1 # Copyright (C) 2001-2007 Python Software Foundation
2 # Author: Barry Warsaw
3 # Contact: email-sig@python.org
4
5 """Basic message object for the email package object model."""
6
7 __all__ = ['Message', 'EmailMessage']
8
9 import binascii
10 import re
11 import quopri
12 from io import BytesIO, StringIO
13
14 # Intrapackage imports
15 from email import utils
16 from email import errors
17 from email._policybase import Policy, compat32
18 from email import charset as _charset
19 from email._encoded_words import decode_b
20 Charset = _charset.Charset
21
22 SEMISPACE = '; '
23
24 # Regular expression that matches `special' characters in parameters, the
25 # existence of which force quoting of the parameter value.
26 tspecials = re.compile(r'[ \(\)<>@,;:\\"/\[\]\?=]')
27
28
29 def _splitparam(param):
30 # Split header parameters. BAW: this may be too simple. It isn't
31 # strictly RFC 2045 (section 5.1) compliant, but it catches most headers
32 # found in the wild. We may eventually need a full fledged parser.
33 # RDM: we might have a Header here; for now just stringify it.
34 a, sep, b = str(param).partition(';')
35 if not sep:
36 return a.strip(), None
37 return a.strip(), b.strip()
38
39 def _formatparam(param, value=None, quote=True):
40 """Convenience function to format and return a key=value pair.
41
42 This will quote the value if needed or if quote is true. If value is a
43 three tuple (charset, language, value), it will be encoded according
44 to RFC2231 rules. If it contains non-ascii characters it will likewise
45 be encoded according to RFC2231 rules, using the utf-8 charset and
46 a null language.
47 """
48 if value is not None and len(value) > 0:
49 # A tuple is used for RFC 2231 encoded parameter values where items
50 # are (charset, language, value). charset is a string, not a Charset
51 # instance. RFC 2231 encoded values are never quoted, per RFC.
52 if isinstance(value, tuple):
53 # Encode as per RFC 2231
54 param += '*'
55 value = utils.encode_rfc2231(value[2], value[0], value[1])
56 return '%s=%s' % (param, value)
57 else:
58 try:
59 value.encode('ascii')
60 except UnicodeEncodeError:
61 param += '*'
62 value = utils.encode_rfc2231(value, 'utf-8', '')
63 return '%s=%s' % (param, value)
64 # BAW: Please check this. I think that if quote is set it should
65 # force quoting even if not necessary.
66 if quote or tspecials.search(value):
67 return '%s="%s"' % (param, utils.quote(value))
68 else:
69 return '%s=%s' % (param, value)
70 else:
71 return param
72
73 def _parseparam(s):
74 # RDM This might be a Header, so for now stringify it.
75 s = ';' + str(s)
76 plist = []
77 while s[:1] == ';':
78 s = s[1:]
79 end = s.find(';')
80 while end > 0 and (s.count('"', 0, end) - s.count('\\"', 0, end)) % 2:
81 end = s.find(';', end + 1)
82 if end < 0:
83 end = len(s)
84 f = s[:end]
85 if '=' in f:
86 i = f.index('=')
87 f = f[:i].strip().lower() + '=' + f[i+1:].strip()
88 plist.append(f.strip())
89 s = s[end:]
90 return plist
91
92
93 def _unquotevalue(value):
94 # This is different than utils.collapse_rfc2231_value() because it doesn't
95 # try to convert the value to a unicode. Message.get_param() and
96 # Message.get_params() are both currently defined to return the tuple in
97 # the face of RFC 2231 parameters.
98 if isinstance(value, tuple):
99 return value[0], value[1], utils.unquote(value[2])
100 else:
101 return utils.unquote(value)
102
103
104 def _decode_uu(encoded):
105 """Decode uuencoded data."""
106 decoded_lines = []
107 encoded_lines_iter = iter(encoded.splitlines())
108 for line in encoded_lines_iter:
109 if line.startswith(b"begin "):
110 mode, _, path = line.removeprefix(b"begin ").partition(b" ")
111 try:
112 int(mode, base=8)
113 except ValueError:
114 continue
115 else:
116 break
117 else:
118 raise ValueError("`begin` line not found")
119 for line in encoded_lines_iter:
120 if not line:
121 raise ValueError("Truncated input")
122 elif line.strip(b' \t\r\n\f') == b'end':
123 break
124 try:
125 decoded_line = binascii.a2b_uu(line)
126 except binascii.Error:
127 # Workaround for broken uuencoders by /Fredrik Lundh
128 nbytes = (((line[0]-32) & 63) * 4 + 5) // 3
129 decoded_line = binascii.a2b_uu(line[:nbytes])
130 decoded_lines.append(decoded_line)
131
132 return b''.join(decoded_lines)
133
134
135 class ESC[4;38;5;81mMessage:
136 """Basic message object.
137
138 A message object is defined as something that has a bunch of RFC 2822
139 headers and a payload. It may optionally have an envelope header
140 (a.k.a. Unix-From or From_ header). If the message is a container (i.e. a
141 multipart or a message/rfc822), then the payload is a list of Message
142 objects, otherwise it is a string.
143
144 Message objects implement part of the `mapping' interface, which assumes
145 there is exactly one occurrence of the header per message. Some headers
146 do in fact appear multiple times (e.g. Received) and for those headers,
147 you must use the explicit API to set or get all the headers. Not all of
148 the mapping methods are implemented.
149 """
150 def __init__(self, policy=compat32):
151 self.policy = policy
152 self._headers = []
153 self._unixfrom = None
154 self._payload = None
155 self._charset = None
156 # Defaults for multipart messages
157 self.preamble = self.epilogue = None
158 self.defects = []
159 # Default content type
160 self._default_type = 'text/plain'
161
162 def __str__(self):
163 """Return the entire formatted message as a string.
164 """
165 return self.as_string()
166
167 def as_string(self, unixfrom=False, maxheaderlen=0, policy=None):
168 """Return the entire formatted message as a string.
169
170 Optional 'unixfrom', when true, means include the Unix From_ envelope
171 header. For backward compatibility reasons, if maxheaderlen is
172 not specified it defaults to 0, so you must override it explicitly
173 if you want a different maxheaderlen. 'policy' is passed to the
174 Generator instance used to serialize the message; if it is not
175 specified the policy associated with the message instance is used.
176
177 If the message object contains binary data that is not encoded
178 according to RFC standards, the non-compliant data will be replaced by
179 unicode "unknown character" code points.
180 """
181 from email.generator import Generator
182 policy = self.policy if policy is None else policy
183 fp = StringIO()
184 g = Generator(fp,
185 mangle_from_=False,
186 maxheaderlen=maxheaderlen,
187 policy=policy)
188 g.flatten(self, unixfrom=unixfrom)
189 return fp.getvalue()
190
191 def __bytes__(self):
192 """Return the entire formatted message as a bytes object.
193 """
194 return self.as_bytes()
195
196 def as_bytes(self, unixfrom=False, policy=None):
197 """Return the entire formatted message as a bytes object.
198
199 Optional 'unixfrom', when true, means include the Unix From_ envelope
200 header. 'policy' is passed to the BytesGenerator instance used to
201 serialize the message; if not specified the policy associated with
202 the message instance is used.
203 """
204 from email.generator import BytesGenerator
205 policy = self.policy if policy is None else policy
206 fp = BytesIO()
207 g = BytesGenerator(fp, mangle_from_=False, policy=policy)
208 g.flatten(self, unixfrom=unixfrom)
209 return fp.getvalue()
210
211 def is_multipart(self):
212 """Return True if the message consists of multiple parts."""
213 return isinstance(self._payload, list)
214
215 #
216 # Unix From_ line
217 #
218 def set_unixfrom(self, unixfrom):
219 self._unixfrom = unixfrom
220
221 def get_unixfrom(self):
222 return self._unixfrom
223
224 #
225 # Payload manipulation.
226 #
227 def attach(self, payload):
228 """Add the given payload to the current payload.
229
230 The current payload will always be a list of objects after this method
231 is called. If you want to set the payload to a scalar object, use
232 set_payload() instead.
233 """
234 if self._payload is None:
235 self._payload = [payload]
236 else:
237 try:
238 self._payload.append(payload)
239 except AttributeError:
240 raise TypeError("Attach is not valid on a message with a"
241 " non-multipart payload")
242
243 def get_payload(self, i=None, decode=False):
244 """Return a reference to the payload.
245
246 The payload will either be a list object or a string. If you mutate
247 the list object, you modify the message's payload in place. Optional
248 i returns that index into the payload.
249
250 Optional decode is a flag indicating whether the payload should be
251 decoded or not, according to the Content-Transfer-Encoding header
252 (default is False).
253
254 When True and the message is not a multipart, the payload will be
255 decoded if this header's value is `quoted-printable' or `base64'. If
256 some other encoding is used, or the header is missing, or if the
257 payload has bogus data (i.e. bogus base64 or uuencoded data), the
258 payload is returned as-is.
259
260 If the message is a multipart and the decode flag is True, then None
261 is returned.
262 """
263 # Here is the logic table for this code, based on the email5.0.0 code:
264 # i decode is_multipart result
265 # ------ ------ ------------ ------------------------------
266 # None True True None
267 # i True True None
268 # None False True _payload (a list)
269 # i False True _payload element i (a Message)
270 # i False False error (not a list)
271 # i True False error (not a list)
272 # None False False _payload
273 # None True False _payload decoded (bytes)
274 # Note that Barry planned to factor out the 'decode' case, but that
275 # isn't so easy now that we handle the 8 bit data, which needs to be
276 # converted in both the decode and non-decode path.
277 if self.is_multipart():
278 if decode:
279 return None
280 if i is None:
281 return self._payload
282 else:
283 return self._payload[i]
284 # For backward compatibility, Use isinstance and this error message
285 # instead of the more logical is_multipart test.
286 if i is not None and not isinstance(self._payload, list):
287 raise TypeError('Expected list, got %s' % type(self._payload))
288 payload = self._payload
289 # cte might be a Header, so for now stringify it.
290 cte = str(self.get('content-transfer-encoding', '')).lower()
291 # payload may be bytes here.
292 if isinstance(payload, str):
293 if utils._has_surrogates(payload):
294 bpayload = payload.encode('ascii', 'surrogateescape')
295 if not decode:
296 try:
297 payload = bpayload.decode(self.get_param('charset', 'ascii'), 'replace')
298 except LookupError:
299 payload = bpayload.decode('ascii', 'replace')
300 elif decode:
301 try:
302 bpayload = payload.encode('ascii')
303 except UnicodeError:
304 # This won't happen for RFC compliant messages (messages
305 # containing only ASCII code points in the unicode input).
306 # If it does happen, turn the string into bytes in a way
307 # guaranteed not to fail.
308 bpayload = payload.encode('raw-unicode-escape')
309 if not decode:
310 return payload
311 if cte == 'quoted-printable':
312 return quopri.decodestring(bpayload)
313 elif cte == 'base64':
314 # XXX: this is a bit of a hack; decode_b should probably be factored
315 # out somewhere, but I haven't figured out where yet.
316 value, defects = decode_b(b''.join(bpayload.splitlines()))
317 for defect in defects:
318 self.policy.handle_defect(self, defect)
319 return value
320 elif cte in ('x-uuencode', 'uuencode', 'uue', 'x-uue'):
321 try:
322 return _decode_uu(bpayload)
323 except ValueError:
324 # Some decoding problem.
325 return bpayload
326 if isinstance(payload, str):
327 return bpayload
328 return payload
329
330 def set_payload(self, payload, charset=None):
331 """Set the payload to the given value.
332
333 Optional charset sets the message's default character set. See
334 set_charset() for details.
335 """
336 if hasattr(payload, 'encode'):
337 if charset is None:
338 self._payload = payload
339 return
340 if not isinstance(charset, Charset):
341 charset = Charset(charset)
342 payload = payload.encode(charset.output_charset)
343 if hasattr(payload, 'decode'):
344 self._payload = payload.decode('ascii', 'surrogateescape')
345 else:
346 self._payload = payload
347 if charset is not None:
348 self.set_charset(charset)
349
350 def set_charset(self, charset):
351 """Set the charset of the payload to a given character set.
352
353 charset can be a Charset instance, a string naming a character set, or
354 None. If it is a string it will be converted to a Charset instance.
355 If charset is None, the charset parameter will be removed from the
356 Content-Type field. Anything else will generate a TypeError.
357
358 The message will be assumed to be of type text/* encoded with
359 charset.input_charset. It will be converted to charset.output_charset
360 and encoded properly, if needed, when generating the plain text
361 representation of the message. MIME headers (MIME-Version,
362 Content-Type, Content-Transfer-Encoding) will be added as needed.
363 """
364 if charset is None:
365 self.del_param('charset')
366 self._charset = None
367 return
368 if not isinstance(charset, Charset):
369 charset = Charset(charset)
370 self._charset = charset
371 if 'MIME-Version' not in self:
372 self.add_header('MIME-Version', '1.0')
373 if 'Content-Type' not in self:
374 self.add_header('Content-Type', 'text/plain',
375 charset=charset.get_output_charset())
376 else:
377 self.set_param('charset', charset.get_output_charset())
378 if charset != charset.get_output_charset():
379 self._payload = charset.body_encode(self._payload)
380 if 'Content-Transfer-Encoding' not in self:
381 cte = charset.get_body_encoding()
382 try:
383 cte(self)
384 except TypeError:
385 # This 'if' is for backward compatibility, it allows unicode
386 # through even though that won't work correctly if the
387 # message is serialized.
388 payload = self._payload
389 if payload:
390 try:
391 payload = payload.encode('ascii', 'surrogateescape')
392 except UnicodeError:
393 payload = payload.encode(charset.output_charset)
394 self._payload = charset.body_encode(payload)
395 self.add_header('Content-Transfer-Encoding', cte)
396
397 def get_charset(self):
398 """Return the Charset instance associated with the message's payload.
399 """
400 return self._charset
401
402 #
403 # MAPPING INTERFACE (partial)
404 #
405 def __len__(self):
406 """Return the total number of headers, including duplicates."""
407 return len(self._headers)
408
409 def __getitem__(self, name):
410 """Get a header value.
411
412 Return None if the header is missing instead of raising an exception.
413
414 Note that if the header appeared multiple times, exactly which
415 occurrence gets returned is undefined. Use get_all() to get all
416 the values matching a header field name.
417 """
418 return self.get(name)
419
420 def __setitem__(self, name, val):
421 """Set the value of a header.
422
423 Note: this does not overwrite an existing header with the same field
424 name. Use __delitem__() first to delete any existing headers.
425 """
426 max_count = self.policy.header_max_count(name)
427 if max_count:
428 lname = name.lower()
429 found = 0
430 for k, v in self._headers:
431 if k.lower() == lname:
432 found += 1
433 if found >= max_count:
434 raise ValueError("There may be at most {} {} headers "
435 "in a message".format(max_count, name))
436 self._headers.append(self.policy.header_store_parse(name, val))
437
438 def __delitem__(self, name):
439 """Delete all occurrences of a header, if present.
440
441 Does not raise an exception if the header is missing.
442 """
443 name = name.lower()
444 newheaders = []
445 for k, v in self._headers:
446 if k.lower() != name:
447 newheaders.append((k, v))
448 self._headers = newheaders
449
450 def __contains__(self, name):
451 return name.lower() in [k.lower() for k, v in self._headers]
452
453 def __iter__(self):
454 for field, value in self._headers:
455 yield field
456
457 def keys(self):
458 """Return a list of all the message's header field names.
459
460 These will be sorted in the order they appeared in the original
461 message, or were added to the message, and may contain duplicates.
462 Any fields deleted and re-inserted are always appended to the header
463 list.
464 """
465 return [k for k, v in self._headers]
466
467 def values(self):
468 """Return a list of all the message's header values.
469
470 These will be sorted in the order they appeared in the original
471 message, or were added to the message, and may contain duplicates.
472 Any fields deleted and re-inserted are always appended to the header
473 list.
474 """
475 return [self.policy.header_fetch_parse(k, v)
476 for k, v in self._headers]
477
478 def items(self):
479 """Get all the message's header fields and values.
480
481 These will be sorted in the order they appeared in the original
482 message, or were added to the message, and may contain duplicates.
483 Any fields deleted and re-inserted are always appended to the header
484 list.
485 """
486 return [(k, self.policy.header_fetch_parse(k, v))
487 for k, v in self._headers]
488
489 def get(self, name, failobj=None):
490 """Get a header value.
491
492 Like __getitem__() but return failobj instead of None when the field
493 is missing.
494 """
495 name = name.lower()
496 for k, v in self._headers:
497 if k.lower() == name:
498 return self.policy.header_fetch_parse(k, v)
499 return failobj
500
501 #
502 # "Internal" methods (public API, but only intended for use by a parser
503 # or generator, not normal application code.
504 #
505
506 def set_raw(self, name, value):
507 """Store name and value in the model without modification.
508
509 This is an "internal" API, intended only for use by a parser.
510 """
511 self._headers.append((name, value))
512
513 def raw_items(self):
514 """Return the (name, value) header pairs without modification.
515
516 This is an "internal" API, intended only for use by a generator.
517 """
518 return iter(self._headers.copy())
519
520 #
521 # Additional useful stuff
522 #
523
524 def get_all(self, name, failobj=None):
525 """Return a list of all the values for the named field.
526
527 These will be sorted in the order they appeared in the original
528 message, and may contain duplicates. Any fields deleted and
529 re-inserted are always appended to the header list.
530
531 If no such fields exist, failobj is returned (defaults to None).
532 """
533 values = []
534 name = name.lower()
535 for k, v in self._headers:
536 if k.lower() == name:
537 values.append(self.policy.header_fetch_parse(k, v))
538 if not values:
539 return failobj
540 return values
541
542 def add_header(self, _name, _value, **_params):
543 """Extended header setting.
544
545 name is the header field to add. keyword arguments can be used to set
546 additional parameters for the header field, with underscores converted
547 to dashes. Normally the parameter will be added as key="value" unless
548 value is None, in which case only the key will be added. If a
549 parameter value contains non-ASCII characters it can be specified as a
550 three-tuple of (charset, language, value), in which case it will be
551 encoded according to RFC2231 rules. Otherwise it will be encoded using
552 the utf-8 charset and a language of ''.
553
554 Examples:
555
556 msg.add_header('content-disposition', 'attachment', filename='bud.gif')
557 msg.add_header('content-disposition', 'attachment',
558 filename=('utf-8', '', Fußballer.ppt'))
559 msg.add_header('content-disposition', 'attachment',
560 filename='Fußballer.ppt'))
561 """
562 parts = []
563 for k, v in _params.items():
564 if v is None:
565 parts.append(k.replace('_', '-'))
566 else:
567 parts.append(_formatparam(k.replace('_', '-'), v))
568 if _value is not None:
569 parts.insert(0, _value)
570 self[_name] = SEMISPACE.join(parts)
571
572 def replace_header(self, _name, _value):
573 """Replace a header.
574
575 Replace the first matching header found in the message, retaining
576 header order and case. If no matching header was found, a KeyError is
577 raised.
578 """
579 _name = _name.lower()
580 for i, (k, v) in zip(range(len(self._headers)), self._headers):
581 if k.lower() == _name:
582 self._headers[i] = self.policy.header_store_parse(k, _value)
583 break
584 else:
585 raise KeyError(_name)
586
587 #
588 # Use these three methods instead of the three above.
589 #
590
591 def get_content_type(self):
592 """Return the message's content type.
593
594 The returned string is coerced to lower case of the form
595 `maintype/subtype'. If there was no Content-Type header in the
596 message, the default type as given by get_default_type() will be
597 returned. Since according to RFC 2045, messages always have a default
598 type this will always return a value.
599
600 RFC 2045 defines a message's default type to be text/plain unless it
601 appears inside a multipart/digest container, in which case it would be
602 message/rfc822.
603 """
604 missing = object()
605 value = self.get('content-type', missing)
606 if value is missing:
607 # This should have no parameters
608 return self.get_default_type()
609 ctype = _splitparam(value)[0].lower()
610 # RFC 2045, section 5.2 says if its invalid, use text/plain
611 if ctype.count('/') != 1:
612 return 'text/plain'
613 return ctype
614
615 def get_content_maintype(self):
616 """Return the message's main content type.
617
618 This is the `maintype' part of the string returned by
619 get_content_type().
620 """
621 ctype = self.get_content_type()
622 return ctype.split('/')[0]
623
624 def get_content_subtype(self):
625 """Returns the message's sub-content type.
626
627 This is the `subtype' part of the string returned by
628 get_content_type().
629 """
630 ctype = self.get_content_type()
631 return ctype.split('/')[1]
632
633 def get_default_type(self):
634 """Return the `default' content type.
635
636 Most messages have a default content type of text/plain, except for
637 messages that are subparts of multipart/digest containers. Such
638 subparts have a default content type of message/rfc822.
639 """
640 return self._default_type
641
642 def set_default_type(self, ctype):
643 """Set the `default' content type.
644
645 ctype should be either "text/plain" or "message/rfc822", although this
646 is not enforced. The default content type is not stored in the
647 Content-Type header.
648 """
649 self._default_type = ctype
650
651 def _get_params_preserve(self, failobj, header):
652 # Like get_params() but preserves the quoting of values. BAW:
653 # should this be part of the public interface?
654 missing = object()
655 value = self.get(header, missing)
656 if value is missing:
657 return failobj
658 params = []
659 for p in _parseparam(value):
660 try:
661 name, val = p.split('=', 1)
662 name = name.strip()
663 val = val.strip()
664 except ValueError:
665 # Must have been a bare attribute
666 name = p.strip()
667 val = ''
668 params.append((name, val))
669 params = utils.decode_params(params)
670 return params
671
672 def get_params(self, failobj=None, header='content-type', unquote=True):
673 """Return the message's Content-Type parameters, as a list.
674
675 The elements of the returned list are 2-tuples of key/value pairs, as
676 split on the `=' sign. The left hand side of the `=' is the key,
677 while the right hand side is the value. If there is no `=' sign in
678 the parameter the value is the empty string. The value is as
679 described in the get_param() method.
680
681 Optional failobj is the object to return if there is no Content-Type
682 header. Optional header is the header to search instead of
683 Content-Type. If unquote is True, the value is unquoted.
684 """
685 missing = object()
686 params = self._get_params_preserve(missing, header)
687 if params is missing:
688 return failobj
689 if unquote:
690 return [(k, _unquotevalue(v)) for k, v in params]
691 else:
692 return params
693
694 def get_param(self, param, failobj=None, header='content-type',
695 unquote=True):
696 """Return the parameter value if found in the Content-Type header.
697
698 Optional failobj is the object to return if there is no Content-Type
699 header, or the Content-Type header has no such parameter. Optional
700 header is the header to search instead of Content-Type.
701
702 Parameter keys are always compared case insensitively. The return
703 value can either be a string, or a 3-tuple if the parameter was RFC
704 2231 encoded. When it's a 3-tuple, the elements of the value are of
705 the form (CHARSET, LANGUAGE, VALUE). Note that both CHARSET and
706 LANGUAGE can be None, in which case you should consider VALUE to be
707 encoded in the us-ascii charset. You can usually ignore LANGUAGE.
708 The parameter value (either the returned string, or the VALUE item in
709 the 3-tuple) is always unquoted, unless unquote is set to False.
710
711 If your application doesn't care whether the parameter was RFC 2231
712 encoded, it can turn the return value into a string as follows:
713
714 rawparam = msg.get_param('foo')
715 param = email.utils.collapse_rfc2231_value(rawparam)
716
717 """
718 if header not in self:
719 return failobj
720 for k, v in self._get_params_preserve(failobj, header):
721 if k.lower() == param.lower():
722 if unquote:
723 return _unquotevalue(v)
724 else:
725 return v
726 return failobj
727
728 def set_param(self, param, value, header='Content-Type', requote=True,
729 charset=None, language='', replace=False):
730 """Set a parameter in the Content-Type header.
731
732 If the parameter already exists in the header, its value will be
733 replaced with the new value.
734
735 If header is Content-Type and has not yet been defined for this
736 message, it will be set to "text/plain" and the new parameter and
737 value will be appended as per RFC 2045.
738
739 An alternate header can be specified in the header argument, and all
740 parameters will be quoted as necessary unless requote is False.
741
742 If charset is specified, the parameter will be encoded according to RFC
743 2231. Optional language specifies the RFC 2231 language, defaulting
744 to the empty string. Both charset and language should be strings.
745 """
746 if not isinstance(value, tuple) and charset:
747 value = (charset, language, value)
748
749 if header not in self and header.lower() == 'content-type':
750 ctype = 'text/plain'
751 else:
752 ctype = self.get(header)
753 if not self.get_param(param, header=header):
754 if not ctype:
755 ctype = _formatparam(param, value, requote)
756 else:
757 ctype = SEMISPACE.join(
758 [ctype, _formatparam(param, value, requote)])
759 else:
760 ctype = ''
761 for old_param, old_value in self.get_params(header=header,
762 unquote=requote):
763 append_param = ''
764 if old_param.lower() == param.lower():
765 append_param = _formatparam(param, value, requote)
766 else:
767 append_param = _formatparam(old_param, old_value, requote)
768 if not ctype:
769 ctype = append_param
770 else:
771 ctype = SEMISPACE.join([ctype, append_param])
772 if ctype != self.get(header):
773 if replace:
774 self.replace_header(header, ctype)
775 else:
776 del self[header]
777 self[header] = ctype
778
779 def del_param(self, param, header='content-type', requote=True):
780 """Remove the given parameter completely from the Content-Type header.
781
782 The header will be re-written in place without the parameter or its
783 value. All values will be quoted as necessary unless requote is
784 False. Optional header specifies an alternative to the Content-Type
785 header.
786 """
787 if header not in self:
788 return
789 new_ctype = ''
790 for p, v in self.get_params(header=header, unquote=requote):
791 if p.lower() != param.lower():
792 if not new_ctype:
793 new_ctype = _formatparam(p, v, requote)
794 else:
795 new_ctype = SEMISPACE.join([new_ctype,
796 _formatparam(p, v, requote)])
797 if new_ctype != self.get(header):
798 del self[header]
799 self[header] = new_ctype
800
801 def set_type(self, type, header='Content-Type', requote=True):
802 """Set the main type and subtype for the Content-Type header.
803
804 type must be a string in the form "maintype/subtype", otherwise a
805 ValueError is raised.
806
807 This method replaces the Content-Type header, keeping all the
808 parameters in place. If requote is False, this leaves the existing
809 header's quoting as is. Otherwise, the parameters will be quoted (the
810 default).
811
812 An alternative header can be specified in the header argument. When
813 the Content-Type header is set, we'll always also add a MIME-Version
814 header.
815 """
816 # BAW: should we be strict?
817 if not type.count('/') == 1:
818 raise ValueError
819 # Set the Content-Type, you get a MIME-Version
820 if header.lower() == 'content-type':
821 del self['mime-version']
822 self['MIME-Version'] = '1.0'
823 if header not in self:
824 self[header] = type
825 return
826 params = self.get_params(header=header, unquote=requote)
827 del self[header]
828 self[header] = type
829 # Skip the first param; it's the old type.
830 for p, v in params[1:]:
831 self.set_param(p, v, header, requote)
832
833 def get_filename(self, failobj=None):
834 """Return the filename associated with the payload if present.
835
836 The filename is extracted from the Content-Disposition header's
837 `filename' parameter, and it is unquoted. If that header is missing
838 the `filename' parameter, this method falls back to looking for the
839 `name' parameter.
840 """
841 missing = object()
842 filename = self.get_param('filename', missing, 'content-disposition')
843 if filename is missing:
844 filename = self.get_param('name', missing, 'content-type')
845 if filename is missing:
846 return failobj
847 return utils.collapse_rfc2231_value(filename).strip()
848
849 def get_boundary(self, failobj=None):
850 """Return the boundary associated with the payload if present.
851
852 The boundary is extracted from the Content-Type header's `boundary'
853 parameter, and it is unquoted.
854 """
855 missing = object()
856 boundary = self.get_param('boundary', missing)
857 if boundary is missing:
858 return failobj
859 # RFC 2046 says that boundaries may begin but not end in w/s
860 return utils.collapse_rfc2231_value(boundary).rstrip()
861
862 def set_boundary(self, boundary):
863 """Set the boundary parameter in Content-Type to 'boundary'.
864
865 This is subtly different than deleting the Content-Type header and
866 adding a new one with a new boundary parameter via add_header(). The
867 main difference is that using the set_boundary() method preserves the
868 order of the Content-Type header in the original message.
869
870 HeaderParseError is raised if the message has no Content-Type header.
871 """
872 missing = object()
873 params = self._get_params_preserve(missing, 'content-type')
874 if params is missing:
875 # There was no Content-Type header, and we don't know what type
876 # to set it to, so raise an exception.
877 raise errors.HeaderParseError('No Content-Type header found')
878 newparams = []
879 foundp = False
880 for pk, pv in params:
881 if pk.lower() == 'boundary':
882 newparams.append(('boundary', '"%s"' % boundary))
883 foundp = True
884 else:
885 newparams.append((pk, pv))
886 if not foundp:
887 # The original Content-Type header had no boundary attribute.
888 # Tack one on the end. BAW: should we raise an exception
889 # instead???
890 newparams.append(('boundary', '"%s"' % boundary))
891 # Replace the existing Content-Type header with the new value
892 newheaders = []
893 for h, v in self._headers:
894 if h.lower() == 'content-type':
895 parts = []
896 for k, v in newparams:
897 if v == '':
898 parts.append(k)
899 else:
900 parts.append('%s=%s' % (k, v))
901 val = SEMISPACE.join(parts)
902 newheaders.append(self.policy.header_store_parse(h, val))
903
904 else:
905 newheaders.append((h, v))
906 self._headers = newheaders
907
908 def get_content_charset(self, failobj=None):
909 """Return the charset parameter of the Content-Type header.
910
911 The returned string is always coerced to lower case. If there is no
912 Content-Type header, or if that header has no charset parameter,
913 failobj is returned.
914 """
915 missing = object()
916 charset = self.get_param('charset', missing)
917 if charset is missing:
918 return failobj
919 if isinstance(charset, tuple):
920 # RFC 2231 encoded, so decode it, and it better end up as ascii.
921 pcharset = charset[0] or 'us-ascii'
922 try:
923 # LookupError will be raised if the charset isn't known to
924 # Python. UnicodeError will be raised if the encoded text
925 # contains a character not in the charset.
926 as_bytes = charset[2].encode('raw-unicode-escape')
927 charset = str(as_bytes, pcharset)
928 except (LookupError, UnicodeError):
929 charset = charset[2]
930 # charset characters must be in us-ascii range
931 try:
932 charset.encode('us-ascii')
933 except UnicodeError:
934 return failobj
935 # RFC 2046, $4.1.2 says charsets are not case sensitive
936 return charset.lower()
937
938 def get_charsets(self, failobj=None):
939 """Return a list containing the charset(s) used in this message.
940
941 The returned list of items describes the Content-Type headers'
942 charset parameter for this message and all the subparts in its
943 payload.
944
945 Each item will either be a string (the value of the charset parameter
946 in the Content-Type header of that part) or the value of the
947 'failobj' parameter (defaults to None), if the part does not have a
948 main MIME type of "text", or the charset is not defined.
949
950 The list will contain one string for each part of the message, plus
951 one for the container message (i.e. self), so that a non-multipart
952 message will still return a list of length 1.
953 """
954 return [part.get_content_charset(failobj) for part in self.walk()]
955
956 def get_content_disposition(self):
957 """Return the message's content-disposition if it exists, or None.
958
959 The return values can be either 'inline', 'attachment' or None
960 according to the rfc2183.
961 """
962 value = self.get('content-disposition')
963 if value is None:
964 return None
965 c_d = _splitparam(value)[0].lower()
966 return c_d
967
968 # I.e. def walk(self): ...
969 from email.iterators import walk
970
971
972 class ESC[4;38;5;81mMIMEPart(ESC[4;38;5;149mMessage):
973
974 def __init__(self, policy=None):
975 if policy is None:
976 from email.policy import default
977 policy = default
978 super().__init__(policy)
979
980
981 def as_string(self, unixfrom=False, maxheaderlen=None, policy=None):
982 """Return the entire formatted message as a string.
983
984 Optional 'unixfrom', when true, means include the Unix From_ envelope
985 header. maxheaderlen is retained for backward compatibility with the
986 base Message class, but defaults to None, meaning that the policy value
987 for max_line_length controls the header maximum length. 'policy' is
988 passed to the Generator instance used to serialize the message; if it
989 is not specified the policy associated with the message instance is
990 used.
991 """
992 policy = self.policy if policy is None else policy
993 if maxheaderlen is None:
994 maxheaderlen = policy.max_line_length
995 return super().as_string(unixfrom, maxheaderlen, policy)
996
997 def __str__(self):
998 return self.as_string(policy=self.policy.clone(utf8=True))
999
1000 def is_attachment(self):
1001 c_d = self.get('content-disposition')
1002 return False if c_d is None else c_d.content_disposition == 'attachment'
1003
1004 def _find_body(self, part, preferencelist):
1005 if part.is_attachment():
1006 return
1007 maintype, subtype = part.get_content_type().split('/')
1008 if maintype == 'text':
1009 if subtype in preferencelist:
1010 yield (preferencelist.index(subtype), part)
1011 return
1012 if maintype != 'multipart' or not self.is_multipart():
1013 return
1014 if subtype != 'related':
1015 for subpart in part.iter_parts():
1016 yield from self._find_body(subpart, preferencelist)
1017 return
1018 if 'related' in preferencelist:
1019 yield (preferencelist.index('related'), part)
1020 candidate = None
1021 start = part.get_param('start')
1022 if start:
1023 for subpart in part.iter_parts():
1024 if subpart['content-id'] == start:
1025 candidate = subpart
1026 break
1027 if candidate is None:
1028 subparts = part.get_payload()
1029 candidate = subparts[0] if subparts else None
1030 if candidate is not None:
1031 yield from self._find_body(candidate, preferencelist)
1032
1033 def get_body(self, preferencelist=('related', 'html', 'plain')):
1034 """Return best candidate mime part for display as 'body' of message.
1035
1036 Do a depth first search, starting with self, looking for the first part
1037 matching each of the items in preferencelist, and return the part
1038 corresponding to the first item that has a match, or None if no items
1039 have a match. If 'related' is not included in preferencelist, consider
1040 the root part of any multipart/related encountered as a candidate
1041 match. Ignore parts with 'Content-Disposition: attachment'.
1042 """
1043 best_prio = len(preferencelist)
1044 body = None
1045 for prio, part in self._find_body(self, preferencelist):
1046 if prio < best_prio:
1047 best_prio = prio
1048 body = part
1049 if prio == 0:
1050 break
1051 return body
1052
1053 _body_types = {('text', 'plain'),
1054 ('text', 'html'),
1055 ('multipart', 'related'),
1056 ('multipart', 'alternative')}
1057 def iter_attachments(self):
1058 """Return an iterator over the non-main parts of a multipart.
1059
1060 Skip the first of each occurrence of text/plain, text/html,
1061 multipart/related, or multipart/alternative in the multipart (unless
1062 they have a 'Content-Disposition: attachment' header) and include all
1063 remaining subparts in the returned iterator. When applied to a
1064 multipart/related, return all parts except the root part. Return an
1065 empty iterator when applied to a multipart/alternative or a
1066 non-multipart.
1067 """
1068 maintype, subtype = self.get_content_type().split('/')
1069 if maintype != 'multipart' or subtype == 'alternative':
1070 return
1071 payload = self.get_payload()
1072 # Certain malformed messages can have content type set to `multipart/*`
1073 # but still have single part body, in which case payload.copy() can
1074 # fail with AttributeError.
1075 try:
1076 parts = payload.copy()
1077 except AttributeError:
1078 # payload is not a list, it is most probably a string.
1079 return
1080
1081 if maintype == 'multipart' and subtype == 'related':
1082 # For related, we treat everything but the root as an attachment.
1083 # The root may be indicated by 'start'; if there's no start or we
1084 # can't find the named start, treat the first subpart as the root.
1085 start = self.get_param('start')
1086 if start:
1087 found = False
1088 attachments = []
1089 for part in parts:
1090 if part.get('content-id') == start:
1091 found = True
1092 else:
1093 attachments.append(part)
1094 if found:
1095 yield from attachments
1096 return
1097 parts.pop(0)
1098 yield from parts
1099 return
1100 # Otherwise we more or less invert the remaining logic in get_body.
1101 # This only really works in edge cases (ex: non-text related or
1102 # alternatives) if the sending agent sets content-disposition.
1103 seen = [] # Only skip the first example of each candidate type.
1104 for part in parts:
1105 maintype, subtype = part.get_content_type().split('/')
1106 if ((maintype, subtype) in self._body_types and
1107 not part.is_attachment() and subtype not in seen):
1108 seen.append(subtype)
1109 continue
1110 yield part
1111
1112 def iter_parts(self):
1113 """Return an iterator over all immediate subparts of a multipart.
1114
1115 Return an empty iterator for a non-multipart.
1116 """
1117 if self.is_multipart():
1118 yield from self.get_payload()
1119
1120 def get_content(self, *args, content_manager=None, **kw):
1121 if content_manager is None:
1122 content_manager = self.policy.content_manager
1123 return content_manager.get_content(self, *args, **kw)
1124
1125 def set_content(self, *args, content_manager=None, **kw):
1126 if content_manager is None:
1127 content_manager = self.policy.content_manager
1128 content_manager.set_content(self, *args, **kw)
1129
1130 def _make_multipart(self, subtype, disallowed_subtypes, boundary):
1131 if self.get_content_maintype() == 'multipart':
1132 existing_subtype = self.get_content_subtype()
1133 disallowed_subtypes = disallowed_subtypes + (subtype,)
1134 if existing_subtype in disallowed_subtypes:
1135 raise ValueError("Cannot convert {} to {}".format(
1136 existing_subtype, subtype))
1137 keep_headers = []
1138 part_headers = []
1139 for name, value in self._headers:
1140 if name.lower().startswith('content-'):
1141 part_headers.append((name, value))
1142 else:
1143 keep_headers.append((name, value))
1144 if part_headers:
1145 # There is existing content, move it to the first subpart.
1146 part = type(self)(policy=self.policy)
1147 part._headers = part_headers
1148 part._payload = self._payload
1149 self._payload = [part]
1150 else:
1151 self._payload = []
1152 self._headers = keep_headers
1153 self['Content-Type'] = 'multipart/' + subtype
1154 if boundary is not None:
1155 self.set_param('boundary', boundary)
1156
1157 def make_related(self, boundary=None):
1158 self._make_multipart('related', ('alternative', 'mixed'), boundary)
1159
1160 def make_alternative(self, boundary=None):
1161 self._make_multipart('alternative', ('mixed',), boundary)
1162
1163 def make_mixed(self, boundary=None):
1164 self._make_multipart('mixed', (), boundary)
1165
1166 def _add_multipart(self, _subtype, *args, _disp=None, **kw):
1167 if (self.get_content_maintype() != 'multipart' or
1168 self.get_content_subtype() != _subtype):
1169 getattr(self, 'make_' + _subtype)()
1170 part = type(self)(policy=self.policy)
1171 part.set_content(*args, **kw)
1172 if _disp and 'content-disposition' not in part:
1173 part['Content-Disposition'] = _disp
1174 self.attach(part)
1175
1176 def add_related(self, *args, **kw):
1177 self._add_multipart('related', *args, _disp='inline', **kw)
1178
1179 def add_alternative(self, *args, **kw):
1180 self._add_multipart('alternative', *args, **kw)
1181
1182 def add_attachment(self, *args, **kw):
1183 self._add_multipart('mixed', *args, _disp='attachment', **kw)
1184
1185 def clear(self):
1186 self._headers = []
1187 self._payload = None
1188
1189 def clear_content(self):
1190 self._headers = [(n, v) for n, v in self._headers
1191 if not n.lower().startswith('content-')]
1192 self._payload = None
1193
1194
1195 class ESC[4;38;5;81mEmailMessage(ESC[4;38;5;149mMIMEPart):
1196
1197 def set_content(self, *args, **kw):
1198 super().set_content(*args, **kw)
1199 if 'MIME-Version' not in self:
1200 self['MIME-Version'] = '1.0'