a found type we
7827 ;; record it as a real one; it might be some sort of
7828 ;; alias for a prefix like "unsigned".
7829 (save-excursion
7830 (goto-char type-start)
7831 (let ((c-promote-possible-types t))
7832 (c-forward-type)))))
7833
7834 (setq backup-at-type at-type
7835 backup-type-start type-start
7836 backup-id-start id-start
7837 backup-kwd-sym kwd-sym
7838 at-type found-type
7839 type-start start
7840 id-start (point)
7841 ;; The previous ambiguous specifier/type turned out
7842 ;; to be a type since we've parsed another one after
7843 ;; it, so clear these backup flags.
7844 backup-at-type-decl nil
7845 backup-maybe-typeless nil))
7846
7847 (if (or kwd-sym noise-start)
7848 (progn
7849 ;; Handle known specifier keywords and
7850 ;; `c-decl-hangon-kwds' which can occur after known
7851 ;; types.
7852
7853 (if (or (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
7854 noise-start)
7855 ;; It's a hang-on keyword or noise clause that can occur
7856 ;; anywhere.
7857 (progn
7858 (if at-type
7859 ;; Move the identifier start position if
7860 ;; we've passed a type.
7861 (setq id-start kwd-clause-end)
7862 ;; Otherwise treat this as a specifier and
7863 ;; move the fallback position.
7864 (setq start-pos kwd-clause-end))
7865 (goto-char kwd-clause-end))
7866
7867 ;; It's an ordinary specifier so we know that
7868 ;; anything before this can't be the type.
7869 (setq backup-at-type nil
7870 start-pos kwd-clause-end)
7871
7872 (if found-type
7873 ;; It's ambiguous whether this keyword is a
7874 ;; specifier or a type prefix, so set the backup
7875 ;; flags. (It's assumed that `c-forward-type'
7876 ;; moved further than `c-forward-keyword-clause'.)
7877 (progn
7878 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
7879 (setq backup-at-type-decl t))
7880 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
7881 (setq backup-maybe-typeless t)))
7882
7883 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
7884 ;; This test only happens after we've scanned a type.
7885 ;; So, with valid syntax, kwd-sym can't be 'typedef.
7886 (setq at-type-decl t))
7887 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
7888 (setq maybe-typeless t))
7889
7890 ;; Haven't matched a type so it's an unambiguous
7891 ;; specifier keyword and we know we're in a
7892 ;; declaration.
7893 (setq at-decl-or-cast t)
7894 (setq prev-kwd-sym kwd-sym)
7895
7896 (goto-char kwd-clause-end))))
7897
7898 ;; If the type isn't known we continue so that we'll jump
7899 ;; over all specifiers and type identifiers. The reason
7900 ;; to do this for a known type prefix is to make things
7901 ;; like "unsigned INT16" work.
7902 (and found-type (not (eq found-type t))))))
7903
7904 (cond
7905 ((eq at-type t)
7906 ;; If a known type was found, we still need to skip over any
7907 ;; hangon keyword clauses after it. Otherwise it has already
7908 ;; been done in the loop above.
7909 (while
7910 (cond ((looking-at c-decl-hangon-key)
7911 (c-forward-keyword-clause 1))
7912 ((and c-opt-cpp-prefix
7913 (looking-at c-noise-macro-with-parens-name-re))
7914 (c-forward-noise-clause))))
7915 (setq id-start (point)))
7916
7917 ((eq at-type 'prefix)
7918 ;; A prefix type is itself a primitive type when it's not
7919 ;; followed by another type.
7920 (setq at-type t))
7921
7922 ((not at-type)
7923 ;; Got no type but set things up to continue anyway to handle
7924 ;; the various cases when a declaration doesn't start with a
7925 ;; type.
7926 (setq id-start start-pos))
7927
7928 ((and (eq at-type 'maybe)
7929 (c-major-mode-is 'c++-mode))
7930 ;; If it's C++ then check if the last "type" ends on the form
7931 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
7932 ;; (con|de)structor.
7933 (save-excursion
7934 (let (name end-2 end-1)
7935 (goto-char id-start)
7936 (c-backward-syntactic-ws)
7937 (setq end-2 (point))
7938 (when (and
7939 (c-simple-skip-symbol-backward)
7940 (progn
7941 (setq name
7942 (buffer-substring-no-properties (point) end-2))
7943 ;; Cheating in the handling of syntactic ws below.
7944 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
7945 (progn
7946 (setq end-1 (point))
7947 (c-simple-skip-symbol-backward))
7948 (>= (point) type-start)
7949 (equal (buffer-substring-no-properties (point) end-1)
7950 name))
7951 ;; It is a (con|de)structor name. In that case the
7952 ;; declaration is typeless so zap out any preceding
7953 ;; identifier(s) that we might have taken as types.
7954 (goto-char type-start)
7955 (setq at-type nil
7956 backup-at-type nil
7957 id-start type-start))))))
7958
7959 ;; Check for and step over a type decl expression after the thing
7960 ;; that is or might be a type. This can't be skipped since we
7961 ;; need the correct end position of the declarator for
7962 ;; `max-type-decl-end-*'.
7963 (let ((start (point)) (paren-depth 0) pos
7964 ;; True if there's a non-open-paren match of
7965 ;; `c-type-decl-prefix-key'.
7966 got-prefix
7967 ;; True if the declarator is surrounded by a parenthesis pair.
7968 got-parens
7969 ;; True if there is an identifier in the declarator.
7970 got-identifier
7971 ;; True if there's a non-close-paren match of
7972 ;; `c-type-decl-suffix-key'.
7973 got-suffix
7974 ;; True if there's a prefix match outside the outermost
7975 ;; paren pair that surrounds the declarator.
7976 got-prefix-before-parens
7977 ;; True if there's a suffix match outside the outermost
7978 ;; paren pair that surrounds the declarator. The value is
7979 ;; the position of the first suffix match.
7980 got-suffix-after-parens
7981 ;; True if we've parsed the type decl to a token that is
7982 ;; known to end declarations in this context.
7983 at-decl-end
7984 ;; The earlier values of `at-type' and `type-start' if we've
7985 ;; shifted the type backwards.
7986 identifier-type identifier-start
7987 ;; If `c-parse-and-markup-<>-arglists' is set we need to
7988 ;; turn it off during the name skipping below to avoid
7989 ;; getting `c-type' properties that might be bogus. That
7990 ;; can happen since we don't know if
7991 ;; `c-restricted-<>-arglists' will be correct inside the
7992 ;; arglist paren that gets entered.
7993 c-parse-and-markup-<>-arglists
7994 ;; Start of the identifier for which `got-identifier' was set.
7995 name-start)
7996
7997 (goto-char id-start)
7998
7999 ;; Skip over type decl prefix operators. (Note similar code in
8000 ;; `c-forward-declarator'.)
8001 (if (and c-recognize-typeless-decls
8002 (equal c-type-decl-prefix-key "\\<\\>"))
8003 (when (eq (char-after) ?\()
8004 (progn
8005 (setq paren-depth (1+ paren-depth))
8006 (forward-char)))
8007 (while (and (looking-at c-type-decl-prefix-key)
8008 (if (and (c-major-mode-is 'c++-mode)
8009 (match-beginning 3))
8010 ;; If the third submatch matches in C++ then
8011 ;; we're looking at an identifier that's a
8012 ;; prefix only if it specifies a member pointer.
8013 (when (progn (setq pos (point))
8014 (setq got-identifier (c-forward-name)))
8015 (setq name-start pos)
8016 (if (looking-at "\\(::\\)")
8017 ;; We only check for a trailing "::" and
8018 ;; let the "*" that should follow be
8019 ;; matched in the next round.
8020 (progn (setq got-identifier nil) t)
8021 ;; It turned out to be the real identifier,
8022 ;; so stop.
8023 nil))
8024 t))
8025
8026 (if (eq (char-after) ?\()
8027 (progn
8028 (setq paren-depth (1+ paren-depth))
8029 (forward-char))
8030 (unless got-prefix-before-parens
8031 (setq got-prefix-before-parens (= paren-depth 0)))
8032 (setq got-prefix t)
8033 (goto-char (match-end 1)))
8034 (c-forward-syntactic-ws)))
8035
8036 (setq got-parens (> paren-depth 0))
8037
8038 ;; Skip over an identifier.
8039 (or got-identifier
8040 (and (looking-at c-identifier-start)
8041 (setq pos (point))
8042 (setq got-identifier (c-forward-name))
8043 (setq name-start pos)))
8044
8045 ;; Skip over type decl suffix operators and trailing noise macros.
8046 (while
8047 (cond
8048 ((and c-opt-cpp-prefix
8049 (looking-at c-noise-macro-with-parens-name-re))
8050 (c-forward-noise-clause))
8051
8052 ((looking-at c-type-decl-suffix-key)
8053 (if (eq (char-after) ?\))
8054 (when (> paren-depth 0)
8055 (setq paren-depth (1- paren-depth))
8056 (forward-char)
8057 t)
8058 (when (if (save-match-data (looking-at "\\s("))
8059 (c-safe (c-forward-sexp 1) t)
8060 (goto-char (match-end 1))
8061 t)
8062 (when (and (not got-suffix-after-parens)
8063 (= paren-depth 0))
8064 (setq got-suffix-after-parens (match-beginning 0)))
8065 (setq got-suffix t))))
8066
8067 (t
8068 ;; No suffix matched. We might have matched the
8069 ;; identifier as a type and the open paren of a
8070 ;; function arglist as a type decl prefix. In that
8071 ;; case we should "backtrack": Reinterpret the last
8072 ;; type as the identifier, move out of the arglist and
8073 ;; continue searching for suffix operators.
8074 ;;
8075 ;; Do this even if there's no preceding type, to cope
8076 ;; with old style function declarations in K&R C,
8077 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
8078 ;; style declarations. That isn't applicable in an
8079 ;; arglist context, though.
8080 (when (and (= paren-depth 1)
8081 (not got-prefix-before-parens)
8082 (not (eq at-type t))
8083 (or backup-at-type
8084 maybe-typeless
8085 backup-maybe-typeless
8086 (when c-recognize-typeless-decls
8087 (not context)))
8088 (setq pos (c-up-list-forward (point)))
8089 (eq (char-before pos) ?\)))
8090 (c-fdoc-shift-type-backward)
8091 (goto-char pos)
8092 t)))
8093
8094 (c-forward-syntactic-ws))
8095
8096 (when (or (and new-style-auto
8097 (looking-at c-auto-ops-re))
8098 (and (or maybe-typeless backup-maybe-typeless)
8099 (not got-identifier)
8100 (not got-prefix)
8101 at-type))
8102 ;; Have found no identifier but `c-typeless-decl-kwds' has
8103 ;; matched so we know we're inside a declaration. The
8104 ;; preceding type must be the identifier instead.
8105 (c-fdoc-shift-type-backward))
8106
8107 ;; Prepare the "-> type;" for fontification later on.
8108 (when (and new-style-auto
8109 (looking-at c-haskell-op-re))
8110 (save-excursion
8111 (goto-char (match-end 0))
8112 (c-forward-syntactic-ws)
8113 (setq type-start (point))
8114 (setq at-type (c-forward-type))))
8115
8116 (setq
8117 at-decl-or-cast
8118 (catch 'at-decl-or-cast
8119
8120 ;; CASE 1
8121 (when (> paren-depth 0)
8122 ;; Encountered something inside parens that isn't matched by
8123 ;; the `c-type-decl-*' regexps, so it's not a type decl
8124 ;; expression. Try to skip out to the same paren depth to
8125 ;; not confuse the cast check below.
8126 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
8127 ;; If we've found a specifier keyword then it's a
8128 ;; declaration regardless.
8129 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
8130
8131 (setq at-decl-end
8132 (looking-at (cond ((eq context '<>) "[,>]")
8133 (context "[,)]")
8134 (t "[,;]"))))
8135
8136 ;; Now we've collected info about various characteristics of
8137 ;; the construct we're looking at. Below follows a decision
8138 ;; tree based on that. It's ordered to check more certain
8139 ;; signs before less certain ones.
8140
8141 (if got-identifier
8142 (progn
8143
8144 ;; CASE 2
8145 (when (and (or at-type maybe-typeless)
8146 (not (or got-prefix got-parens)))
8147 ;; Got another identifier directly after the type, so it's a
8148 ;; declaration.
8149 (throw 'at-decl-or-cast t))
8150
8151 (when (and got-parens
8152 (not got-prefix)
8153 ;; (not got-suffix-after-parens)
8154 (or backup-at-type
8155 maybe-typeless
8156 backup-maybe-typeless
8157 (eq at-decl-or-cast t)
8158 ;; Check whether we have "bar (gnu);" where we
8159 ;; are directly inside a class (etc.) called "bar".
8160 (save-excursion
8161 (and
8162 (progn
8163 (goto-char name-start)
8164 (not (memq (c-forward-type) '(nil maybe))))
8165 (progn
8166 (goto-char id-start)
8167 (c-directly-in-class-called-p
8168 (buffer-substring
8169 type-start
8170 (progn
8171 (goto-char type-start)
8172 (c-forward-type)
8173 (c-backward-syntactic-ws)
8174 (point)))))))))
8175 ;; Got a declaration of the form "foo bar (gnu);" or "bar
8176 ;; (gnu);" where we've recognized "bar" as the type and "gnu"
8177 ;; as the declarator, and in the latter case, checked that
8178 ;; "bar (gnu)" appears directly inside the class "bar". In
8179 ;; this case it's however more likely that "bar" is the
8180 ;; declarator and "gnu" a function argument or initializer
8181 ;; (if `c-recognize-paren-inits' is set), since the parens
8182 ;; around "gnu" would be superfluous if it's a declarator.
8183 ;; Shift the type one step backward.
8184 (c-fdoc-shift-type-backward)))
8185
8186 ;; Found no identifier.
8187
8188 (if backup-at-type
8189 (progn
8190
8191 ;; CASE 3
8192 (when (= (point) start)
8193 ;; Got a plain list of identifiers. If a colon follows it's
8194 ;; a valid label, or maybe a bitfield. Otherwise the last
8195 ;; one probably is the declared identifier and we should
8196 ;; back up to the previous type, providing it isn't a cast.
8197 (if (and (eq (char-after) ?:)
8198 (not (c-major-mode-is 'java-mode)))
8199 (cond
8200 ;; If we've found a specifier keyword then it's a
8201 ;; declaration regardless.
8202 ((eq at-decl-or-cast t)
8203 (throw 'at-decl-or-cast t))
8204 ((and c-has-bitfields
8205 (eq at-decl-or-cast 'ids)) ; bitfield.
8206 (setq backup-if-not-cast t)
8207 (throw 'at-decl-or-cast t)))
8208
8209 (setq backup-if-not-cast t)
8210 (throw 'at-decl-or-cast t)))
8211
8212 ;; CASE 4
8213 (when (and got-suffix
8214 (not got-prefix)
8215 (not got-parens))
8216 ;; Got a plain list of identifiers followed by some suffix.
8217 ;; If this isn't a cast then the last identifier probably is
8218 ;; the declared one and we should back up to the previous
8219 ;; type.
8220 (setq backup-if-not-cast t)
8221 (throw 'at-decl-or-cast t)))
8222
8223 ;; CASE 5
8224 (when (eq at-type t)
8225 ;; If the type is known we know that there can't be any
8226 ;; identifier somewhere else, and it's only in declarations in
8227 ;; e.g. function prototypes and in casts that the identifier may
8228 ;; be left out.
8229 (throw 'at-decl-or-cast t))
8230
8231 (when (= (point) start)
8232 ;; Only got a single identifier (parsed as a type so far).
8233 ;; CASE 6
8234 (if (and
8235 ;; Check that the identifier isn't at the start of an
8236 ;; expression.
8237 at-decl-end
8238 (cond
8239 ((eq context 'decl)
8240 ;; Inside an arglist that contains declarations. If K&R
8241 ;; style declarations and parenthesis style initializers
8242 ;; aren't allowed then the single identifier must be a
8243 ;; type, else we require that it's known or found
8244 ;; (primitive types are handled above).
8245 (or (and (not c-recognize-knr-p)
8246 (not c-recognize-paren-inits))
8247 (memq at-type '(known found))))
8248 ((eq context '<>)
8249 ;; Inside a template arglist. Accept known and found
8250 ;; types; other identifiers could just as well be
8251 ;; constants in C++.
8252 (memq at-type '(known found)))))
8253 (throw 'at-decl-or-cast t)
8254 ;; CASE 7
8255 ;; Can't be a valid declaration or cast, but if we've found a
8256 ;; specifier it can't be anything else either, so treat it as
8257 ;; an invalid/unfinished declaration or cast.
8258 (throw 'at-decl-or-cast at-decl-or-cast))))
8259
8260 (if (and got-parens
8261 (not got-prefix)
8262 (not context)
8263 (not (eq at-type t))
8264 (or backup-at-type
8265 maybe-typeless
8266 backup-maybe-typeless
8267 (when c-recognize-typeless-decls
8268 (or (not got-suffix)
8269 (not (looking-at
8270 c-after-suffixed-type-maybe-decl-key))))))
8271 ;; Got an empty paren pair and a preceding type that probably
8272 ;; really is the identifier. Shift the type backwards to make
8273 ;; the last one the identifier. This is analogous to the
8274 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
8275 ;; above.
8276 ;;
8277 ;; Exception: In addition to the conditions in that
8278 ;; "backtracking" code, do not shift backward if we're not
8279 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
8280 ;; Since there's no preceding type, the shift would mean that
8281 ;; the declaration is typeless. But if the regexp doesn't match
8282 ;; then we will simply fall through in the tests below and not
8283 ;; recognize it at all, so it's better to try it as an abstract
8284 ;; declarator instead.
8285 (c-fdoc-shift-type-backward)
8286
8287 ;; Still no identifier.
8288 ;; CASE 8
8289 (when (and got-prefix (or got-parens got-suffix))
8290 ;; Require `got-prefix' together with either `got-parens' or
8291 ;; `got-suffix' to recognize it as an abstract declarator:
8292 ;; `got-parens' only is probably an empty function call.
8293 ;; `got-suffix' only can build an ordinary expression together
8294 ;; with the preceding identifier which we've taken as a type.
8295 ;; We could actually accept on `got-prefix' only, but that can
8296 ;; easily occur temporarily while writing an expression so we
8297 ;; avoid that case anyway. We could do a better job if we knew
8298 ;; the point when the fontification was invoked.
8299 (throw 'at-decl-or-cast t))
8300
8301 ;; CASE 9
8302 (when (and at-type
8303 (not got-prefix)
8304 (not got-parens)
8305 got-suffix-after-parens
8306 (eq (char-after got-suffix-after-parens) ?\())
8307 ;; Got a type, no declarator but a paren suffix. I.e. it's a
8308 ;; normal function call after all (or perhaps a C++ style object
8309 ;; instantiation expression).
8310 (throw 'at-decl-or-cast nil))))
8311
8312 ;; CASE 10
8313 (when at-decl-or-cast
8314 ;; By now we've located the type in the declaration that we know
8315 ;; we're in.
8316 (throw 'at-decl-or-cast t))
8317
8318 ;; CASE 11
8319 (when (and got-identifier
8320 (not context)
8321 (looking-at c-after-suffixed-type-decl-key)
8322 (if (and got-parens
8323 (not got-prefix)
8324 (not got-suffix)
8325 (not (eq at-type t)))
8326 ;; Shift the type backward in the case that there's a
8327 ;; single identifier inside parens. That can only
8328 ;; occur in K&R style function declarations so it's
8329 ;; more likely that it really is a function call.
8330 ;; Therefore we only do this after
8331 ;; `c-after-suffixed-type-decl-key' has matched.
8332 (progn (c-fdoc-shift-type-backward) t)
8333 got-suffix-after-parens))
8334 ;; A declaration according to `c-after-suffixed-type-decl-key'.
8335 (throw 'at-decl-or-cast t))
8336
8337 ;; CASE 12
8338 (when (and (or got-prefix (not got-parens))
8339 (memq at-type '(t known)))
8340 ;; It's a declaration if a known type precedes it and it can't be a
8341 ;; function call.
8342 (throw 'at-decl-or-cast t))
8343
8344 ;; If we get here we can't tell if this is a type decl or a normal
8345 ;; expression by looking at it alone. (That's under the assumption
8346 ;; that normal expressions always can look like type decl expressions,
8347 ;; which isn't really true but the cases where it doesn't hold are so
8348 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
8349 ;; the effort to look for them.)
8350
8351 ;;; 2008-04-16: commented out the next form, to allow the function to recognize
8352 ;;; "foo (int bar)" in CC (an implicit type (in class foo) without a semicolon)
8353 ;;; as a(n almost complete) declaration, enabling it to be fontified.
8354 ;; CASE 13
8355 ;; (unless (or at-decl-end (looking-at "=[^=]"))
8356 ;; If this is a declaration it should end here or its initializer(*)
8357 ;; should start here, so check for allowed separation tokens. Note
8358 ;; that this rule doesn't work e.g. with a K&R arglist after a
8359 ;; function header.
8360 ;;
8361 ;; *) Don't check for C++ style initializers using parens
8362 ;; since those already have been matched as suffixes.
8363 ;;
8364 ;; If `at-decl-or-cast' is then we've found some other sign that
8365 ;; it's a declaration or cast, so then it's probably an
8366 ;; invalid/unfinished one.
8367 ;; (throw 'at-decl-or-cast at-decl-or-cast))
8368
8369 ;; Below are tests that only should be applied when we're certain to
8370 ;; not have parsed halfway through an expression.
8371
8372 ;; CASE 14
8373 (when (memq at-type '(t known))
8374 ;; The expression starts with a known type so treat it as a
8375 ;; declaration.
8376 (throw 'at-decl-or-cast t))
8377
8378 ;; CASE 15
8379 (when (and (c-major-mode-is 'c++-mode)
8380 ;; In C++ we check if the identifier is a known type, since
8381 ;; (con|de)structors use the class name as identifier.
8382 ;; We've always shifted over the identifier as a type and
8383 ;; then backed up again in this case.
8384 identifier-type
8385 (or (memq identifier-type '(found known))
8386 (and (eq (char-after identifier-start) ?~)
8387 ;; `at-type' probably won't be 'found for
8388 ;; destructors since the "~" is then part of the
8389 ;; type name being checked against the list of
8390 ;; known types, so do a check without that
8391 ;; operator.
8392 (or (save-excursion
8393 (goto-char (1+ identifier-start))
8394 (c-forward-syntactic-ws)
8395 (c-with-syntax-table
8396 c-identifier-syntax-table
8397 (looking-at c-known-type-key)))
8398 (save-excursion
8399 (goto-char (1+ identifier-start))
8400 ;; We have already parsed the type earlier,
8401 ;; so it'd be possible to cache the end
8402 ;; position instead of redoing it here, but
8403 ;; then we'd need to keep track of another
8404 ;; position everywhere.
8405 (c-check-type (point)
8406 (progn (c-forward-type)
8407 (point))))))))
8408 (throw 'at-decl-or-cast t))
8409
8410 (if got-identifier
8411 (progn
8412 ;; CASE 16
8413 (when (and got-prefix-before-parens
8414 at-type
8415 (or at-decl-end (looking-at "=[^=]"))
8416 (not context)
8417 (or (not got-suffix)
8418 at-decl-start))
8419 ;; Got something like "foo * bar;". Since we're not inside
8420 ;; an arglist it would be a meaningless expression because
8421 ;; the result isn't used. We therefore choose to recognize
8422 ;; it as a declaration. We only allow a suffix (which makes
8423 ;; the construct look like a function call) when
8424 ;; `at-decl-start' provides additional evidence that we do
8425 ;; have a declaration.
8426 (setq maybe-expression t)
8427 (throw 'at-decl-or-cast t))
8428
8429 ;; CASE 17
8430 (when (and (or got-suffix-after-parens
8431 (looking-at "=[^=]"))
8432 (eq at-type 'found)
8433 (not (eq context 'arglist)))
8434 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
8435 ;; be an odd expression or it could be a declaration. Treat
8436 ;; it as a declaration if "a" has been used as a type
8437 ;; somewhere else (if it's a known type we won't get here).
8438 (setq maybe-expression t)
8439 (throw 'at-decl-or-cast t)))
8440
8441 ;; CASE 18
8442 (when (and context
8443 (or got-prefix
8444 (and (eq context 'decl)
8445 (not c-recognize-paren-inits)
8446 (or got-parens got-suffix))))
8447 ;; Got a type followed by an abstract declarator. If `got-prefix'
8448 ;; is set it's something like "a *" without anything after it. If
8449 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
8450 ;; or similar, which we accept only if the context rules out
8451 ;; expressions.
8452 (throw 'at-decl-or-cast t)))
8453
8454 ;; If we had a complete symbol table here (which rules out
8455 ;; `c-found-types') we should return t due to the disambiguation rule
8456 ;; (in at least C++) that anything that can be parsed as a declaration
8457 ;; is a declaration. Now we're being more defensive and prefer to
8458 ;; highlight things like "foo (bar);" as a declaration only if we're
8459 ;; inside an arglist that contains declarations.
8460 ;; CASE 19
8461 (eq context 'decl))))
8462
8463 ;; The point is now after the type decl expression.
8464
8465 (cond
8466 ;; Check for a cast.
8467 ((save-excursion
8468 (and
8469 c-cast-parens
8470
8471 ;; Should be the first type/identifier in a cast paren.
8472 (> preceding-token-end (point-min))
8473 (memq (char-before preceding-token-end) c-cast-parens)
8474
8475 ;; The closing paren should follow.
8476 (progn
8477 (c-forward-syntactic-ws)
8478 (looking-at "\\s)"))
8479
8480 ;; There should be a primary expression after it.
8481 (let (pos)
8482 (forward-char)
8483 (c-forward-syntactic-ws)
8484 (setq cast-end (point))
8485 (and (looking-at c-primary-expr-regexp)
8486 (progn
8487 (setq pos (match-end 0))
8488 (or
8489 ;; Check if the expression begins with a prefix keyword.
8490 (match-beginning 2)
8491 (if (match-beginning 1)
8492 ;; Expression begins with an ambiguous operator. Treat
8493 ;; it as a cast if it's a type decl or if we've
8494 ;; recognized the type somewhere else.
8495 (or at-decl-or-cast
8496 (memq at-type '(t known found)))
8497 ;; Unless it's a keyword, it's the beginning of a primary
8498 ;; expression.
8499 (not (looking-at c-keywords-regexp)))))
8500 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
8501 ;; that it matched a whole one so that we don't e.g. confuse
8502 ;; the operator '-' with '->'. It's ok if it matches further,
8503 ;; though, since it e.g. can match the float '.5' while the
8504 ;; operator regexp only matches '.'.
8505 (or (not (looking-at c-nonsymbol-token-regexp))
8506 (<= (match-end 0) pos))))
8507
8508 ;; There should either be a cast before it or something that isn't an
8509 ;; identifier or close paren.
8510 (> preceding-token-end (point-min))
8511 (progn
8512 (goto-char (1- preceding-token-end))
8513 (or (eq (point) last-cast-end)
8514 (progn
8515 (c-backward-syntactic-ws)
8516 (if (< (skip-syntax-backward "w_") 0)
8517 ;; It's a symbol. Accept it only if it's one of the
8518 ;; keywords that can precede an expression (without
8519 ;; surrounding parens).
8520 (looking-at c-simple-stmt-key)
8521 (and
8522 ;; Check that it isn't a close paren (block close is ok,
8523 ;; though).
8524 (not (memq (char-before) '(?\) ?\])))
8525 ;; Check that it isn't a nonsymbol identifier.
8526 (not (c-on-identifier)))))))))
8527
8528 ;; Handle the cast.
8529 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
8530 (let ((c-promote-possible-types t))
8531 (goto-char type-start)
8532 (c-forward-type)))
8533
8534 (goto-char cast-end)
8535 'cast)
8536
8537 (at-decl-or-cast
8538 ;; We're at a declaration. Highlight the type and the following
8539 ;; declarators.
8540
8541 (when backup-if-not-cast
8542 (c-fdoc-shift-type-backward t))
8543
8544 (when (and (eq context 'decl) (looking-at ","))
8545 ;; Make sure to propagate the `c-decl-arg-start' property to
8546 ;; the next argument if it's set in this one, to cope with
8547 ;; interactive refontification.
8548 (c-put-c-type-property (point) 'c-decl-arg-start))
8549
8550 ;; Record the type's coordinates in `c-record-type-identifiers' for
8551 ;; later fontification.
8552 (when (and c-record-type-identifiers at-type ;; (not (eq at-type t))
8553 ;; There seems no reason to exclude a token from
8554 ;; fontification just because it's "a known type that can't
8555 ;; be a name or other expression". 2013-09-18.
8556 )
8557 (let ((c-promote-possible-types t))
8558 (save-excursion
8559 (goto-char type-start)
8560 (c-forward-type))))
8561
8562 (list id-start
8563 (and (or at-type-decl at-typedef)
8564 (cons at-type-decl at-typedef))
8565 maybe-expression
8566 type-start))
8567
8568 (t
8569 ;; False alarm. Restore the recorded ranges.
8570 (setq c-record-type-identifiers save-rec-type-ids
8571 c-record-ref-identifiers save-rec-ref-ids)
8572 nil))))
8573
8574 (defun c-forward-label (&optional assume-markup preceding-token-end limit)
8575 ;; Assuming that point is at the beginning of a token, check if it starts a
8576 ;; label and if so move over it and return non-nil (t in default situations,
8577 ;; specific symbols (see below) for interesting situations), otherwise don't
8578 ;; move and return nil. "Label" here means "most things with a colon".
8579 ;;
8580 ;; More precisely, a "label" is regarded as one of:
8581 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
8582 ;; (ii) A case label - either the entire construct "case FOO:", or just the
8583 ;; bare "case", should the colon be missing. We return t;
8584 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
8585 ;; return t;
8586 ;; (iv) One of QT's "extended" C++ variants of
8587 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
8588 ;; Returns the symbol `qt-2kwds-colon'.
8589 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
8590 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
8591 ;; colon). Currently (2006-03), this applies only to Objective C's
8592 ;; keywords "@private", "@protected", and "@public". Returns t.
8593 ;;
8594 ;; One of the things which will NOT be recognized as a label is a bit-field
8595 ;; element of a struct, something like "int foo:5".
8596 ;;
8597 ;; The end of the label is taken to be just after the colon, or the end of
8598 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
8599 ;; after the end on return. The terminating char gets marked with
8600 ;; `c-decl-end' to improve recognition of the following declaration or
8601 ;; statement.
8602 ;;
8603 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
8604 ;; label, if any, has already been marked up like that.
8605 ;;
8606 ;; If PRECEDING-TOKEN-END is given, it should be the first position
8607 ;; after the preceding token, i.e. on the other side of the
8608 ;; syntactic ws from the point. Use a value less than or equal to
8609 ;; (point-min) if the point is at the first token in (the visible
8610 ;; part of) the buffer.
8611 ;;
8612 ;; The optional LIMIT limits the forward scan for the colon.
8613 ;;
8614 ;; This function records the ranges of the label symbols on
8615 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
8616 ;; non-nil.
8617 ;;
8618 ;; This function might do hidden buffer changes.
8619
8620 (let ((start (point))
8621 label-end
8622 qt-symbol-idx
8623 macro-start ; if we're in one.
8624 label-type
8625 kwd)
8626 (cond
8627 ;; "case" or "default" (Doesn't apply to AWK).
8628 ((looking-at c-label-kwds-regexp)
8629 (let ((kwd-end (match-end 1)))
8630 ;; Record only the keyword itself for fontification, since in
8631 ;; case labels the following is a constant expression and not
8632 ;; a label.
8633 (when c-record-type-identifiers
8634 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
8635
8636 ;; Find the label end.
8637 (goto-char kwd-end)
8638 (setq label-type
8639 (if (and (c-syntactic-re-search-forward
8640 ;; Stop on chars that aren't allowed in expressions,
8641 ;; and on operator chars that would be meaningless
8642 ;; there. FIXME: This doesn't cope with ?: operators.
8643 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
8644 limit t t nil 1)
8645 (match-beginning 2))
8646
8647 (progn ; there's a proper :
8648 (goto-char (match-beginning 2)) ; just after the :
8649 (c-put-c-type-property (1- (point)) 'c-decl-end)
8650 t)
8651
8652 ;; It's an unfinished label. We consider the keyword enough
8653 ;; to recognize it as a label, so that it gets fontified.
8654 ;; Leave the point at the end of it, but don't put any
8655 ;; `c-decl-end' marker.
8656 (goto-char kwd-end)
8657 t))))
8658
8659 ;; @private, @protected, @public, in Objective C, or similar.
8660 ((and c-opt-extra-label-key
8661 (looking-at c-opt-extra-label-key))
8662 ;; For a `c-opt-extra-label-key' match, we record the whole
8663 ;; thing for fontification. That's to get the leading '@' in
8664 ;; Objective-C protection labels fontified.
8665 (goto-char (match-end 1))
8666 (when c-record-type-identifiers
8667 (c-record-ref-id (cons (match-beginning 1) (point))))
8668 (c-put-c-type-property (1- (point)) 'c-decl-end)
8669 (setq label-type t))
8670
8671 ;; All other cases of labels.
8672 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
8673
8674 ;; A colon label must have something before the colon.
8675 (not (eq (char-after) ?:))
8676
8677 ;; Check that we're not after a token that can't precede a label.
8678 (or
8679 ;; Trivially succeeds when there's no preceding token.
8680 ;; Succeeds when we're at a virtual semicolon.
8681 (if preceding-token-end
8682 (<= preceding-token-end (point-min))
8683 (save-excursion
8684 (c-backward-syntactic-ws)
8685 (setq preceding-token-end (point))
8686 (or (bobp)
8687 (c-at-vsemi-p))))
8688
8689 ;; Check if we're after a label, if we're after a closing
8690 ;; paren that belong to statement, and with
8691 ;; `c-label-prefix-re'. It's done in different order
8692 ;; depending on `assume-markup' since the checks have
8693 ;; different expensiveness.
8694 (if assume-markup
8695 (or
8696 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
8697 'c-decl-end)
8698
8699 (save-excursion
8700 (goto-char (1- preceding-token-end))
8701 (c-beginning-of-current-token)
8702 (or (looking-at c-label-prefix-re)
8703 (looking-at c-block-stmt-1-key)))
8704
8705 (and (eq (char-before preceding-token-end) ?\))
8706 (c-after-conditional)))
8707
8708 (or
8709 (save-excursion
8710 (goto-char (1- preceding-token-end))
8711 (c-beginning-of-current-token)
8712 (or (looking-at c-label-prefix-re)
8713 (looking-at c-block-stmt-1-key)))
8714
8715 (cond
8716 ((eq (char-before preceding-token-end) ?\))
8717 (c-after-conditional))
8718
8719 ((eq (char-before preceding-token-end) ?:)
8720 ;; Might be after another label, so check it recursively.
8721 (save-restriction
8722 (save-excursion
8723 (goto-char (1- preceding-token-end))
8724 ;; Essentially the same as the
8725 ;; `c-syntactic-re-search-forward' regexp below.
8726 (setq macro-start
8727 (save-excursion (and (c-beginning-of-macro)
8728 (point))))
8729 (if macro-start (narrow-to-region macro-start (point-max)))
8730 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
8731 ;; Note: the following should work instead of the
8732 ;; narrow-to-region above. Investigate why not,
8733 ;; sometime. ACM, 2006-03-31.
8734 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
8735 ;; macro-start t)
8736 (let ((pte (point))
8737 ;; If the caller turned on recording for us,
8738 ;; it shouldn't apply when we check the
8739 ;; preceding label.
8740 c-record-type-identifiers)
8741 ;; A label can't start at a cpp directive. Check for
8742 ;; this, since c-forward-syntactic-ws would foul up on it.
8743 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
8744 (c-forward-syntactic-ws)
8745 (c-forward-label nil pte start))))))))))
8746
8747 ;; Point is still at the beginning of the possible label construct.
8748 ;;
8749 ;; Check that the next nonsymbol token is ":", or that we're in one
8750 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
8751 ;; arguments. FIXME: Should build this regexp from the language
8752 ;; constants.
8753 (cond
8754 ;; public: protected: private:
8755 ((and
8756 (c-major-mode-is 'c++-mode)
8757 (search-forward-regexp
8758 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
8759 (progn (backward-char)
8760 (c-forward-syntactic-ws limit)
8761 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
8762 (forward-char)
8763 (setq label-type t))
8764 ;; QT double keyword like "protected slots:" or goto target.
8765 ((progn (goto-char start) nil))
8766 ((when (c-syntactic-re-search-forward
8767 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
8768 (backward-char)
8769 (setq label-end (point))
8770 (setq qt-symbol-idx
8771 (and (c-major-mode-is 'c++-mode)
8772 (string-match
8773 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
8774 (buffer-substring start (point)))))
8775 (c-forward-syntactic-ws limit)
8776 (cond
8777 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
8778 (forward-char)
8779 (setq label-type
8780 (if (or (string= "signals" ; Special QT macro
8781 (setq kwd (buffer-substring-no-properties start label-end)))
8782 (string= "Q_SIGNALS" kwd))
8783 'qt-1kwd-colon
8784 'goto-target)))
8785 ((and qt-symbol-idx
8786 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
8787 (progn (c-forward-syntactic-ws limit)
8788 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
8789 (forward-char)
8790 (setq label-type 'qt-2kwds-colon)))))))
8791
8792 (save-restriction
8793 (narrow-to-region start (point))
8794
8795 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
8796 (catch 'check-label
8797 (goto-char start)
8798 (while (progn
8799 (when (looking-at c-nonlabel-token-key)
8800 (goto-char start)
8801 (setq label-type nil)
8802 (throw 'check-label nil))
8803 (and (c-safe (c-forward-sexp)
8804 (c-forward-syntactic-ws)
8805 t)
8806 (not (eobp)))))
8807
8808 ;; Record the identifiers in the label for fontification, unless
8809 ;; it begins with `c-label-kwds' in which case the following
8810 ;; identifiers are part of a (constant) expression that
8811 ;; shouldn't be fontified.
8812 (when (and c-record-type-identifiers
8813 (progn (goto-char start)
8814 (not (looking-at c-label-kwds-regexp))))
8815 (while (c-syntactic-re-search-forward c-symbol-key nil t)
8816 (c-record-ref-id (cons (match-beginning 0)
8817 (match-end 0)))))
8818
8819 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
8820 (goto-char (point-max)))))
8821
8822 (t
8823 ;; Not a label.
8824 (goto-char start)))
8825 label-type))
8826
8827 (defun c-forward-objc-directive ()
8828 ;; Assuming the point is at the beginning of a token, try to move
8829 ;; forward to the end of the Objective-C directive that starts
8830 ;; there. Return t if a directive was fully recognized, otherwise
8831 ;; the point is moved as far as one could be successfully parsed and
8832 ;; nil is returned.
8833 ;;
8834 ;; This function records identifier ranges on
8835 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
8836 ;; `c-record-type-identifiers' is non-nil.
8837 ;;
8838 ;; This function might do hidden buffer changes.
8839
8840 (let ((start (point))
8841 start-char
8842 (c-promote-possible-types t)
8843 lim
8844 ;; Turn off recognition of angle bracket arglists while parsing
8845 ;; types here since the protocol reference list might then be
8846 ;; considered part of the preceding name or superclass-name.
8847 c-recognize-<>-arglists)
8848
8849 (if (or
8850 (when (looking-at
8851 (eval-when-compile
8852 (c-make-keywords-re t
8853 (append (c-lang-const c-protection-kwds objc)
8854 '("@end"))
8855 'objc-mode)))
8856 (goto-char (match-end 1))
8857 t)
8858
8859 (and
8860 (looking-at
8861 (eval-when-compile
8862 (c-make-keywords-re t
8863 '("@interface" "@implementation" "@protocol")
8864 'objc-mode)))
8865
8866 ;; Handle the name of the class itself.
8867 (progn
8868 ;; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
8869 ;; at EOB.
8870 (goto-char (match-end 0))
8871 (setq lim (point))
8872 (c-skip-ws-forward)
8873 (c-forward-type))
8874
8875 (catch 'break
8876 ;; Look for ": superclass-name" or "( category-name )".
8877 (when (looking-at "[:(]")
8878 (setq start-char (char-after))
8879 (forward-char)
8880 (c-forward-syntactic-ws)
8881 (unless (c-forward-type) (throw 'break nil))
8882 (when (eq start-char ?\()
8883 (unless (eq (char-after) ?\)) (throw 'break nil))
8884 (forward-char)
8885 (c-forward-syntactic-ws)))
8886
8887 ;; Look for a protocol reference list.
8888 (if (eq (char-after) ?<)
8889 (let ((c-recognize-<>-arglists t)
8890 (c-parse-and-markup-<>-arglists t)
8891 c-restricted-<>-arglists)
8892 (c-forward-<>-arglist t))
8893 t))))
8894
8895 (progn
8896 (c-backward-syntactic-ws lim)
8897 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
8898 (c-put-c-type-property (1- (point)) 'c-decl-end)
8899 t)
8900
8901 (c-clear-c-type-property start (point) 'c-decl-end)
8902 nil)))
8903
8904 (defun c-beginning-of-inheritance-list (&optional lim)
8905 ;; Go to the first non-whitespace after the colon that starts a
8906 ;; multiple inheritance introduction. Optional LIM is the farthest
8907 ;; back we should search.
8908 ;;
8909 ;; This function might do hidden buffer changes.
8910 (c-with-syntax-table c++-template-syntax-table
8911 (c-backward-token-2 0 t lim)
8912 (while (and (or (looking-at c-symbol-start)
8913 (looking-at "[<,]\\|::"))
8914 (zerop (c-backward-token-2 1 t lim))))))
8915
8916 (defun c-in-method-def-p ()
8917 ;; Return nil if we aren't in a method definition, otherwise the
8918 ;; position of the initial [+-].
8919 ;;
8920 ;; This function might do hidden buffer changes.
8921 (save-excursion
8922 (beginning-of-line)
8923 (and c-opt-method-key
8924 (looking-at c-opt-method-key)
8925 (point))
8926 ))
8927
8928 ;; Contributed by Kevin Ryde <user42@zip.com.au>.
8929 (defun c-in-gcc-asm-p ()
8930 ;; Return non-nil if point is within a gcc \"asm\" block.
8931 ;;
8932 ;; This should be called with point inside an argument list.
8933 ;;
8934 ;; Only one level of enclosing parentheses is considered, so for
8935 ;; instance nil is returned when in a function call within an asm
8936 ;; operand.
8937 ;;
8938 ;; This function might do hidden buffer changes.
8939
8940 (and c-opt-asm-stmt-key
8941 (save-excursion
8942 (beginning-of-line)
8943 (backward-up-list 1)
8944 (c-beginning-of-statement-1 (point-min) nil t)
8945 (looking-at c-opt-asm-stmt-key))))
8946
8947 (defun c-at-toplevel-p ()
8948 "Return a determination as to whether point is \"at the top level\".
8949 Informally, \"at the top level\" is anywhere where you can write
8950 a function.
8951
8952 More precisely, being at the top-level means that point is either
8953 outside any enclosing block (such as a function definition), or
8954 directly inside a class, namespace or other block that contains
8955 another declaration level.
8956
8957 If point is not at the top-level (e.g. it is inside a method
8958 definition), then nil is returned. Otherwise, if point is at a
8959 top-level not enclosed within a class definition, t is returned.
8960 Otherwise, a 2-vector is returned where the zeroth element is the
8961 buffer position of the start of the class declaration, and the first
8962 element is the buffer position of the enclosing class's opening
8963 brace.
8964
8965 Note that this function might do hidden buffer changes. See the
8966 comment at the start of cc-engine.el for more info."
8967 ;; Note to maintainers: this function consumes a great mass of CPU cycles.
8968 ;; Its use should thus be minimized as far as possible.
8969 (let ((paren-state (c-parse-state)))
8970 (or (not (c-most-enclosing-brace paren-state))
8971 (c-search-uplist-for-classkey paren-state))))
8972
8973 (defun c-just-after-func-arglist-p (&optional lim)
8974 ;; Return non-nil if the point is in the region after the argument
8975 ;; list of a function and its opening brace (or semicolon in case it
8976 ;; got no body). If there are K&R style argument declarations in
8977 ;; that region, the point has to be inside the first one for this
8978 ;; function to recognize it.
8979 ;;
8980 ;; If successful, the point is moved to the first token after the
8981 ;; function header (see `c-forward-decl-or-cast-1' for details) and
8982 ;; the position of the opening paren of the function arglist is
8983 ;; returned.
8984 ;;
8985 ;; The point is clobbered if not successful.
8986 ;;
8987 ;; LIM is used as bound for backward buffer searches.
8988 ;;
8989 ;; This function might do hidden buffer changes.
8990
8991 (let ((beg (point)) id-start)
8992 (and
8993 (eq (c-beginning-of-statement-1 lim) 'same)
8994
8995 (not (and (c-major-mode-is 'objc-mode)
8996 (c-forward-objc-directive)))
8997
8998 (setq id-start
8999 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
9000 (< id-start beg)
9001
9002 ;; There should not be a '=' or ',' between beg and the
9003 ;; start of the declaration since that means we were in the
9004 ;; "expression part" of the declaration.
9005 (or (> (point) beg)
9006 (not (looking-at "[=,]")))
9007
9008 (save-excursion
9009 ;; Check that there's an arglist paren in the
9010 ;; declaration.
9011 (goto-char id-start)
9012 (cond ((eq (char-after) ?\()
9013 ;; The declarator is a paren expression, so skip past it
9014 ;; so that we don't get stuck on that instead of the
9015 ;; function arglist.
9016 (c-forward-sexp))
9017 ((and c-opt-op-identifier-prefix
9018 (looking-at c-opt-op-identifier-prefix))
9019 ;; Don't trip up on "operator ()".
9020 (c-forward-token-2 2 t)))
9021 (and (< (point) beg)
9022 (c-syntactic-re-search-forward "(" beg t t)
9023 (1- (point)))))))
9024
9025 (defun c-in-knr-argdecl (&optional lim)
9026 ;; Return the position of the first argument declaration if point is
9027 ;; inside a K&R style argument declaration list, nil otherwise.
9028 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
9029 ;; position that bounds the backward search for the argument list. This
9030 ;; function doesn't move point.
9031 ;;
9032 ;; Point must be within a possible K&R region, e.g. just before a top-level
9033 ;; "{". It must be outside of parens and brackets. The test can return
9034 ;; false positives otherwise.
9035 ;;
9036 ;; This function might do hidden buffer changes.
9037 (save-excursion
9038 (save-restriction
9039 ;; If we're in a macro, our search range is restricted to it. Narrow to
9040 ;; the searchable range.
9041 (let* ((macro-start (save-excursion (and (c-beginning-of-macro) (point))))
9042 (macro-end (save-excursion (and macro-start (c-end-of-macro) (point))))
9043 (low-lim (max (or lim (point-min)) (or macro-start (point-min))))
9044 before-lparen after-rparen
9045 (here (point))
9046 (pp-count-out 20) ; Max number of paren/brace constructs before
9047 ; we give up.
9048 ids ; List of identifiers in the parenthesized list.
9049 id-start after-prec-token decl-or-cast decl-res
9050 c-last-identifier-range identifier-ok)
9051 (narrow-to-region low-lim (or macro-end (point-max)))
9052
9053 ;; Search backwards for the defun's argument list. We give up if we
9054 ;; encounter a "}" (end of a previous defun) an "=" (which can't be in
9055 ;; a knr region) or BOB.
9056 ;;
9057 ;; The criterion for a paren structure being the arg list is:
9058 ;; o - there is non-WS stuff after it but before any "{"; AND
9059 ;; o - the token after it isn't a ";" AND
9060 ;; o - it is preceded by either an identifier (the function name) or
9061 ;; a macro expansion like "DEFUN (...)"; AND
9062 ;; o - its content is a non-empty comma-separated list of identifiers
9063 ;; (an empty arg list won't have a knr region).
9064 ;;
9065 ;; The following snippet illustrates these rules:
9066 ;; int foo (bar, baz, yuk)
9067 ;; int bar [] ;
9068 ;; int (*baz) (my_type) ;
9069 ;; int (*(* yuk) (void)) (void) ;
9070 ;; {
9071 ;;
9072 ;; Additionally, for a knr list to be recognized:
9073 ;; o - The identifier of each declarator up to and including the
9074 ;; one "near" point must be contained in the arg list.
9075
9076 (catch 'knr
9077 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
9078 (setq pp-count-out (1- pp-count-out))
9079 (c-syntactic-skip-backward "^)]}=")
9080 (cond ((eq (char-before) ?\))
9081 (setq after-rparen (point)))
9082 ((eq (char-before) ?\])
9083 (setq after-rparen nil))
9084 (t ; either } (hit previous defun) or = or no more
9085 ; parens/brackets.
9086 (throw 'knr nil)))
9087
9088 (if after-rparen
9089 ;; We're inside a paren. Could it be our argument list....?
9090 (if
9091 (and
9092 (progn
9093 (goto-char after-rparen)
9094 (unless (c-go-list-backward) (throw 'knr nil)) ;
9095 ;; FIXME!!! What about macros between the parens? 2007/01/20
9096 (setq before-lparen (point)))
9097
9098 ;; It can't be the arg list if next token is ; or {
9099 (progn (goto-char after-rparen)
9100 (c-forward-syntactic-ws)
9101 (not (memq (char-after) '(?\; ?\{ ?\=))))
9102
9103 ;; Is the thing preceding the list an identifier (the
9104 ;; function name), or a macro expansion?
9105 (progn
9106 (goto-char before-lparen)
9107 (eq (c-backward-token-2) 0)
9108 (or (eq (c-on-identifier) (point))
9109 (and (eq (char-after) ?\))
9110 (c-go-up-list-backward)
9111 (eq (c-backward-token-2) 0)
9112 (eq (c-on-identifier) (point)))))
9113
9114 ;; Have we got a non-empty list of comma-separated
9115 ;; identifiers?
9116 (progn
9117 (goto-char before-lparen)
9118 (c-forward-token-2) ; to first token inside parens
9119 (and
9120 (setq id-start (c-on-identifier)) ; Must be at least one.
9121 (catch 'id-list
9122 (while
9123 (progn
9124 (forward-char)
9125 (c-end-of-current-token)
9126 (push (buffer-substring-no-properties id-start
9127 (point))
9128 ids)
9129 (c-forward-syntactic-ws)
9130 (eq (char-after) ?\,))
9131 (c-forward-token-2)
9132 (unless (setq id-start (c-on-identifier))
9133 (throw 'id-list nil)))
9134 (eq (char-after) ?\)))))
9135
9136 ;; Are all the identifiers in the k&r list up to the
9137 ;; current one also in the argument list?
9138 (progn
9139 (forward-char) ; over the )
9140 (setq after-prec-token after-rparen)
9141 (c-forward-syntactic-ws)
9142 (while (and
9143 (or (consp (setq decl-or-cast
9144 (c-forward-decl-or-cast-1
9145 after-prec-token
9146 nil ; Or 'arglist ???
9147 nil)))
9148 (progn
9149 (goto-char after-prec-token)
9150 (c-forward-syntactic-ws)
9151 (setq identifier-ok (eq (char-after) ?{))
9152 nil))
9153 (eq (char-after) ?\;)
9154 (setq after-prec-token (1+ (point)))
9155 (goto-char (car decl-or-cast))
9156 (setq decl-res (c-forward-declarator))
9157 (setq identifier-ok
9158 (member (buffer-substring-no-properties
9159 (car decl-res) (cadr decl-res))
9160 ids))
9161 (progn
9162 (goto-char after-prec-token)
9163 (prog1 (< (point) here)
9164 (c-forward-syntactic-ws))))
9165 (setq identifier-ok nil))
9166 identifier-ok))
9167 ;; ...Yes. We've identified the function's argument list.
9168 (throw 'knr
9169 (progn (goto-char after-rparen)
9170 (c-forward-syntactic-ws)
9171 (point)))
9172 ;; ...No. The current parens aren't the function's arg list.
9173 (goto-char before-lparen))
9174
9175 (or (c-go-list-backward) ; backwards over [ .... ]
9176 (throw 'knr nil)))))))))
9177
9178 (defun c-skip-conditional ()
9179 ;; skip forward over conditional at point, including any predicate
9180 ;; statements in parentheses. No error checking is performed.
9181 ;;
9182 ;; This function might do hidden buffer changes.
9183 (c-forward-sexp (cond
9184 ;; else if()
9185 ((looking-at (concat "\\<else"
9186 "\\([ \t\n]\\|\\\\\n\\)+"
9187 "if\\>\\([^_]\\|$\\)"))
9188 3)
9189 ;; do, else, try, finally
9190 ((looking-at (concat "\\<\\("
9191 "do\\|else\\|try\\|finally"
9192 "\\)\\>\\([^_]\\|$\\)"))
9193 1)
9194 ;; for, if, while, switch, catch, synchronized, foreach
9195 (t 2))))
9196
9197 (defun c-after-conditional (&optional lim)
9198 ;; If looking at the token after a conditional then return the
9199 ;; position of its start, otherwise return nil.
9200 ;;
9201 ;; This function might do hidden buffer changes.
9202 (save-excursion
9203 (and (zerop (c-backward-token-2 1 t lim))
9204 (or (looking-at c-block-stmt-1-key)
9205 (and (eq (char-after) ?\()
9206 (zerop (c-backward-token-2 1 t lim))
9207 (or (looking-at c-block-stmt-2-key)
9208 (looking-at c-block-stmt-1-2-key))))
9209 (point))))
9210
9211 (defun c-after-special-operator-id (&optional lim)
9212 ;; If the point is after an operator identifier that isn't handled
9213 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
9214 ;; position of the start of that identifier is returned. nil is
9215 ;; returned otherwise. The point may be anywhere in the syntactic
9216 ;; whitespace after the last token of the operator identifier.
9217 ;;
9218 ;; This function might do hidden buffer changes.
9219 (save-excursion
9220 (and c-overloadable-operators-regexp
9221 (zerop (c-backward-token-2 1 nil lim))
9222 (looking-at c-overloadable-operators-regexp)
9223 (or (not c-opt-op-identifier-prefix)
9224 (and
9225 (zerop (c-backward-token-2 1 nil lim))
9226 (looking-at c-opt-op-identifier-prefix)))
9227 (point))))
9228
9229 (defsubst c-backward-to-block-anchor (&optional lim)
9230 ;; Assuming point is at a brace that opens a statement block of some
9231 ;; kind, move to the proper anchor point for that block. It might
9232 ;; need to be adjusted further by c-add-stmt-syntax, but the
9233 ;; position at return is suitable as start position for that
9234 ;; function.
9235 ;;
9236 ;; This function might do hidden buffer changes.
9237 (unless (= (point) (c-point 'boi))
9238 (let ((start (c-after-conditional lim)))
9239 (if start
9240 (goto-char start)))))
9241
9242 (defsubst c-backward-to-decl-anchor (&optional lim)
9243 ;; Assuming point is at a brace that opens the block of a top level
9244 ;; declaration of some kind, move to the proper anchor point for
9245 ;; that block.
9246 ;;
9247 ;; This function might do hidden buffer changes.
9248 (unless (= (point) (c-point 'boi))
9249 (c-beginning-of-statement-1 lim)))
9250
9251 (defun c-search-decl-header-end ()
9252 ;; Search forward for the end of the "header" of the current
9253 ;; declaration. That's the position where the definition body
9254 ;; starts, or the first variable initializer, or the ending
9255 ;; semicolon. I.e. search forward for the closest following
9256 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
9257 ;; _after_ the first found token, or at point-max if none is found.
9258 ;;
9259 ;; This function might do hidden buffer changes.
9260
9261 (let ((base (point)))
9262 (if (c-major-mode-is 'c++-mode)
9263
9264 ;; In C++ we need to take special care to handle operator
9265 ;; tokens and those pesky template brackets.
9266 (while (and
9267 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
9268 (or
9269 (c-end-of-current-token base)
9270 ;; Handle operator identifiers, i.e. ignore any
9271 ;; operator token preceded by "operator".
9272 (save-excursion
9273 (and (c-safe (c-backward-sexp) t)
9274 (looking-at c-opt-op-identifier-prefix)))
9275 (and (eq (char-before) ?<)
9276 (c-with-syntax-table c++-template-syntax-table
9277 (if (c-safe (goto-char (c-up-list-forward (point))))
9278 t
9279 (goto-char (point-max))
9280 nil)))))
9281 (setq base (point)))
9282
9283 (while (and
9284 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
9285 (c-end-of-current-token base))
9286 (setq base (point))))))
9287
9288 (defun c-beginning-of-decl-1 (&optional lim)
9289 ;; Go to the beginning of the current declaration, or the beginning
9290 ;; of the previous one if already at the start of it. Point won't
9291 ;; be moved out of any surrounding paren. Return a cons cell of the
9292 ;; form (MOVE . KNR-POS). MOVE is like the return value from
9293 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
9294 ;; style argument declarations (and they are to be recognized) then
9295 ;; KNR-POS is set to the start of the first such argument
9296 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
9297 ;; position that bounds the backward search.
9298 ;;
9299 ;; NB: Cases where the declaration continues after the block, as in
9300 ;; "struct foo { ... } bar;", are currently recognized as two
9301 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
9302 ;;
9303 ;; This function might do hidden buffer changes.
9304 (catch 'return
9305 (let* ((start (point))
9306 (last-stmt-start (point))
9307 (move (c-beginning-of-statement-1 lim nil t)))
9308
9309 ;; `c-beginning-of-statement-1' stops at a block start, but we
9310 ;; want to continue if the block doesn't begin a top level
9311 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
9312 ;; or an open paren.
9313 (let ((beg (point)) tentative-move)
9314 ;; Go back one "statement" each time round the loop until we're just
9315 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
9316 ;; an ObjC method. This will move over a multiple declaration whose
9317 ;; components are comma separated.
9318 (while (and
9319 ;; Must check with c-opt-method-key in ObjC mode.
9320 (not (and c-opt-method-key
9321 (looking-at c-opt-method-key)))
9322 (/= last-stmt-start (point))
9323 (progn
9324 (c-backward-syntactic-ws lim)
9325 (not (or (memq (char-before) '(?\; ?} ?: nil))
9326 (c-at-vsemi-p))))
9327 (save-excursion
9328 (backward-char)
9329 (not (looking-at "\\s(")))
9330 ;; Check that we don't move from the first thing in a
9331 ;; macro to its header.
9332 (not (eq (setq tentative-move
9333 (c-beginning-of-statement-1 lim nil t))
9334 'macro)))
9335 (setq last-stmt-start beg
9336 beg (point)
9337 move tentative-move))
9338 (goto-char beg))
9339
9340 (when c-recognize-knr-p
9341 (let ((fallback-pos (point)) knr-argdecl-start)
9342 ;; Handle K&R argdecls. Back up after the "statement" jumped
9343 ;; over by `c-beginning-of-statement-1', unless it was the
9344 ;; function body, in which case we're sitting on the opening
9345 ;; brace now. Then test if we're in a K&R argdecl region and
9346 ;; that we started at the other side of the first argdecl in
9347 ;; it.
9348 (unless (eq (char-after) ?{)
9349 (goto-char last-stmt-start))
9350 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
9351 (< knr-argdecl-start start)
9352 (progn
9353 (goto-char knr-argdecl-start)
9354 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
9355 (throw 'return
9356 (cons (if (eq (char-after fallback-pos) ?{)
9357 'previous
9358 'same)
9359 knr-argdecl-start))
9360 (goto-char fallback-pos))))
9361
9362 ;; `c-beginning-of-statement-1' counts each brace block as a separate
9363 ;; statement, so the result will be 'previous if we've moved over any.
9364 ;; So change our result back to 'same if necessary.
9365 ;;
9366 ;; If they were brace list initializers we might not have moved over a
9367 ;; declaration boundary though, so change it to 'same if we've moved
9368 ;; past a '=' before '{', but not ';'. (This ought to be integrated
9369 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
9370 ;; potentially can search over a large amount of text.). Take special
9371 ;; pains not to get mislead by C++'s "operator=", and the like.
9372 (if (and (eq move 'previous)
9373 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
9374 c++-template-syntax-table
9375 (syntax-table))
9376 (save-excursion
9377 (and
9378 (progn
9379 (while ; keep going back to "[;={"s until we either find
9380 ; no more, or get to one which isn't an "operator ="
9381 (and (c-syntactic-re-search-forward "[;={]" start t t t)
9382 (eq (char-before) ?=)
9383 c-overloadable-operators-regexp
9384 c-opt-op-identifier-prefix
9385 (save-excursion
9386 (eq (c-backward-token-2) 0)
9387 (looking-at c-overloadable-operators-regexp)
9388 (eq (c-backward-token-2) 0)
9389 (looking-at c-opt-op-identifier-prefix))))
9390 (eq (char-before) ?=))
9391 (c-syntactic-re-search-forward "[;{]" start t t)
9392 (eq (char-before) ?{)
9393 (c-safe (goto-char (c-up-list-forward (point))) t)
9394 (not (c-syntactic-re-search-forward ";" start t t))))))
9395 (cons 'same nil)
9396 (cons move nil)))))
9397
9398 (defun c-end-of-decl-1 ()
9399 ;; Assuming point is at the start of a declaration (as detected by
9400 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
9401 ;; `c-beginning-of-decl-1', this function handles the case when a
9402 ;; block is followed by identifiers in e.g. struct declarations in C
9403 ;; or C++. If a proper end was found then t is returned, otherwise
9404 ;; point is moved as far as possible within the current sexp and nil
9405 ;; is returned. This function doesn't handle macros; use
9406 ;; `c-end-of-macro' instead in those cases.
9407 ;;
9408 ;; This function might do hidden buffer changes.
9409 (let ((start (point))
9410 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
9411 c++-template-syntax-table
9412 (syntax-table))))
9413 (catch 'return
9414 (c-search-decl-header-end)
9415
9416 (when (and c-recognize-knr-p
9417 (eq (char-before) ?\;)
9418 (c-in-knr-argdecl start))
9419 ;; Stopped at the ';' in a K&R argdecl section which is
9420 ;; detected using the same criteria as in
9421 ;; `c-beginning-of-decl-1'. Move to the following block
9422 ;; start.
9423 (c-syntactic-re-search-forward "{" nil 'move t))
9424
9425 (when (eq (char-before) ?{)
9426 ;; Encountered a block in the declaration. Jump over it.
9427 (condition-case nil
9428 (goto-char (c-up-list-forward (point)))
9429 (error (goto-char (point-max))
9430 (throw 'return nil)))
9431 (if (or (not c-opt-block-decls-with-vars-key)
9432 (save-excursion
9433 (c-with-syntax-table decl-syntax-table
9434 (let ((lim (point)))
9435 (goto-char start)
9436 (not (and
9437 ;; Check for `c-opt-block-decls-with-vars-key'
9438 ;; before the first paren.
9439 (c-syntactic-re-search-forward
9440 (concat "[;=([{]\\|\\("
9441 c-opt-block-decls-with-vars-key
9442 "\\)")
9443 lim t t t)
9444 (match-beginning 1)
9445 (not (eq (char-before) ?_))
9446 ;; Check that the first following paren is
9447 ;; the block.
9448 (c-syntactic-re-search-forward "[;=([{]"
9449 lim t t t)
9450 (eq (char-before) ?{)))))))
9451 ;; The declaration doesn't have any of the
9452 ;; `c-opt-block-decls-with-vars' keywords in the
9453 ;; beginning, so it ends here at the end of the block.
9454 (throw 'return t)))
9455
9456 (c-with-syntax-table decl-syntax-table
9457 (while (progn
9458 (if (eq (char-before) ?\;)
9459 (throw 'return t))
9460 (c-syntactic-re-search-forward ";" nil 'move t))))
9461 nil)))
9462
9463 (defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
9464 ;; Assuming the point is at an open brace, check if it starts a
9465 ;; block that contains another declaration level, i.e. that isn't a
9466 ;; statement block or a brace list, and if so return non-nil.
9467 ;;
9468 ;; If the check is successful, the return value is the start of the
9469 ;; keyword that tells what kind of construct it is, i.e. typically
9470 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
9471 ;; the point will be at the start of the construct, before any
9472 ;; leading specifiers, otherwise it's at the returned position.
9473 ;;
9474 ;; The point is clobbered if the check is unsuccessful.
9475 ;;
9476 ;; CONTAINING-SEXP is the position of the open of the surrounding
9477 ;; paren, or nil if none.
9478 ;;
9479 ;; The optional LIMIT limits the backward search for the start of
9480 ;; the construct. It's assumed to be at a syntactically relevant
9481 ;; position.
9482 ;;
9483 ;; If any template arglists are found in the searched region before
9484 ;; the open brace, they get marked with paren syntax.
9485 ;;
9486 ;; This function might do hidden buffer changes.
9487
9488 (let ((open-brace (point)) kwd-start first-specifier-pos)
9489 (c-syntactic-skip-backward c-block-prefix-charset limit t)
9490
9491 (when (and c-recognize-<>-arglists
9492 (eq (char-before) ?>))
9493 ;; Could be at the end of a template arglist.
9494 (let ((c-parse-and-markup-<>-arglists t))
9495 (while (and
9496 (c-backward-<>-arglist nil limit)
9497 (progn
9498 (c-syntactic-skip-backward c-block-prefix-charset limit t)
9499 (eq (char-before) ?>))))))
9500
9501 ;; Skip back over noise clauses.
9502 (while (and
9503 c-opt-cpp-prefix
9504 (eq (char-before) ?\))
9505 (let ((after-paren (point)))
9506 (if (and (c-go-list-backward)
9507 (progn (c-backward-syntactic-ws)
9508 (c-simple-skip-symbol-backward))
9509 (or (looking-at c-paren-nontype-key)
9510 (looking-at c-noise-macro-with-parens-name-re)))
9511 (progn
9512 (c-syntactic-skip-backward c-block-prefix-charset limit t)
9513 t)
9514 (goto-char after-paren)
9515 nil))))
9516
9517 ;; Note: Can't get bogus hits inside template arglists below since they
9518 ;; have gotten paren syntax above.
9519 (when (and
9520 ;; If `goto-start' is set we begin by searching for the
9521 ;; first possible position of a leading specifier list.
9522 ;; The `c-decl-block-key' search continues from there since
9523 ;; we know it can't match earlier.
9524 (if goto-start
9525 (when (c-syntactic-re-search-forward c-symbol-start
9526 open-brace t t)
9527 (goto-char (setq first-specifier-pos (match-beginning 0)))
9528 t)
9529 t)
9530
9531 (cond
9532 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
9533 (goto-char (setq kwd-start (match-beginning 0)))
9534 (and
9535 ;; Exclude cases where we matched what would ordinarily
9536 ;; be a block declaration keyword, except where it's not
9537 ;; legal because it's part of a "compound keyword" like
9538 ;; "enum class". Of course, if c-after-brace-list-key
9539 ;; is nil, we can skip the test.
9540 (or (equal c-after-brace-list-key "\\<\\>")
9541 (save-match-data
9542 (save-excursion
9543 (not
9544 (and
9545 (looking-at c-after-brace-list-key)
9546 (= (c-backward-token-2 1 t) 0)
9547 (looking-at c-brace-list-key))))))
9548 (or
9549 ;; Found a keyword that can't be a type?
9550 (match-beginning 1)
9551
9552 ;; Can be a type too, in which case it's the return type of a
9553 ;; function (under the assumption that no declaration level
9554 ;; block construct starts with a type).
9555 (not (c-forward-type))
9556
9557 ;; Jumped over a type, but it could be a declaration keyword
9558 ;; followed by the declared identifier that we've jumped over
9559 ;; instead (e.g. in "class Foo {"). If it indeed is a type
9560 ;; then we should be at the declarator now, so check for a
9561 ;; valid declarator start.
9562 ;;
9563 ;; Note: This doesn't cope with the case when a declared
9564 ;; identifier is followed by e.g. '(' in a language where '('
9565 ;; also might be part of a declarator expression. Currently
9566 ;; there's no such language.
9567 (not (or (looking-at c-symbol-start)
9568 (looking-at c-type-decl-prefix-key))))))
9569
9570 ;; In Pike a list of modifiers may be followed by a brace
9571 ;; to make them apply to many identifiers. Note that the
9572 ;; match data will be empty on return in this case.
9573 ((and (c-major-mode-is 'pike-mode)
9574 (progn
9575 (goto-char open-brace)
9576 (= (c-backward-token-2) 0))
9577 (looking-at c-specifier-key)
9578 ;; Use this variant to avoid yet another special regexp.
9579 (c-keyword-member (c-keyword-sym (match-string 1))
9580 'c-modifier-kwds))
9581 (setq kwd-start (point))
9582 t)))
9583
9584 ;; Got a match.
9585
9586 (if goto-start
9587 ;; Back up over any preceding specifiers and their clauses
9588 ;; by going forward from `first-specifier-pos', which is the
9589 ;; earliest possible position where the specifier list can
9590 ;; start.
9591 (progn
9592 (goto-char first-specifier-pos)
9593
9594 (while (< (point) kwd-start)
9595 (if (looking-at c-symbol-key)
9596 ;; Accept any plain symbol token on the ground that
9597 ;; it's a specifier masked through a macro (just
9598 ;; like `c-forward-decl-or-cast-1' skip forward over
9599 ;; such tokens).
9600 ;;
9601 ;; Could be more restrictive wrt invalid keywords,
9602 ;; but that'd only occur in invalid code so there's
9603 ;; no use spending effort on it.
9604 (let ((end (match-end 0)))
9605 (unless (c-forward-keyword-clause 0)
9606 (goto-char end)
9607 (c-forward-syntactic-ws)))
9608
9609 ;; Can't parse a declaration preamble and is still
9610 ;; before `kwd-start'. That means `first-specifier-pos'
9611 ;; was in some earlier construct. Search again.
9612 (if (c-syntactic-re-search-forward c-symbol-start
9613 kwd-start 'move t)
9614 (goto-char (setq first-specifier-pos (match-beginning 0)))
9615 ;; Got no preamble before the block declaration keyword.
9616 (setq first-specifier-pos kwd-start))))
9617
9618 (goto-char first-specifier-pos))
9619 (goto-char kwd-start))
9620
9621 kwd-start)))
9622
9623 (defun c-directly-in-class-called-p (name)
9624 ;; Check whether point is directly inside a brace block which is the brace
9625 ;; block of a class, struct, or union which is called NAME, a string.
9626 (let* ((paren-state (c-parse-state))
9627 (brace-pos (c-pull-open-brace paren-state))
9628 )
9629 (when (eq (char-after brace-pos) ?{)
9630 (goto-char brace-pos)
9631 (save-excursion
9632 ; *c-looking-at-decl-block
9633 ; containing-sexp goto-start &optional
9634 ; limit)
9635 (when (and (c-looking-at-decl-block
9636 (c-pull-open-brace paren-state)
9637 nil)
9638 (looking-at c-class-key))
9639 (goto-char (match-end 1))
9640 (c-forward-syntactic-ws)
9641 (looking-at name))))))
9642
9643 (defun c-search-uplist-for-classkey (paren-state)
9644 ;; Check if the closest containing paren sexp is a declaration
9645 ;; block, returning a 2 element vector in that case. Aref 0
9646 ;; contains the bufpos at boi of the class key line, and aref 1
9647 ;; contains the bufpos of the open brace. This function is an
9648 ;; obsolete wrapper for `c-looking-at-decl-block'.
9649 ;;
9650 ;; This function might do hidden buffer changes.
9651 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
9652 (when open-paren-pos
9653 (save-excursion
9654 (goto-char open-paren-pos)
9655 (when (and (eq (char-after) ?{)
9656 (c-looking-at-decl-block
9657 (c-safe-position open-paren-pos paren-state)
9658 nil))
9659 (back-to-indentation)
9660 (vector (point) open-paren-pos))))))
9661
9662 (defun c-most-enclosing-decl-block (paren-state)
9663 ;; Return the buffer position of the most enclosing decl-block brace (in the
9664 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
9665 ;; none was found.
9666 (let* ((open-brace (c-pull-open-brace paren-state))
9667 (next-open-brace (c-pull-open-brace paren-state)))
9668 (while (and open-brace
9669 (save-excursion
9670 (goto-char open-brace)
9671 (not (c-looking-at-decl-block next-open-brace nil))))
9672 (setq open-brace next-open-brace
9673 next-open-brace (c-pull-open-brace paren-state)))
9674 open-brace))
9675
9676 (defun c-cheap-inside-bracelist-p (paren-state)
9677 ;; Return the position of the L-brace if point is inside a brace list
9678 ;; initialization of an array, etc. This is an approximate function,
9679 ;; designed for speed over accuracy. It will not find every bracelist, but
9680 ;; a non-nil result is reliable. We simply search for "= {" (naturally with
9681 ;; syntactic whitespace allowed). PAREN-STATE is the normal thing that it
9682 ;; is everywhere else.
9683 (let (b-pos)
9684 (save-excursion
9685 (while
9686 (and (setq b-pos (c-pull-open-brace paren-state))
9687 (progn (goto-char b-pos)
9688 (c-backward-sws)
9689 (c-backward-token-2)
9690 (not (looking-at "=")))))
9691 b-pos)))
9692
9693 (defun c-backward-typed-enum-colon ()
9694 ;; We're at a "{" which might be the opening brace of a enum which is
9695 ;; strongly typed (by a ":" followed by a type). If this is the case, leave
9696 ;; point before the colon and return t. Otherwise leave point unchanged and return nil.
9697 ;; Match data will be clobbered.
9698 (let ((here (point))
9699 (colon-pos nil))
9700 (save-excursion
9701 (while
9702 (and (eql (c-backward-token-2) 0)
9703 (or (not (looking-at "\\s)"))
9704 (c-go-up-list-backward))
9705 (cond
9706 ((and (eql (char-after) ?:)
9707 (save-excursion
9708 (c-backward-syntactic-ws)
9709 (c-on-identifier)))
9710 (setq colon-pos (point))
9711 (forward-char)
9712 (c-forward-syntactic-ws)
9713 (or (and (c-forward-type)
9714 (progn (c-forward-syntactic-ws)
9715 (eq (point) here)))
9716 (setq colon-pos nil))
9717 nil)
9718 ((eql (char-after) ?\()
9719 t)
9720 ((looking-at c-symbol-key)
9721 t)
9722 (t nil)))))
9723 (when colon-pos
9724 (goto-char colon-pos)
9725 t)))
9726
9727 (defun c-backward-over-enum-header ()
9728 ;; We're at a "{". Move back to the enum-like keyword that starts this
9729 ;; declaration and return t, otherwise don't move and return nil.
9730 (let ((here (point))
9731 up-sexp-pos before-identifier)
9732 (when c-recognize-post-brace-list-type-p
9733 (c-backward-typed-enum-colon))
9734 (while
9735 (and
9736 (eq (c-backward-token-2) 0)
9737 (or (not (looking-at "\\s)"))
9738 (c-go-up-list-backward))
9739 (cond
9740 ((and (looking-at c-symbol-key) (c-on-identifier)
9741 (not before-identifier))
9742 (setq before-identifier t))
9743 ((and before-identifier
9744 (or (eql (char-after) ?,)
9745 (looking-at c-postfix-decl-spec-key)))
9746 (setq before-identifier nil)
9747 t)
9748 ((looking-at c-after-brace-list-key) t)
9749 ((looking-at c-brace-list-key) nil)
9750 ((eq (char-after) ?\()
9751 (and (eq (c-backward-token-2) 0)
9752 (or (looking-at c-decl-hangon-key)
9753 (and c-opt-cpp-prefix
9754 (looking-at c-noise-macro-with-parens-name-re)))))
9755
9756 ((and c-recognize-<>-arglists
9757 (eq (char-after) ?<)
9758 (looking-at "\\s("))
9759 t)
9760 (t nil))))
9761 (or (looking-at c-brace-list-key)
9762 (progn (goto-char here) nil))))
9763
9764 (defun c-inside-bracelist-p (containing-sexp paren-state)
9765 ;; return the buffer position of the beginning of the brace list
9766 ;; statement if we're inside a brace list, otherwise return nil.
9767 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
9768 ;; paren. PAREN-STATE is the remainder of the state of enclosing
9769 ;; braces
9770 ;;
9771 ;; N.B.: This algorithm can potentially get confused by cpp macros
9772 ;; placed in inconvenient locations. It's a trade-off we make for
9773 ;; speed.
9774 ;;
9775 ;; This function might do hidden buffer changes.
9776 (or
9777 ;; This will pick up brace list declarations.
9778 (save-excursion
9779 (goto-char containing-sexp)
9780 (c-backward-over-enum-header))
9781 ;; this will pick up array/aggregate init lists, even if they are nested.
9782 (save-excursion
9783 (let ((class-key
9784 ;; Pike can have class definitions anywhere, so we must
9785 ;; check for the class key here.
9786 (and (c-major-mode-is 'pike-mode)
9787 c-decl-block-key))
9788 bufpos braceassignp lim next-containing macro-start)
9789 (while (and (not bufpos)
9790 containing-sexp)
9791 (when paren-state
9792 (if (consp (car paren-state))
9793 (setq lim (cdr (car paren-state))
9794 paren-state (cdr paren-state))
9795 (setq lim (car paren-state)))
9796 (when paren-state
9797 (setq next-containing (car paren-state)
9798 paren-state (cdr paren-state))))
9799 (goto-char containing-sexp)
9800 (if (c-looking-at-inexpr-block next-containing next-containing)
9801 ;; We're in an in-expression block of some kind. Do not
9802 ;; check nesting. We deliberately set the limit to the
9803 ;; containing sexp, so that c-looking-at-inexpr-block
9804 ;; doesn't check for an identifier before it.
9805 (setq containing-sexp nil)
9806 ;; see if the open brace is preceded by = or [...] in
9807 ;; this statement, but watch out for operator=
9808 (setq braceassignp 'dontknow)
9809 (c-backward-token-2 1 t lim)
9810 ;; Checks to do only on the first sexp before the brace.
9811 (when (and c-opt-inexpr-brace-list-key
9812 (eq (char-after) ?\[))
9813 ;; In Java, an initialization brace list may follow
9814 ;; directly after "new Foo[]", so check for a "new"
9815 ;; earlier.
9816 (while (eq braceassignp 'dontknow)
9817 (setq braceassignp
9818 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
9819 ((looking-at c-opt-inexpr-brace-list-key) t)
9820 ((looking-at "\\sw\\|\\s_\\|[.[]")
9821 ;; Carry on looking if this is an
9822 ;; identifier (may contain "." in Java)
9823 ;; or another "[]" sexp.
9824 'dontknow)
9825 (t nil)))))
9826 ;; Checks to do on all sexps before the brace, up to the
9827 ;; beginning of the statement.
9828 (while (eq braceassignp 'dontknow)
9829 (cond ((eq (char-after) ?\;)
9830 (setq braceassignp nil))
9831 ((and class-key
9832 (looking-at class-key))
9833 (setq braceassignp nil))
9834 ((eq (char-after) ?=)
9835 ;; We've seen a =, but must check earlier tokens so
9836 ;; that it isn't something that should be ignored.
9837 (setq braceassignp 'maybe)
9838 (while (and (eq braceassignp 'maybe)
9839 (zerop (c-backward-token-2 1 t lim)))
9840 (setq braceassignp
9841 (cond
9842 ;; Check for operator =
9843 ((and c-opt-op-identifier-prefix
9844 (looking-at c-opt-op-identifier-prefix))
9845 nil)
9846 ;; Check for `<opchar>= in Pike.
9847 ((and (c-major-mode-is 'pike-mode)
9848 (or (eq (char-after) ?`)
9849 ;; Special case for Pikes
9850 ;; `[]=, since '[' is not in
9851 ;; the punctuation class.
9852 (and (eq (char-after) ?\[)
9853 (eq (char-before) ?`))))
9854 nil)
9855 ((looking-at "\\s.") 'maybe)
9856 ;; make sure we're not in a C++ template
9857 ;; argument assignment
9858 ((and
9859 (c-major-mode-is 'c++-mode)
9860 (save-excursion
9861 (let ((here (point))
9862 (pos< (progn
9863 (skip-chars-backward "^<>")
9864 (point))))
9865 (and (eq (char-before) ?<)
9866 (not (c-crosses-statement-barrier-p
9867 pos< here))
9868 (not (c-in-literal))
9869 ))))
9870 nil)
9871 (t t))))))
9872 (if (and (eq braceassignp 'dontknow)
9873 (/= (c-backward-token-2 1 t lim) 0))
9874 (setq braceassignp nil)))
9875 (cond
9876 (braceassignp
9877 ;; We've hit the beginning of the aggregate list.
9878 (c-beginning-of-statement-1
9879 (c-most-enclosing-brace paren-state))
9880 (setq bufpos (point)))
9881 ((eq (char-after) ?\;)
9882 ;; Brace lists can't contain a semicolon, so we're done.
9883 (setq containing-sexp nil))
9884 ((and (setq macro-start (point))
9885 (c-forward-to-cpp-define-body)
9886 (eq (point) containing-sexp))
9887 ;; We've a macro whose expansion starts with the '{'.
9888 ;; Heuristically, if we have a ';' in it we've not got a
9889 ;; brace list, otherwise we have.
9890 (let ((macro-end (progn (c-end-of-macro) (point))))
9891 (goto-char containing-sexp)
9892 (forward-char)
9893 (if (and (c-syntactic-re-search-forward "[;,]" macro-end t t)
9894 (eq (char-before) ?\;))
9895 (setq bufpos nil
9896 containing-sexp nil)
9897 (setq bufpos macro-start))))
9898 (t
9899 ;; Go up one level
9900 (setq containing-sexp next-containing
9901 lim nil
9902 next-containing nil)))))
9903
9904 bufpos))
9905 ))
9906
9907 (defun c-looking-at-special-brace-list (&optional lim)
9908 ;; If we're looking at the start of a pike-style list, i.e., `({ })',
9909 ;; `([ ])', `(< >)', etc., a cons of a cons of its starting and ending
9910 ;; positions and its entry in c-special-brace-lists is returned, nil
9911 ;; otherwise. The ending position is nil if the list is still open.
9912 ;; LIM is the limit for forward search. The point may either be at
9913 ;; the `(' or at the following paren character. Tries to check the
9914 ;; matching closer, but assumes it's correct if no balanced paren is
9915 ;; found (i.e. the case `({ ... } ... )' is detected as _not_ being
9916 ;; a special brace list).
9917 ;;
9918 ;; This function might do hidden buffer changes.
9919 (if c-special-brace-lists
9920 (condition-case ()
9921 (save-excursion
9922 (let ((beg (point))
9923 inner-beg end type)
9924 (c-forward-syntactic-ws)
9925 (if (eq (char-after) ?\()
9926 (progn
9927 (forward-char 1)
9928 (c-forward-syntactic-ws)
9929 (setq inner-beg (point))
9930 (setq type (assq (char-after) c-special-brace-lists)))
9931 (if (setq type (assq (char-after) c-special-brace-lists))
9932 (progn
9933 (setq inner-beg (point))
9934 (c-backward-syntactic-ws)
9935 (forward-char -1)
9936 (setq beg (if (eq (char-after) ?\()
9937 (point)
9938 nil)))))
9939 (if (and beg type)
9940 (if (and (c-safe
9941 (goto-char beg)
9942 (c-forward-sexp 1)
9943 (setq end (point))
9944 (= (char-before) ?\)))
9945 (c-safe
9946 (goto-char inner-beg)
9947 (if (looking-at "\\s(")
9948 ;; Check balancing of the inner paren
9949 ;; below.
9950 (progn
9951 (c-forward-sexp 1)
9952 t)
9953 ;; If the inner char isn't a paren then
9954 ;; we can't check balancing, so just
9955 ;; check the char before the outer
9956 ;; closing paren.
9957 (goto-char end)
9958 (backward-char)
9959 (c-backward-syntactic-ws)
9960 (= (char-before) (cdr type)))))
9961 (if (or (/= (char-syntax (char-before)) ?\))
9962 (= (progn
9963 (c-forward-syntactic-ws)
9964 (point))
9965 (1- end)))
9966 (cons (cons beg end) type))
9967 (cons (list beg) type)))))
9968 (error nil))))
9969
9970 (defun c-looking-at-bos (&optional lim)
9971 ;; Return non-nil if between two statements or declarations, assuming
9972 ;; point is not inside a literal or comment.
9973 ;;
9974 ;; Obsolete - `c-at-statement-start-p' or `c-at-expression-start-p'
9975 ;; are recommended instead.
9976 ;;
9977 ;; This function might do hidden buffer changes.
9978 (c-at-statement-start-p))
9979 (make-obsolete 'c-looking-at-bos 'c-at-statement-start-p "22.1")
9980
9981 (defun c-looking-at-inexpr-block (lim containing-sexp &optional check-at-end)
9982 ;; Return non-nil if we're looking at the beginning of a block
9983 ;; inside an expression. The value returned is actually a cons of
9984 ;; either 'inlambda, 'inexpr-statement or 'inexpr-class and the
9985 ;; position of the beginning of the construct.
9986 ;;
9987 ;; LIM limits the backward search. CONTAINING-SEXP is the start
9988 ;; position of the closest containing list. If it's nil, the
9989 ;; containing paren isn't used to decide whether we're inside an
9990 ;; expression or not. If both LIM and CONTAINING-SEXP are used, LIM
9991 ;; needs to be farther back.
9992 ;;
9993 ;; If CHECK-AT-END is non-nil then extra checks at the end of the
9994 ;; brace block might be done. It should only be used when the
9995 ;; construct can be assumed to be complete, i.e. when the original
9996 ;; starting position was further down than that.
9997 ;;
9998 ;; This function might do hidden buffer changes.
9999
10000 (save-excursion
10001 (let ((res 'maybe) passed-paren
10002 (closest-lim (or containing-sexp lim (point-min)))
10003 ;; Look at the character after point only as a last resort
10004 ;; when we can't disambiguate.
10005 (block-follows (and (eq (char-after) ?{) (point))))
10006
10007 (while (and (eq res 'maybe)
10008 (progn (c-backward-syntactic-ws)
10009 (> (point) closest-lim))
10010 (not (bobp))
10011 (progn (backward-char)
10012 (looking-at "[]).]\\|\\w\\|\\s_"))
10013 (c-safe (forward-char)
10014 (goto-char (scan-sexps (point) -1))))
10015
10016 (setq res
10017 (if (looking-at c-keywords-regexp)
10018 (let ((kw-sym (c-keyword-sym (match-string 1))))
10019 (cond
10020 ((and block-follows
10021 (c-keyword-member kw-sym 'c-inexpr-class-kwds))
10022 (and (not (eq passed-paren ?\[))
10023 (or (not (looking-at c-class-key))
10024 ;; If the class definition is at the start of
10025 ;; a statement, we don't consider it an
10026 ;; in-expression class.
10027 (let ((prev (point)))
10028 (while (and
10029 (= (c-backward-token-2 1 nil closest-lim) 0)
10030 (eq (char-syntax (char-after)) ?w))
10031 (setq prev (point)))
10032 (goto-char prev)
10033 (not (c-at-statement-start-p)))
10034 ;; Also, in Pike we treat it as an
10035 ;; in-expression class if it's used in an
10036 ;; object clone expression.
10037 (save-excursion
10038 (and check-at-end
10039 (c-major-mode-is 'pike-mode)
10040 (progn (goto-char block-follows)
10041 (zerop (c-forward-token-2 1 t)))
10042 (eq (char-after) ?\())))
10043 (cons 'inexpr-class (point))))
10044 ((c-keyword-member kw-sym 'c-inexpr-block-kwds)
10045 (when (not passed-paren)
10046 (cons 'inexpr-statement (point))))
10047 ((c-keyword-member kw-sym 'c-lambda-kwds)
10048 (when (or (not passed-paren)
10049 (eq passed-paren ?\())
10050 (cons 'inlambda (point))))
10051 ((c-keyword-member kw-sym 'c-block-stmt-kwds)
10052 nil)
10053 (t
10054 'maybe)))
10055
10056 (if (looking-at "\\s(")
10057 (if passed-paren
10058 (if (and (eq passed-paren ?\[)
10059 (eq (char-after) ?\[))
10060 ;; Accept several square bracket sexps for
10061 ;; Java array initializations.
10062 'maybe)
10063 (setq passed-paren (char-after))
10064 'maybe)
10065 'maybe))))
10066
10067 (if (eq res 'maybe)
10068 (when (and c-recognize-paren-inexpr-blocks
10069 block-follows
10070 containing-sexp
10071 (eq (char-after containing-sexp) ?\())
10072 (goto-char containing-sexp)
10073 (if (or (save-excursion
10074 (c-backward-syntactic-ws lim)
10075 (while (and (eq (char-before) ?>)
10076 (c-get-char-property (1- (point))
10077 'syntax-table)
10078 (c-go-list-backward nil lim))
10079 (c-backward-syntactic-ws lim))
10080 (and (> (point) (or lim (point-min)))
10081 (c-on-identifier)))
10082 (and c-special-brace-lists
10083 (c-looking-at-special-brace-list)))
10084 nil
10085 (cons 'inexpr-statement (point))))
10086
10087 res))))
10088
10089 (defun c-looking-at-inexpr-block-backward (paren-state)
10090 ;; Returns non-nil if we're looking at the end of an in-expression
10091 ;; block, otherwise the same as `c-looking-at-inexpr-block'.
10092 ;; PAREN-STATE is the paren state relevant at the current position.
10093 ;;
10094 ;; This function might do hidden buffer changes.
10095 (save-excursion
10096 ;; We currently only recognize a block.
10097 (let ((here (point))
10098 (elem (car-safe paren-state))
10099 containing-sexp)
10100 (when (and (consp elem)
10101 (progn (goto-char (cdr elem))
10102 (c-forward-syntactic-ws here)
10103 (= (point) here)))
10104 (goto-char (car elem))
10105 (if (setq paren-state (cdr paren-state))
10106 (setq containing-sexp (car-safe paren-state)))
10107 (c-looking-at-inexpr-block (c-safe-position containing-sexp
10108 paren-state)
10109 containing-sexp)))))
10110
10111 (defun c-at-macro-vsemi-p (&optional pos)
10112 ;; Is there a "virtual semicolon" at POS or point?
10113 ;; (See cc-defs.el for full details of "virtual semicolons".)
10114 ;;
10115 ;; This is true when point is at the last non syntactic WS position on the
10116 ;; line, there is a macro call last on the line, and this particular macro's
10117 ;; name is defined by the regexp `c-vs-macro-regexp' as not needing a
10118 ;; semicolon.
10119 (save-excursion
10120 (save-restriction
10121 (widen)
10122 (if pos
10123 (goto-char pos)
10124 (setq pos (point)))
10125 (and
10126 c-macro-with-semi-re
10127 (eq (skip-chars-backward " \t") 0)
10128
10129 ;; Check we've got nothing after this except comments and empty lines
10130 ;; joined by escaped EOLs.
10131 (skip-chars-forward " \t") ; always returns non-nil.
10132 (progn
10133 (while ; go over 1 block comment per iteration.
10134 (and
10135 (looking-at "\\(\\\\[\n\r][ \t]*\\)*")
10136 (goto-char (match-end 0))
10137 (cond
10138 ((looking-at c-block-comment-start-regexp)
10139 (and (forward-comment 1)
10140 (skip-chars-forward " \t"))) ; always returns non-nil
10141 ((looking-at c-line-comment-start-regexp)
10142 (end-of-line)
10143 nil)
10144 (t nil))))
10145 (eolp))
10146
10147 (goto-char pos)
10148 (progn (c-backward-syntactic-ws)
10149 (eq (point) pos))
10150
10151 ;; Check for one of the listed macros being before point.
10152 (or (not (eq (char-before) ?\)))
10153 (when (c-go-list-backward)
10154 (c-backward-syntactic-ws)
10155 t))
10156 (c-simple-skip-symbol-backward)
10157 (looking-at c-macro-with-semi-re)
10158 (goto-char pos)
10159 (not (c-in-literal)))))) ; The most expensive check last.
10160
10161 (defun c-macro-vsemi-status-unknown-p () t) ; See cc-defs.el.
10162
10163 \f
10164 ;; `c-guess-basic-syntax' and the functions that precedes it below
10165 ;; implements the main decision tree for determining the syntactic
10166 ;; analysis of the current line of code.
10167
10168 ;; Dynamically bound to t when `c-guess-basic-syntax' is called during
10169 ;; auto newline analysis.
10170 (defvar c-auto-newline-analysis nil)
10171
10172 (defun c-brace-anchor-point (bracepos)
10173 ;; BRACEPOS is the position of a brace in a construct like "namespace
10174 ;; Bar {". Return the anchor point in this construct; this is the
10175 ;; earliest symbol on the brace's line which isn't earlier than
10176 ;; "namespace".
10177 ;;
10178 ;; Currently (2007-08-17), "like namespace" means "matches
10179 ;; c-other-block-decl-kwds". It doesn't work with "class" or "struct"
10180 ;; or anything like that.
10181 (save-excursion
10182 (let ((boi (c-point 'boi bracepos)))
10183 (goto-char bracepos)
10184 (while (and (> (point) boi)
10185 (not (looking-at c-other-decl-block-key)))
10186 (c-backward-token-2))
10187 (if (> (point) boi) (point) boi))))
10188
10189 (defsubst c-add-syntax (symbol &rest args)
10190 ;; A simple function to prepend a new syntax element to
10191 ;; `c-syntactic-context'. Using `setq' on it is unsafe since it
10192 ;; should always be dynamically bound but since we read it first
10193 ;; we'll fail properly anyway if this function is misused.
10194 (setq c-syntactic-context (cons (cons symbol args)
10195 c-syntactic-context)))
10196
10197 (defsubst c-append-syntax (symbol &rest args)
10198 ;; Like `c-add-syntax' but appends to the end of the syntax list.
10199 ;; (Normally not necessary.)
10200 (setq c-syntactic-context (nconc c-syntactic-context
10201 (list (cons symbol args)))))
10202
10203 (defun c-add-stmt-syntax (syntax-symbol
10204 syntax-extra-args
10205 stop-at-boi-only
10206 containing-sexp
10207 paren-state)
10208 ;; Add the indicated SYNTAX-SYMBOL to `c-syntactic-context', extending it as
10209 ;; needed with further syntax elements of the types `substatement',
10210 ;; `inexpr-statement', `arglist-cont-nonempty', `statement-block-intro', and
10211 ;; `defun-block-intro'.
10212 ;;
10213 ;; Do the generic processing to anchor the given syntax symbol on
10214 ;; the preceding statement: Skip over any labels and containing
10215 ;; statements on the same line, and then search backward until we
10216 ;; find a statement or block start that begins at boi without a
10217 ;; label or comment.
10218 ;;
10219 ;; Point is assumed to be at the prospective anchor point for the
10220 ;; given SYNTAX-SYMBOL. More syntax entries are added if we need to
10221 ;; skip past open parens and containing statements. Most of the added
10222 ;; syntax elements will get the same anchor point - the exception is
10223 ;; for an anchor in a construct like "namespace"[*] - this is as early
10224 ;; as possible in the construct but on the same line as the {.
10225 ;;
10226 ;; [*] i.e. with a keyword matching c-other-block-decl-kwds.
10227 ;;
10228 ;; SYNTAX-EXTRA-ARGS are a list of the extra arguments for the
10229 ;; syntax symbol. They are appended after the anchor point.
10230 ;;
10231 ;; If STOP-AT-BOI-ONLY is nil, we can stop in the middle of the line
10232 ;; if the current statement starts there.
10233 ;;
10234 ;; Note: It's not a problem if PAREN-STATE "overshoots"
10235 ;; CONTAINING-SEXP, i.e. contains info about parens further down.
10236 ;;
10237 ;; This function might do hidden buffer changes.
10238
10239 (if (= (point) (c-point 'boi))
10240 ;; This is by far the most common case, so let's give it special
10241 ;; treatment.
10242 (apply 'c-add-syntax syntax-symbol (point) syntax-extra-args)
10243
10244 (let ((syntax-last c-syntactic-context)
10245 (boi (c-point 'boi))
10246 ;; Set when we're on a label, so that we don't stop there.
10247 ;; FIXME: To be complete we should check if we're on a label
10248 ;; now at the start.
10249 on-label)
10250
10251 ;; Use point as the anchor point for "namespace", "extern", etc.
10252 (apply 'c-add-syntax syntax-symbol
10253 (if (rassq syntax-symbol c-other-decl-block-key-in-symbols-alist)
10254 (point) nil)
10255 syntax-extra-args)
10256
10257 ;; Loop while we have to back out of containing blocks.
10258 (while
10259 (and
10260 (catch 'back-up-block
10261
10262 ;; Loop while we have to back up statements.
10263 (while (or (/= (point) boi)
10264 on-label
10265 (looking-at c-comment-start-regexp))
10266
10267 ;; Skip past any comments that stands between the
10268 ;; statement start and boi.
10269 (let ((savepos (point)))
10270 (while (and (/= savepos boi)
10271 (c-backward-single-comment))
10272 (setq savepos (point)
10273 boi (c-point 'boi)))
10274 (goto-char savepos))
10275
10276 ;; Skip to the beginning of this statement or backward
10277 ;; another one.
10278 (let ((old-pos (point))
10279 (old-boi boi)
10280 (step-type (c-beginning-of-statement-1 containing-sexp)))
10281 (setq boi (c-point 'boi)
10282 on-label (eq step-type 'label))
10283
10284 (cond ((= (point) old-pos)
10285 ;; If we didn't move we're at the start of a block and
10286 ;; have to continue outside it.
10287 (throw 'back-up-block t))
10288
10289 ((and (eq step-type 'up)
10290 (>= (point) old-boi)
10291 (looking-at "else\\>[^_]")
10292 (save-excursion
10293 (goto-char old-pos)
10294 (looking-at "if\\>[^_]")))
10295 ;; Special case to avoid deeper and deeper indentation
10296 ;; of "else if" clauses.
10297 )
10298
10299 ((and (not stop-at-boi-only)
10300 (/= old-pos old-boi)
10301 (memq step-type '(up previous)))
10302 ;; If stop-at-boi-only is nil, we shouldn't back up
10303 ;; over previous or containing statements to try to
10304 ;; reach boi, so go back to the last position and
10305 ;; exit.
10306 (goto-char old-pos)
10307 (throw 'back-up-block nil))
10308
10309 (t
10310 (if (and (not stop-at-boi-only)
10311 (memq step-type '(up previous beginning)))
10312 ;; If we've moved into another statement then we
10313 ;; should no longer try to stop in the middle of a
10314 ;; line.
10315 (setq stop-at-boi-only t))
10316
10317 ;; Record this as a substatement if we skipped up one
10318 ;; level.
10319 (when (eq step-type 'up)
10320 (c-add-syntax 'substatement nil))))
10321 )))
10322
10323 containing-sexp)
10324
10325 ;; Now we have to go out of this block.
10326 (goto-char containing-sexp)
10327
10328 ;; Don't stop in the middle of a special brace list opener
10329 ;; like "({".
10330 (when c-special-brace-lists
10331 (let ((special-list (c-looking-at-special-brace-list)))
10332 (when (and special-list
10333 (< (car (car special-list)) (point)))
10334 (setq containing-sexp (car (car special-list)))
10335 (goto-char containing-sexp))))
10336
10337 (setq paren-state (c-whack-state-after containing-sexp paren-state)
10338 containing-sexp (c-most-enclosing-brace paren-state)
10339 boi (c-point 'boi))
10340
10341 ;; Analyze the construct in front of the block we've stepped out
10342 ;; from and add the right syntactic element for it.
10343 (let ((paren-pos (point))
10344 (paren-char (char-after))
10345 step-type)
10346
10347 (if (eq paren-char ?\()
10348 ;; Stepped out of a parenthesis block, so we're in an
10349 ;; expression now.
10350 (progn
10351 (when (/= paren-pos boi)
10352 (if (and c-recognize-paren-inexpr-blocks
10353 (progn
10354 (c-backward-syntactic-ws containing-sexp)
10355 (or (not (looking-at "\\>"))
10356 (not (c-on-identifier))))
10357 (save-excursion
10358 (goto-char (1+ paren-pos))
10359 (c-forward-syntactic-ws)
10360 (eq (char-after) ?{)))
10361 ;; Stepped out of an in-expression statement. This
10362 ;; syntactic element won't get an anchor pos.
10363 (c-add-syntax 'inexpr-statement)
10364
10365 ;; A parenthesis normally belongs to an arglist.
10366 (c-add-syntax 'arglist-cont-nonempty nil paren-pos)))
10367
10368 (goto-char (max boi
10369 (if containing-sexp
10370 (1+ containing-sexp)
10371 (point-min))))
10372 (setq step-type 'same
10373 on-label nil))
10374
10375 ;; Stepped out of a brace block.
10376 (setq step-type (c-beginning-of-statement-1 containing-sexp)
10377 on-label (eq step-type 'label))
10378
10379 (if (and (eq step-type 'same)
10380 (/= paren-pos (point)))
10381 (let (inexpr)
10382 (cond
10383 ((save-excursion
10384 (goto-char paren-pos)
10385 (setq inexpr (c-looking-at-inexpr-block
10386 (c-safe-position containing-sexp paren-state)
10387 containing-sexp)))
10388 (c-add-syntax (if (eq (car inexpr) 'inlambda)
10389 'defun-block-intro
10390 'statement-block-intro)
10391 nil))
10392 ((looking-at c-other-decl-block-key)
10393 (c-add-syntax
10394 (cdr (assoc (match-string 1)
10395 c-other-decl-block-key-in-symbols-alist))
10396 (max (c-point 'boi paren-pos) (point))))
10397 (t (c-add-syntax 'defun-block-intro nil))))
10398
10399 (c-add-syntax 'statement-block-intro nil)))
10400
10401 (if (= paren-pos boi)
10402 ;; Always done if the open brace was at boi. The
10403 ;; c-beginning-of-statement-1 call above is necessary
10404 ;; anyway, to decide the type of block-intro to add.
10405 (goto-char paren-pos)
10406 (setq boi (c-point 'boi)))
10407 ))
10408
10409 ;; Fill in the current point as the anchor for all the symbols
10410 ;; added above.
10411 (let ((p c-syntactic-context) q)
10412 (while (not (eq p syntax-last))
10413 (setq q (cdr (car p))) ; e.g. (nil 28) [from (arglist-cont-nonempty nil 28)]
10414 (while q
10415 (unless (car q)
10416 (setcar q (point)))
10417 (setq q (cdr q)))
10418 (setq p (cdr p))))
10419 )))
10420
10421 (defun c-add-class-syntax (symbol
10422 containing-decl-open
10423 containing-decl-start
10424 containing-decl-kwd
10425 paren-state)
10426 ;; The inclass and class-close syntactic symbols are added in
10427 ;; several places and some work is needed to fix everything.
10428 ;; Therefore it's collected here.
10429 ;;
10430 ;; This function might do hidden buffer changes.
10431 (goto-char containing-decl-open)
10432 (if (and (eq symbol 'inclass) (= (point) (c-point 'boi)))
10433 (progn
10434 (c-add-syntax symbol containing-decl-open)
10435 containing-decl-open)
10436 (goto-char containing-decl-start)
10437 ;; Ought to use `c-add-stmt-syntax' instead of backing up to boi
10438 ;; here, but we have to do like this for compatibility.
10439 (back-to-indentation)
10440 (c-add-syntax symbol (point))
10441 (if (and (c-keyword-member containing-decl-kwd
10442 'c-inexpr-class-kwds)
10443 (/= containing-decl-start (c-point 'boi containing-decl-start)))
10444 (c-add-syntax 'inexpr-class))
10445 (point)))
10446
10447 (defun c-guess-continued-construct (indent-point
10448 char-after-ip
10449 beg-of-same-or-containing-stmt
10450 containing-sexp
10451 paren-state)
10452 ;; This function contains the decision tree reached through both
10453 ;; cases 18 and 10. It's a continued statement or top level
10454 ;; construct of some kind.
10455 ;;
10456 ;; This function might do hidden buffer changes.
10457
10458 (let (special-brace-list placeholder)
10459 (goto-char indent-point)
10460 (skip-chars-forward " \t")
10461
10462 (cond
10463 ;; (CASE A removed.)
10464 ;; CASE B: open braces for class or brace-lists
10465 ((setq special-brace-list
10466 (or (and c-special-brace-lists
10467 (c-looking-at-special-brace-list))
10468 (eq char-after-ip ?{)))
10469
10470 (cond
10471 ;; CASE B.1: class-open
10472 ((save-excursion
10473 (and (eq (char-after) ?{)
10474 (c-looking-at-decl-block containing-sexp t)
10475 (setq beg-of-same-or-containing-stmt (point))))
10476 (c-add-syntax 'class-open beg-of-same-or-containing-stmt))
10477
10478 ;; CASE B.2: brace-list-open
10479 ((or (consp special-brace-list)
10480 (save-excursion
10481 (goto-char beg-of-same-or-containing-stmt)
10482 (c-syntactic-re-search-forward "=\\([^=]\\|$\\)"
10483 indent-point t t t)))
10484 ;; The most semantically accurate symbol here is
10485 ;; brace-list-open, but we normally report it simply as a
10486 ;; statement-cont. The reason is that one normally adjusts
10487 ;; brace-list-open for brace lists as top-level constructs,
10488 ;; and brace lists inside statements is a completely different
10489 ;; context. C.f. case 5A.3.
10490 (c-beginning-of-statement-1 containing-sexp)
10491 (c-add-stmt-syntax (if c-auto-newline-analysis
10492 ;; Turn off the dwim above when we're
10493 ;; analyzing the nature of the brace
10494 ;; for the auto newline feature.
10495 'brace-list-open
10496 'statement-cont)
10497 nil nil
10498 containing-sexp paren-state))
10499
10500 ;; CASE B.3: The body of a function declared inside a normal
10501 ;; block. Can occur e.g. in Pike and when using gcc
10502 ;; extensions, but watch out for macros followed by blocks.
10503 ;; C.f. cases E, 16F and 17G.
10504 ((and (not (c-at-statement-start-p))
10505 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
10506 'same)
10507 (save-excursion
10508 (let ((c-recognize-typeless-decls nil))
10509 ;; Turn off recognition of constructs that lacks a
10510 ;; type in this case, since that's more likely to be
10511 ;; a macro followed by a block.
10512 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10513 (c-add-stmt-syntax 'defun-open nil t
10514 containing-sexp paren-state))
10515
10516 ;; CASE B.4: Continued statement with block open. The most
10517 ;; accurate analysis is perhaps `statement-cont' together with
10518 ;; `block-open' but we play DWIM and use `substatement-open'
10519 ;; instead. The rationale is that this typically is a macro
10520 ;; followed by a block which makes it very similar to a
10521 ;; statement with a substatement block.
10522 (t
10523 (c-add-stmt-syntax 'substatement-open nil nil
10524 containing-sexp paren-state))
10525 ))
10526
10527 ;; CASE C: iostream insertion or extraction operator
10528 ((and (looking-at "\\(<<\\|>>\\)\\([^=]\\|$\\)")
10529 (save-excursion
10530 (goto-char beg-of-same-or-containing-stmt)
10531 ;; If there is no preceding streamop in the statement
10532 ;; then indent this line as a normal statement-cont.
10533 (when (c-syntactic-re-search-forward
10534 "\\(<<\\|>>\\)\\([^=]\\|$\\)" indent-point 'move t t)
10535 (c-add-syntax 'stream-op (c-point 'boi))
10536 t))))
10537
10538 ;; CASE E: In the "K&R region" of a function declared inside a
10539 ;; normal block. C.f. case B.3.
10540 ((and (save-excursion
10541 ;; Check that the next token is a '{'. This works as
10542 ;; long as no language that allows nested function
10543 ;; definitions allows stuff like member init lists, K&R
10544 ;; declarations or throws clauses there.
10545 ;;
10546 ;; Note that we do a forward search for something ahead
10547 ;; of the indentation line here. That's not good since
10548 ;; the user might not have typed it yet. Unfortunately
10549 ;; it's exceedingly tricky to recognize a function
10550 ;; prototype in a code block without resorting to this.
10551 (c-forward-syntactic-ws)
10552 (eq (char-after) ?{))
10553 (not (c-at-statement-start-p))
10554 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
10555 'same)
10556 (save-excursion
10557 (let ((c-recognize-typeless-decls nil))
10558 ;; Turn off recognition of constructs that lacks a
10559 ;; type in this case, since that's more likely to be
10560 ;; a macro followed by a block.
10561 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10562 (c-add-stmt-syntax 'func-decl-cont nil t
10563 containing-sexp paren-state))
10564
10565 ;;CASE F: continued statement and the only preceding items are
10566 ;;annotations.
10567 ((and (c-major-mode-is 'java-mode)
10568 (setq placeholder (point))
10569 (c-beginning-of-statement-1)
10570 (progn
10571 (while (and (c-forward-annotation)
10572 (< (point) placeholder))
10573 (c-forward-syntactic-ws))
10574 t)
10575 (prog1
10576 (>= (point) placeholder)
10577 (goto-char placeholder)))
10578 (c-beginning-of-statement-1 containing-sexp)
10579 (c-add-syntax 'annotation-var-cont (point)))
10580
10581 ;; CASE G: a template list continuation?
10582 ;; Mostly a duplication of case 5D.3 to fix templates-19:
10583 ((and (c-major-mode-is 'c++-mode)
10584 (save-excursion
10585 (goto-char indent-point)
10586 (c-with-syntax-table c++-template-syntax-table
10587 (setq placeholder (c-up-list-backward)))
10588 (and placeholder
10589 (eq (char-after placeholder) ?<)
10590 (/= (char-before placeholder) ?<)
10591 (progn
10592 (goto-char (1+ placeholder))
10593 (not (looking-at c-<-op-cont-regexp))))))
10594 (c-with-syntax-table c++-template-syntax-table
10595 (goto-char placeholder)
10596 (c-beginning-of-statement-1 containing-sexp t))
10597 (if (save-excursion
10598 (c-backward-syntactic-ws containing-sexp)
10599 (eq (char-before) ?<))
10600 ;; In a nested template arglist.
10601 (progn
10602 (goto-char placeholder)
10603 (c-syntactic-skip-backward "^,;" containing-sexp t)
10604 (c-forward-syntactic-ws))
10605 (back-to-indentation))
10606 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
10607 ;; template aware.
10608 (c-add-syntax 'template-args-cont (point) placeholder))
10609
10610 ;; CASE D: continued statement.
10611 (t
10612 (c-beginning-of-statement-1 containing-sexp)
10613 (c-add-stmt-syntax 'statement-cont nil nil
10614 containing-sexp paren-state))
10615 )))
10616
10617 ;; The next autoload was added by RMS on 2005/8/9 - don't know why (ACM,
10618 ;; 2005/11/29).
10619 ;;;###autoload
10620 (defun c-guess-basic-syntax ()
10621 "Return the syntactic context of the current line."
10622 (save-excursion
10623 (beginning-of-line)
10624 (c-save-buffer-state
10625 ((indent-point (point))
10626 (case-fold-search nil)
10627 ;; A whole ugly bunch of various temporary variables. Have
10628 ;; to declare them here since it's not possible to declare
10629 ;; a variable with only the scope of a cond test and the
10630 ;; following result clauses, and most of this function is a
10631 ;; single gigantic cond. :P
10632 literal char-before-ip before-ws-ip char-after-ip macro-start
10633 in-macro-expr c-syntactic-context placeholder c-in-literal-cache
10634 step-type tmpsymbol keyword injava-inher special-brace-list tmp-pos
10635 containing-<
10636 ;; The following record some positions for the containing
10637 ;; declaration block if we're directly within one:
10638 ;; `containing-decl-open' is the position of the open
10639 ;; brace. `containing-decl-start' is the start of the
10640 ;; declaration. `containing-decl-kwd' is the keyword
10641 ;; symbol of the keyword that tells what kind of block it
10642 ;; is.
10643 containing-decl-open
10644 containing-decl-start
10645 containing-decl-kwd
10646 ;; The open paren of the closest surrounding sexp or nil if
10647 ;; there is none.
10648 containing-sexp
10649 ;; The position after the closest preceding brace sexp
10650 ;; (nested sexps are ignored), or the position after
10651 ;; `containing-sexp' if there is none, or (point-min) if
10652 ;; `containing-sexp' is nil.
10653 lim
10654 ;; The paren state outside `containing-sexp', or at
10655 ;; `indent-point' if `containing-sexp' is nil.
10656 (paren-state (c-parse-state))
10657 ;; There's always at most one syntactic element which got
10658 ;; an anchor pos. It's stored in syntactic-relpos.
10659 syntactic-relpos
10660 (c-stmt-delim-chars c-stmt-delim-chars))
10661
10662 ;; Check if we're directly inside an enclosing declaration
10663 ;; level block.
10664 (when (and (setq containing-sexp
10665 (c-most-enclosing-brace paren-state))
10666 (progn
10667 (goto-char containing-sexp)
10668 (eq (char-after) ?{))
10669 (setq placeholder
10670 (c-looking-at-decl-block
10671 (c-most-enclosing-brace paren-state
10672 containing-sexp)
10673 t)))
10674 (setq containing-decl-open containing-sexp
10675 containing-decl-start (point)
10676 containing-sexp nil)
10677 (goto-char placeholder)
10678 (setq containing-decl-kwd (and (looking-at c-keywords-regexp)
10679 (c-keyword-sym (match-string 1)))))
10680
10681 ;; Init some position variables.
10682 (if paren-state
10683 (progn
10684 (setq containing-sexp (car paren-state)
10685 paren-state (cdr paren-state))
10686 (if (consp containing-sexp)
10687 (save-excursion
10688 (goto-char (cdr containing-sexp))
10689 (if (and (c-major-mode-is 'c++-mode)
10690 (c-back-over-member-initializer-braces))
10691 (c-syntactic-skip-backward "^}" nil t))
10692 (setq lim (point))
10693 (if paren-state
10694 ;; Ignore balanced paren. The next entry
10695 ;; can't be another one.
10696 (setq containing-sexp (car paren-state)
10697 paren-state (cdr paren-state))
10698 ;; If there is no surrounding open paren then
10699 ;; put the last balanced pair back on paren-state.
10700 (setq paren-state (cons containing-sexp paren-state)
10701 containing-sexp nil)))
10702 (setq lim (1+ containing-sexp))))
10703 (setq lim (point-min)))
10704
10705 ;; If we're in a parenthesis list then ',' delimits the
10706 ;; "statements" rather than being an operator (with the
10707 ;; exception of the "for" clause). This difference is
10708 ;; typically only noticeable when statements are used in macro
10709 ;; arglists.
10710 (when (and containing-sexp
10711 (eq (char-after containing-sexp) ?\())
10712 (setq c-stmt-delim-chars c-stmt-delim-chars-with-comma))
10713 ;; cache char before and after indent point, and move point to
10714 ;; the most likely position to perform the majority of tests
10715 (goto-char indent-point)
10716 (c-backward-syntactic-ws lim)
10717 (setq before-ws-ip (point)
10718 char-before-ip (char-before))
10719 (goto-char indent-point)
10720 (skip-chars-forward " \t")
10721 (setq char-after-ip (char-after))
10722
10723 ;; are we in a literal?
10724 (setq literal (c-in-literal lim))
10725
10726 ;; now figure out syntactic qualities of the current line
10727 (cond
10728
10729 ;; CASE 1: in a string.
10730 ((eq literal 'string)
10731 (c-add-syntax 'string (c-point 'bopl)))
10732
10733 ;; CASE 2: in a C or C++ style comment.
10734 ((and (memq literal '(c c++))
10735 ;; This is a kludge for XEmacs where we use
10736 ;; `buffer-syntactic-context', which doesn't correctly
10737 ;; recognize "\*/" to end a block comment.
10738 ;; `parse-partial-sexp' which is used by
10739 ;; `c-literal-limits' will however do that in most
10740 ;; versions, which results in that we get nil from
10741 ;; `c-literal-limits' even when `c-in-literal' claims
10742 ;; we're inside a comment.
10743 (setq placeholder (c-literal-start lim)))
10744 (c-add-syntax literal placeholder))
10745
10746 ;; CASE 3: in a cpp preprocessor macro continuation.
10747 ((and (save-excursion
10748 (when (c-beginning-of-macro)
10749 (setq macro-start (point))))
10750 (/= macro-start (c-point 'boi))
10751 (progn
10752 (setq tmpsymbol 'cpp-macro-cont)
10753 (or (not c-syntactic-indentation-in-macros)
10754 (save-excursion
10755 (goto-char macro-start)
10756 ;; If at the beginning of the body of a #define
10757 ;; directive then analyze as cpp-define-intro
10758 ;; only. Go on with the syntactic analysis
10759 ;; otherwise. in-macro-expr is set if we're in a
10760 ;; cpp expression, i.e. before the #define body
10761 ;; or anywhere in a non-#define directive.
10762 (if (c-forward-to-cpp-define-body)
10763 (let ((indent-boi (c-point 'boi indent-point)))
10764 (setq in-macro-expr (> (point) indent-boi)
10765 tmpsymbol 'cpp-define-intro)
10766 (= (point) indent-boi))
10767 (setq in-macro-expr t)
10768 nil)))))
10769 (c-add-syntax tmpsymbol macro-start)
10770 (setq macro-start nil))
10771
10772 ;; CASE 11: an else clause?
10773 ((looking-at "else\\>[^_]")
10774 (c-beginning-of-statement-1 containing-sexp)
10775 (c-add-stmt-syntax 'else-clause nil t
10776 containing-sexp paren-state))
10777
10778 ;; CASE 12: while closure of a do/while construct?
10779 ((and (looking-at "while\\>[^_]")
10780 (save-excursion
10781 (prog1 (eq (c-beginning-of-statement-1 containing-sexp)
10782 'beginning)
10783 (setq placeholder (point)))))
10784 (goto-char placeholder)
10785 (c-add-stmt-syntax 'do-while-closure nil t
10786 containing-sexp paren-state))
10787
10788 ;; CASE 13: A catch or finally clause? This case is simpler
10789 ;; than if-else and do-while, because a block is required
10790 ;; after every try, catch and finally.
10791 ((save-excursion
10792 (and (cond ((c-major-mode-is 'c++-mode)
10793 (looking-at "catch\\>[^_]"))
10794 ((c-major-mode-is 'java-mode)
10795 (looking-at "\\(catch\\|finally\\)\\>[^_]")))
10796 (and (c-safe (c-backward-syntactic-ws)
10797 (c-backward-sexp)
10798 t)
10799 (eq (char-after) ?{)
10800 (c-safe (c-backward-syntactic-ws)
10801 (c-backward-sexp)
10802 t)
10803 (if (eq (char-after) ?\()
10804 (c-safe (c-backward-sexp) t)
10805 t))
10806 (looking-at "\\(try\\|catch\\)\\>[^_]")
10807 (setq placeholder (point))))
10808 (goto-char placeholder)
10809 (c-add-stmt-syntax 'catch-clause nil t
10810 containing-sexp paren-state))
10811
10812 ;; CASE 18: A substatement we can recognize by keyword.
10813 ((save-excursion
10814 (and c-opt-block-stmt-key
10815 (not (eq char-before-ip ?\;))
10816 (not (c-at-vsemi-p before-ws-ip))
10817 (not (memq char-after-ip '(?\) ?\] ?,)))
10818 (or (not (eq char-before-ip ?}))
10819 (c-looking-at-inexpr-block-backward c-state-cache))
10820 (> (point)
10821 (progn
10822 ;; Ought to cache the result from the
10823 ;; c-beginning-of-statement-1 calls here.
10824 (setq placeholder (point))
10825 (while (eq (setq step-type
10826 (c-beginning-of-statement-1 lim))
10827 'label))
10828 (if (eq step-type 'previous)
10829 (goto-char placeholder)
10830 (setq placeholder (point))
10831 (if (and (eq step-type 'same)
10832 (not (looking-at c-opt-block-stmt-key)))
10833 ;; Step up to the containing statement if we
10834 ;; stayed in the same one.
10835 (let (step)
10836 (while (eq
10837 (setq step
10838 (c-beginning-of-statement-1 lim))
10839 'label))
10840 (if (eq step 'up)
10841 (setq placeholder (point))
10842 ;; There was no containing statement after all.
10843 (goto-char placeholder)))))
10844 placeholder))
10845 (if (looking-at c-block-stmt-2-key)
10846 ;; Require a parenthesis after these keywords.
10847 ;; Necessary to catch e.g. synchronized in Java,
10848 ;; which can be used both as statement and
10849 ;; modifier.
10850 (and (zerop (c-forward-token-2 1 nil))
10851 (eq (char-after) ?\())
10852 (looking-at c-opt-block-stmt-key))))
10853
10854 (if (eq step-type 'up)
10855 ;; CASE 18A: Simple substatement.
10856 (progn
10857 (goto-char placeholder)
10858 (cond
10859 ((eq char-after-ip ?{)
10860 (c-add-stmt-syntax 'substatement-open nil nil
10861 containing-sexp paren-state))
10862 ((save-excursion
10863 (goto-char indent-point)
10864 (back-to-indentation)
10865 (c-forward-label))
10866 (c-add-stmt-syntax 'substatement-label nil nil
10867 containing-sexp paren-state))
10868 (t
10869 (c-add-stmt-syntax 'substatement nil nil
10870 containing-sexp paren-state))))
10871
10872 ;; CASE 18B: Some other substatement. This is shared
10873 ;; with case 10.
10874 (c-guess-continued-construct indent-point
10875 char-after-ip
10876 placeholder
10877 lim
10878 paren-state)))
10879
10880 ;; CASE 14: A case or default label
10881 ((save-excursion
10882 (and (looking-at c-label-kwds-regexp)
10883 (or (c-major-mode-is 'idl-mode)
10884 (and
10885 containing-sexp
10886 (goto-char containing-sexp)
10887 (eq (char-after) ?{)
10888 (progn (c-backward-syntactic-ws) t)
10889 (eq (char-before) ?\))
10890 (c-go-list-backward)
10891 (progn (c-backward-syntactic-ws) t)
10892 (c-simple-skip-symbol-backward)
10893 (looking-at c-block-stmt-2-key)))))
10894 (if containing-sexp
10895 (progn
10896 (goto-char containing-sexp)
10897 (setq lim (c-most-enclosing-brace c-state-cache
10898 containing-sexp))
10899 (c-backward-to-block-anchor lim)
10900 (c-add-stmt-syntax 'case-label nil t lim paren-state))
10901 ;; Got a bogus label at the top level. In lack of better
10902 ;; alternatives, anchor it on (point-min).
10903 (c-add-syntax 'case-label (point-min))))
10904
10905 ;; CASE 15: any other label
10906 ((save-excursion
10907 (back-to-indentation)
10908 (and (not (looking-at c-syntactic-ws-start))
10909 (not (looking-at c-label-kwds-regexp))
10910 (c-forward-label)))
10911 (cond (containing-decl-open
10912 (setq placeholder (c-add-class-syntax 'inclass
10913 containing-decl-open
10914 containing-decl-start
10915 containing-decl-kwd
10916 paren-state))
10917 ;; Append access-label with the same anchor point as
10918 ;; inclass gets.
10919 (c-append-syntax 'access-label placeholder))
10920
10921 (containing-sexp
10922 (goto-char containing-sexp)
10923 (setq lim (c-most-enclosing-brace c-state-cache
10924 containing-sexp))
10925 (save-excursion
10926 (setq tmpsymbol
10927 (if (and (eq (c-beginning-of-statement-1 lim) 'up)
10928 (looking-at "switch\\>[^_]"))
10929 ;; If the surrounding statement is a switch then
10930 ;; let's analyze all labels as switch labels, so
10931 ;; that they get lined up consistently.
10932 'case-label
10933 'label)))
10934 (c-backward-to-block-anchor lim)
10935 (c-add-stmt-syntax tmpsymbol nil t lim paren-state))
10936
10937 (t
10938 ;; A label on the top level. Treat it as a class
10939 ;; context. (point-min) is the closest we get to the
10940 ;; class open brace.
10941 (c-add-syntax 'access-label (point-min)))))
10942
10943 ;; CASE 4: In-expression statement. C.f. cases 7B, 16A and
10944 ;; 17E.
10945 ((setq placeholder (c-looking-at-inexpr-block
10946 (c-safe-position containing-sexp paren-state)
10947 containing-sexp
10948 ;; Have to turn on the heuristics after
10949 ;; the point even though it doesn't work
10950 ;; very well. C.f. test case class-16.pike.
10951 t))
10952 (setq tmpsymbol (assq (car placeholder)
10953 '((inexpr-class . class-open)
10954 (inexpr-statement . block-open))))
10955 (if tmpsymbol
10956 ;; It's a statement block or an anonymous class.
10957 (setq tmpsymbol (cdr tmpsymbol))
10958 ;; It's a Pike lambda. Check whether we are between the
10959 ;; lambda keyword and the argument list or at the defun
10960 ;; opener.
10961 (setq tmpsymbol (if (eq char-after-ip ?{)
10962 'inline-open
10963 'lambda-intro-cont)))
10964 (goto-char (cdr placeholder))
10965 (back-to-indentation)
10966 (c-add-stmt-syntax tmpsymbol nil t
10967 (c-most-enclosing-brace c-state-cache (point))
10968 paren-state)
10969 (unless (eq (point) (cdr placeholder))
10970 (c-add-syntax (car placeholder))))
10971
10972 ;; CASE 5: Line is inside a declaration level block or at top level.
10973 ((or containing-decl-open (null containing-sexp))
10974 (cond
10975
10976 ;; CASE 5A: we are looking at a defun, brace list, class,
10977 ;; or inline-inclass method opening brace
10978 ((setq special-brace-list
10979 (or (and c-special-brace-lists
10980 (c-looking-at-special-brace-list))
10981 (eq char-after-ip ?{)))
10982 (cond
10983
10984 ;; CASE 5A.1: Non-class declaration block open.
10985 ((save-excursion
10986 (let (tmp)
10987 (and (eq char-after-ip ?{)
10988 (setq tmp (c-looking-at-decl-block containing-sexp t))
10989 (progn
10990 (setq placeholder (point))
10991 (goto-char tmp)
10992 (looking-at c-symbol-key))
10993 (c-keyword-member
10994 (c-keyword-sym (setq keyword (match-string 0)))
10995 'c-other-block-decl-kwds))))
10996 (goto-char placeholder)
10997 (c-add-stmt-syntax
10998 (if (string-equal keyword "extern")
10999 ;; Special case for extern-lang-open.
11000 'extern-lang-open
11001 (intern (concat keyword "-open")))
11002 nil t containing-sexp paren-state))
11003
11004 ;; CASE 5A.2: we are looking at a class opening brace
11005 ((save-excursion
11006 (goto-char indent-point)
11007 (skip-chars-forward " \t")
11008 (and (eq (char-after) ?{)
11009 (c-looking-at-decl-block containing-sexp t)
11010 (setq placeholder (point))))
11011 (c-add-syntax 'class-open placeholder))
11012
11013 ;; CASE 5A.3: brace list open
11014 ((save-excursion
11015 (c-beginning-of-decl-1 lim)
11016 (while (cond
11017 ((looking-at c-specifier-key)
11018 (c-forward-keyword-clause 1))
11019 ((and c-opt-cpp-prefix
11020 (looking-at c-noise-macro-with-parens-name-re))
11021 (c-forward-noise-clause))))
11022 (setq placeholder (c-point 'boi))
11023 (or (consp special-brace-list)
11024 (and (or (save-excursion
11025 (goto-char indent-point)
11026 (setq tmpsymbol nil)
11027 (while (and (> (point) placeholder)
11028 (zerop (c-backward-token-2 1 t))
11029 (not (looking-at "=\\([^=]\\|$\\)")))
11030 (and c-opt-inexpr-brace-list-key
11031 (not tmpsymbol)
11032 (looking-at c-opt-inexpr-brace-list-key)
11033 (setq tmpsymbol 'topmost-intro-cont)))
11034 (looking-at "=\\([^=]\\|$\\)"))
11035 (looking-at c-brace-list-key))
11036 (save-excursion
11037 (while (and (< (point) indent-point)
11038 (zerop (c-forward-token-2 1 t))
11039 (not (memq (char-after) '(?\; ?\()))))
11040 (not (memq (char-after) '(?\; ?\()))
11041 ))))
11042 (if (and (not c-auto-newline-analysis)
11043 (c-major-mode-is 'java-mode)
11044 (eq tmpsymbol 'topmost-intro-cont))
11045 ;; We're in Java and have found that the open brace
11046 ;; belongs to a "new Foo[]" initialization list,
11047 ;; which means the brace list is part of an
11048 ;; expression and not a top level definition. We
11049 ;; therefore treat it as any topmost continuation
11050 ;; even though the semantically correct symbol still
11051 ;; is brace-list-open, on the same grounds as in
11052 ;; case B.2.
11053 (progn
11054 (c-beginning-of-statement-1 lim)
11055 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
11056 (c-add-syntax 'brace-list-open placeholder)))
11057
11058 ;; CASE 5A.4: inline defun open
11059 ((and containing-decl-open
11060 (not (c-keyword-member containing-decl-kwd
11061 'c-other-block-decl-kwds)))
11062 (c-add-syntax 'inline-open)
11063 (c-add-class-syntax 'inclass
11064 containing-decl-open
11065 containing-decl-start
11066 containing-decl-kwd
11067 paren-state))
11068
11069 ;; CASE 5A.5: ordinary defun open
11070 (t
11071 (save-excursion
11072 (c-beginning-of-decl-1 lim)
11073 (while (cond
11074 ((looking-at c-specifier-key)
11075 (c-forward-keyword-clause 1))
11076 ((and c-opt-cpp-prefix
11077 (looking-at c-noise-macro-with-parens-name-re))
11078 (c-forward-noise-clause))))
11079 (c-add-syntax 'defun-open (c-point 'boi))
11080 ;; Bogus to use bol here, but it's the legacy. (Resolved,
11081 ;; 2007-11-09)
11082 ))))
11083
11084 ;; CASE 5R: Member init list. (Used to be part of CASE 5B.1)
11085 ;; Note there is no limit on the backward search here, since member
11086 ;; init lists can, in practice, be very large.
11087 ((save-excursion
11088 (when (and (c-major-mode-is 'c++-mode)
11089 (setq placeholder (c-back-over-member-initializers)))
11090 (setq tmp-pos (point))))
11091 (if (= (c-point 'bosws) (1+ tmp-pos))
11092 (progn
11093 ;; There is no preceding member init clause.
11094 ;; Indent relative to the beginning of indentation
11095 ;; for the topmost-intro line that contains the
11096 ;; prototype's open paren.
11097 (goto-char placeholder)
11098 (c-add-syntax 'member-init-intro (c-point 'boi)))
11099 ;; Indent relative to the first member init clause.
11100 (goto-char (1+ tmp-pos))
11101 (c-forward-syntactic-ws)
11102 (c-add-syntax 'member-init-cont (point))))
11103
11104 ;; CASE 5B: After a function header but before the body (or
11105 ;; the ending semicolon if there's no body).
11106 ((save-excursion
11107 (when (setq placeholder (c-just-after-func-arglist-p
11108 (max lim (c-determine-limit 500))))
11109 (setq tmp-pos (point))))
11110 (cond
11111
11112 ;; CASE 5B.1: Member init list.
11113 ((eq (char-after tmp-pos) ?:)
11114 ;; There is no preceding member init clause.
11115 ;; Indent relative to the beginning of indentation
11116 ;; for the topmost-intro line that contains the
11117 ;; prototype's open paren.
11118 (goto-char placeholder)
11119 (c-add-syntax 'member-init-intro (c-point 'boi)))
11120
11121 ;; CASE 5B.2: K&R arg decl intro
11122 ((and c-recognize-knr-p
11123 (c-in-knr-argdecl lim))
11124 (c-beginning-of-statement-1 lim)
11125 (c-add-syntax 'knr-argdecl-intro (c-point 'boi))
11126 (if containing-decl-open
11127 (c-add-class-syntax 'inclass
11128 containing-decl-open
11129 containing-decl-start
11130 containing-decl-kwd
11131 paren-state)))
11132
11133 ;; CASE 5B.4: Nether region after a C++ or Java func
11134 ;; decl, which could include a `throws' declaration.
11135 (t
11136 (c-beginning-of-statement-1 lim)
11137 (c-add-syntax 'func-decl-cont (c-point 'boi))
11138 )))
11139
11140 ;; CASE 5C: inheritance line. could be first inheritance
11141 ;; line, or continuation of a multiple inheritance
11142 ((or (and (c-major-mode-is 'c++-mode)
11143 (progn
11144 (when (eq char-after-ip ?,)
11145 (skip-chars-forward " \t")
11146 (forward-char))
11147 (looking-at c-opt-postfix-decl-spec-key)))
11148 (and (or (eq char-before-ip ?:)
11149 ;; watch out for scope operator
11150 (save-excursion
11151 (and (eq char-after-ip ?:)
11152 (c-safe (forward-char 1) t)
11153 (not (eq (char-after) ?:))
11154 )))
11155 (save-excursion
11156 (c-beginning-of-statement-1 lim)
11157 (when (looking-at c-opt-<>-sexp-key)
11158 (goto-char (match-end 1))
11159 (c-forward-syntactic-ws)
11160 (c-forward-<>-arglist nil)
11161 (c-forward-syntactic-ws))
11162 (looking-at c-class-key)))
11163 ;; for Java
11164 (and (c-major-mode-is 'java-mode)
11165 (let ((fence (save-excursion
11166 (c-beginning-of-statement-1 lim)
11167 (point)))
11168 cont done)
11169 (save-excursion
11170 (while (not done)
11171 (cond ((looking-at c-opt-postfix-decl-spec-key)
11172 (setq injava-inher (cons cont (point))
11173 done t))
11174 ((or (not (c-safe (c-forward-sexp -1) t))
11175 (<= (point) fence))
11176 (setq done t))
11177 )
11178 (setq cont t)))
11179 injava-inher)
11180 (not (c-crosses-statement-barrier-p (cdr injava-inher)
11181 (point)))
11182 ))
11183 (cond
11184
11185 ;; CASE 5C.1: non-hanging colon on an inher intro
11186 ((eq char-after-ip ?:)
11187 (c-beginning-of-statement-1 lim)
11188 (c-add-syntax 'inher-intro (c-point 'boi))
11189 ;; don't add inclass symbol since relative point already
11190 ;; contains any class offset
11191 )
11192
11193 ;; CASE 5C.2: hanging colon on an inher intro
11194 ((eq char-before-ip ?:)
11195 (c-beginning-of-statement-1 lim)
11196 (c-add-syntax 'inher-intro (c-point 'boi))
11197 (if containing-decl-open
11198 (c-add-class-syntax 'inclass
11199 containing-decl-open
11200 containing-decl-start
11201 containing-decl-kwd
11202 paren-state)))
11203
11204 ;; CASE 5C.3: in a Java implements/extends
11205 (injava-inher
11206 (let ((where (cdr injava-inher))
11207 (cont (car injava-inher)))
11208 (goto-char where)
11209 (cond ((looking-at "throws\\>[^_]")
11210 (c-add-syntax 'func-decl-cont
11211 (progn (c-beginning-of-statement-1 lim)
11212 (c-point 'boi))))
11213 (cont (c-add-syntax 'inher-cont where))
11214 (t (c-add-syntax 'inher-intro
11215 (progn (goto-char (cdr injava-inher))
11216 (c-beginning-of-statement-1 lim)
11217 (point))))
11218 )))
11219
11220 ;; CASE 5C.4: a continued inheritance line
11221 (t
11222 (c-beginning-of-inheritance-list lim)
11223 (c-add-syntax 'inher-cont (point))
11224 ;; don't add inclass symbol since relative point already
11225 ;; contains any class offset
11226 )))
11227
11228 ;; CASE 5P: AWK pattern or function or continuation
11229 ;; thereof.
11230 ((c-major-mode-is 'awk-mode)
11231 (setq placeholder (point))
11232 (c-add-stmt-syntax
11233 (if (and (eq (c-beginning-of-statement-1) 'same)
11234 (/= (point) placeholder))
11235 'topmost-intro-cont
11236 'topmost-intro)
11237 nil nil
11238 containing-sexp paren-state))
11239
11240 ;; CASE 5D: this could be a top-level initialization, a
11241 ;; member init list continuation, or a template argument
11242 ;; list continuation.
11243 ((save-excursion
11244 ;; Note: We use the fact that lim is always after any
11245 ;; preceding brace sexp.
11246 (if c-recognize-<>-arglists
11247 (while (and
11248 (progn
11249 (c-syntactic-skip-backward "^;,=<>" lim t)
11250 (> (point) lim))
11251 (or
11252 (when c-overloadable-operators-regexp
11253 (when (setq placeholder (c-after-special-operator-id lim))
11254 (goto-char placeholder)
11255 t))
11256 (cond
11257 ((eq (char-before) ?>)
11258 (or (c-backward-<>-arglist nil lim)
11259 (backward-char))
11260 t)
11261 ((eq (char-before) ?<)
11262 (backward-char)
11263 (if (save-excursion
11264 (c-forward-<>-arglist nil))
11265 (progn (forward-char)
11266 nil)
11267 t))
11268 (t nil)))))
11269 ;; NB: No c-after-special-operator-id stuff in this
11270 ;; clause - we assume only C++ needs it.
11271 (c-syntactic-skip-backward "^;,=" lim t))
11272 (memq (char-before) '(?, ?= ?<)))
11273 (cond
11274
11275 ;; CASE 5D.3: perhaps a template list continuation?
11276 ((and (c-major-mode-is 'c++-mode)
11277 (save-excursion
11278 (save-restriction
11279 (c-with-syntax-table c++-template-syntax-table
11280 (goto-char indent-point)
11281 (setq placeholder (c-up-list-backward))
11282 (and placeholder
11283 (eq (char-after placeholder) ?<))))))
11284 (c-with-syntax-table c++-template-syntax-table
11285 (goto-char placeholder)
11286 (c-beginning-of-statement-1 lim t))
11287 (if (save-excursion
11288 (c-backward-syntactic-ws lim)
11289 (eq (char-before) ?<))
11290 ;; In a nested template arglist.
11291 (progn
11292 (goto-char placeholder)
11293 (c-syntactic-skip-backward "^,;" lim t)
11294 (c-forward-syntactic-ws))
11295 (back-to-indentation))
11296 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
11297 ;; template aware.
11298 (c-add-syntax 'template-args-cont (point) placeholder))
11299
11300 ;; CASE 5D.4: perhaps a multiple inheritance line?
11301 ((and (c-major-mode-is 'c++-mode)
11302 (save-excursion
11303 (c-beginning-of-statement-1 lim)
11304 (setq placeholder (point))
11305 (if (looking-at "static\\>[^_]")
11306 (c-forward-token-2 1 nil indent-point))
11307 (and (looking-at c-class-key)
11308 (zerop (c-forward-token-2 2 nil indent-point))
11309 (if (eq (char-after) ?<)
11310 (c-with-syntax-table c++-template-syntax-table
11311 (zerop (c-forward-token-2 1 t indent-point)))
11312 t)
11313 (eq (char-after) ?:))))
11314 (goto-char placeholder)
11315 (c-add-syntax 'inher-cont (c-point 'boi)))
11316
11317 ;; CASE 5D.5: Continuation of the "expression part" of a
11318 ;; top level construct. Or, perhaps, an unrecognized construct.
11319 (t
11320 (while (and (setq placeholder (point))
11321 (eq (car (c-beginning-of-decl-1 containing-sexp)) ; Can't use `lim' here.
11322 'same)
11323 (save-excursion
11324 (c-backward-syntactic-ws)
11325 (eq (char-before) ?}))
11326 (< (point) placeholder)))
11327 (c-add-stmt-syntax
11328 (cond
11329 ((eq (point) placeholder) 'statement) ; unrecognized construct
11330 ;; A preceding comma at the top level means that a
11331 ;; new variable declaration starts here. Use
11332 ;; topmost-intro-cont for it, for consistency with
11333 ;; the first variable declaration. C.f. case 5N.
11334 ((eq char-before-ip ?,) 'topmost-intro-cont)
11335 (t 'statement-cont))
11336 nil nil containing-sexp paren-state))
11337 ))
11338
11339 ;; CASE 5F: Close of a non-class declaration level block.
11340 ((and (eq char-after-ip ?})
11341 (c-keyword-member containing-decl-kwd
11342 'c-other-block-decl-kwds))
11343 ;; This is inconsistent: Should use `containing-decl-open'
11344 ;; here if it's at boi, like in case 5J.
11345 (goto-char containing-decl-start)
11346 (c-add-stmt-syntax
11347 (if (string-equal (symbol-name containing-decl-kwd) "extern")
11348 ;; Special case for compatibility with the
11349 ;; extern-lang syntactic symbols.
11350 'extern-lang-close
11351 (intern (concat (symbol-name containing-decl-kwd)
11352 "-close")))
11353 nil t
11354 (c-most-enclosing-brace paren-state (point))
11355 paren-state))
11356
11357 ;; CASE 5G: we are looking at the brace which closes the
11358 ;; enclosing nested class decl
11359 ((and containing-sexp
11360 (eq char-after-ip ?})
11361 (eq containing-decl-open containing-sexp))
11362 (c-add-class-syntax 'class-close
11363 containing-decl-open
11364 containing-decl-start
11365 containing-decl-kwd
11366 paren-state))
11367
11368 ;; CASE 5H: we could be looking at subsequent knr-argdecls
11369 ((and c-recognize-knr-p
11370 (not containing-sexp) ; can't be knr inside braces.
11371 (not (eq char-before-ip ?}))
11372 (save-excursion
11373 (setq placeholder (cdr (c-beginning-of-decl-1 lim)))
11374 (and placeholder
11375 ;; Do an extra check to avoid tripping up on
11376 ;; statements that occur in invalid contexts
11377 ;; (e.g. in macro bodies where we don't really
11378 ;; know the context of what we're looking at).
11379 (not (and c-opt-block-stmt-key
11380 (looking-at c-opt-block-stmt-key)))))
11381 (< placeholder indent-point))
11382 (goto-char placeholder)
11383 (c-add-syntax 'knr-argdecl (point)))
11384
11385 ;; CASE 5I: ObjC method definition.
11386 ((and c-opt-method-key
11387 (looking-at c-opt-method-key))
11388 (c-beginning-of-statement-1 nil t)
11389 (if (= (point) indent-point)
11390 ;; Handle the case when it's the first (non-comment)
11391 ;; thing in the buffer. Can't look for a 'same return
11392 ;; value from cbos1 since ObjC directives currently
11393 ;; aren't recognized fully, so that we get 'same
11394 ;; instead of 'previous if it moved over a preceding
11395 ;; directive.
11396 (goto-char (point-min)))
11397 (c-add-syntax 'objc-method-intro (c-point 'boi)))
11398
11399 ;; CASE 5N: At a variable declaration that follows a class
11400 ;; definition or some other block declaration that doesn't
11401 ;; end at the closing '}'. C.f. case 5D.5.
11402 ((progn
11403 (c-backward-syntactic-ws lim)
11404 (and (eq (char-before) ?})
11405 (save-excursion
11406 (let ((start (point)))
11407 (if (and c-state-cache
11408 (consp (car c-state-cache))
11409 (eq (cdar c-state-cache) (point)))
11410 ;; Speed up the backward search a bit.
11411 (goto-char (caar c-state-cache)))
11412 (c-beginning-of-decl-1 containing-sexp) ; Can't use `lim' here.
11413 (setq placeholder (point))
11414 (if (= start (point))
11415 ;; The '}' is unbalanced.
11416 nil
11417 (c-end-of-decl-1)
11418 (>= (point) indent-point))))))
11419 (goto-char placeholder)
11420 (c-add-stmt-syntax 'topmost-intro-cont nil nil
11421 containing-sexp paren-state))
11422
11423 ;; NOTE: The point is at the end of the previous token here.
11424
11425 ;; CASE 5J: we are at the topmost level, make
11426 ;; sure we skip back past any access specifiers
11427 ((and
11428 ;; A macro continuation line is never at top level.
11429 (not (and macro-start
11430 (> indent-point macro-start)))
11431 (save-excursion
11432 (setq placeholder (point))
11433 (or (memq char-before-ip '(?\; ?{ ?} nil))
11434 (c-at-vsemi-p before-ws-ip)
11435 (when (and (eq char-before-ip ?:)
11436 (eq (c-beginning-of-statement-1 lim)
11437 'label))
11438 (c-backward-syntactic-ws lim)
11439 (setq placeholder (point)))
11440 (and (c-major-mode-is 'objc-mode)
11441 (catch 'not-in-directive
11442 (c-beginning-of-statement-1 lim)
11443 (setq placeholder (point))
11444 (while (and (c-forward-objc-directive)
11445 (< (point) indent-point))
11446 (c-forward-syntactic-ws)
11447 (if (>= (point) indent-point)
11448 (throw 'not-in-directive t))
11449 (setq placeholder (point)))
11450 nil)))))
11451 ;; For historic reasons we anchor at bol of the last
11452 ;; line of the previous declaration. That's clearly
11453 ;; highly bogus and useless, and it makes our lives hard
11454 ;; to remain compatible. :P
11455 (goto-char placeholder)
11456 (c-add-syntax 'topmost-intro (c-point 'bol))
11457 (if containing-decl-open
11458 (if (c-keyword-member containing-decl-kwd
11459 'c-other-block-decl-kwds)
11460 (progn
11461 (goto-char (c-brace-anchor-point containing-decl-open))
11462 (c-add-stmt-syntax
11463 (if (string-equal (symbol-name containing-decl-kwd)
11464 "extern")
11465 ;; Special case for compatibility with the
11466 ;; extern-lang syntactic symbols.
11467 'inextern-lang
11468 (intern (concat "in"
11469 (symbol-name containing-decl-kwd))))
11470 nil t
11471 (c-most-enclosing-brace paren-state (point))
11472 paren-state))
11473 (c-add-class-syntax 'inclass
11474 containing-decl-open
11475 containing-decl-start
11476 containing-decl-kwd
11477 paren-state)))
11478 (when (and c-syntactic-indentation-in-macros
11479 macro-start
11480 (/= macro-start (c-point 'boi indent-point)))
11481 (c-add-syntax 'cpp-define-intro)
11482 (setq macro-start nil)))
11483
11484 ;; CASE 5K: we are at an ObjC method definition
11485 ;; continuation line.
11486 ((and c-opt-method-key
11487 (save-excursion
11488 (c-beginning-of-statement-1 lim)
11489 (beginning-of-line)
11490 (when (looking-at c-opt-method-key)
11491 (setq placeholder (point)))))
11492 (c-add-syntax 'objc-method-args-cont placeholder))
11493
11494 ;; CASE 5L: we are at the first argument of a template
11495 ;; arglist that begins on the previous line.
11496 ((and c-recognize-<>-arglists
11497 (eq (char-before) ?<)
11498 (not (and c-overloadable-operators-regexp
11499 (c-after-special-operator-id lim))))
11500 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
11501 (c-add-syntax 'template-args-cont (c-point 'boi)))
11502
11503 ;; CASE 5Q: we are at a statement within a macro.
11504 (macro-start
11505 (c-beginning-of-statement-1 containing-sexp)
11506 (c-add-stmt-syntax 'statement nil t containing-sexp paren-state))
11507
11508 ;;CASE 5N: We are at a topmost continuation line and the only
11509 ;;preceding items are annotations.
11510 ((and (c-major-mode-is 'java-mode)
11511 (setq placeholder (point))
11512 (c-beginning-of-statement-1)
11513 (progn
11514 (while (and (c-forward-annotation))
11515 (c-forward-syntactic-ws))
11516 t)
11517 (prog1
11518 (>= (point) placeholder)
11519 (goto-char placeholder)))
11520 (c-add-syntax 'annotation-top-cont (c-point 'boi)))
11521
11522 ;; CASE 5M: we are at a topmost continuation line
11523 (t
11524 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
11525 (when (c-major-mode-is 'objc-mode)
11526 (setq placeholder (point))
11527 (while (and (c-forward-objc-directive)
11528 (< (point) indent-point))
11529 (c-forward-syntactic-ws)
11530 (setq placeholder (point)))
11531 (goto-char placeholder))
11532 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
11533 ))
11534
11535 ;; (CASE 6 has been removed.)
11536
11537 ;; CASE 7: line is an expression, not a statement. Most
11538 ;; likely we are either in a function prototype or a function
11539 ;; call argument list
11540 ((not (or (and c-special-brace-lists
11541 (save-excursion
11542 (goto-char containing-sexp)
11543 (c-looking-at-special-brace-list)))
11544 (eq (char-after containing-sexp) ?{)))
11545 (cond
11546
11547 ;; CASE 7A: we are looking at the arglist closing paren.
11548 ;; C.f. case 7F.
11549 ((memq char-after-ip '(?\) ?\]))
11550 (goto-char containing-sexp)
11551 (setq placeholder (c-point 'boi))
11552 (if (and (c-safe (backward-up-list 1) t)
11553 (>= (point) placeholder))
11554 (progn
11555 (forward-char)
11556 (skip-chars-forward " \t"))
11557 (goto-char placeholder))
11558 (c-add-stmt-syntax 'arglist-close (list containing-sexp) t
11559 (c-most-enclosing-brace paren-state (point))
11560 paren-state))
11561
11562 ;; CASE 7B: Looking at the opening brace of an
11563 ;; in-expression block or brace list. C.f. cases 4, 16A
11564 ;; and 17E.
11565 ((and (eq char-after-ip ?{)
11566 (progn
11567 (setq placeholder (c-inside-bracelist-p (point)
11568 paren-state))
11569 (if placeholder
11570 (setq tmpsymbol '(brace-list-open . inexpr-class))
11571 (setq tmpsymbol '(block-open . inexpr-statement)
11572 placeholder
11573 (cdr-safe (c-looking-at-inexpr-block
11574 (c-safe-position containing-sexp
11575 paren-state)
11576 containing-sexp)))
11577 ;; placeholder is nil if it's a block directly in
11578 ;; a function arglist. That makes us skip out of
11579 ;; this case.
11580 )))
11581 (goto-char placeholder)
11582 (back-to-indentation)
11583 (c-add-stmt-syntax (car tmpsymbol) nil t
11584 (c-most-enclosing-brace paren-state (point))
11585 paren-state)
11586 (if (/= (point) placeholder)
11587 (c-add-syntax (cdr tmpsymbol))))
11588
11589 ;; CASE 7C: we are looking at the first argument in an empty
11590 ;; argument list. Use arglist-close if we're actually
11591 ;; looking at a close paren or bracket.
11592 ((memq char-before-ip '(?\( ?\[))
11593 (goto-char containing-sexp)
11594 (setq placeholder (c-point 'boi))
11595 (if (and (c-safe (backward-up-list 1) t)
11596 (>= (point) placeholder))
11597 (progn
11598 (forward-char)
11599 (skip-chars-forward " \t"))
11600 (goto-char placeholder))
11601 (c-add-stmt-syntax 'arglist-intro (list containing-sexp) t
11602 (c-most-enclosing-brace paren-state (point))
11603 paren-state))
11604
11605 ;; CASE 7D: we are inside a conditional test clause. treat
11606 ;; these things as statements
11607 ((progn
11608 (goto-char containing-sexp)
11609 (and (c-safe (c-forward-sexp -1) t)
11610 (looking-at "\\<for\\>[^_]")))
11611 (goto-char (1+ containing-sexp))
11612 (c-forward-syntactic-ws indent-point)
11613 (if (eq char-before-ip ?\;)
11614 (c-add-syntax 'statement (point))
11615 (c-add-syntax 'statement-cont (point))
11616 ))
11617
11618 ;; CASE 7E: maybe a continued ObjC method call. This is the
11619 ;; case when we are inside a [] bracketed exp, and what
11620 ;; precede the opening bracket is not an identifier.
11621 ((and c-opt-method-key
11622 (eq (char-after containing-sexp) ?\[)
11623 (progn
11624 (goto-char (1- containing-sexp))
11625 (c-backward-syntactic-ws (c-point 'bod))
11626 (if (not (looking-at c-symbol-key))
11627 (c-add-syntax 'objc-method-call-cont containing-sexp))
11628 )))
11629
11630 ;; CASE 7F: we are looking at an arglist continuation line,
11631 ;; but the preceding argument is on the same line as the
11632 ;; opening paren. This case includes multi-line
11633 ;; mathematical paren groupings, but we could be on a
11634 ;; for-list continuation line. C.f. case 7A.
11635 ((progn
11636 (goto-char (1+ containing-sexp))
11637 (< (save-excursion
11638 (c-forward-syntactic-ws)
11639 (point))
11640 (c-point 'bonl)))
11641 (goto-char containing-sexp) ; paren opening the arglist
11642 (setq placeholder (c-point 'boi))
11643 (if (and (c-safe (backward-up-list 1) t)
11644 (>= (point) placeholder))
11645 (progn
11646 (forward-char)
11647 (skip-chars-forward " \t"))
11648 (goto-char placeholder))
11649 (c-add-stmt-syntax 'arglist-cont-nonempty (list containing-sexp) t
11650 (c-most-enclosing-brace c-state-cache (point))
11651 paren-state))
11652
11653 ;; CASE 7G: we are looking at just a normal arglist
11654 ;; continuation line
11655 (t (c-forward-syntactic-ws indent-point)
11656 (c-add-syntax 'arglist-cont (c-point 'boi)))
11657 ))
11658
11659 ;; CASE 8: func-local multi-inheritance line
11660 ((and (c-major-mode-is 'c++-mode)
11661 (save-excursion
11662 (goto-char indent-point)
11663 (skip-chars-forward " \t")
11664 (looking-at c-opt-postfix-decl-spec-key)))
11665 (goto-char indent-point)
11666 (skip-chars-forward " \t")
11667 (cond
11668
11669 ;; CASE 8A: non-hanging colon on an inher intro
11670 ((eq char-after-ip ?:)
11671 (c-backward-syntactic-ws lim)
11672 (c-add-syntax 'inher-intro (c-point 'boi)))
11673
11674 ;; CASE 8B: hanging colon on an inher intro
11675 ((eq char-before-ip ?:)
11676 (c-add-syntax 'inher-intro (c-point 'boi)))
11677
11678 ;; CASE 8C: a continued inheritance line
11679 (t
11680 (c-beginning-of-inheritance-list lim)
11681 (c-add-syntax 'inher-cont (point))
11682 )))
11683
11684 ;; CASE 9: we are inside a brace-list
11685 ((and (not (c-major-mode-is 'awk-mode)) ; Maybe this isn't needed (ACM, 2002/3/29)
11686 (setq special-brace-list
11687 (or (and c-special-brace-lists ;;;; ALWAYS NIL FOR AWK!!
11688 (save-excursion
11689 (goto-char containing-sexp)
11690 (c-looking-at-special-brace-list)))
11691 (c-inside-bracelist-p containing-sexp paren-state))))
11692 (cond
11693
11694 ;; CASE 9A: In the middle of a special brace list opener.
11695 ((and (consp special-brace-list)
11696 (save-excursion
11697 (goto-char containing-sexp)
11698 (eq (char-after) ?\())
11699 (eq char-after-ip (car (cdr special-brace-list))))
11700 (goto-char (car (car special-brace-list)))
11701 (skip-chars-backward " \t")
11702 (if (and (bolp)
11703 (assoc 'statement-cont
11704 (setq placeholder (c-guess-basic-syntax))))
11705 (setq c-syntactic-context placeholder)
11706 (c-beginning-of-statement-1
11707 (c-safe-position (1- containing-sexp) paren-state))
11708 (c-forward-token-2 0)
11709 (while (cond
11710 ((looking-at c-specifier-key)
11711 (c-forward-keyword-clause 1))
11712 ((and c-opt-cpp-prefix
11713 (looking-at c-noise-macro-with-parens-name-re))
11714 (c-forward-noise-clause))))
11715 (c-add-syntax 'brace-list-open (c-point 'boi))))
11716
11717 ;; CASE 9B: brace-list-close brace
11718 ((if (consp special-brace-list)
11719 ;; Check special brace list closer.
11720 (progn
11721 (goto-char (car (car special-brace-list)))
11722 (save-excursion
11723 (goto-char indent-point)
11724 (back-to-indentation)
11725 (or
11726 ;; We were between the special close char and the `)'.
11727 (and (eq (char-after) ?\))
11728 (eq (1+ (point)) (cdr (car special-brace-list))))
11729 ;; We were before the special close char.
11730 (and (eq (char-after) (cdr (cdr special-brace-list)))
11731 (zerop (c-forward-token-2))
11732 (eq (1+ (point)) (cdr (car special-brace-list)))))))
11733 ;; Normal brace list check.
11734 (and (eq char-after-ip ?})
11735 (c-safe (goto-char (c-up-list-backward (point))) t)
11736 (= (point) containing-sexp)))
11737 (if (eq (point) (c-point 'boi))
11738 (c-add-syntax 'brace-list-close (point))
11739 (setq lim (c-most-enclosing-brace c-state-cache (point)))
11740 (c-beginning-of-statement-1 lim nil nil t)
11741 (c-add-stmt-syntax 'brace-list-close nil t lim paren-state)))
11742
11743 (t
11744 ;; Prepare for the rest of the cases below by going to the
11745 ;; token following the opening brace
11746 (if (consp special-brace-list)
11747 (progn
11748 (goto-char (car (car special-brace-list)))
11749 (c-forward-token-2 1 nil indent-point))
11750 (goto-char containing-sexp))
11751 (forward-char)
11752 (let ((start (point)))
11753 (c-forward-syntactic-ws indent-point)
11754 (goto-char (max start (c-point 'bol))))
11755 (c-skip-ws-forward indent-point)
11756 (cond
11757
11758 ;; CASE 9C: we're looking at the first line in a brace-list
11759 ((= (point) indent-point)
11760 (if (consp special-brace-list)
11761 (goto-char (car (car special-brace-list)))
11762 (goto-char containing-sexp))
11763 (if (eq (point) (c-point 'boi))
11764 (c-add-syntax 'brace-list-intro (point))
11765 (setq lim (c-most-enclosing-brace c-state-cache (point)))
11766 (c-beginning-of-statement-1 lim)
11767 (c-add-stmt-syntax 'brace-list-intro nil t lim paren-state)))
11768
11769 ;; CASE 9D: this is just a later brace-list-entry or
11770 ;; brace-entry-open
11771 (t (if (or (eq char-after-ip ?{)
11772 (and c-special-brace-lists
11773 (save-excursion
11774 (goto-char indent-point)
11775 (c-forward-syntactic-ws (c-point 'eol))
11776 (c-looking-at-special-brace-list (point)))))
11777 (c-add-syntax 'brace-entry-open (point))
11778 (c-add-syntax 'brace-list-entry (point))
11779 ))
11780 ))))
11781
11782 ;; CASE 10: A continued statement or top level construct.
11783 ((and (not (memq char-before-ip '(?\; ?:)))
11784 (not (c-at-vsemi-p before-ws-ip))
11785 (or (not (eq char-before-ip ?}))
11786 (c-looking-at-inexpr-block-backward c-state-cache))
11787 (> (point)
11788 (save-excursion
11789 (c-beginning-of-statement-1 containing-sexp)
11790 (setq placeholder (point))))
11791 (/= placeholder containing-sexp))
11792 ;; This is shared with case 18.
11793 (c-guess-continued-construct indent-point
11794 char-after-ip
11795 placeholder
11796 containing-sexp
11797 paren-state))
11798
11799 ;; CASE 16: block close brace, possibly closing the defun or
11800 ;; the class
11801 ((eq char-after-ip ?})
11802 ;; From here on we have the next containing sexp in lim.
11803 (setq lim (c-most-enclosing-brace paren-state))
11804 (goto-char containing-sexp)
11805 (cond
11806
11807 ;; CASE 16E: Closing a statement block? This catches
11808 ;; cases where it's preceded by a statement keyword,
11809 ;; which works even when used in an "invalid" context,
11810 ;; e.g. a macro argument.
11811 ((c-after-conditional)
11812 (c-backward-to-block-anchor lim)
11813 (c-add-stmt-syntax 'block-close nil t lim paren-state))
11814
11815 ;; CASE 16A: closing a lambda defun or an in-expression
11816 ;; block? C.f. cases 4, 7B and 17E.
11817 ((setq placeholder (c-looking-at-inexpr-block
11818 (c-safe-position containing-sexp paren-state)
11819 nil))
11820 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
11821 'inline-close
11822 'block-close))
11823 (goto-char containing-sexp)
11824 (back-to-indentation)
11825 (if (= containing-sexp (point))
11826 (c-add-syntax tmpsymbol (point))
11827 (goto-char (cdr placeholder))
11828 (back-to-indentation)
11829 (c-add-stmt-syntax tmpsymbol nil t
11830 (c-most-enclosing-brace paren-state (point))
11831 paren-state)
11832 (if (/= (point) (cdr placeholder))
11833 (c-add-syntax (car placeholder)))))
11834
11835 ;; CASE 16B: does this close an inline or a function in
11836 ;; a non-class declaration level block?
11837 ((save-excursion
11838 (and lim
11839 (progn
11840 (goto-char lim)
11841 (c-looking-at-decl-block
11842 (c-most-enclosing-brace paren-state lim)
11843 nil))
11844 (setq placeholder (point))))
11845 (c-backward-to-decl-anchor lim)
11846 (back-to-indentation)
11847 (if (save-excursion
11848 (goto-char placeholder)
11849 (looking-at c-other-decl-block-key))
11850 (c-add-syntax 'defun-close (point))
11851 (c-add-syntax 'inline-close (point))))
11852
11853 ;; CASE 16F: Can be a defun-close of a function declared
11854 ;; in a statement block, e.g. in Pike or when using gcc
11855 ;; extensions, but watch out for macros followed by
11856 ;; blocks. Let it through to be handled below.
11857 ;; C.f. cases B.3 and 17G.
11858 ((save-excursion
11859 (and (not (c-at-statement-start-p))
11860 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
11861 (setq placeholder (point))
11862 (let ((c-recognize-typeless-decls nil))
11863 ;; Turn off recognition of constructs that
11864 ;; lacks a type in this case, since that's more
11865 ;; likely to be a macro followed by a block.
11866 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
11867 (back-to-indentation)
11868 (if (/= (point) containing-sexp)
11869 (goto-char placeholder))
11870 (c-add-stmt-syntax 'defun-close nil t lim paren-state))
11871
11872 ;; CASE 16C: If there is an enclosing brace then this is
11873 ;; a block close since defun closes inside declaration
11874 ;; level blocks have been handled above.
11875 (lim
11876 ;; If the block is preceded by a case/switch label on
11877 ;; the same line, we anchor at the first preceding label
11878 ;; at boi. The default handling in c-add-stmt-syntax
11879 ;; really fixes it better, but we do like this to keep
11880 ;; the indentation compatible with version 5.28 and
11881 ;; earlier. C.f. case 17H.
11882 (while (and (/= (setq placeholder (point)) (c-point 'boi))
11883 (eq (c-beginning-of-statement-1 lim) 'label)))
11884 (goto-char placeholder)
11885 (if (looking-at c-label-kwds-regexp)
11886 (c-add-syntax 'block-close (point))
11887 (goto-char containing-sexp)
11888 ;; c-backward-to-block-anchor not necessary here; those
11889 ;; situations are handled in case 16E above.
11890 (c-add-stmt-syntax 'block-close nil t lim paren-state)))
11891
11892 ;; CASE 16D: Only top level defun close left.
11893 (t
11894 (goto-char containing-sexp)
11895 (c-backward-to-decl-anchor lim)
11896 (c-add-stmt-syntax 'defun-close nil nil
11897 (c-most-enclosing-brace paren-state)
11898 paren-state))
11899 ))
11900
11901 ;; CASE 19: line is an expression, not a statement, and is directly
11902 ;; contained by a template delimiter. Most likely, we are in a
11903 ;; template arglist within a statement. This case is based on CASE
11904 ;; 7. At some point in the future, we may wish to create more
11905 ;; syntactic symbols such as `template-intro',
11906 ;; `template-cont-nonempty', etc., and distinguish between them as we
11907 ;; do for `arglist-intro' etc. (2009-12-07).
11908 ((and c-recognize-<>-arglists
11909 (setq containing-< (c-up-list-backward indent-point containing-sexp))
11910 (eq (char-after containing-<) ?\<))
11911 (setq placeholder (c-point 'boi containing-<))
11912 (goto-char containing-sexp) ; Most nested Lbrace/Lparen (but not
11913 ; '<') before indent-point.
11914 (if (>= (point) placeholder)
11915 (progn
11916 (forward-char)
11917 (skip-chars-forward " \t"))
11918 (goto-char placeholder))
11919 (c-add-stmt-syntax 'template-args-cont (list containing-<) t
11920 (c-most-enclosing-brace c-state-cache (point))
11921 paren-state))
11922
11923 ;; CASE 17: Statement or defun catchall.
11924 (t
11925 (goto-char indent-point)
11926 ;; Back up statements until we find one that starts at boi.
11927 (while (let* ((prev-point (point))
11928 (last-step-type (c-beginning-of-statement-1
11929 containing-sexp)))
11930 (if (= (point) prev-point)
11931 (progn
11932 (setq step-type (or step-type last-step-type))
11933 nil)
11934 (setq step-type last-step-type)
11935 (/= (point) (c-point 'boi)))))
11936 (cond
11937
11938 ;; CASE 17B: continued statement
11939 ((and (eq step-type 'same)
11940 (/= (point) indent-point))
11941 (c-add-stmt-syntax 'statement-cont nil nil
11942 containing-sexp paren-state))
11943
11944 ;; CASE 17A: After a case/default label?
11945 ((progn
11946 (while (and (eq step-type 'label)
11947 (not (looking-at c-label-kwds-regexp)))
11948 (setq step-type
11949 (c-beginning-of-statement-1 containing-sexp)))
11950 (eq step-type 'label))
11951 (c-add-stmt-syntax (if (eq char-after-ip ?{)
11952 'statement-case-open
11953 'statement-case-intro)
11954 nil t containing-sexp paren-state))
11955
11956 ;; CASE 17D: any old statement
11957 ((progn
11958 (while (eq step-type 'label)
11959 (setq step-type
11960 (c-beginning-of-statement-1 containing-sexp)))
11961 (eq step-type 'previous))
11962 (c-add-stmt-syntax 'statement nil t
11963 containing-sexp paren-state)
11964 (if (eq char-after-ip ?{)
11965 (c-add-syntax 'block-open)))
11966
11967 ;; CASE 17I: Inside a substatement block.
11968 ((progn
11969 ;; The following tests are all based on containing-sexp.
11970 (goto-char containing-sexp)
11971 ;; From here on we have the next containing sexp in lim.
11972 (setq lim (c-most-enclosing-brace paren-state containing-sexp))
11973 (c-after-conditional))
11974 (c-backward-to-block-anchor lim)
11975 (c-add-stmt-syntax 'statement-block-intro nil t
11976 lim paren-state)
11977 (if (eq char-after-ip ?{)
11978 (c-add-syntax 'block-open)))
11979
11980 ;; CASE 17E: first statement in an in-expression block.
11981 ;; C.f. cases 4, 7B and 16A.
11982 ((setq placeholder (c-looking-at-inexpr-block
11983 (c-safe-position containing-sexp paren-state)
11984 nil))
11985 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
11986 'defun-block-intro
11987 'statement-block-intro))
11988 (back-to-indentation)
11989 (if (= containing-sexp (point))
11990 (c-add-syntax tmpsymbol (point))
11991 (goto-char (cdr placeholder))
11992 (back-to-indentation)
11993 (c-add-stmt-syntax tmpsymbol nil t
11994 (c-most-enclosing-brace c-state-cache (point))
11995 paren-state)
11996 (if (/= (point) (cdr placeholder))
11997 (c-add-syntax (car placeholder))))
11998 (if (eq char-after-ip ?{)
11999 (c-add-syntax 'block-open)))
12000
12001 ;; CASE 17F: first statement in an inline, or first
12002 ;; statement in a top-level defun. we can tell this is it
12003 ;; if there are no enclosing braces that haven't been
12004 ;; narrowed out by a class (i.e. don't use bod here).
12005 ((save-excursion
12006 (or (not (setq placeholder (c-most-enclosing-brace
12007 paren-state)))
12008 (and (progn
12009 (goto-char placeholder)
12010 (eq (char-after) ?{))
12011 (c-looking-at-decl-block (c-most-enclosing-brace
12012 paren-state (point))
12013 nil))))
12014 (c-backward-to-decl-anchor lim)
12015 (back-to-indentation)
12016 (c-add-syntax 'defun-block-intro (point)))
12017
12018 ;; CASE 17G: First statement in a function declared inside
12019 ;; a normal block. This can occur in Pike and with
12020 ;; e.g. the gcc extensions, but watch out for macros
12021 ;; followed by blocks. C.f. cases B.3 and 16F.
12022 ((save-excursion
12023 (and (not (c-at-statement-start-p))
12024 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
12025 (setq placeholder (point))
12026 (let ((c-recognize-typeless-decls nil))
12027 ;; Turn off recognition of constructs that lacks
12028 ;; a type in this case, since that's more likely
12029 ;; to be a macro followed by a block.
12030 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
12031 (back-to-indentation)
12032 (if (/= (point) containing-sexp)
12033 (goto-char placeholder))
12034 (c-add-stmt-syntax 'defun-block-intro nil t
12035 lim paren-state))
12036
12037 ;; CASE 17H: First statement in a block.
12038 (t
12039 ;; If the block is preceded by a case/switch label on the
12040 ;; same line, we anchor at the first preceding label at
12041 ;; boi. The default handling in c-add-stmt-syntax is
12042 ;; really fixes it better, but we do like this to keep the
12043 ;; indentation compatible with version 5.28 and earlier.
12044 ;; C.f. case 16C.
12045 (while (and (/= (setq placeholder (point)) (c-point 'boi))
12046 (eq (c-beginning-of-statement-1 lim) 'label)))
12047 (goto-char placeholder)
12048 (if (looking-at c-label-kwds-regexp)
12049 (c-add-syntax 'statement-block-intro (point))
12050 (goto-char containing-sexp)
12051 ;; c-backward-to-block-anchor not necessary here; those
12052 ;; situations are handled in case 17I above.
12053 (c-add-stmt-syntax 'statement-block-intro nil t
12054 lim paren-state))
12055 (if (eq char-after-ip ?{)
12056 (c-add-syntax 'block-open)))
12057 ))
12058 )
12059
12060 ;; now we need to look at any modifiers
12061 (goto-char indent-point)
12062 (skip-chars-forward " \t")
12063
12064 ;; are we looking at a comment only line?
12065 (when (and (looking-at c-comment-start-regexp)
12066 (/= (c-forward-token-2 0 nil (c-point 'eol)) 0))
12067 (c-append-syntax 'comment-intro))
12068
12069 ;; we might want to give additional offset to friends (in C++).
12070 (when (and c-opt-friend-key
12071 (looking-at c-opt-friend-key))
12072 (c-append-syntax 'friend))
12073
12074 ;; Set syntactic-relpos.
12075 (let ((p c-syntactic-context))
12076 (while (and p
12077 (if (integerp (c-langelem-pos (car p)))
12078 (progn
12079 (setq syntactic-relpos (c-langelem-pos (car p)))
12080 nil)
12081 t))
12082 (setq p (cdr p))))
12083
12084 ;; Start of or a continuation of a preprocessor directive?
12085 (if (and macro-start
12086 (eq macro-start (c-point 'boi))
12087 (not (and (c-major-mode-is 'pike-mode)
12088 (eq (char-after (1+ macro-start)) ?\"))))
12089 (c-append-syntax 'cpp-macro)
12090 (when (and c-syntactic-indentation-in-macros macro-start)
12091 (if in-macro-expr
12092 (when (or
12093 (< syntactic-relpos macro-start)
12094 (not (or
12095 (assq 'arglist-intro c-syntactic-context)
12096 (assq 'arglist-cont c-syntactic-context)
12097 (assq 'arglist-cont-nonempty c-syntactic-context)
12098 (assq 'arglist-close c-syntactic-context))))
12099 ;; If inside a cpp expression, i.e. anywhere in a
12100 ;; cpp directive except a #define body, we only let
12101 ;; through the syntactic analysis that is internal
12102 ;; in the expression. That means the arglist
12103 ;; elements, if they are anchored inside the cpp
12104 ;; expression.
12105 (setq c-syntactic-context nil)
12106 (c-add-syntax 'cpp-macro-cont macro-start))
12107 (when (and (eq macro-start syntactic-relpos)
12108 (not (assq 'cpp-define-intro c-syntactic-context))
12109 (save-excursion
12110 (goto-char macro-start)
12111 (or (not (c-forward-to-cpp-define-body))
12112 (<= (point) (c-point 'boi indent-point)))))
12113 ;; Inside a #define body and the syntactic analysis is
12114 ;; anchored on the start of the #define. In this case
12115 ;; we add cpp-define-intro to get the extra
12116 ;; indentation of the #define body.
12117 (c-add-syntax 'cpp-define-intro)))))
12118
12119 ;; return the syntax
12120 c-syntactic-context)))
12121
12122 \f
12123 ;; Indentation calculation.
12124
12125 (defun c-evaluate-offset (offset langelem symbol)
12126 ;; offset can be a number, a function, a variable, a list, or one of
12127 ;; the symbols + or -
12128 ;;
12129 ;; This function might do hidden buffer changes.
12130 (let ((res
12131 (cond
12132 ((numberp offset) offset)
12133 ((vectorp offset) offset)
12134 ((null offset) nil)
12135
12136 ((eq offset '+) c-basic-offset)
12137 ((eq offset '-) (- c-basic-offset))
12138 ((eq offset '++) (* 2 c-basic-offset))
12139 ((eq offset '--) (* 2 (- c-basic-offset)))
12140 ((eq offset '*) (/ c-basic-offset 2))
12141 ((eq offset '/) (/ (- c-basic-offset) 2))
12142
12143 ((functionp offset)
12144 (c-evaluate-offset
12145 (funcall offset
12146 (cons (c-langelem-sym langelem)
12147 (c-langelem-pos langelem)))
12148 langelem symbol))
12149
12150 ((listp offset)
12151 (cond
12152 ((eq (car offset) 'quote)
12153 (c-benign-error "The offset %S for %s was mistakenly quoted"
12154 offset symbol)
12155 nil)
12156
12157 ((memq (car offset) '(min max))
12158 (let (res val (method (car offset)))
12159 (setq offset (cdr offset))
12160 (while offset
12161 (setq val (c-evaluate-offset (car offset) langelem symbol))
12162 (cond
12163 ((not val))
12164 ((not res)
12165 (setq res val))
12166 ((integerp val)
12167 (if (vectorp res)
12168 (c-benign-error "\
12169 Error evaluating offset %S for %s: \
12170 Cannot combine absolute offset %S with relative %S in `%s' method"
12171 (car offset) symbol res val method)
12172 (setq res (funcall method res val))))
12173 (t
12174 (if (integerp res)
12175 (c-benign-error "\
12176 Error evaluating offset %S for %s: \
12177 Cannot combine relative offset %S with absolute %S in `%s' method"
12178 (car offset) symbol res val method)
12179 (setq res (vector (funcall method (aref res 0)
12180 (aref val 0)))))))
12181 (setq offset (cdr offset)))
12182 res))
12183
12184 ((eq (car offset) 'add)
12185 (let (res val)
12186 (setq offset (cdr offset))
12187 (while offset
12188 (setq val (c-evaluate-offset (car offset) langelem symbol))
12189 (cond
12190 ((not val))
12191 ((not res)
12192 (setq res val))
12193 ((integerp val)
12194 (if (vectorp res)
12195 (setq res (vector (+ (aref res 0) val)))
12196 (setq res (+ res val))))
12197 (t
12198 (if (vectorp res)
12199 (c-benign-error "\
12200 Error evaluating offset %S for %s: \
12201 Cannot combine absolute offsets %S and %S in `add' method"
12202 (car offset) symbol res val)
12203 (setq res val)))) ; Override.
12204 (setq offset (cdr offset)))
12205 res))
12206
12207 (t
12208 (let (res)
12209 (when (eq (car offset) 'first)
12210 (setq offset (cdr offset)))
12211 (while (and (not res) offset)
12212 (setq res (c-evaluate-offset (car offset) langelem symbol)
12213 offset (cdr offset)))
12214 res))))
12215
12216 ((and (symbolp offset) (boundp offset))
12217 (symbol-value offset))
12218
12219 (t
12220 (c-benign-error "Unknown offset format %S for %s" offset symbol)
12221 nil))))
12222
12223 (if (or (null res) (integerp res)
12224 (and (vectorp res) (= (length res) 1) (integerp (aref res 0))))
12225 res
12226 (c-benign-error "Error evaluating offset %S for %s: Got invalid value %S"
12227 offset symbol res)
12228 nil)))
12229
12230 (defun c-calc-offset (langelem)
12231 ;; Get offset from LANGELEM which is a list beginning with the
12232 ;; syntactic symbol and followed by any analysis data it provides.
12233 ;; That data may be zero or more elements, but if at least one is
12234 ;; given then the first is the anchor position (or nil). The symbol
12235 ;; is matched against `c-offsets-alist' and the offset calculated
12236 ;; from that is returned.
12237 ;;
12238 ;; This function might do hidden buffer changes.
12239 (let* ((symbol (c-langelem-sym langelem))
12240 (match (assq symbol c-offsets-alist))
12241 (offset (cdr-safe match)))
12242 (if match
12243 (setq offset (c-evaluate-offset offset langelem symbol))
12244 (if c-strict-syntax-p
12245 (c-benign-error "No offset found for syntactic symbol %s" symbol))
12246 (setq offset 0))
12247 (if (vectorp offset)
12248 offset
12249 (or (and (numberp offset) offset)
12250 (and (symbolp offset) (symbol-value offset))
12251 0))
12252 ))
12253
12254 (defun c-get-offset (langelem)
12255 ;; This is a compatibility wrapper for `c-calc-offset' in case
12256 ;; someone is calling it directly. It takes an old style syntactic
12257 ;; element on the form (SYMBOL . ANCHOR-POS) and converts it to the
12258 ;; new list form.
12259 ;;
12260 ;; This function might do hidden buffer changes.
12261 (if (c-langelem-pos langelem)
12262 (c-calc-offset (list (c-langelem-sym langelem)
12263 (c-langelem-pos langelem)))
12264 (c-calc-offset langelem)))
12265
12266 (defun c-get-syntactic-indentation (langelems)
12267 ;; Calculate the syntactic indentation from a syntactic description
12268 ;; as returned by `c-guess-syntax'.
12269 ;;
12270 ;; Note that topmost-intro always has an anchor position at bol, for
12271 ;; historical reasons. It's often used together with other symbols
12272 ;; that has more sane positions. Since we always use the first
12273 ;; found anchor position, we rely on that these other symbols always
12274 ;; precede topmost-intro in the LANGELEMS list.
12275 ;;
12276 ;; This function might do hidden buffer changes.
12277 (let ((indent 0) anchor)
12278
12279 (while langelems
12280 (let* ((c-syntactic-element (car langelems))
12281 (res (c-calc-offset c-syntactic-element)))
12282
12283 (if (vectorp res)
12284 ;; Got an absolute column that overrides any indentation
12285 ;; we've collected so far, but not the relative
12286 ;; indentation we might get for the nested structures
12287 ;; further down the langelems list.
12288 (setq indent (elt res 0)
12289 anchor (point-min)) ; A position at column 0.
12290
12291 ;; Got a relative change of the current calculated
12292 ;; indentation.
12293 (setq indent (+ indent res))
12294
12295 ;; Use the anchor position from the first syntactic
12296 ;; element with one.
12297 (unless anchor
12298 (setq anchor (c-langelem-pos (car langelems)))))
12299
12300 (setq langelems (cdr langelems))))
12301
12302 (if anchor
12303 (+ indent (save-excursion
12304 (goto-char anchor)
12305 (current-column)))
12306 indent)))
12307
12308 \f
12309 (cc-provide 'cc-engine)
12310
12311 ;; Local Variables:
12312 ;; indent-tabs-mode: t
12313 ;; tab-width: 8
12314 ;; End:
12315 ;;; cc-engine.el ends here