1 ;;; rfc822.el --- hairy rfc822 parser for mail and news and suchlike
3 ;; Copyright (C) 1986, 87, 1990 Free Software Foundation, Inc.
5 ;; Author: Richard Mlynarik <mly@eddie.mit.edu>
9 ;; This file is part of GNU Emacs.
11 ;; GNU Emacs is free software; you can redistribute it and/or modify
12 ;; it under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation; either version 2, or (at your option)
16 ;; GNU Emacs is distributed in the hope that it will be useful,
17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 ;; GNU General Public License for more details.
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with GNU Emacs; see the file COPYING. If not, write to the
23 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
24 ;; Boston, MA 02111-1307, USA.
28 ;; Support functions for parsing RFC-822 headers, used by mail and news
33 (defvar rfc822-address-start)
35 ;; uses rfc822-address-start free, throws to address
36 (defun rfc822-bad-address (reason)
39 (narrow-to-region rfc822-address-start
40 (if (re-search-forward "[,;]" nil t)
41 (max (point-min) (1- (point)))
43 ;; make the error string be suitable for inclusion in (...)
44 (let ((losers '("\\" "(" ")" "\n")))
46 (goto-char (point-min))
47 (while (search-forward (car losers) nil t)
51 (setq losers (cdr losers))))
52 (goto-char (point-min)) (insert "(Unparsable address -- "
55 (goto-char (point-max)) (insert "\")"))
56 (rfc822-nuke-whitespace)
57 (throw 'address (buffer-substring rfc822-address-start (point))))
59 (defun rfc822-nuke-whitespace (&optional leave-space)
63 ((= (setq ch (following-char)) ?\()
66 (rfc822-bad-address "Unbalanced comment (...)")
67 (/= (setq ch (following-char)) ?\)))
68 (cond ((looking-at "[^()\\]+")
71 (rfc822-nuke-whitespace))
72 ((< (point) (1- (point-max)))
75 (rfc822-bad-address "orphaned backslash"))))
76 ;; delete remaining "()"
80 ((memq ch '(?\ ?\t ?\n))
81 (delete-region (point)
82 (progn (skip-chars-forward " \t\n") (point)))
89 (= (preceding-char) ?\ )
92 (defun rfc822-looking-at (regex &optional leave-space)
93 (if (cond ((stringp regex)
94 (if (looking-at regex)
95 (progn (goto-char (match-end 0))
99 (= (following-char) regex))
100 (progn (forward-char 1)
102 (let ((tem (match-data)))
103 (rfc822-nuke-whitespace leave-space)
107 (defun rfc822-snarf-word ()
108 ;; word is atom | quoted-string
109 (cond ((= (following-char) ?\")
111 (or (rfc822-looking-at "\"\\([^\"\\\n]\\|\\\\.\\|\\\\\n\\)*\"")
112 (rfc822-bad-address "Unterminated quoted string")))
113 ((rfc822-looking-at "[^][\000-\037 ()<>@,;:\\\".]+")
117 (rfc822-bad-address "Rubbish in address"))))
119 (defun rfc822-snarf-words ()
121 (while (rfc822-looking-at ?.)
122 (rfc822-snarf-word)))
124 (defun rfc822-snarf-subdomain ()
125 ;; sub-domain is domain-ref | domain-literal
126 (cond ((= (following-char) ?\[)
128 (or (rfc822-looking-at "\\[\\([^][\\\n]\\|\\\\.\\|\\\\\n\\)*\\]")
129 (rfc822-bad-address "Unterminated domain literal [...]")))
130 ((rfc822-looking-at "[^][\000-\037 ()<>@,;:\\\".]+")
131 ;; domain-literal = atom
134 (rfc822-bad-address "Rubbish in host/domain specification"))))
136 (defun rfc822-snarf-domain ()
137 (rfc822-snarf-subdomain)
138 (while (rfc822-looking-at ?.)
139 (rfc822-snarf-subdomain)))
141 (defun rfc822-snarf-frob-list (name separator terminator snarfer
148 (format "End of addresses in middle of %s" name)))
149 ((rfc822-looking-at terminator)
151 ((rfc822-looking-at separator)
152 ;; multiple separators are allowed and do nothing.
153 (while (rfc822-looking-at separator))
159 (format "Gubbish in middle of %s" name))))
160 (setq tem (funcall snarfer)
163 (setq list (if (listp tem)
164 (nconc (reverse tem) list)
168 ;; return either an address (a string) or a list of addresses
169 (defun rfc822-addresses-1 (&optional allow-groups)
170 ;; Looking for an rfc822 `address'
171 ;; Either a group (1*word ":" [#mailbox] ";")
172 ;; or a mailbox (addr-spec | 1*word route-addr)
173 ;; addr-spec is (local-part "@" domain)
174 ;; route-addr is ("<" [1#("@" domain) ":"] addr-spec ">")
175 ;; local-part is (word *("." word))
176 ;; word is (atom | quoted-string)
177 ;; quoted-string is ("\([^\"\\n]\|\\.\|\\\n\)")
178 ;; atom is [^\000-\037\177 ()<>@,;:\".[]]+
179 ;; domain is sub-domain *("." sub-domain)
180 ;; sub-domain is domain-ref | domain-literal
181 ;; domain-literal is "[" *(dtext | quoted-pair) "]"
182 ;; dtext is "[^][\\n"
183 ;; domain-ref is atom
184 (let ((rfc822-address-start (point))
187 ;; optimize common cases:
190 ;; followed by "\\'\\|,\\|([^()\\]*)\\'"
191 ;; other common cases are:
192 ;; foo bar <foo.bar@baz.zap>
193 ;; "foo bar" <foo.bar@baz.zap>
194 ;; those aren't hacked yet.
195 (if (and (rfc822-looking-at "[^][\000-\037 ()<>@,;:\\\"]+\\(\\|@[^][\000-\037 ()<>@,;:\\\"]+\\)" t)
197 (rfc822-looking-at ?,))))
199 ;; rfc822-looking-at may have inserted a space
200 (or (bobp) (/= (preceding-char) ?\ ) (delete-char -1))
201 ;; relying on the fact that rfc822-looking-at <char>
202 ;; doesn't mung match-data
203 (throw 'address (buffer-substring rfc822-address-start (match-end 0)))))
204 (goto-char rfc822-address-start)
206 (cond ((and (= n 1) (rfc822-looking-at ?@))
208 (rfc822-snarf-domain)
210 (buffer-substring rfc822-address-start (point))))
211 ((rfc822-looking-at ?:)
212 (cond ((not allow-groups)
213 (rfc822-bad-address "A group name may not appear here"))
215 (rfc822-bad-address "No name for :...; group")))
218 ;; return a list of addresses
219 (rfc822-snarf-frob-list ":...; group" ?\, ?\;
220 'rfc822-addresses-1 t)))
221 ((rfc822-looking-at ?<)
222 (let ((start (point))
224 (cond ((rfc822-looking-at ?>)
227 ((and (not (eobp)) (= (following-char) ?\@))
228 ;; <@foo.bar,@baz:quux@abcd.efg>
229 (rfc822-snarf-frob-list "<...> address" ?\, ?\:
231 (if (rfc822-looking-at ?\@)
232 (rfc822-snarf-domain)
234 "Gubbish in route-addr")))))
236 (or (rfc822-looking-at ?@)
237 (rfc822-bad-address "Malformed <..@..> address"))
238 (rfc822-snarf-domain)
240 ((progn (rfc822-snarf-words) (rfc822-looking-at ?@))
241 ; allow <foo> (losing unix seems to do this)
242 (rfc822-snarf-domain)))
244 (if (rfc822-looking-at ?\>)
246 (buffer-substring (if strip start (1- start))
247 (if strip end (1+ end))))
248 (rfc822-bad-address "Unterminated <...> address")))))
249 ((looking-at "[^][\000-\037 ()<>@,;:\\.]")
250 ;; this allows "." to be part of the words preceding
251 ;; an addr-spec, since many broken mailers output
252 ;; "Hern K. Herklemeyer III
253 ;; <yank@megadeath.dod.gods-own-country>"
256 (or (= n 0) (bobp) (= (preceding-char) ?\ )
260 (setq again (or (rfc822-looking-at ?.)
261 (looking-at "[^][\000-\037 ()<>@,;:\\.]"))))))
263 (throw 'address nil))
264 ((= n 1) ; allow "foo" (losing unix seems to do this)
266 (buffer-substring rfc822-address-start (point))))
268 (rfc822-bad-address "Missing comma between addresses or badly-formatted address"))
269 ((or (eobp) (= (following-char) ?,))
270 (rfc822-bad-address "Missing comma or route-spec"))
272 (rfc822-bad-address "Strange character or missing comma")))))))
275 (defun rfc822-addresses (header-text)
276 (if (string-match "\\`[ \t]*\\([^][\000-\037 ()<>@,;:\\\".]+\\)[ \t]*\\'"
278 ;; Make very simple case moderately fast.
279 (list (substring header-text (match-beginning 1) (match-end 1)))
280 (let ((buf (generate-new-buffer " rfc822")))
284 (make-local-variable 'case-fold-search)
285 (setq case-fold-search nil) ;For speed(?)
287 ;; unfold continuation lines
288 (goto-char (point-min))
290 (while (re-search-forward "\\([^\\]\\(\\\\\\\\\\)*\\)\n[ \t]" nil t)
291 (replace-match "\\1 " t))
293 (goto-char (point-min))
296 rfc822-address-start); this is for rfc822-bad-address
297 (rfc822-nuke-whitespace)
299 (setq rfc822-address-start (point))
301 (catch 'address ; this is for rfc822-bad-address
302 (cond ((rfc822-looking-at ?\,)
304 ((looking-at "[][\000-\037@;:\\.>)]")
307 (format "Strange character \\%c found"
310 (rfc822-addresses-1 t)))))
313 (setq list (cons tem list)))
315 (setq list (nconc (nreverse tem) list)))))
317 (and buf (kill-buffer buf))))))
321 ;;; arch-tag: 5d388a24-e173-40fb-9b8e-85269de44b37
322 ;;; rfc822.el ends here