svn-gvsig-desktop / tags / v1_1_Build_1003 / extensions / extScripting / scripts / jython / Lib / rfc822.py @ 12271
History | View | Annotate | Download (30.5 KB)
1 |
"""RFC-822 message manipulation class.
|
---|---|
2 |
|
3 |
XXX This is only a very rough sketch of a full RFC-822 parser;
|
4 |
in particular the tokenizing of addresses does not adhere to all the
|
5 |
quoting rules.
|
6 |
|
7 |
Directions for use:
|
8 |
|
9 |
To create a Message object: first open a file, e.g.:
|
10 |
fp = open(file, 'r')
|
11 |
You can use any other legal way of getting an open file object, e.g. use
|
12 |
sys.stdin or call os.popen().
|
13 |
Then pass the open file object to the Message() constructor:
|
14 |
m = Message(fp)
|
15 |
|
16 |
This class can work with any input object that supports a readline
|
17 |
method. If the input object has seek and tell capability, the
|
18 |
rewindbody method will work; also illegal lines will be pushed back
|
19 |
onto the input stream. If the input object lacks seek but has an
|
20 |
`unread' method that can push back a line of input, Message will use
|
21 |
that to push back illegal lines. Thus this class can be used to parse
|
22 |
messages coming from a buffered stream.
|
23 |
|
24 |
The optional `seekable' argument is provided as a workaround for
|
25 |
certain stdio libraries in which tell() discards buffered data before
|
26 |
discovering that the lseek() system call doesn't work. For maximum
|
27 |
portability, you should set the seekable argument to zero to prevent
|
28 |
that initial \code{tell} when passing in an unseekable object such as
|
29 |
a a file object created from a socket object. If it is 1 on entry --
|
30 |
which it is by default -- the tell() method of the open file object is
|
31 |
called once; if this raises an exception, seekable is reset to 0. For
|
32 |
other nonzero values of seekable, this test is not made.
|
33 |
|
34 |
To get the text of a particular header there are several methods:
|
35 |
str = m.getheader(name)
|
36 |
str = m.getrawheader(name)
|
37 |
where name is the name of the header, e.g. 'Subject'.
|
38 |
The difference is that getheader() strips the leading and trailing
|
39 |
whitespace, while getrawheader() doesn't. Both functions retain
|
40 |
embedded whitespace (including newlines) exactly as they are
|
41 |
specified in the header, and leave the case of the text unchanged.
|
42 |
|
43 |
For addresses and address lists there are functions
|
44 |
realname, mailaddress = m.getaddr(name) and
|
45 |
list = m.getaddrlist(name)
|
46 |
where the latter returns a list of (realname, mailaddr) tuples.
|
47 |
|
48 |
There is also a method
|
49 |
time = m.getdate(name)
|
50 |
which parses a Date-like field and returns a time-compatible tuple,
|
51 |
i.e. a tuple such as returned by time.localtime() or accepted by
|
52 |
time.mktime().
|
53 |
|
54 |
See the class definition for lower level access methods.
|
55 |
|
56 |
There are also some utility functions here.
|
57 |
"""
|
58 |
# Cleanup and extensions by Eric S. Raymond <esr@thyrsus.com>
|
59 |
|
60 |
import time |
61 |
|
62 |
__all__ = ["Message","AddressList","parsedate","parsedate_tz","mktime_tz"] |
63 |
|
64 |
_blanklines = ('\r\n', '\n') # Optimization for islast() |
65 |
|
66 |
|
67 |
class Message: |
68 |
"""Represents a single RFC-822-compliant message."""
|
69 |
|
70 |
def __init__(self, fp, seekable = 1): |
71 |
"""Initialize the class instance and read the headers."""
|
72 |
if seekable == 1: |
73 |
# Exercise tell() to make sure it works
|
74 |
# (and then assume seek() works, too)
|
75 |
try:
|
76 |
fp.tell() |
77 |
except:
|
78 |
seekable = 0
|
79 |
else:
|
80 |
seekable = 1
|
81 |
self.fp = fp
|
82 |
self.seekable = seekable
|
83 |
self.startofheaders = None |
84 |
self.startofbody = None |
85 |
#
|
86 |
if self.seekable: |
87 |
try:
|
88 |
self.startofheaders = self.fp.tell() |
89 |
except IOError: |
90 |
self.seekable = 0 |
91 |
#
|
92 |
self.readheaders()
|
93 |
#
|
94 |
if self.seekable: |
95 |
try:
|
96 |
self.startofbody = self.fp.tell() |
97 |
except IOError: |
98 |
self.seekable = 0 |
99 |
|
100 |
def rewindbody(self): |
101 |
"""Rewind the file to the start of the body (if seekable)."""
|
102 |
if not self.seekable: |
103 |
raise IOError, "unseekable file" |
104 |
self.fp.seek(self.startofbody) |
105 |
|
106 |
def readheaders(self): |
107 |
"""Read header lines.
|
108 |
|
109 |
Read header lines up to the entirely blank line that
|
110 |
terminates them. The (normally blank) line that ends the
|
111 |
headers is skipped, but not included in the returned list.
|
112 |
If a non-header line ends the headers, (which is an error),
|
113 |
an attempt is made to backspace over it; it is never
|
114 |
included in the returned list.
|
115 |
|
116 |
The variable self.status is set to the empty string if all
|
117 |
went well, otherwise it is an error message.
|
118 |
The variable self.headers is a completely uninterpreted list
|
119 |
of lines contained in the header (so printing them will
|
120 |
reproduce the header exactly as it appears in the file).
|
121 |
"""
|
122 |
self.dict = {}
|
123 |
self.unixfrom = '' |
124 |
self.headers = list = []
|
125 |
self.status = '' |
126 |
headerseen = ""
|
127 |
firstline = 1
|
128 |
startofline = unread = tell = None
|
129 |
if hasattr(self.fp, 'unread'): |
130 |
unread = self.fp.unread
|
131 |
elif self.seekable: |
132 |
tell = self.fp.tell
|
133 |
while 1: |
134 |
if tell:
|
135 |
try:
|
136 |
startofline = tell() |
137 |
except IOError: |
138 |
startofline = tell = None
|
139 |
self.seekable = 0 |
140 |
line = self.fp.readline()
|
141 |
if not line: |
142 |
self.status = 'EOF in headers' |
143 |
break
|
144 |
# Skip unix From name time lines
|
145 |
if firstline and line.startswith('From '): |
146 |
self.unixfrom = self.unixfrom + line |
147 |
continue
|
148 |
firstline = 0
|
149 |
if headerseen and line[0] in ' \t': |
150 |
# It's a continuation line.
|
151 |
list.append(line)
|
152 |
x = (self.dict[headerseen] + "\n " + line.strip()) |
153 |
self.dict[headerseen] = x.strip()
|
154 |
continue
|
155 |
elif self.iscomment(line): |
156 |
# It's a comment. Ignore it.
|
157 |
continue
|
158 |
elif self.islast(line): |
159 |
# Note! No pushback here! The delimiter line gets eaten.
|
160 |
break
|
161 |
headerseen = self.isheader(line)
|
162 |
if headerseen:
|
163 |
# It's a legal header line, save it.
|
164 |
list.append(line)
|
165 |
self.dict[headerseen] = line[len(headerseen)+1:].strip() |
166 |
continue
|
167 |
else:
|
168 |
# It's not a header line; throw it back and stop here.
|
169 |
if not self.dict: |
170 |
self.status = 'No headers' |
171 |
else:
|
172 |
self.status = 'Non-header line where header expected' |
173 |
# Try to undo the read.
|
174 |
if unread:
|
175 |
unread(line) |
176 |
elif tell:
|
177 |
self.fp.seek(startofline)
|
178 |
else:
|
179 |
self.status = self.status + '; bad seek' |
180 |
break
|
181 |
|
182 |
def isheader(self, line): |
183 |
"""Determine whether a given line is a legal header.
|
184 |
|
185 |
This method should return the header name, suitably canonicalized.
|
186 |
You may override this method in order to use Message parsing
|
187 |
on tagged data in RFC822-like formats with special header formats.
|
188 |
"""
|
189 |
i = line.find(':')
|
190 |
if i > 0: |
191 |
return line[:i].lower()
|
192 |
else:
|
193 |
return None |
194 |
|
195 |
def islast(self, line): |
196 |
"""Determine whether a line is a legal end of RFC-822 headers.
|
197 |
|
198 |
You may override this method if your application wants
|
199 |
to bend the rules, e.g. to strip trailing whitespace,
|
200 |
or to recognize MH template separators ('--------').
|
201 |
For convenience (e.g. for code reading from sockets) a
|
202 |
line consisting of \r\n also matches.
|
203 |
"""
|
204 |
return line in _blanklines |
205 |
|
206 |
def iscomment(self, line): |
207 |
"""Determine whether a line should be skipped entirely.
|
208 |
|
209 |
You may override this method in order to use Message parsing
|
210 |
on tagged data in RFC822-like formats that support embedded
|
211 |
comments or free-text data.
|
212 |
"""
|
213 |
return None |
214 |
|
215 |
def getallmatchingheaders(self, name): |
216 |
"""Find all header lines matching a given header name.
|
217 |
|
218 |
Look through the list of headers and find all lines
|
219 |
matching a given header name (and their continuation
|
220 |
lines). A list of the lines is returned, without
|
221 |
interpretation. If the header does not occur, an
|
222 |
empty list is returned. If the header occurs multiple
|
223 |
times, all occurrences are returned. Case is not
|
224 |
important in the header name.
|
225 |
"""
|
226 |
name = name.lower() + ':'
|
227 |
n = len(name)
|
228 |
list = [] |
229 |
hit = 0
|
230 |
for line in self.headers: |
231 |
if line[:n].lower() == name:
|
232 |
hit = 1
|
233 |
elif not line[:1].isspace(): |
234 |
hit = 0
|
235 |
if hit:
|
236 |
list.append(line)
|
237 |
return list |
238 |
|
239 |
def getfirstmatchingheader(self, name): |
240 |
"""Get the first header line matching name.
|
241 |
|
242 |
This is similar to getallmatchingheaders, but it returns
|
243 |
only the first matching header (and its continuation
|
244 |
lines).
|
245 |
"""
|
246 |
name = name.lower() + ':'
|
247 |
n = len(name)
|
248 |
list = [] |
249 |
hit = 0
|
250 |
for line in self.headers: |
251 |
if hit:
|
252 |
if not line[:1].isspace(): |
253 |
break
|
254 |
elif line[:n].lower() == name:
|
255 |
hit = 1
|
256 |
if hit:
|
257 |
list.append(line)
|
258 |
return list |
259 |
|
260 |
def getrawheader(self, name): |
261 |
"""A higher-level interface to getfirstmatchingheader().
|
262 |
|
263 |
Return a string containing the literal text of the
|
264 |
header but with the keyword stripped. All leading,
|
265 |
trailing and embedded whitespace is kept in the
|
266 |
string, however.
|
267 |
Return None if the header does not occur.
|
268 |
"""
|
269 |
|
270 |
list = self.getfirstmatchingheader(name)
|
271 |
if not list: |
272 |
return None |
273 |
list[0] = list[0][len(name) + 1:] |
274 |
return ''.join(list) |
275 |
|
276 |
def getheader(self, name, default=None): |
277 |
"""Get the header value for a name.
|
278 |
|
279 |
This is the normal interface: it returns a stripped
|
280 |
version of the header value for a given header name,
|
281 |
or None if it doesn't exist. This uses the dictionary
|
282 |
version which finds the *last* such header.
|
283 |
"""
|
284 |
try:
|
285 |
return self.dict[name.lower()] |
286 |
except KeyError: |
287 |
return default
|
288 |
get = getheader |
289 |
|
290 |
def getheaders(self, name): |
291 |
"""Get all values for a header.
|
292 |
|
293 |
This returns a list of values for headers given more than once;
|
294 |
each value in the result list is stripped in the same way as the
|
295 |
result of getheader(). If the header is not given, return an
|
296 |
empty list.
|
297 |
"""
|
298 |
result = [] |
299 |
current = ''
|
300 |
have_header = 0
|
301 |
for s in self.getallmatchingheaders(name): |
302 |
if s[0].isspace(): |
303 |
if current:
|
304 |
current = "%s\n %s" % (current, s.strip())
|
305 |
else:
|
306 |
current = s.strip() |
307 |
else:
|
308 |
if have_header:
|
309 |
result.append(current) |
310 |
current = s[s.find(":") + 1:].strip() |
311 |
have_header = 1
|
312 |
if have_header:
|
313 |
result.append(current) |
314 |
return result
|
315 |
|
316 |
def getaddr(self, name): |
317 |
"""Get a single address from a header, as a tuple.
|
318 |
|
319 |
An example return value:
|
320 |
('Guido van Rossum', 'guido@cwi.nl')
|
321 |
"""
|
322 |
# New, by Ben Escoto
|
323 |
alist = self.getaddrlist(name)
|
324 |
if alist:
|
325 |
return alist[0] |
326 |
else:
|
327 |
return (None, None) |
328 |
|
329 |
def getaddrlist(self, name): |
330 |
"""Get a list of addresses from a header.
|
331 |
|
332 |
Retrieves a list of addresses from a header, where each address is a
|
333 |
tuple as returned by getaddr(). Scans all named headers, so it works
|
334 |
properly with multiple To: or Cc: headers for example.
|
335 |
|
336 |
"""
|
337 |
raw = [] |
338 |
for h in self.getallmatchingheaders(name): |
339 |
if h[0] in ' \t': |
340 |
raw.append(h) |
341 |
else:
|
342 |
if raw:
|
343 |
raw.append(', ')
|
344 |
i = h.find(':')
|
345 |
if i > 0: |
346 |
addr = h[i+1:]
|
347 |
raw.append(addr) |
348 |
alladdrs = ''.join(raw)
|
349 |
a = AddrlistClass(alladdrs) |
350 |
return a.getaddrlist()
|
351 |
|
352 |
def getdate(self, name): |
353 |
"""Retrieve a date field from a header.
|
354 |
|
355 |
Retrieves a date field from the named header, returning
|
356 |
a tuple compatible with time.mktime().
|
357 |
"""
|
358 |
try:
|
359 |
data = self[name]
|
360 |
except KeyError: |
361 |
return None |
362 |
return parsedate(data)
|
363 |
|
364 |
def getdate_tz(self, name): |
365 |
"""Retrieve a date field from a header as a 10-tuple.
|
366 |
|
367 |
The first 9 elements make up a tuple compatible with
|
368 |
time.mktime(), and the 10th is the offset of the poster's
|
369 |
time zone from GMT/UTC.
|
370 |
"""
|
371 |
try:
|
372 |
data = self[name]
|
373 |
except KeyError: |
374 |
return None |
375 |
return parsedate_tz(data)
|
376 |
|
377 |
|
378 |
# Access as a dictionary (only finds *last* header of each type):
|
379 |
|
380 |
def __len__(self): |
381 |
"""Get the number of headers in a message."""
|
382 |
return len(self.dict) |
383 |
|
384 |
def __getitem__(self, name): |
385 |
"""Get a specific header, as from a dictionary."""
|
386 |
return self.dict[name.lower()] |
387 |
|
388 |
def __setitem__(self, name, value): |
389 |
"""Set the value of a header.
|
390 |
|
391 |
Note: This is not a perfect inversion of __getitem__, because
|
392 |
any changed headers get stuck at the end of the raw-headers list
|
393 |
rather than where the altered header was.
|
394 |
"""
|
395 |
del self[name] # Won't fail if it doesn't exist |
396 |
self.dict[name.lower()] = value
|
397 |
text = name + ": " + value
|
398 |
lines = text.split("\n")
|
399 |
for line in lines: |
400 |
self.headers.append(line + "\n") |
401 |
|
402 |
def __delitem__(self, name): |
403 |
"""Delete all occurrences of a specific header, if it is present."""
|
404 |
name = name.lower() |
405 |
if not self.dict.has_key(name): |
406 |
return
|
407 |
del self.dict[name] |
408 |
name = name + ':'
|
409 |
n = len(name)
|
410 |
list = [] |
411 |
hit = 0
|
412 |
for i in range(len(self.headers)): |
413 |
line = self.headers[i]
|
414 |
if line[:n].lower() == name:
|
415 |
hit = 1
|
416 |
elif not line[:1].isspace(): |
417 |
hit = 0
|
418 |
if hit:
|
419 |
list.append(i)
|
420 |
list.reverse()
|
421 |
for i in list: |
422 |
del self.headers[i] |
423 |
|
424 |
def has_key(self, name): |
425 |
"""Determine whether a message contains the named header."""
|
426 |
return self.dict.has_key(name.lower()) |
427 |
|
428 |
def keys(self): |
429 |
"""Get all of a message's header field names."""
|
430 |
return self.dict.keys() |
431 |
|
432 |
def values(self): |
433 |
"""Get all of a message's header field values."""
|
434 |
return self.dict.values() |
435 |
|
436 |
def items(self): |
437 |
"""Get all of a message's headers.
|
438 |
|
439 |
Returns a list of name, value tuples.
|
440 |
"""
|
441 |
return self.dict.items() |
442 |
|
443 |
def __str__(self): |
444 |
str = ''
|
445 |
for hdr in self.headers: |
446 |
str = str + hdr
|
447 |
return str |
448 |
|
449 |
|
450 |
# Utility functions
|
451 |
# -----------------
|
452 |
|
453 |
# XXX Should fix unquote() and quote() to be really conformant.
|
454 |
# XXX The inverses of the parse functions may also be useful.
|
455 |
|
456 |
|
457 |
def unquote(str): |
458 |
"""Remove quotes from a string."""
|
459 |
if len(str) > 1: |
460 |
if str[0] == '"' and str[-1:] == '"': |
461 |
return str[1:-1] |
462 |
if str[0] == '<' and str[-1:] == '>': |
463 |
return str[1:-1] |
464 |
return str |
465 |
|
466 |
|
467 |
def quote(str): |
468 |
"""Add quotes around a string."""
|
469 |
return str.replace('\\', '\\\\').replace('"', '\\"') |
470 |
|
471 |
|
472 |
def parseaddr(address): |
473 |
"""Parse an address into a (realname, mailaddr) tuple."""
|
474 |
a = AddrlistClass(address) |
475 |
list = a.getaddrlist() |
476 |
if not list: |
477 |
return (None, None) |
478 |
else:
|
479 |
return list[0] |
480 |
|
481 |
|
482 |
class AddrlistClass: |
483 |
"""Address parser class by Ben Escoto.
|
484 |
|
485 |
To understand what this class does, it helps to have a copy of
|
486 |
RFC-822 in front of you.
|
487 |
|
488 |
Note: this class interface is deprecated and may be removed in the future.
|
489 |
Use rfc822.AddressList instead.
|
490 |
"""
|
491 |
|
492 |
def __init__(self, field): |
493 |
"""Initialize a new instance.
|
494 |
|
495 |
`field' is an unparsed address header field, containing
|
496 |
one or more addresses.
|
497 |
"""
|
498 |
self.specials = '()<>@,:;.\"[]' |
499 |
self.pos = 0 |
500 |
self.LWS = ' \t' |
501 |
self.CR = '\r\n' |
502 |
self.atomends = self.specials + self.LWS + self.CR |
503 |
self.field = field
|
504 |
self.commentlist = []
|
505 |
|
506 |
def gotonext(self): |
507 |
"""Parse up to the start of the next address."""
|
508 |
while self.pos < len(self.field): |
509 |
if self.field[self.pos] in self.LWS + '\n\r': |
510 |
self.pos = self.pos + 1 |
511 |
elif self.field[self.pos] == '(': |
512 |
self.commentlist.append(self.getcomment()) |
513 |
else: break |
514 |
|
515 |
def getaddrlist(self): |
516 |
"""Parse all addresses.
|
517 |
|
518 |
Returns a list containing all of the addresses.
|
519 |
"""
|
520 |
ad = self.getaddress()
|
521 |
if ad:
|
522 |
return ad + self.getaddrlist() |
523 |
else: return [] |
524 |
|
525 |
def getaddress(self): |
526 |
"""Parse the next address."""
|
527 |
self.commentlist = []
|
528 |
self.gotonext()
|
529 |
|
530 |
oldpos = self.pos
|
531 |
oldcl = self.commentlist
|
532 |
plist = self.getphraselist()
|
533 |
|
534 |
self.gotonext()
|
535 |
returnlist = [] |
536 |
|
537 |
if self.pos >= len(self.field): |
538 |
# Bad email address technically, no domain.
|
539 |
if plist:
|
540 |
returnlist = [(' '.join(self.commentlist), plist[0])] |
541 |
|
542 |
elif self.field[self.pos] in '.@': |
543 |
# email address is just an addrspec
|
544 |
# this isn't very efficient since we start over
|
545 |
self.pos = oldpos
|
546 |
self.commentlist = oldcl
|
547 |
addrspec = self.getaddrspec()
|
548 |
returnlist = [(' '.join(self.commentlist), addrspec)] |
549 |
|
550 |
elif self.field[self.pos] == ':': |
551 |
# address is a group
|
552 |
returnlist = [] |
553 |
|
554 |
fieldlen = len(self.field) |
555 |
self.pos = self.pos + 1 |
556 |
while self.pos < len(self.field): |
557 |
self.gotonext()
|
558 |
if self.pos < fieldlen and self.field[self.pos] == ';': |
559 |
self.pos = self.pos + 1 |
560 |
break
|
561 |
returnlist = returnlist + self.getaddress()
|
562 |
|
563 |
elif self.field[self.pos] == '<': |
564 |
# Address is a phrase then a route addr
|
565 |
routeaddr = self.getrouteaddr()
|
566 |
|
567 |
if self.commentlist: |
568 |
returnlist = [(' '.join(plist) + ' (' + \ |
569 |
' '.join(self.commentlist) + ')', routeaddr)] |
570 |
else: returnlist = [(' '.join(plist), routeaddr)] |
571 |
|
572 |
else:
|
573 |
if plist:
|
574 |
returnlist = [(' '.join(self.commentlist), plist[0])] |
575 |
elif self.field[self.pos] in self.specials: |
576 |
self.pos = self.pos + 1 |
577 |
|
578 |
self.gotonext()
|
579 |
if self.pos < len(self.field) and self.field[self.pos] == ',': |
580 |
self.pos = self.pos + 1 |
581 |
return returnlist
|
582 |
|
583 |
def getrouteaddr(self): |
584 |
"""Parse a route address (Return-path value).
|
585 |
|
586 |
This method just skips all the route stuff and returns the addrspec.
|
587 |
"""
|
588 |
if self.field[self.pos] != '<': |
589 |
return
|
590 |
|
591 |
expectroute = 0
|
592 |
self.pos = self.pos + 1 |
593 |
self.gotonext()
|
594 |
adlist = None
|
595 |
while self.pos < len(self.field): |
596 |
if expectroute:
|
597 |
self.getdomain()
|
598 |
expectroute = 0
|
599 |
elif self.field[self.pos] == '>': |
600 |
self.pos = self.pos + 1 |
601 |
break
|
602 |
elif self.field[self.pos] == '@': |
603 |
self.pos = self.pos + 1 |
604 |
expectroute = 1
|
605 |
elif self.field[self.pos] == ':': |
606 |
self.pos = self.pos + 1 |
607 |
expectaddrspec = 1
|
608 |
else:
|
609 |
adlist = self.getaddrspec()
|
610 |
self.pos = self.pos + 1 |
611 |
break
|
612 |
self.gotonext()
|
613 |
|
614 |
return adlist
|
615 |
|
616 |
def getaddrspec(self): |
617 |
"""Parse an RFC-822 addr-spec."""
|
618 |
aslist = [] |
619 |
|
620 |
self.gotonext()
|
621 |
while self.pos < len(self.field): |
622 |
if self.field[self.pos] == '.': |
623 |
aslist.append('.')
|
624 |
self.pos = self.pos + 1 |
625 |
elif self.field[self.pos] == '"': |
626 |
aslist.append('"%s"' % self.getquote()) |
627 |
elif self.field[self.pos] in self.atomends: |
628 |
break
|
629 |
else: aslist.append(self.getatom()) |
630 |
self.gotonext()
|
631 |
|
632 |
if self.pos >= len(self.field) or self.field[self.pos] != '@': |
633 |
return ''.join(aslist) |
634 |
|
635 |
aslist.append('@')
|
636 |
self.pos = self.pos + 1 |
637 |
self.gotonext()
|
638 |
return ''.join(aslist) + self.getdomain() |
639 |
|
640 |
def getdomain(self): |
641 |
"""Get the complete domain name from an address."""
|
642 |
sdlist = [] |
643 |
while self.pos < len(self.field): |
644 |
if self.field[self.pos] in self.LWS: |
645 |
self.pos = self.pos + 1 |
646 |
elif self.field[self.pos] == '(': |
647 |
self.commentlist.append(self.getcomment()) |
648 |
elif self.field[self.pos] == '[': |
649 |
sdlist.append(self.getdomainliteral())
|
650 |
elif self.field[self.pos] == '.': |
651 |
self.pos = self.pos + 1 |
652 |
sdlist.append('.')
|
653 |
elif self.field[self.pos] in self.atomends: |
654 |
break
|
655 |
else: sdlist.append(self.getatom()) |
656 |
return ''.join(sdlist) |
657 |
|
658 |
def getdelimited(self, beginchar, endchars, allowcomments = 1): |
659 |
"""Parse a header fragment delimited by special characters.
|
660 |
|
661 |
`beginchar' is the start character for the fragment.
|
662 |
If self is not looking at an instance of `beginchar' then
|
663 |
getdelimited returns the empty string.
|
664 |
|
665 |
`endchars' is a sequence of allowable end-delimiting characters.
|
666 |
Parsing stops when one of these is encountered.
|
667 |
|
668 |
If `allowcomments' is non-zero, embedded RFC-822 comments
|
669 |
are allowed within the parsed fragment.
|
670 |
"""
|
671 |
if self.field[self.pos] != beginchar: |
672 |
return '' |
673 |
|
674 |
slist = ['']
|
675 |
quote = 0
|
676 |
self.pos = self.pos + 1 |
677 |
while self.pos < len(self.field): |
678 |
if quote == 1: |
679 |
slist.append(self.field[self.pos]) |
680 |
quote = 0
|
681 |
elif self.field[self.pos] in endchars: |
682 |
self.pos = self.pos + 1 |
683 |
break
|
684 |
elif allowcomments and self.field[self.pos] == '(': |
685 |
slist.append(self.getcomment())
|
686 |
elif self.field[self.pos] == '\\': |
687 |
quote = 1
|
688 |
else:
|
689 |
slist.append(self.field[self.pos]) |
690 |
self.pos = self.pos + 1 |
691 |
|
692 |
return ''.join(slist) |
693 |
|
694 |
def getquote(self): |
695 |
"""Get a quote-delimited fragment from self's field."""
|
696 |
return self.getdelimited('"', '"\r', 0) |
697 |
|
698 |
def getcomment(self): |
699 |
"""Get a parenthesis-delimited fragment from self's field."""
|
700 |
return self.getdelimited('(', ')\r', 1) |
701 |
|
702 |
def getdomainliteral(self): |
703 |
"""Parse an RFC-822 domain-literal."""
|
704 |
return '[%s]' % self.getdelimited('[', ']\r', 0) |
705 |
|
706 |
def getatom(self): |
707 |
"""Parse an RFC-822 atom."""
|
708 |
atomlist = ['']
|
709 |
|
710 |
while self.pos < len(self.field): |
711 |
if self.field[self.pos] in self.atomends: |
712 |
break
|
713 |
else: atomlist.append(self.field[self.pos]) |
714 |
self.pos = self.pos + 1 |
715 |
|
716 |
return ''.join(atomlist) |
717 |
|
718 |
def getphraselist(self): |
719 |
"""Parse a sequence of RFC-822 phrases.
|
720 |
|
721 |
A phrase is a sequence of words, which are in turn either
|
722 |
RFC-822 atoms or quoted-strings. Phrases are canonicalized
|
723 |
by squeezing all runs of continuous whitespace into one space.
|
724 |
"""
|
725 |
plist = [] |
726 |
|
727 |
while self.pos < len(self.field): |
728 |
if self.field[self.pos] in self.LWS: |
729 |
self.pos = self.pos + 1 |
730 |
elif self.field[self.pos] == '"': |
731 |
plist.append(self.getquote())
|
732 |
elif self.field[self.pos] == '(': |
733 |
self.commentlist.append(self.getcomment()) |
734 |
elif self.field[self.pos] in self.atomends: |
735 |
break
|
736 |
else: plist.append(self.getatom()) |
737 |
|
738 |
return plist
|
739 |
|
740 |
class AddressList(AddrlistClass): |
741 |
"""An AddressList encapsulates a list of parsed RFC822 addresses."""
|
742 |
def __init__(self, field): |
743 |
AddrlistClass.__init__(self, field)
|
744 |
if field:
|
745 |
self.addresslist = self.getaddrlist() |
746 |
else:
|
747 |
self.addresslist = []
|
748 |
|
749 |
def __len__(self): |
750 |
return len(self.addresslist) |
751 |
|
752 |
def __str__(self): |
753 |
return ", ".join(map(dump_address_pair, self.addresslist)) |
754 |
|
755 |
def __add__(self, other): |
756 |
# Set union
|
757 |
newaddr = AddressList(None)
|
758 |
newaddr.addresslist = self.addresslist[:]
|
759 |
for x in other.addresslist: |
760 |
if not x in self.addresslist: |
761 |
newaddr.addresslist.append(x) |
762 |
return newaddr
|
763 |
|
764 |
def __iadd__(self, other): |
765 |
# Set union, in-place
|
766 |
for x in other.addresslist: |
767 |
if not x in self.addresslist: |
768 |
self.addresslist.append(x)
|
769 |
return self |
770 |
|
771 |
def __sub__(self, other): |
772 |
# Set difference
|
773 |
newaddr = AddressList(None)
|
774 |
for x in self.addresslist: |
775 |
if not x in other.addresslist: |
776 |
newaddr.addresslist.append(x) |
777 |
return newaddr
|
778 |
|
779 |
def __isub__(self, other): |
780 |
# Set difference, in-place
|
781 |
for x in other.addresslist: |
782 |
if x in self.addresslist: |
783 |
self.addresslist.remove(x)
|
784 |
return self |
785 |
|
786 |
def __getitem__(self, index): |
787 |
# Make indexing, slices, and 'in' work
|
788 |
return self.addresslist[index] |
789 |
|
790 |
def dump_address_pair(pair): |
791 |
"""Dump a (name, address) pair in a canonicalized form."""
|
792 |
if pair[0]: |
793 |
return '"' + pair[0] + '" <' + pair[1] + '>' |
794 |
else:
|
795 |
return pair[1] |
796 |
|
797 |
# Parse a date field
|
798 |
|
799 |
_monthnames = ['jan', 'feb', 'mar', 'apr', 'may', 'jun', 'jul', |
800 |
'aug', 'sep', 'oct', 'nov', 'dec', |
801 |
'january', 'february', 'march', 'april', 'may', 'june', 'july', |
802 |
'august', 'september', 'october', 'november', 'december'] |
803 |
_daynames = ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun'] |
804 |
|
805 |
# The timezone table does not include the military time zones defined
|
806 |
# in RFC822, other than Z. According to RFC1123, the description in
|
807 |
# RFC822 gets the signs wrong, so we can't rely on any such time
|
808 |
# zones. RFC1123 recommends that numeric timezone indicators be used
|
809 |
# instead of timezone names.
|
810 |
|
811 |
_timezones = {'UT':0, 'UTC':0, 'GMT':0, 'Z':0, |
812 |
'AST': -400, 'ADT': -300, # Atlantic (used in Canada) |
813 |
'EST': -500, 'EDT': -400, # Eastern |
814 |
'CST': -600, 'CDT': -500, # Central |
815 |
'MST': -700, 'MDT': -600, # Mountain |
816 |
'PST': -800, 'PDT': -700 # Pacific |
817 |
} |
818 |
|
819 |
|
820 |
def parsedate_tz(data): |
821 |
"""Convert a date string to a time tuple.
|
822 |
|
823 |
Accounts for military timezones.
|
824 |
"""
|
825 |
data = data.split() |
826 |
if data[0][-1] in (',', '.') or data[0].lower() in _daynames: |
827 |
# There's a dayname here. Skip it
|
828 |
del data[0] |
829 |
if len(data) == 3: # RFC 850 date, deprecated |
830 |
stuff = data[0].split('-') |
831 |
if len(stuff) == 3: |
832 |
data = stuff + data[1:]
|
833 |
if len(data) == 4: |
834 |
s = data[3]
|
835 |
i = s.find('+')
|
836 |
if i > 0: |
837 |
data[3:] = [s[:i], s[i+1:]] |
838 |
else:
|
839 |
data.append('') # Dummy tz |
840 |
if len(data) < 5: |
841 |
return None |
842 |
data = data[:5]
|
843 |
[dd, mm, yy, tm, tz] = data |
844 |
mm = mm.lower() |
845 |
if not mm in _monthnames: |
846 |
dd, mm = mm, dd.lower() |
847 |
if not mm in _monthnames: |
848 |
return None |
849 |
mm = _monthnames.index(mm)+1
|
850 |
if mm > 12: mm = mm - 12 |
851 |
if dd[-1] == ',': |
852 |
dd = dd[:-1]
|
853 |
i = yy.find(':')
|
854 |
if i > 0: |
855 |
yy, tm = tm, yy |
856 |
if yy[-1] == ',': |
857 |
yy = yy[:-1]
|
858 |
if not yy[0].isdigit(): |
859 |
yy, tz = tz, yy |
860 |
if tm[-1] == ',': |
861 |
tm = tm[:-1]
|
862 |
tm = tm.split(':')
|
863 |
if len(tm) == 2: |
864 |
[thh, tmm] = tm |
865 |
tss = '0'
|
866 |
elif len(tm) == 3: |
867 |
[thh, tmm, tss] = tm |
868 |
else:
|
869 |
return None |
870 |
try:
|
871 |
yy = int(yy)
|
872 |
dd = int(dd)
|
873 |
thh = int(thh)
|
874 |
tmm = int(tmm)
|
875 |
tss = int(tss)
|
876 |
except ValueError: |
877 |
return None |
878 |
tzoffset = None
|
879 |
tz = tz.upper() |
880 |
if _timezones.has_key(tz):
|
881 |
tzoffset = _timezones[tz] |
882 |
else:
|
883 |
try:
|
884 |
tzoffset = int(tz)
|
885 |
except ValueError: |
886 |
pass
|
887 |
# Convert a timezone offset into seconds ; -0500 -> -18000
|
888 |
if tzoffset:
|
889 |
if tzoffset < 0: |
890 |
tzsign = -1
|
891 |
tzoffset = -tzoffset |
892 |
else:
|
893 |
tzsign = 1
|
894 |
tzoffset = tzsign * ( (tzoffset/100)*3600 + (tzoffset % 100)*60) |
895 |
tuple = (yy, mm, dd, thh, tmm, tss, 0, 0, 0, tzoffset) |
896 |
return tuple |
897 |
|
898 |
|
899 |
def parsedate(data): |
900 |
"""Convert a time string to a time tuple."""
|
901 |
t = parsedate_tz(data) |
902 |
if type(t) == type( () ): |
903 |
return t[:9] |
904 |
else: return t |
905 |
|
906 |
|
907 |
def mktime_tz(data): |
908 |
"""Turn a 10-tuple as returned by parsedate_tz() into a UTC timestamp."""
|
909 |
if data[9] is None: |
910 |
# No zone info, so localtime is better assumption than GMT
|
911 |
return time.mktime(data[:8] + (-1,)) |
912 |
else:
|
913 |
t = time.mktime(data[:8] + (0,)) |
914 |
return t - data[9] - time.timezone |
915 |
|
916 |
def formatdate(timeval=None): |
917 |
"""Returns time format preferred for Internet standards.
|
918 |
|
919 |
Sun, 06 Nov 1994 08:49:37 GMT ; RFC 822, updated by RFC 1123
|
920 |
"""
|
921 |
if timeval is None: |
922 |
timeval = time.time() |
923 |
return "%s" % time.strftime('%a, %d %b %Y %H:%M:%S GMT', |
924 |
time.gmtime(timeval)) |
925 |
|
926 |
|
927 |
# When used as script, run a small test program.
|
928 |
# The first command line argument must be a filename containing one
|
929 |
# message in RFC-822 format.
|
930 |
|
931 |
if __name__ == '__main__': |
932 |
import sys, os |
933 |
file = os.path.join(os.environ['HOME'], 'Mail/inbox/1') |
934 |
if sys.argv[1:]: file = sys.argv[1] |
935 |
f = open(file, 'r') |
936 |
m = Message(f) |
937 |
print 'From:', m.getaddr('from') |
938 |
print 'To:', m.getaddrlist('to') |
939 |
print 'Subject:', m.getheader('subject') |
940 |
print 'Date:', m.getheader('date') |
941 |
date = m.getdate_tz('date')
|
942 |
tz = date[-1]
|
943 |
date = time.localtime(mktime_tz(date)) |
944 |
if date:
|
945 |
print 'ParsedDate:', time.asctime(date), |
946 |
hhmmss = tz |
947 |
hhmm, ss = divmod(hhmmss, 60) |
948 |
hh, mm = divmod(hhmm, 60) |
949 |
print "%+03d%02d" % (hh, mm), |
950 |
if ss: print ".%02d" % ss, |
951 |
print
|
952 |
else:
|
953 |
print 'ParsedDate:', None |
954 |
m.rewindbody() |
955 |
n = 0
|
956 |
while f.readline():
|
957 |
n = n + 1
|
958 |
print 'Lines:', n |
959 |
print '-'*70 |
960 |
print 'len =', len(m) |
961 |
if m.has_key('Date'): print 'Date =', m['Date'] |
962 |
if m.has_key('X-Nonsense'): pass |
963 |
print 'keys =', m.keys() |
964 |
print 'values =', m.values() |
965 |
print 'items =', m.items() |