GNU Emacs is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
-the Free Software Foundation; either version 1, or (at your option)
+the Free Software Foundation; either version 2, or (at your option)
any later version.
GNU Emacs is distributed in the hope that it will be useful,
You should have received a copy of the GNU General Public License
along with GNU Emacs; see the file COPYING. If not, write to
-the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. */
+the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+Boston, MA 02111-1307, USA. */
#include <config.h>
#include "lisp.h"
#include "syntax.h"
+#include "category.h"
#include "buffer.h"
+#include "charset.h"
#include "region-cache.h"
#include "commands.h"
#include "blockinput.h"
+#include "intervals.h"
#include <sys/types.h>
#include "regex.h"
-#define REGEXP_CACHE_SIZE 5
+#define REGEXP_CACHE_SIZE 20
/* If the regexp is non-nil, then the buffer contains the compiled form
of that regexp, suitable for searching. */
-struct regexp_cache {
+struct regexp_cache
+{
struct regexp_cache *next;
Lisp_Object regexp;
struct re_pattern_buffer buf;
Qnil if no searching has been done yet. */
static Lisp_Object last_thing_searched;
-/* error condition signalled when regexp compile_pattern fails */
+/* error condition signaled when regexp compile_pattern fails */
Lisp_Object Qinvalid_regexp;
If it is 0, we should compile the pattern not to record any
subexpression bounds.
POSIX is nonzero if we want full backtracking (POSIX style)
- for this pattern. 0 means backtrack only enough to get a valid match. */
+ for this pattern. 0 means backtrack only enough to get a valid match.
+ MULTIBYTE is nonzero if we want to handle multibyte characters in
+ PATTERN. 0 means all multibyte characters are recognized just as
+ sequences of binary data. */
static void
-compile_pattern_1 (cp, pattern, translate, regp, posix)
+compile_pattern_1 (cp, pattern, translate, regp, posix, multibyte)
struct regexp_cache *cp;
Lisp_Object pattern;
- char *translate;
+ Lisp_Object *translate;
struct re_registers *regp;
int posix;
+ int multibyte;
{
- CONST char *val;
+ char *val;
reg_syntax_t old;
cp->regexp = Qnil;
cp->buf.translate = translate;
cp->posix = posix;
+ cp->buf.multibyte = multibyte;
BLOCK_INPUT;
old = re_set_syntax (RE_SYNTAX_EMACS
| (posix ? 0 : RE_NO_POSIX_BACKTRACKING));
- val = (CONST char *) re_compile_pattern ((char *) XSTRING (pattern)->data,
- XSTRING (pattern)->size, &cp->buf);
+ val = (char *) re_compile_pattern ((char *) XSTRING (pattern)->data,
+ XSTRING (pattern)->size, &cp->buf);
re_set_syntax (old);
UNBLOCK_INPUT;
if (val)
compile_pattern (pattern, regp, translate, posix)
Lisp_Object pattern;
struct re_registers *regp;
- char *translate;
+ Lisp_Object *translate;
int posix;
{
struct regexp_cache *cp, **cpp;
+ /* Should we check it here, or add an argument `multibyte' to this
+ function? */
+ int multibyte = !NILP (current_buffer->enable_multibyte_characters);
for (cpp = &searchbuf_head; ; cpp = &cp->next)
{
cp = *cpp;
- if (!NILP (Fstring_equal (cp->regexp, pattern))
+ if (XSTRING (cp->regexp)->size == XSTRING (pattern)->size
+ && !NILP (Fstring_equal (cp->regexp, pattern))
&& cp->buf.translate == translate
- && cp->posix == posix)
+ && cp->posix == posix
+ && cp->buf.multibyte == multibyte)
break;
/* If we're at the end of the cache, compile into the last cell. */
if (cp->next == 0)
{
- compile_pattern_1 (cp, pattern, translate, regp, posix);
+ compile_pattern_1 (cp, pattern, translate, regp, posix, multibyte);
break;
}
}
CHECK_STRING (string, 0);
bufp = compile_pattern (string, &search_regs,
(!NILP (current_buffer->case_fold_search)
- ? DOWNCASE_TABLE : 0),
+ ? XCHAR_TABLE (DOWNCASE_TABLE)->contents : 0),
posix);
immediate_quit = 1;
that make up the visible portion of the buffer. */
p1 = BEGV_ADDR;
- s1 = GPT - BEGV;
+ s1 = GPT_BYTE - BEGV_BYTE;
p2 = GAP_END_ADDR;
- s2 = ZV - GPT;
+ s2 = ZV_BYTE - GPT_BYTE;
if (s1 < 0)
{
p2 = p1;
- s2 = ZV - BEGV;
+ s2 = ZV_BYTE - BEGV_BYTE;
s1 = 0;
}
if (s2 < 0)
{
- s1 = ZV - BEGV;
+ s1 = ZV_BYTE - BEGV_BYTE;
s2 = 0;
}
+
+ re_match_object = Qnil;
i = re_match_2 (bufp, (char *) p1, s1, (char *) p2, s2,
- point - BEGV, &search_regs,
- ZV - BEGV);
+ PT_BYTE - BEGV_BYTE, &search_regs,
+ ZV_BYTE - BEGV_BYTE);
if (i == -2)
matcher_overflow ();
val = (0 <= i ? Qt : Qnil);
- for (i = 0; i < search_regs.num_regs; i++)
- if (search_regs.start[i] >= 0)
- {
- search_regs.start[i] += BEGV;
- search_regs.end[i] += BEGV;
- }
+ if (i >= 0)
+ for (i = 0; i < search_regs.num_regs; i++)
+ if (search_regs.start[i] >= 0)
+ {
+ search_regs.start[i]
+ = BYTE_TO_CHAR (search_regs.start[i] + BEGV_BYTE);
+ search_regs.end[i]
+ = BYTE_TO_CHAR (search_regs.end[i] + BEGV_BYTE);
+ }
XSETBUFFER (last_thing_searched, current_buffer);
immediate_quit = 0;
return val;
bufp = compile_pattern (regexp, &search_regs,
(!NILP (current_buffer->case_fold_search)
- ? DOWNCASE_TABLE : 0),
- 0);
+ ? XCHAR_TABLE (DOWNCASE_TABLE)->contents : 0),
+ posix);
immediate_quit = 1;
+ re_match_object = string;
+
val = re_search (bufp, (char *) XSTRING (string)->data,
XSTRING (string)->size, s, XSTRING (string)->size - s,
&search_regs);
bufp = compile_pattern (regexp, 0, 0, 0);
immediate_quit = 1;
+ re_match_object = string;
+
val = re_search (bufp, (char *) XSTRING (string)->data,
XSTRING (string)->size, 0, XSTRING (string)->size,
0);
immediate_quit = 0;
return val;
}
+
+/* Match REGEXP against STRING, searching all of STRING ignoring case,
+ and return the index of the match, or negative on failure.
+ This does not clobber the match data. */
+
+extern Lisp_Object Vascii_downcase_table;
+
+int
+fast_c_string_match_ignore_case (regexp, string)
+ Lisp_Object regexp;
+ char *string;
+{
+ int val;
+ struct re_pattern_buffer *bufp;
+ int len = strlen (string);
+
+ re_match_object = Qt;
+ bufp = compile_pattern (regexp, 0,
+ XCHAR_TABLE (Vascii_downcase_table)->contents, 0);
+ immediate_quit = 1;
+ val = re_search (bufp, string, len, 0, len, 0);
+ immediate_quit = 0;
+ return val;
+}
\f
/* max and min. */
the position of the last character before the next such
obstacle --- the last character the dumb search loop should
examine. */
- register int ceiling = end - 1;
+ int ceiling_byte = CHAR_TO_BYTE (end) - 1;
+ int start_byte = CHAR_TO_BYTE (start);
/* If we're looking for a newline, consult the newline cache
to see where we can avoid some scanning. */
int next_change;
immediate_quit = 0;
while (region_cache_forward
- (current_buffer, newline_cache, start, &next_change))
- start = next_change;
+ (current_buffer, newline_cache, start_byte, &next_change))
+ start_byte = next_change;
immediate_quit = allow_quit;
- /* start should never be after end. */
- if (start >= end)
- start = end - 1;
+ /* START should never be after END. */
+ if (start_byte > ceiling_byte)
+ start_byte = ceiling_byte;
/* Now the text after start is an unknown region, and
next_change is the position of the next known region. */
- ceiling = min (next_change - 1, ceiling);
+ ceiling_byte = min (next_change - 1, ceiling_byte);
}
/* The dumb loop can only scan text stored in contiguous
bytes. BUFFER_CEILING_OF returns the last character
position that is contiguous, so the ceiling is the
position after that. */
- ceiling = min (BUFFER_CEILING_OF (start), ceiling);
+ ceiling_byte = min (BUFFER_CEILING_OF (start_byte), ceiling_byte);
{
/* The termination address of the dumb loop. */
- register unsigned char *ceiling_addr = &FETCH_CHAR (ceiling) + 1;
- register unsigned char *cursor = &FETCH_CHAR (start);
+ register unsigned char *ceiling_addr
+ = BYTE_POS_ADDR (ceiling_byte) + 1;
+ register unsigned char *cursor
+ = BYTE_POS_ADDR (start_byte);
unsigned char *base = cursor;
while (cursor < ceiling_addr)
the region from start to cursor is free of them. */
if (target == '\n' && newline_cache)
know_region_cache (current_buffer, newline_cache,
- start + scan_start - base,
- start + cursor - base);
+ start_byte + scan_start - base,
+ start_byte + cursor - base);
/* Did we find the target character? */
if (cursor < ceiling_addr)
if (--count == 0)
{
immediate_quit = 0;
- return (start + cursor - base + 1);
+ return BYTE_TO_CHAR (start_byte + cursor - base + 1);
}
cursor++;
}
}
- start += cursor - base;
+ start = BYTE_TO_CHAR (start_byte + cursor - base);
}
}
else
while (start > end)
{
/* The last character to check before the next obstacle. */
- register int ceiling = end;
+ int ceiling_byte = CHAR_TO_BYTE (end);
+ int start_byte = CHAR_TO_BYTE (start);
/* Consult the newline cache, if appropriate. */
if (target == '\n' && newline_cache)
int next_change;
immediate_quit = 0;
while (region_cache_backward
- (current_buffer, newline_cache, start, &next_change))
- start = next_change;
+ (current_buffer, newline_cache, start_byte, &next_change))
+ start_byte = next_change;
immediate_quit = allow_quit;
/* Start should never be at or before end. */
- if (start <= end)
- start = end + 1;
+ if (start_byte <= ceiling_byte)
+ start_byte = ceiling_byte + 1;
/* Now the text before start is an unknown region, and
next_change is the position of the next known region. */
- ceiling = max (next_change, ceiling);
+ ceiling_byte = max (next_change, ceiling_byte);
}
/* Stop scanning before the gap. */
- ceiling = max (BUFFER_FLOOR_OF (start - 1), ceiling);
+ ceiling_byte = max (BUFFER_FLOOR_OF (start_byte - 1), ceiling_byte);
{
/* The termination address of the dumb loop. */
- register unsigned char *ceiling_addr = &FETCH_CHAR (ceiling);
- register unsigned char *cursor = &FETCH_CHAR (start - 1);
+ register unsigned char *ceiling_addr = BYTE_POS_ADDR (ceiling_byte);
+ register unsigned char *cursor = BYTE_POS_ADDR (start_byte - 1);
unsigned char *base = cursor;
while (cursor >= ceiling_addr)
the region from after the cursor to start is free of them. */
if (target == '\n' && newline_cache)
know_region_cache (current_buffer, newline_cache,
- start + cursor - base,
- start + scan_start - base);
+ start_byte + cursor - base,
+ start_byte + scan_start - base);
/* Did we find the target character? */
if (cursor >= ceiling_addr)
if (++count >= 0)
{
immediate_quit = 0;
- return (start + cursor - base);
+ return BYTE_TO_CHAR (start_byte + cursor - base);
}
cursor--;
}
}
- start += cursor - base;
+ start = BYTE_TO_CHAR (start_byte + cursor - base);
}
}
*shortage = count * direction;
return start;
}
-
-int
-find_next_newline_no_quit (from, cnt)
- register int from, cnt;
-{
- return scan_buffer ('\n', from, 0, cnt, (int *) 0, 0);
-}
-
-int
-find_next_newline (from, cnt)
- register int from, cnt;
-{
- return scan_buffer ('\n', from, 0, cnt, (int *) 0, 1);
-}
-
-
-/* Like find_next_newline, but returns position before the newline,
- not after, and only search up to TO. This isn't just
- find_next_newline (...)-1, because you might hit TO. */
-int
-find_before_next_newline (from, to, cnt)
- int from, to, cnt;
-{
- int shortage;
- int pos = scan_buffer ('\n', from, to, cnt, &shortage, 1);
-
- if (shortage == 0)
- pos--;
-
- return pos;
-}
\f
-Lisp_Object skip_chars ();
-
-DEFUN ("skip-chars-forward", Fskip_chars_forward, Sskip_chars_forward, 1, 2, 0,
- "Move point forward, stopping before a char not in STRING, or at pos LIM.\n\
-STRING is like the inside of a `[...]' in a regular expression\n\
-except that `]' is never special and `\\' quotes `^', `-' or `\\'.\n\
-Thus, with arg \"a-zA-Z\", this skips letters stopping before first nonletter.\n\
-With arg \"^a-zA-Z\", skips nonletters stopping before first letter.\n\
-Returns the distance traveled, either zero or positive.")
- (string, lim)
- Lisp_Object string, lim;
-{
- return skip_chars (1, 0, string, lim);
-}
+/* Search for COUNT instances of a line boundary, which means either a
+ newline or (if selective display enabled) a carriage return.
+ Start at START. If COUNT is negative, search backwards.
-DEFUN ("skip-chars-backward", Fskip_chars_backward, Sskip_chars_backward, 1, 2, 0,
- "Move point backward, stopping after a char not in STRING, or at pos LIM.\n\
-See `skip-chars-forward' for details.\n\
-Returns the distance traveled, either zero or negative.")
- (string, lim)
- Lisp_Object string, lim;
-{
- return skip_chars (0, 0, string, lim);
-}
+ We report the resulting position by calling TEMP_SET_PT_BOTH.
-DEFUN ("skip-syntax-forward", Fskip_syntax_forward, Sskip_syntax_forward, 1, 2, 0,
- "Move point forward across chars in specified syntax classes.\n\
-SYNTAX is a string of syntax code characters.\n\
-Stop before a char whose syntax is not in SYNTAX, or at position LIM.\n\
-If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX.\n\
-This function returns the distance traveled, either zero or positive.")
- (syntax, lim)
- Lisp_Object syntax, lim;
-{
- return skip_chars (1, 1, syntax, lim);
-}
+ If we find COUNT instances. we position after (always after,
+ even if scanning backwards) the COUNTth match, and return 0.
-DEFUN ("skip-syntax-backward", Fskip_syntax_backward, Sskip_syntax_backward, 1, 2, 0,
- "Move point backward across chars in specified syntax classes.\n\
-SYNTAX is a string of syntax code characters.\n\
-Stop on reaching a char whose syntax is not in SYNTAX, or at position LIM.\n\
-If SYNTAX starts with ^, skip characters whose syntax is NOT in SYNTAX.\n\
-This function returns the distance traveled, either zero or negative.")
- (syntax, lim)
- Lisp_Object syntax, lim;
-{
- return skip_chars (0, 1, syntax, lim);
-}
+ If we don't find COUNT instances before reaching the end of the
+ buffer (or the beginning, if scanning backwards), we return
+ the number of line boundaries left unfound, and position at
+ the limit we bumped up against.
-Lisp_Object
-skip_chars (forwardp, syntaxp, string, lim)
- int forwardp, syntaxp;
- Lisp_Object string, lim;
+ If ALLOW_QUIT is non-zero, set immediate_quit. That's good to do
+ except when inside redisplay. */
+
+int
+scan_newline (start, start_byte, limit, limit_byte, count, allow_quit)
+ int start, start_byte;
+ int limit, limit_byte;
+ register int count;
+ int allow_quit;
{
- register unsigned char *p, *pend;
- register unsigned char c;
- unsigned char fastmap[0400];
- int negate = 0;
- register int i;
+ int direction = ((count > 0) ? 1 : -1);
- CHECK_STRING (string, 0);
+ register unsigned char *cursor;
+ unsigned char *base;
- if (NILP (lim))
- XSETINT (lim, forwardp ? ZV : BEGV);
- else
- CHECK_NUMBER_COERCE_MARKER (lim, 1);
+ register int ceiling;
+ register unsigned char *ceiling_addr;
- /* In any case, don't allow scan outside bounds of buffer. */
- /* jla turned this off, for no known reason.
- bfox turned the ZV part on, and rms turned the
- BEGV part back on. */
- if (XINT (lim) > ZV)
- XSETFASTINT (lim, ZV);
- if (XINT (lim) < BEGV)
- XSETFASTINT (lim, BEGV);
+ /* If we are not in selective display mode,
+ check only for newlines. */
+ int selective_display = (!NILP (current_buffer->selective_display)
+ && !INTEGERP (current_buffer->selective_display));
- p = XSTRING (string)->data;
- pend = p + XSTRING (string)->size;
- bzero (fastmap, sizeof fastmap);
+ /* The code that follows is like scan_buffer
+ but checks for either newline or carriage return. */
- if (p != pend && *p == '^')
- {
- negate = 1; p++;
- }
+ immediate_quit = allow_quit;
- /* Find the characters specified and set their elements of fastmap.
- If syntaxp, each character counts as itself.
- Otherwise, handle backslashes and ranges specially */
+ start_byte = CHAR_TO_BYTE (start);
- while (p != pend)
+ if (count > 0)
{
- c = *p++;
- if (syntaxp)
- fastmap[c] = 1;
- else
+ while (start_byte < limit_byte)
{
- if (c == '\\')
+ ceiling = BUFFER_CEILING_OF (start_byte);
+ ceiling = min (limit_byte - 1, ceiling);
+ ceiling_addr = BYTE_POS_ADDR (ceiling) + 1;
+ base = (cursor = BYTE_POS_ADDR (start_byte));
+ while (1)
{
- if (p == pend) break;
- c = *p++;
+ while (*cursor != '\n' && ++cursor != ceiling_addr)
+ ;
+
+ if (cursor != ceiling_addr)
+ {
+ if (--count == 0)
+ {
+ immediate_quit = 0;
+ start_byte = start_byte + cursor - base + 1;
+ start = BYTE_TO_CHAR (start_byte);
+ TEMP_SET_PT_BOTH (start, start_byte);
+ return 0;
+ }
+ else
+ if (++cursor == ceiling_addr)
+ break;
+ }
+ else
+ break;
}
- if (p != pend && *p == '-')
+ start_byte += cursor - base;
+ }
+ }
+ else
+ {
+ int start_byte = CHAR_TO_BYTE (start);
+ while (start_byte > limit_byte)
+ {
+ ceiling = BUFFER_FLOOR_OF (start_byte - 1);
+ ceiling = max (limit_byte, ceiling);
+ ceiling_addr = BYTE_POS_ADDR (ceiling) - 1;
+ base = (cursor = BYTE_POS_ADDR (start_byte - 1) + 1);
+ while (1)
{
- p++;
- if (p == pend) break;
- while (c <= *p)
+ while (--cursor != ceiling_addr && *cursor != '\n')
+ ;
+
+ if (cursor != ceiling_addr)
{
- fastmap[c] = 1;
- c++;
+ if (++count == 0)
+ {
+ immediate_quit = 0;
+ /* Return the position AFTER the match we found. */
+ start_byte = start_byte + cursor - base + 1;
+ start = BYTE_TO_CHAR (start_byte);
+ TEMP_SET_PT_BOTH (start, start_byte);
+ return 0;
+ }
}
- p++;
+ else
+ break;
}
- else
- fastmap[c] = 1;
+ /* Here we add 1 to compensate for the last decrement
+ of CURSOR, which took it past the valid range. */
+ start_byte += cursor - base + 1;
}
}
- if (syntaxp && fastmap['-'] != 0)
- fastmap[' '] = 1;
+ TEMP_SET_PT_BOTH (limit, limit_byte);
- /* If ^ was the first character, complement the fastmap. */
-
- if (negate)
- for (i = 0; i < sizeof fastmap; i++)
- fastmap[i] ^= 1;
+ return count * direction;
+}
- {
- int start_point = point;
+int
+find_next_newline_no_quit (from, cnt)
+ register int from, cnt;
+{
+ return scan_buffer ('\n', from, 0, cnt, (int *) 0, 0);
+}
- immediate_quit = 1;
- if (syntaxp)
- {
+/* Like find_next_newline, but returns position before the newline,
+ not after, and only search up to TO. This isn't just
+ find_next_newline (...)-1, because you might hit TO. */
- if (forwardp)
- {
- while (point < XINT (lim)
- && fastmap[(unsigned char) syntax_code_spec[(int) SYNTAX (FETCH_CHAR (point))]])
- SET_PT (point + 1);
- }
- else
- {
- while (point > XINT (lim)
- && fastmap[(unsigned char) syntax_code_spec[(int) SYNTAX (FETCH_CHAR (point - 1))]])
- SET_PT (point - 1);
- }
- }
- else
- {
- if (forwardp)
- {
- while (point < XINT (lim) && fastmap[FETCH_CHAR (point)])
- SET_PT (point + 1);
- }
- else
- {
- while (point > XINT (lim) && fastmap[FETCH_CHAR (point - 1)])
- SET_PT (point - 1);
- }
- }
- immediate_quit = 0;
+int
+find_before_next_newline (from, to, cnt)
+ int from, to, cnt;
+{
+ int shortage;
+ int pos = scan_buffer ('\n', from, to, cnt, &shortage, 1);
- return make_number (point - start_point);
- }
+ if (shortage == 0)
+ pos--;
+
+ return pos;
}
\f
/* Subroutines of Lisp buffer search functions. */
{
CHECK_NUMBER_COERCE_MARKER (bound, 1);
lim = XINT (bound);
- if (n > 0 ? lim < point : lim > point)
+ if (n > 0 ? lim < PT : lim > PT)
error ("Invalid search bound (wrong side of point)");
if (lim > ZV)
lim = ZV;
lim = BEGV;
}
- np = search_buffer (string, point, lim, n, RE,
+ np = search_buffer (string, PT, lim, n, RE,
(!NILP (current_buffer->case_fold_search)
- ? XSTRING (current_buffer->case_canon_table)->data : 0),
+ ? XCHAR_TABLE (current_buffer->case_canon_table)->contents
+ : 0),
(!NILP (current_buffer->case_fold_search)
- ? XSTRING (current_buffer->case_eqv_table)->data : 0),
+ ? XCHAR_TABLE (current_buffer->case_eqv_table)->contents
+ : 0),
posix);
if (np <= 0)
{
return make_number (np);
}
\f
+/* Return 1 if REGEXP it matches just one constant string. */
+
static int
trivial_regexp_p (regexp)
Lisp_Object regexp;
{
case '|': case '(': case ')': case '`': case '\'': case 'b':
case 'B': case '<': case '>': case 'w': case 'W': case 's':
- case 'S': case '1': case '2': case '3': case '4': case '5':
+ case 'S': case '=':
+ case 'c': case 'C': /* for categoryspec and notcategoryspec */
+ case '1': case '2': case '3': case '4': case '5':
case '6': case '7': case '8': case '9':
return 0;
}
int lim;
int n;
int RE;
- register unsigned char *trt;
- register unsigned char *inverse_trt;
+ Lisp_Object *trt;
+ Lisp_Object *inverse_trt;
int posix;
{
int len = XSTRING (string)->size;
{
struct re_pattern_buffer *bufp;
- bufp = compile_pattern (string, &search_regs, (char *) trt, posix);
+ bufp = compile_pattern (string, &search_regs, trt, posix);
immediate_quit = 1; /* Quit immediately if user types ^G,
because letting this function finish
that make up the visible portion of the buffer. */
p1 = BEGV_ADDR;
- s1 = GPT - BEGV;
+ s1 = GPT_BYTE - BEGV_BYTE;
p2 = GAP_END_ADDR;
- s2 = ZV - GPT;
+ s2 = ZV_BYTE - GPT_BYTE;
if (s1 < 0)
{
p2 = p1;
- s2 = ZV - BEGV;
+ s2 = ZV_BYTE - BEGV_BYTE;
s1 = 0;
}
if (s2 < 0)
{
- s1 = ZV - BEGV;
+ s1 = ZV_BYTE - BEGV_BYTE;
s2 = 0;
}
+ re_match_object = Qnil;
+
while (n < 0)
{
int val;
}
if (val >= 0)
{
- j = BEGV;
for (i = 0; i < search_regs.num_regs; i++)
if (search_regs.start[i] >= 0)
{
- search_regs.start[i] += j;
- search_regs.end[i] += j;
+ search_regs.start[i]
+ = BYTE_TO_CHAR (search_regs.start[i] + BEGV_BYTE);
+ search_regs.end[i]
+ = BYTE_TO_CHAR (search_regs.end[i] + BEGV_BYTE);
}
XSETBUFFER (last_thing_searched, current_buffer);
/* Set pos to the new position. */
}
if (val >= 0)
{
- j = BEGV;
for (i = 0; i < search_regs.num_regs; i++)
if (search_regs.start[i] >= 0)
{
- search_regs.start[i] += j;
- search_regs.end[i] += j;
+ search_regs.start[i]
+ = BYTE_TO_CHAR (search_regs.start[i] + BEGV_BYTE);
+ search_regs.end[i]
+ = BYTE_TO_CHAR (search_regs.end[i] + BEGV_BYTE);
}
XSETBUFFER (last_thing_searched, current_buffer);
pos = search_regs.end[0];
}
else /* non-RE case */
{
+ int pos_byte = CHAR_TO_BYTE (pos);
+ int lim_byte = CHAR_TO_BYTE (lim);
#ifdef C_ALLOCA
int BM_tab_space[0400];
BM_tab = &BM_tab_space[0];
len--;
base_pat++;
}
- *pat++ = (trt ? trt[*base_pat++] : *base_pat++);
+ *pat++ = (trt ? XINT (trt[*base_pat++]) : *base_pat++);
}
len = pat - patbuf;
pat = base_pat = patbuf;
/* it entirely if there is none. */
dirlen = len * direction;
- infinity = dirlen - (lim + pos + len + len) * direction;
+ infinity = dirlen - (lim_byte + pos_byte + len + len) * direction;
if (direction < 0)
pat = (base_pat += len - 1);
BM_tab_base = BM_tab;
if (i == dirlen) i = infinity;
if (trt != 0)
{
- k = (j = trt[j]);
+ k = (j = XINT (trt[j]));
if (i == infinity)
stride_for_teases = BM_tab[j];
BM_tab[j] = dirlen - i;
/* A translation table is accompanied by its inverse -- see */
/* comment following downcase_table for details */
- while ((j = inverse_trt[j]) != k)
+ while ((j = (unsigned char) XINT (inverse_trt[j])) != k)
BM_tab[j] = dirlen - i;
}
else
/* different. */
}
infinity = dirlen - infinity;
- pos += dirlen - ((direction > 0) ? direction : 0);
- /* loop invariant - pos points at where last char (first char if reverse)
- of pattern would align in a possible match. */
+ pos_byte += dirlen - ((direction > 0) ? direction : 0);
+ /* loop invariant - POS_BYTE points at where last char (first
+ char if reverse) of pattern would align in a possible match. */
while (n != 0)
{
/* It's been reported that some (broken) compiler thinks that
Boolean expressions in an arithmetic context are unsigned.
Using an explicit ?1:0 prevents this. */
- if ((lim - pos - ((direction > 0) ? 1 : 0)) * direction < 0)
+ if ((lim_byte - pos_byte - ((direction > 0) ? 1 : 0)) * direction
+ < 0)
return (n * (0 - direction));
/* First we do the part we can by pointers (maybe nothing) */
QUIT;
pat = base_pat;
- limit = pos - dirlen + direction;
+ limit = pos_byte - dirlen + direction;
limit = ((direction > 0)
? BUFFER_CEILING_OF (limit)
: BUFFER_FLOOR_OF (limit));
- /* LIMIT is now the last (not beyond-last!) value
- POS can take on without hitting edge of buffer or the gap. */
+ /* LIMIT is now the last (not beyond-last!) value POS_BYTE
+ can take on without hitting edge of buffer or the gap. */
limit = ((direction > 0)
- ? min (lim - 1, min (limit, pos + 20000))
- : max (lim, max (limit, pos - 20000)));
- if ((limit - pos) * direction > 20)
+ ? min (lim_byte - 1, min (limit, pos_byte + 20000))
+ : max (lim_byte, max (limit, pos_byte - 20000)));
+ if ((limit - pos_byte) * direction > 20)
{
- p_limit = &FETCH_CHAR (limit);
- p2 = (cursor = &FETCH_CHAR (pos));
+ p_limit = BYTE_POS_ADDR (limit);
+ p2 = (cursor = BYTE_POS_ADDR (pos_byte));
/* In this loop, pos + cursor - p2 is the surrogate for pos */
while (1) /* use one cursor setting as long as i can */
{
while ((EMACS_INT) cursor <= (EMACS_INT) p_limit)
cursor += BM_tab[*cursor];
else
- while ((unsigned EMACS_INT) cursor <= (unsigned EMACS_INT) p_limit)
+ while ((EMACS_UINT) cursor <= (EMACS_UINT) p_limit)
cursor += BM_tab[*cursor];
}
else
while ((EMACS_INT) cursor >= (EMACS_INT) p_limit)
cursor += BM_tab[*cursor];
else
- while ((unsigned EMACS_INT) cursor >= (unsigned EMACS_INT) p_limit)
+ while ((EMACS_UINT) cursor >= (EMACS_UINT) p_limit)
cursor += BM_tab[*cursor];
}
/* If you are here, cursor is beyond the end of the searched region. */
if (trt != 0)
{
while ((i -= direction) + direction != 0)
- if (pat[i] != trt[*(cursor -= direction)])
+ if (pat[i] != XINT (trt[*(cursor -= direction)]))
break;
}
else
{
cursor -= direction;
- set_search_regs (pos + cursor - p2 + ((direction > 0)
+ set_search_regs (pos_byte + cursor - p2 + ((direction > 0)
? 1 - len : 0),
len);
else
cursor += stride_for_teases; /* <sigh> we lose - */
}
- pos += cursor - p2;
+ pos_byte += cursor - p2;
}
else
/* Now we'll pick up a clump that has to be done the hard */
/* way because it covers a discontinuity */
{
limit = ((direction > 0)
- ? BUFFER_CEILING_OF (pos - dirlen + 1)
- : BUFFER_FLOOR_OF (pos - dirlen - 1));
+ ? BUFFER_CEILING_OF (pos_byte - dirlen + 1)
+ : BUFFER_FLOOR_OF (pos_byte - dirlen - 1));
limit = ((direction > 0)
- ? min (limit + len, lim - 1)
- : max (limit - len, lim));
- /* LIMIT is now the last value POS can have
+ ? min (limit + len, lim_byte - 1)
+ : max (limit - len, lim_byte));
+ /* LIMIT is now the last value POS_BYTE can have
and still be valid for a possible match. */
while (1)
{
/* speed because it will usually run only once. */
/* (the reach is at most len + 21, and typically */
/* does not exceed len) */
- while ((limit - pos) * direction >= 0)
- pos += BM_tab[FETCH_CHAR(pos)];
+ while ((limit - pos_byte) * direction >= 0)
+ pos_byte += BM_tab[FETCH_BYTE (pos_byte)];
/* now run the same tests to distinguish going off the */
/* end, a match or a phony match. */
- if ((pos - limit) * direction <= len)
+ if ((pos_byte - limit) * direction <= len)
break; /* ran off the end */
/* Found what might be a match.
- Set POS back to last (first if reverse) char pos. */
- pos -= infinity;
+ Set POS_BYTE back to last (first if reverse) pos. */
+ pos_byte -= infinity;
i = dirlen - direction;
while ((i -= direction) + direction != 0)
{
- pos -= direction;
+ pos_byte -= direction;
if (pat[i] != (trt != 0
- ? trt[FETCH_CHAR(pos)]
- : FETCH_CHAR (pos)))
+ ? XINT (trt[FETCH_BYTE (pos_byte)])
+ : FETCH_BYTE (pos_byte)))
break;
}
- /* Above loop has moved POS part or all the way
- back to the first char pos (last char pos if reverse).
+ /* Above loop has moved POS_BYTE part or all the way
+ back to the first pos (last pos if reverse).
Set it once again at the last (first if reverse) char. */
- pos += dirlen - i- direction;
+ pos_byte += dirlen - i- direction;
if (i + direction == 0)
{
- pos -= direction;
+ pos_byte -= direction;
- set_search_regs (pos + ((direction > 0) ? 1 - len : 0),
+ set_search_regs (pos_byte + ((direction > 0) ? 1 - len : 0),
len);
if ((n -= direction) != 0)
- pos += dirlen; /* to resume search */
+ pos_byte += dirlen; /* to resume search */
else
return ((direction > 0)
? search_regs.end[0] : search_regs.start[0]);
}
else
- pos += stride_for_teases;
+ pos_byte += stride_for_teases;
}
}
/* We have done one clump. Can we continue? */
- if ((lim - pos) * direction < 0)
+ if ((lim_byte - pos_byte) * direction < 0)
return ((0 - n) * direction);
}
- return pos;
+ return BYTE_TO_CHAR (pos_byte);
}
}
-/* Record beginning BEG and end BEG + LEN
+/* Record beginning BEG_BYTE and end BEG_BYTE + NBYTES
for a match just found in the current buffer. */
static void
-set_search_regs (beg, len)
- int beg, len;
+set_search_regs (beg_byte, nbytes)
+ int beg_byte, nbytes;
{
/* Make sure we have registers in which to store
the match position. */
search_regs.num_regs = 2;
}
- search_regs.start[0] = beg;
- search_regs.end[0] = beg + len;
+ search_regs.start[0] = BYTE_TO_CHAR (beg_byte);
+ search_regs.end[0] = BYTE_TO_CHAR (beg_byte + nbytes);
XSETBUFFER (last_thing_searched, current_buffer);
}
\f
}
\f
DEFUN ("search-backward", Fsearch_backward, Ssearch_backward, 1, 4,
- "sSearch backward: ",
+ "MSearch backward: ",
"Search backward from point for STRING.\n\
Set point to the beginning of the occurrence found, and return point.\n\
An optional second argument bounds the search; it is a buffer position.\n\
return search_command (string, bound, noerror, count, -1, 0, 0);
}
-DEFUN ("search-forward", Fsearch_forward, Ssearch_forward, 1, 4, "sSearch: ",
+DEFUN ("search-forward", Fsearch_forward, Ssearch_forward, 1, 4, "MSearch: ",
"Search forward from point for STRING.\n\
Set point to the end of the occurrence found, and return point.\n\
An optional second argument bounds the search; it is a buffer position.\n\
return search_command (regexp, bound, noerror, count, 1, 1, 1);
}
\f
-DEFUN ("replace-match", Freplace_match, Sreplace_match, 1, 4, 0,
+DEFUN ("replace-match", Freplace_match, Sreplace_match, 1, 5, 0,
"Replace text matched by last search with NEWTEXT.\n\
If second arg FIXEDCASE is non-nil, do not alter case of replacement text.\n\
Otherwise maybe capitalize the whole text, or maybe just word initials,\n\
\n\
The optional fourth argument STRING can be a string to modify.\n\
In that case, this function creates and returns a new string\n\
-which is made by replacing the part of STRING that was matched.")
- (newtext, fixedcase, literal, string)
- Lisp_Object newtext, fixedcase, literal, string;
+which is made by replacing the part of STRING that was matched.\n\
+\n\
+The optional fifth argument SUBEXP specifies a subexpression of the match.\n\
+It says to replace just that subexpression instead of the whole match.\n\
+This is useful only after a regular expression search or match\n\
+since only regular expressions have distinguished subexpressions.")
+ (newtext, fixedcase, literal, string, subexp)
+ Lisp_Object newtext, fixedcase, literal, string, subexp;
{
enum { nochange, all_caps, cap_initial } case_action;
register int pos, last;
int some_nonuppercase_initial;
register int c, prevc;
int inslen;
+ int sub;
+ int opoint, newpoint;
CHECK_STRING (newtext, 0);
if (search_regs.num_regs <= 0)
error ("replace-match called before any match found");
+ if (NILP (subexp))
+ sub = 0;
+ else
+ {
+ CHECK_NUMBER (subexp, 3);
+ sub = XINT (subexp);
+ if (sub < 0 || sub >= search_regs.num_regs)
+ args_out_of_range (subexp, make_number (search_regs.num_regs));
+ }
+
if (NILP (string))
{
- if (search_regs.start[0] < BEGV
- || search_regs.start[0] > search_regs.end[0]
- || search_regs.end[0] > ZV)
- args_out_of_range (make_number (search_regs.start[0]),
- make_number (search_regs.end[0]));
+ if (search_regs.start[sub] < BEGV
+ || search_regs.start[sub] > search_regs.end[sub]
+ || search_regs.end[sub] > ZV)
+ args_out_of_range (make_number (search_regs.start[sub]),
+ make_number (search_regs.end[sub]));
}
else
{
- if (search_regs.start[0] < 0
- || search_regs.start[0] > search_regs.end[0]
- || search_regs.end[0] > XSTRING (string)->size)
- args_out_of_range (make_number (search_regs.start[0]),
- make_number (search_regs.end[0]));
+ if (search_regs.start[sub] < 0
+ || search_regs.start[sub] > search_regs.end[sub]
+ || search_regs.end[sub] > XSTRING (string)->size)
+ args_out_of_range (make_number (search_regs.start[sub]),
+ make_number (search_regs.end[sub]));
}
if (NILP (fixedcase))
{
+ int beg;
/* Decide how to casify by examining the matched text. */
- last = search_regs.end[0];
+ if (NILP (string))
+ last = CHAR_TO_BYTE (search_regs.end[sub]);
+ else
+ last = search_regs.end[sub];
+
+ if (NILP (string))
+ beg = CHAR_TO_BYTE (search_regs.start[sub]);
+ else
+ beg = search_regs.start[sub];
+
prevc = '\n';
case_action = all_caps;
some_nonuppercase_initial = 0;
some_uppercase = 0;
- for (pos = search_regs.start[0]; pos < last; pos++)
+ for (pos = beg; pos < last; pos++)
{
if (NILP (string))
- c = FETCH_CHAR (pos);
+ c = FETCH_BYTE (pos);
else
c = XSTRING (string)->data[pos];
Lisp_Object before, after;
before = Fsubstring (string, make_number (0),
- make_number (search_regs.start[0]));
- after = Fsubstring (string, make_number (search_regs.end[0]), Qnil);
+ make_number (search_regs.start[sub]));
+ after = Fsubstring (string, make_number (search_regs.end[sub]), Qnil);
- /* Do case substitution into NEWTEXT if desired. */
+ /* Substitute parts of the match into NEWTEXT
+ if desired. */
if (NILP (literal))
{
int lastpos = -1;
{
int substart = -1;
int subend;
+ int delbackslash = 0;
c = XSTRING (newtext)->data[pos];
if (c == '\\')
c = XSTRING (newtext)->data[++pos];
if (c == '&')
{
- substart = search_regs.start[0];
- subend = search_regs.end[0];
+ substart = search_regs.start[sub];
+ subend = search_regs.end[sub];
}
else if (c >= '1' && c <= '9' && c <= search_regs.num_regs + '0')
{
- if (search_regs.start[c - '0'] >= 1)
+ if (search_regs.start[c - '0'] >= 0)
{
substart = search_regs.start[c - '0'];
subend = search_regs.end[c - '0'];
}
}
+ else if (c == '\\')
+ delbackslash = 1;
+ else
+ error ("Invalid use of `\\' in replacement text");
}
if (substart >= 0)
{
if (pos - 1 != lastpos + 1)
- middle = Fsubstring (newtext, lastpos + 1, pos - 1);
+ middle = Fsubstring (newtext,
+ make_number (lastpos + 1),
+ make_number (pos - 1));
else
middle = Qnil;
accum = concat3 (accum, middle,
make_number (subend)));
lastpos = pos;
}
+ else if (delbackslash)
+ {
+ middle = Fsubstring (newtext, make_number (lastpos + 1),
+ make_number (pos));
+ accum = concat2 (accum, middle);
+ lastpos = pos;
+ }
}
if (pos != lastpos + 1)
- middle = Fsubstring (newtext, lastpos + 1, pos);
+ middle = Fsubstring (newtext, make_number (lastpos + 1),
+ make_number (pos));
else
middle = Qnil;
newtext = concat2 (accum, middle);
}
+ /* Do case substitution in NEWTEXT if desired. */
if (case_action == all_caps)
newtext = Fupcase (newtext);
else if (case_action == cap_initial)
- newtext = upcase_initials (newtext);
+ newtext = Fupcase_initials (newtext);
return concat3 (before, newtext, after);
}
+ /* Record point, the move (quietly) to the start of the match. */
+ if (PT > search_regs.start[sub])
+ opoint = PT - ZV;
+ else
+ opoint = PT;
+
+ TEMP_SET_PT (search_regs.start[sub]);
+
/* We insert the replacement text before the old text, and then
delete the original text. This means that markers at the
beginning or end of the original will float to the corresponding
position in the replacement. */
- SET_PT (search_regs.start[0]);
if (!NILP (literal))
Finsert_and_inherit (1, &newtext);
else
for (pos = 0; pos < XSTRING (newtext)->size; pos++)
{
- int offset = point - search_regs.start[0];
+ int offset = PT - search_regs.start[sub];
c = XSTRING (newtext)->data[pos];
if (c == '\\')
if (c == '&')
Finsert_buffer_substring
(Fcurrent_buffer (),
- make_number (search_regs.start[0] + offset),
- make_number (search_regs.end[0] + offset));
+ make_number (search_regs.start[sub] + offset),
+ make_number (search_regs.end[sub] + offset));
else if (c >= '1' && c <= '9' && c <= search_regs.num_regs + '0')
{
if (search_regs.start[c - '0'] >= 1)
make_number (search_regs.start[c - '0'] + offset),
make_number (search_regs.end[c - '0'] + offset));
}
- else
+ else if (c == '\\')
insert_char (c);
+ else
+ error ("Invalid use of `\\' in replacement text");
}
else
insert_char (c);
UNGCPRO;
}
- inslen = point - (search_regs.start[0]);
- del_range (search_regs.start[0] + inslen, search_regs.end[0] + inslen);
+ inslen = PT - (search_regs.start[sub]);
+ del_range (search_regs.start[sub] + inslen, search_regs.end[sub] + inslen);
if (case_action == all_caps)
- Fupcase_region (make_number (point - inslen), make_number (point));
+ Fupcase_region (make_number (PT - inslen), make_number (PT));
else if (case_action == cap_initial)
- upcase_initials_region (make_number (point - inslen), make_number (point));
+ Fupcase_initials_region (make_number (PT - inslen), make_number (PT));
+
+ newpoint = PT;
+
+ /* Put point back where it was in the text. */
+ if (opoint <= 0)
+ TEMP_SET_PT (opoint + ZV);
+ else
+ TEMP_SET_PT (opoint);
+
+ /* Now move point "officially" to the start of the inserted replacement. */
+ move_if_not_intangible (newpoint);
+
return Qnil;
}
\f
DEFUN ("match-beginning", Fmatch_beginning, Smatch_beginning, 1, 1, 0,
"Return position of start of text matched by last search.\n\
-NUM specifies which parenthesized expression in the last regexp.\n\
- Value is nil if NUMth pair didn't match, or there were less than NUM pairs.\n\
+SUBEXP, a number, specifies which parenthesized expression in the last\n\
+ regexp.\n\
+Value is nil if SUBEXPth pair didn't match, or there were less than\n\
+ SUBEXP pairs.\n\
Zero means the entire text matched by the whole regexp or whole string.")
- (num)
- Lisp_Object num;
+ (subexp)
+ Lisp_Object subexp;
{
- return match_limit (num, 1);
+ return match_limit (subexp, 1);
}
DEFUN ("match-end", Fmatch_end, Smatch_end, 1, 1, 0,
"Return position of end of text matched by last search.\n\
-ARG, a number, specifies which parenthesized expression in the last regexp.\n\
- Value is nil if ARGth pair didn't match, or there were less than ARG pairs.\n\
+SUBEXP, a number, specifies which parenthesized expression in the last\n\
+ regexp.\n\
+Value is nil if SUBEXPth pair didn't match, or there were less than\n\
+ SUBEXP pairs.\n\
Zero means the entire text matched by the whole regexp or whole string.")
- (num)
- Lisp_Object num;
+ (subexp)
+ Lisp_Object subexp;
{
- return match_limit (num, 0);
+ return match_limit (subexp, 0);
}
-DEFUN ("match-data", Fmatch_data, Smatch_data, 0, 0, 0,
+DEFUN ("match-data", Fmatch_data, Smatch_data, 0, 2, 0,
"Return a list containing all info on what the last search matched.\n\
Element 2N is `(match-beginning N)'; element 2N + 1 is `(match-end N)'.\n\
All the elements are markers or nil (nil if the Nth pair didn't match)\n\
if the last match was on a buffer; integers or nil if a string was matched.\n\
-Use `store-match-data' to reinstate the data in this list.")
- ()
+Use `store-match-data' to reinstate the data in this list.\n\
+\n\
+If INTEGERS (the optional first argument) is non-nil, always use integers\n\
+\(rather than markers) to represent buffer positions.\n\
+If REUSE is a list, reuse it as part of the value. If REUSE is long enough\n\
+to hold all the values, and if INTEGERS is non-nil, no consing is done.")
+ (integers, reuse)
+ Lisp_Object integers, reuse;
{
+ Lisp_Object tail, prev;
Lisp_Object *data;
int i, len;
if (NILP (last_thing_searched))
- error ("match-data called before any match found");
+ return Qnil;
data = (Lisp_Object *) alloca ((2 * search_regs.num_regs)
* sizeof (Lisp_Object));
int start = search_regs.start[i];
if (start >= 0)
{
- if (EQ (last_thing_searched, Qt))
+ if (EQ (last_thing_searched, Qt)
+ || ! NILP (integers))
{
XSETFASTINT (data[2 * i], start);
XSETFASTINT (data[2 * i + 1], search_regs.end[i]);
else
data[2 * i] = data [2 * i + 1] = Qnil;
}
- return Flist (2 * len + 2, data);
+
+ /* If REUSE is not usable, cons up the values and return them. */
+ if (! CONSP (reuse))
+ return Flist (2 * len + 2, data);
+
+ /* If REUSE is a list, store as many value elements as will fit
+ into the elements of REUSE. */
+ for (i = 0, tail = reuse; CONSP (tail);
+ i++, tail = XCONS (tail)->cdr)
+ {
+ if (i < 2 * len + 2)
+ XCONS (tail)->car = data[i];
+ else
+ XCONS (tail)->car = Qnil;
+ prev = tail;
+ }
+
+ /* If we couldn't fit all value elements into REUSE,
+ cons up the rest of them and add them to the end of REUSE. */
+ if (i < 2 * len + 2)
+ XCONS (prev)->cdr = Flist (2 * len + 2 - i, data + i);
+
+ return reuse;
}
DEFUN ("regexp-quote", Fregexp_quote, Sregexp_quote, 1, 1, 0,
"Return a regexp string which matches exactly STRING and nothing else.")
- (str)
- Lisp_Object str;
+ (string)
+ Lisp_Object string;
{
register unsigned char *in, *out, *end;
register unsigned char *temp;
- CHECK_STRING (str, 0);
+ CHECK_STRING (string, 0);
- temp = (unsigned char *) alloca (XSTRING (str)->size * 2);
+ temp = (unsigned char *) alloca (XSTRING (string)->size * 2);
/* Now copy the data into the new string, inserting escapes. */
- in = XSTRING (str)->data;
- end = in + XSTRING (str)->size;
+ in = XSTRING (string)->data;
+ end = in + XSTRING (string)->size;
out = temp;
for (; in != end; in++)
defsubr (&Sposix_looking_at);
defsubr (&Sstring_match);
defsubr (&Sposix_string_match);
- defsubr (&Sskip_chars_forward);
- defsubr (&Sskip_chars_backward);
- defsubr (&Sskip_syntax_forward);
- defsubr (&Sskip_syntax_backward);
defsubr (&Ssearch_forward);
defsubr (&Ssearch_backward);
defsubr (&Sword_search_forward);