*** empty log message ***
[bpt/emacs.git] / lisp / mail / rfc822.el
CommitLineData
cedaf3aa 1;; Hairy rfc822 parser for mail and news and suchlike
c82d7590 2;; Copyright (C) 1986-1990 Free Software Foundation, Inc.
cedaf3aa
RS
3;; Author Richard Mlynarik.
4
5;; This file is part of GNU Emacs.
6
7;; GNU Emacs is free software; you can redistribute it and/or modify
8;; it under the terms of the GNU General Public License as published by
9;; the Free Software Foundation; either version 1, or (at your option)
10;; any later version.
11
12;; GNU Emacs is distributed in the hope that it will be useful,
13;; but WITHOUT ANY WARRANTY; without even the implied warranty of
14;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15;; GNU General Public License for more details.
16
17;; You should have received a copy of the GNU General Public License
18;; along with GNU Emacs; see the file COPYING. If not, write to
19;; the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
20
cedaf3aa
RS
21;; uses address-start free, throws to address
22(defun rfc822-bad-address (reason)
23 (save-restriction
24 (insert "_^_")
25 (narrow-to-region address-start
26 (if (re-search-forward "[,;]" nil t)
27 (max (point-min) (1- (point)))
28 (point-max)))
29 ;; make the error string be suitable for inclusion in (...)
30 (let ((losers '("\\" "(" ")" "\n")))
31 (while losers
32 (goto-char (point-min))
33 (while (search-forward (car losers) nil t)
34 (backward-char 1)
35 (insert ?\\)
36 (forward-char 1))
37 (setq losers (cdr losers))))
38 (goto-char (point-min)) (insert "(Unparsable address -- "
39 reason
40 ":\n\t \"")
41 (goto-char (point-max)) (insert "\")"))
42 (rfc822-nuke-whitespace)
43 (throw 'address (buffer-substring address-start (point))))
44
45(defun rfc822-nuke-whitespace (&optional leave-space)
46 (let (ch)
47 (while (cond ((eobp)
48 nil)
49 ((= (setq ch (following-char)) ?\()
50 (forward-char 1)
51 (while (if (eobp)
52 (rfc822-bad-address "Unbalanced comment (...)")
53 (/= (setq ch (following-char)) ?\)))
54 (cond ((looking-at "[^()\\]+")
55 (replace-match ""))
56 ((= ch ?\()
57 (rfc822-nuke-whitespace))
58 ((< (point) (1- (point-max)))
59 (delete-char 2))
60 (t
61 (rfc822-bad-address "orphaned backslash"))))
62 ;; delete remaining "()"
63 (forward-char -1)
64 (delete-char 2)
65 t)
66 ((memq ch '(?\ ?\t ?\n))
67 (delete-region (point)
68 (progn (skip-chars-forward " \t\n") (point)))
69 t)
70 (t
71 nil)))
72 (or (not leave-space)
73 (eobp)
74 (bobp)
75 (= (preceding-char) ?\ )
76 (insert ?\ ))))
77
78(defun rfc822-looking-at (regex &optional leave-space)
79 (if (cond ((stringp regex)
80 (if (looking-at regex)
81 (progn (goto-char (match-end 0))
82 t)))
83 (t
84 (if (and (not (eobp))
85 (= (following-char) regex))
86 (progn (forward-char 1)
87 t))))
88 (let ((tem (match-data)))
89 (rfc822-nuke-whitespace leave-space)
90 (store-match-data tem)
91 t)))
92
93(defun rfc822-snarf-word ()
94 ;; word is atom | quoted-string
95 (cond ((= (following-char) ?\")
96 ;; quoted-string
97 (or (rfc822-looking-at "\"\\([^\"\\\n]\\|\\\\.\\|\\\\\n\\)*\"")
98 (rfc822-bad-address "Unterminated quoted string")))
99 ((rfc822-looking-at "[^][\000-\037\177-\377 ()<>@,;:\\\".]+")
100 ;; atom
101 )
102 (t
103 (rfc822-bad-address "Rubbish in address"))))
104
105(defun rfc822-snarf-words ()
106 (rfc822-snarf-word)
107 (while (rfc822-looking-at ?.)
108 (rfc822-snarf-word)))
109
110(defun rfc822-snarf-subdomain ()
111 ;; sub-domain is domain-ref | domain-literal
112 (cond ((= (following-char) ?\[)
113 ;; domain-ref
114 (or (rfc822-looking-at "\\[\\([^][\\\n]\\|\\\\.\\|\\\\\n\\)*\\]")
115 (rfc822-bad-address "Unterminated domain literal [...]")))
116 ((rfc822-looking-at "[^][\000-\037\177-\377 ()<>@,;:\\\".]+")
117 ;; domain-literal = atom
118 )
119 (t
120 (rfc822-bad-address "Rubbish in host/domain specification"))))
121
122(defun rfc822-snarf-domain ()
123 (rfc822-snarf-subdomain)
124 (while (rfc822-looking-at ?.)
125 (rfc822-snarf-subdomain)))
126
127(defun rfc822-snarf-frob-list (name separator terminator snarfer
128 &optional return)
129 (let ((first t)
130 (list ())
131 tem)
132 (while (cond ((eobp)
133 (rfc822-bad-address
134 (format "End of addresses in middle of %s" name)))
135 ((rfc822-looking-at terminator)
136 nil)
137 ((rfc822-looking-at separator)
138 ;; multiple separators are allowed and do nothing.
139 (while (rfc822-looking-at separator))
140 t)
141 (first
142 t)
143 (t
144 (rfc822-bad-address
145 (format "Gubbish in middle of %s" name))))
146 (setq tem (funcall snarfer)
147 first nil)
148 (and return tem
149 (setq list (if (listp tem)
150 (nconc (reverse tem) list)
151 (cons tem list)))))
152 (nreverse list)))
153
154;; return either an address (a string) or a list of addresses
155(defun rfc822-addresses-1 (&optional allow-groups)
156 ;; Looking for an rfc822 `address'
157 ;; Either a group (1*word ":" [#mailbox] ";")
158 ;; or a mailbox (addr-spec | 1*word route-addr)
159 ;; addr-spec is (local-part "@" domain)
160 ;; route-addr is ("<" [1#("@" domain) ":"] addr-spec ">")
161 ;; local-part is (word *("." word))
162 ;; word is (atom | quoted-string)
163 ;; quoted-string is ("\([^\"\\n]\|\\.\|\\\n\)")
164 ;; atom is [^\000-\037\177 ()<>@,;:\".[]]+
165 ;; domain is sub-domain *("." sub-domain)
166 ;; sub-domain is domain-ref | domain-literal
167 ;; domain-literal is "[" *(dtext | quoted-pair) "]"
168 ;; dtext is "[^][\\n"
169 ;; domain-ref is atom
170 (let ((address-start (point))
171 (n 0))
172 (catch 'address
173 ;; optimize common cases:
174 ;; foo
175 ;; foo.bar@bar.zap
176 ;; followed by "\\'\\|,\\|([^()\\]*)\\'"
177 ;; other common cases are:
178 ;; foo bar <foo.bar@baz.zap>
179 ;; "foo bar" <foo.bar@baz.zap>
180 ;; those aren't hacked yet.
181 (if (and (rfc822-looking-at "[^][\000-\037\177-\377 ()<>@,;:\\\"]+\\(\\|@[^][\000-\037\177-\377 ()<>@,;:\\\"]+\\)" t)
182 (progn (or (eobp)
183 (rfc822-looking-at ?,))))
184 (progn
185 ;; rfc822-looking-at may have inserted a space
186 (or (bobp) (/= (preceding-char) ?\ ) (delete-char -1))
187 ;; relying on the fact that rfc822-looking-at <char>
188 ;; doesn't mung match-data
189 (throw 'address (buffer-substring address-start (match-end 0)))))
190 (goto-char address-start)
191 (while t
192 (cond ((and (= n 1) (rfc822-looking-at ?@))
193 ;; local-part@domain
194 (rfc822-snarf-domain)
195 (throw 'address
196 (buffer-substring address-start (point))))
197 ((rfc822-looking-at ?:)
198 (cond ((not allow-groups)
199 (rfc822-bad-address "A group name may not appear here"))
200 ((= n 0)
201 (rfc822-bad-address "No name for :...; group")))
202 ;; group
203 (throw 'address
204 ;; return a list of addresses
205 (rfc822-snarf-frob-list ":...; group" ?\, ?\;
206 'rfc822-addresses-1 t)))
207 ((rfc822-looking-at ?<)
208 (let ((start (point))
209 (strip t))
210 (cond ((rfc822-looking-at ?>)
211 ;; empty path
212 ())
213 ((and (not (eobp)) (= (following-char) ?\@))
214 ;; <@foo.bar,@baz:quux@abcd.efg>
215 (rfc822-snarf-frob-list "<...> address" ?\, ?\:
216 (function (lambda ()
217 (if (rfc822-looking-at ?\@)
218 (rfc822-snarf-domain)
219 (rfc822-bad-address
220 "Gubbish in route-addr")))))
221 (rfc822-snarf-words)
222 (or (rfc822-looking-at ?@)
223 (rfc822-bad-address "Malformed <..@..> address"))
224 (rfc822-snarf-domain)
225 (setq strip nil))
226 ((progn (rfc822-snarf-words) (rfc822-looking-at ?@))
227 ; allow <foo> (losing unix seems to do this)
228 (rfc822-snarf-domain)))
229 (let ((end (point)))
230 (if (rfc822-looking-at ?\>)
231 (throw 'address
232 (buffer-substring (if strip start (1- start))
233 (if strip end (1+ end))))
234 (rfc822-bad-address "Unterminated <...> address")))))
235 ((looking-at "[^][\000-\037\177-\377 ()<>@,;:\\.]")
236 ;; this allows "." to be part of the words preceding
237 ;; an addr-spec, since many broken mailers output
238 ;; "Hern K. Herklemeyer III
239 ;; <yank@megadeath.dod.gods-own-country>"
278fc732 240 (let ((again t))
241 (while again
242 (or (= n 0) (bobp) (= (preceding-char) ?\ )
243 (insert ?\ ))
48aefb44 244 (rfc822-snarf-words)
278fc732 245 (setq n (1+ n))
246 (setq again (or (rfc822-looking-at ?.)
247 (looking-at "[^][\000-\037\177-\377 ()<>@,;:\\.]"))))))
cedaf3aa
RS
248 ((= n 0)
249 (throw 'address nil))
250 ((= n 1) ; allow "foo" (losing unix seems to do this)
251 (throw 'address
252 (buffer-substring address-start (point))))
278fc732 253 ((> n 1)
254 (rfc822-bad-address "Missing comma between addresses or badly-formatted address"))
255 ((or (eobp) (= (following-char) ?,))
cedaf3aa
RS
256 (rfc822-bad-address "Missing comma or route-spec"))
257 (t
258 (rfc822-bad-address "Strange character or missing comma")))))))
259
260
261(defun rfc822-addresses (header-text)
262 (if (string-match "\\`[ \t]*\\([^][\000-\037\177-\377 ()<>@,;:\\\".]+\\)[ \t]*\\'"
263 header-text)
264 ;; Make very simple case moderately fast.
265 (list (substring header-text (match-beginning 1) (match-end 1)))
266 (let ((buf (generate-new-buffer " rfc822")))
267 (unwind-protect
268 (save-excursion
269 (set-buffer buf)
270 (make-local-variable 'case-fold-search)
271 (setq case-fold-search nil) ;For speed(?)
272 (insert header-text)
273 ;; unfold continuation lines
274 (goto-char (point-min))
275
276 (while (re-search-forward "\\([^\\]\\(\\\\\\\\\\)*\\)\n[ \t]" nil t)
277 (replace-match "\\1 " t))
278
279 (goto-char (point-min))
280 (rfc822-nuke-whitespace)
281 (let ((list ())
282 tem
283 address-start); this is for rfc822-bad-address
284 (while (not (eobp))
285 (setq address-start (point))
286 (setq tem
287 (catch 'address ; this is for rfc822-bad-address
288 (cond ((rfc822-looking-at ?\,)
289 nil)
c82d7590 290 ((looking-at "[][\000-\037\177-\377@;:\\.>)]")
cedaf3aa
RS
291 (forward-char)
292 (rfc822-bad-address
293 (format "Strange character \\%c found"
294 (preceding-char))))
295 (t
296 (rfc822-addresses-1 t)))))
297 (cond ((null tem))
298 ((stringp tem)
299 (setq list (cons tem list)))
300 (t
301 (setq list (nconc (nreverse tem) list)))))
302 (nreverse list)))
303 (and buf (kill-buffer buf))))))
304
49116ac0
JB
305(provide 'rfc822)
306