Merge lp:~james-page/ubuntu/hardy/mailman/fix-659975 into lp:ubuntu/hardy/mailman
- Hardy (8.04)
- fix-659975
- Merge into hardy
Proposed by
James Page
Status: | Merged |
---|---|
Merge reported by: | Mathias Gug |
Merged at revision: | not available |
Proposed branch: | lp:~james-page/ubuntu/hardy/mailman/fix-659975 |
Merge into: | lp:ubuntu/hardy/mailman |
Diff against target: |
513 lines (+488/-0) 3 files modified
debian/changelog (+7/-0) debian/email-2.5.8-patches/email/_parseaddr.py (+478/-0) debian/rules (+3/-0) |
To merge this branch: | bzr merge lp:~james-page/ubuntu/hardy/mailman/fix-659975 |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Mathias Gug | Approve | ||
Chuck Short | Pending | ||
Ubuntu Development Team | Pending | ||
Review via email:
|
Commit message
Description of the change
This branch fixes a minor email parsing issue in mailman 2.1.9 which actually resides in the bundled version of the Python email package distributed with this release.
The patched file is overlaid ontop of the extracted .tar.gz during the package build process.
Later versions of Ubuntu do not have this issue as the bundling was removed in later versions.
To post a comment you must log in.
Preview Diff
[H/L] Next/Prev Comment, [J/K] Next/Prev File, [N/P] Next/Prev Hunk
1 | === modified file 'debian/changelog' |
2 | --- debian/changelog 2008-03-07 02:55:22 +0000 |
3 | +++ debian/changelog 2010-10-18 11:31:08 +0000 |
4 | @@ -1,3 +1,10 @@ |
5 | +mailman (1:2.1.9-9ubuntu1.1) hardy-proposed; urgency=low |
6 | + |
7 | + * Patched bundled python-2.5.8 email package to correctly parse |
8 | + email addresses (LP: #659975) |
9 | + |
10 | + -- James Page <james.page@canonical.com> Thu, 14 Oct 2010 15:43:11 +0100 |
11 | + |
12 | mailman (1:2.1.9-9ubuntu1) hardy; urgency=low |
13 | |
14 | * debian/control: |
15 | |
16 | === added directory 'debian/email-2.5.8-patches' |
17 | === added directory 'debian/email-2.5.8-patches/email' |
18 | === added file 'debian/email-2.5.8-patches/email/_parseaddr.py' |
19 | --- debian/email-2.5.8-patches/email/_parseaddr.py 1970-01-01 00:00:00 +0000 |
20 | +++ debian/email-2.5.8-patches/email/_parseaddr.py 2010-10-18 11:31:08 +0000 |
21 | @@ -0,0 +1,478 @@ |
22 | +# Copyright (C) 2002-2006 Python Software Foundation |
23 | + |
24 | +"""Email address parsing code. |
25 | + |
26 | +Lifted directly from rfc822.py. This should eventually be rewritten. |
27 | +""" |
28 | + |
29 | +import time |
30 | +from types import TupleType |
31 | + |
32 | +try: |
33 | + True, False |
34 | +except NameError: |
35 | + True = 1 |
36 | + False = 0 |
37 | + |
38 | +SPACE = ' ' |
39 | +EMPTYSTRING = '' |
40 | +COMMASPACE = ', ' |
41 | + |
42 | +# Parse a date field |
43 | +_monthnames = ['jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', |
44 | + 'aug', 'sep', 'oct', 'nov', 'dec', |
45 | + 'january', 'february', 'march', 'april', 'may', 'june', 'july', |
46 | + 'august', 'september', 'october', 'november', 'december'] |
47 | + |
48 | +_daynames = ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun'] |
49 | + |
50 | +# The timezone table does not include the military time zones defined |
51 | +# in RFC822, other than Z. According to RFC1123, the description in |
52 | +# RFC822 gets the signs wrong, so we can't rely on any such time |
53 | +# zones. RFC1123 recommends that numeric timezone indicators be used |
54 | +# instead of timezone names. |
55 | + |
56 | +_timezones = {'UT':0, 'UTC':0, 'GMT':0, 'Z':0, |
57 | + 'AST': -400, 'ADT': -300, # Atlantic (used in Canada) |
58 | + 'EST': -500, 'EDT': -400, # Eastern |
59 | + 'CST': -600, 'CDT': -500, # Central |
60 | + 'MST': -700, 'MDT': -600, # Mountain |
61 | + 'PST': -800, 'PDT': -700 # Pacific |
62 | + } |
63 | + |
64 | + |
65 | +def parsedate_tz(data): |
66 | + """Convert a date string to a time tuple. |
67 | + |
68 | + Accounts for military timezones. |
69 | + """ |
70 | + data = data.split() |
71 | + # The FWS after the comma after the day-of-week is optional, so search and |
72 | + # adjust for this. |
73 | + if data[0].endswith(',') or data[0].lower() in _daynames: |
74 | + # There's a dayname here. Skip it |
75 | + del data[0] |
76 | + else: |
77 | + i = data[0].rfind(',') |
78 | + if i >= 0: |
79 | + data[0] = data[0][i+1:] |
80 | + if len(data) == 3: # RFC 850 date, deprecated |
81 | + stuff = data[0].split('-') |
82 | + if len(stuff) == 3: |
83 | + data = stuff + data[1:] |
84 | + if len(data) == 4: |
85 | + s = data[3] |
86 | + i = s.find('+') |
87 | + if i > 0: |
88 | + data[3:] = [s[:i], s[i+1:]] |
89 | + else: |
90 | + data.append('') # Dummy tz |
91 | + if len(data) < 5: |
92 | + return None |
93 | + data = data[:5] |
94 | + [dd, mm, yy, tm, tz] = data |
95 | + mm = mm.lower() |
96 | + if mm not in _monthnames: |
97 | + dd, mm = mm, dd.lower() |
98 | + if mm not in _monthnames: |
99 | + return None |
100 | + mm = _monthnames.index(mm) + 1 |
101 | + if mm > 12: |
102 | + mm -= 12 |
103 | + if dd[-1] == ',': |
104 | + dd = dd[:-1] |
105 | + i = yy.find(':') |
106 | + if i > 0: |
107 | + yy, tm = tm, yy |
108 | + if yy[-1] == ',': |
109 | + yy = yy[:-1] |
110 | + if not yy[0].isdigit(): |
111 | + yy, tz = tz, yy |
112 | + if tm[-1] == ',': |
113 | + tm = tm[:-1] |
114 | + tm = tm.split(':') |
115 | + if len(tm) == 2: |
116 | + [thh, tmm] = tm |
117 | + tss = '0' |
118 | + elif len(tm) == 3: |
119 | + [thh, tmm, tss] = tm |
120 | + else: |
121 | + return None |
122 | + try: |
123 | + yy = int(yy) |
124 | + dd = int(dd) |
125 | + thh = int(thh) |
126 | + tmm = int(tmm) |
127 | + tss = int(tss) |
128 | + except ValueError: |
129 | + return None |
130 | + tzoffset = None |
131 | + tz = tz.upper() |
132 | + if _timezones.has_key(tz): |
133 | + tzoffset = _timezones[tz] |
134 | + else: |
135 | + try: |
136 | + tzoffset = int(tz) |
137 | + except ValueError: |
138 | + pass |
139 | + # Convert a timezone offset into seconds ; -0500 -> -18000 |
140 | + if tzoffset: |
141 | + if tzoffset < 0: |
142 | + tzsign = -1 |
143 | + tzoffset = -tzoffset |
144 | + else: |
145 | + tzsign = 1 |
146 | + tzoffset = tzsign * ( (tzoffset/100)*3600 + (tzoffset % 100)*60) |
147 | + return yy, mm, dd, thh, tmm, tss, 0, 1, 0, tzoffset |
148 | + |
149 | + |
150 | +def parsedate(data): |
151 | + """Convert a time string to a time tuple.""" |
152 | + t = parsedate_tz(data) |
153 | + if isinstance(t, TupleType): |
154 | + return t[:9] |
155 | + else: |
156 | + return t |
157 | + |
158 | + |
159 | +def mktime_tz(data): |
160 | + """Turn a 10-tuple as returned by parsedate_tz() into a UTC timestamp.""" |
161 | + if data[9] is None: |
162 | + # No zone info, so localtime is better assumption than GMT |
163 | + return time.mktime(data[:8] + (-1,)) |
164 | + else: |
165 | + t = time.mktime(data[:8] + (0,)) |
166 | + return t - data[9] - time.timezone |
167 | + |
168 | + |
169 | +def quote(str): |
170 | + """Add quotes around a string.""" |
171 | + return str.replace('\\', '\\\\').replace('"', '\\"') |
172 | + |
173 | + |
174 | +class AddrlistClass: |
175 | + """Address parser class by Ben Escoto. |
176 | + |
177 | + To understand what this class does, it helps to have a copy of RFC 2822 in |
178 | + front of you. |
179 | + |
180 | + Note: this class interface is deprecated and may be removed in the future. |
181 | + Use rfc822.AddressList instead. |
182 | + """ |
183 | + |
184 | + def __init__(self, field): |
185 | + """Initialize a new instance. |
186 | + |
187 | + `field' is an unparsed address header field, containing |
188 | + one or more addresses. |
189 | + """ |
190 | + self.specials = '()<>@,:;.\"[]' |
191 | + self.pos = 0 |
192 | + self.LWS = ' \t' |
193 | + self.CR = '\r\n' |
194 | + self.FWS = self.LWS + self.CR |
195 | + self.atomends = self.specials + self.LWS + self.CR |
196 | + # Note that RFC 2822 now specifies `.' as obs-phrase, meaning that it |
197 | + # is obsolete syntax. RFC 2822 requires that we recognize obsolete |
198 | + # syntax, so allow dots in phrases. |
199 | + self.phraseends = self.atomends.replace('.', '') |
200 | + self.field = field |
201 | + self.commentlist = [] |
202 | + |
203 | + def gotonext(self): |
204 | + """Parse up to the start of the next address.""" |
205 | + while self.pos < len(self.field): |
206 | + if self.field[self.pos] in self.LWS + '\n\r': |
207 | + self.pos += 1 |
208 | + elif self.field[self.pos] == '(': |
209 | + self.commentlist.append(self.getcomment()) |
210 | + else: |
211 | + break |
212 | + |
213 | + def getaddrlist(self): |
214 | + """Parse all addresses. |
215 | + |
216 | + Returns a list containing all of the addresses. |
217 | + """ |
218 | + result = [] |
219 | + while self.pos < len(self.field): |
220 | + ad = self.getaddress() |
221 | + if ad: |
222 | + result += ad |
223 | + else: |
224 | + result.append(('', '')) |
225 | + return result |
226 | + |
227 | + def getaddress(self): |
228 | + """Parse the next address.""" |
229 | + self.commentlist = [] |
230 | + self.gotonext() |
231 | + |
232 | + oldpos = self.pos |
233 | + oldcl = self.commentlist |
234 | + plist = self.getphraselist() |
235 | + |
236 | + self.gotonext() |
237 | + returnlist = [] |
238 | + |
239 | + if self.pos >= len(self.field): |
240 | + # Bad email address technically, no domain. |
241 | + if plist: |
242 | + returnlist = [(SPACE.join(self.commentlist), plist[0])] |
243 | + |
244 | + elif self.field[self.pos] in '.@': |
245 | + # email address is just an addrspec |
246 | + # this isn't very efficient since we start over |
247 | + self.pos = oldpos |
248 | + self.commentlist = oldcl |
249 | + addrspec = self.getaddrspec() |
250 | + returnlist = [(SPACE.join(self.commentlist), addrspec)] |
251 | + |
252 | + elif self.field[self.pos] == ':': |
253 | + # address is a group |
254 | + returnlist = [] |
255 | + |
256 | + fieldlen = len(self.field) |
257 | + self.pos += 1 |
258 | + while self.pos < len(self.field): |
259 | + self.gotonext() |
260 | + if self.pos < fieldlen and self.field[self.pos] == ';': |
261 | + self.pos += 1 |
262 | + break |
263 | + returnlist = returnlist + self.getaddress() |
264 | + |
265 | + elif self.field[self.pos] == '<': |
266 | + # Address is a phrase then a route addr |
267 | + routeaddr = self.getrouteaddr() |
268 | + |
269 | + if self.commentlist: |
270 | + returnlist = [(SPACE.join(plist) + ' (' + |
271 | + ' '.join(self.commentlist) + ')', routeaddr)] |
272 | + else: |
273 | + returnlist = [(SPACE.join(plist), routeaddr)] |
274 | + |
275 | + else: |
276 | + if plist: |
277 | + returnlist = [(SPACE.join(self.commentlist), plist[0])] |
278 | + elif self.field[self.pos] in self.specials: |
279 | + self.pos += 1 |
280 | + |
281 | + self.gotonext() |
282 | + if self.pos < len(self.field) and self.field[self.pos] == ',': |
283 | + self.pos += 1 |
284 | + return returnlist |
285 | + |
286 | + def getrouteaddr(self): |
287 | + """Parse a route address (Return-path value). |
288 | + |
289 | + This method just skips all the route stuff and returns the addrspec. |
290 | + """ |
291 | + if self.field[self.pos] != '<': |
292 | + return |
293 | + |
294 | + expectroute = False |
295 | + self.pos += 1 |
296 | + self.gotonext() |
297 | + adlist = '' |
298 | + while self.pos < len(self.field): |
299 | + if expectroute: |
300 | + self.getdomain() |
301 | + expectroute = False |
302 | + elif self.field[self.pos] == '>': |
303 | + self.pos += 1 |
304 | + break |
305 | + elif self.field[self.pos] == '@': |
306 | + self.pos += 1 |
307 | + expectroute = True |
308 | + elif self.field[self.pos] == ':': |
309 | + self.pos += 1 |
310 | + else: |
311 | + adlist = self.getaddrspec() |
312 | + self.pos += 1 |
313 | + break |
314 | + self.gotonext() |
315 | + |
316 | + return adlist |
317 | + |
318 | + def getaddrspec(self): |
319 | + """Parse an RFC 2822 addr-spec.""" |
320 | + aslist = [] |
321 | + |
322 | + self.gotonext() |
323 | + while self.pos < len(self.field): |
324 | + if self.field[self.pos] == '.': |
325 | + aslist.append('.') |
326 | + self.pos += 1 |
327 | + elif self.field[self.pos] == '"': |
328 | + aslist.append('"%s"' % self.getquote()) |
329 | + elif self.field[self.pos] in self.atomends: |
330 | + break |
331 | + else: |
332 | + aslist.append(self.getatom()) |
333 | + self.gotonext() |
334 | + |
335 | + if self.pos >= len(self.field) or self.field[self.pos] != '@': |
336 | + return EMPTYSTRING.join(aslist) |
337 | + |
338 | + aslist.append('@') |
339 | + self.pos += 1 |
340 | + self.gotonext() |
341 | + return EMPTYSTRING.join(aslist) + self.getdomain() |
342 | + |
343 | + def getdomain(self): |
344 | + """Get the complete domain name from an address.""" |
345 | + sdlist = [] |
346 | + while self.pos < len(self.field): |
347 | + if self.field[self.pos] in self.LWS: |
348 | + self.pos += 1 |
349 | + elif self.field[self.pos] == '(': |
350 | + self.commentlist.append(self.getcomment()) |
351 | + elif self.field[self.pos] == '[': |
352 | + sdlist.append(self.getdomainliteral()) |
353 | + elif self.field[self.pos] == '.': |
354 | + self.pos += 1 |
355 | + sdlist.append('.') |
356 | + elif self.field[self.pos] in self.atomends: |
357 | + break |
358 | + else: |
359 | + sdlist.append(self.getatom()) |
360 | + return EMPTYSTRING.join(sdlist) |
361 | + |
362 | + def getdelimited(self, beginchar, endchars, allowcomments=True): |
363 | + """Parse a header fragment delimited by special characters. |
364 | + |
365 | + `beginchar' is the start character for the fragment. |
366 | + If self is not looking at an instance of `beginchar' then |
367 | + getdelimited returns the empty string. |
368 | + |
369 | + `endchars' is a sequence of allowable end-delimiting characters. |
370 | + Parsing stops when one of these is encountered. |
371 | + |
372 | + If `allowcomments' is non-zero, embedded RFC 2822 comments are allowed |
373 | + within the parsed fragment. |
374 | + """ |
375 | + if self.field[self.pos] != beginchar: |
376 | + return '' |
377 | + |
378 | + slist = [''] |
379 | + quote = False |
380 | + self.pos += 1 |
381 | + while self.pos < len(self.field): |
382 | + if quote: |
383 | + slist.append(self.field[self.pos]) |
384 | + quote = False |
385 | + elif self.field[self.pos] in endchars: |
386 | + self.pos += 1 |
387 | + break |
388 | + elif allowcomments and self.field[self.pos] == '(': |
389 | + slist.append(self.getcomment()) |
390 | + continue # have already advanced pos from getcomment |
391 | + elif self.field[self.pos] == '\\': |
392 | + quote = True |
393 | + else: |
394 | + slist.append(self.field[self.pos]) |
395 | + self.pos += 1 |
396 | + |
397 | + return EMPTYSTRING.join(slist) |
398 | + |
399 | + def getquote(self): |
400 | + """Get a quote-delimited fragment from self's field.""" |
401 | + return self.getdelimited('"', '"\r', False) |
402 | + |
403 | + def getcomment(self): |
404 | + """Get a parenthesis-delimited fragment from self's field.""" |
405 | + return self.getdelimited('(', ')\r', True) |
406 | + |
407 | + def getdomainliteral(self): |
408 | + """Parse an RFC 2822 domain-literal.""" |
409 | + return '[%s]' % self.getdelimited('[', ']\r', False) |
410 | + |
411 | + def getatom(self, atomends=None): |
412 | + """Parse an RFC 2822 atom. |
413 | + |
414 | + Optional atomends specifies a different set of end token delimiters |
415 | + (the default is to use self.atomends). This is used e.g. in |
416 | + getphraselist() since phrase endings must not include the `.' (which |
417 | + is legal in phrases).""" |
418 | + atomlist = [''] |
419 | + if atomends is None: |
420 | + atomends = self.atomends |
421 | + |
422 | + while self.pos < len(self.field): |
423 | + if self.field[self.pos] in atomends: |
424 | + break |
425 | + else: |
426 | + atomlist.append(self.field[self.pos]) |
427 | + self.pos += 1 |
428 | + |
429 | + return EMPTYSTRING.join(atomlist) |
430 | + |
431 | + def getphraselist(self): |
432 | + """Parse a sequence of RFC 2822 phrases. |
433 | + |
434 | + A phrase is a sequence of words, which are in turn either RFC 2822 |
435 | + atoms or quoted-strings. Phrases are canonicalized by squeezing all |
436 | + runs of continuous whitespace into one space. |
437 | + """ |
438 | + plist = [] |
439 | + |
440 | + while self.pos < len(self.field): |
441 | + if self.field[self.pos] in self.FWS: |
442 | + self.pos += 1 |
443 | + elif self.field[self.pos] == '"': |
444 | + plist.append(self.getquote()) |
445 | + elif self.field[self.pos] == '(': |
446 | + self.commentlist.append(self.getcomment()) |
447 | + elif self.field[self.pos] in self.phraseends: |
448 | + break |
449 | + else: |
450 | + plist.append(self.getatom(self.phraseends)) |
451 | + |
452 | + return plist |
453 | + |
454 | +class AddressList(AddrlistClass): |
455 | + """An AddressList encapsulates a list of parsed RFC 2822 addresses.""" |
456 | + def __init__(self, field): |
457 | + AddrlistClass.__init__(self, field) |
458 | + if field: |
459 | + self.addresslist = self.getaddrlist() |
460 | + else: |
461 | + self.addresslist = [] |
462 | + |
463 | + def __len__(self): |
464 | + return len(self.addresslist) |
465 | + |
466 | + def __add__(self, other): |
467 | + # Set union |
468 | + newaddr = AddressList(None) |
469 | + newaddr.addresslist = self.addresslist[:] |
470 | + for x in other.addresslist: |
471 | + if not x in self.addresslist: |
472 | + newaddr.addresslist.append(x) |
473 | + return newaddr |
474 | + |
475 | + def __iadd__(self, other): |
476 | + # Set union, in-place |
477 | + for x in other.addresslist: |
478 | + if not x in self.addresslist: |
479 | + self.addresslist.append(x) |
480 | + return self |
481 | + |
482 | + def __sub__(self, other): |
483 | + # Set difference |
484 | + newaddr = AddressList(None) |
485 | + for x in self.addresslist: |
486 | + if not x in other.addresslist: |
487 | + newaddr.addresslist.append(x) |
488 | + return newaddr |
489 | + |
490 | + def __isub__(self, other): |
491 | + # Set difference, in-place |
492 | + for x in other.addresslist: |
493 | + if x in self.addresslist: |
494 | + self.addresslist.remove(x) |
495 | + return self |
496 | + |
497 | + def __getitem__(self, index): |
498 | + # Make indexing, slices, and 'in' work |
499 | + return self.addresslist[index] |
500 | |
501 | === modified file 'debian/rules' |
502 | --- debian/rules 2007-12-04 09:12:39 +0000 |
503 | +++ debian/rules 2010-10-18 11:31:08 +0000 |
504 | @@ -147,6 +147,9 @@ |
505 | # apache default config |
506 | install -m 0644 debian/contrib/apache.conf debian/mailman/etc/mailman |
507 | |
508 | +# Patch mailman supplied fork of email-2.5.8 to fix bug parsing email addresses |
509 | + install -m 0644 debian/email-2.5.8-patches/email/_parseaddr.py debian/mailman/usr/lib/mailman/pythonlib/email/_parseaddr.py |
510 | + |
511 | # Move templates |
512 | mv `find debian/mailman/etc/mailman -mindepth 1 -maxdepth 1 -type d` debian/mailman/usr/share/mailman |
513 | cp build/contrib/qmail-to-mailman.py debian/mailman/usr/share/mailman |
Looks good to me. Uploaded to hardy-proposed.