X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=mimedecode.py;h=aa915434f09574b8fd9d4ec5ef3d2ab40f355549;hb=1b8fe309964ed2c651f3453ef98896a0301fe463;hp=48a6aba65ff01450825798e128319e0479d4e917;hpb=b4bcf7032c5374be2c52de0dd88c16bd36d9f040;p=mimedecode.git diff --git a/mimedecode.py b/mimedecode.py index 48a6aba..aa91543 100755 --- a/mimedecode.py +++ b/mimedecode.py @@ -2,8 +2,13 @@ """Decode MIME message""" import sys, os +import subprocess from mimedecode_version import __version__, \ __author__, __copyright__, __license__ +if sys.version_info[0] >= 3: + # Replace email.message._formatparam with _formatparam from Python 2.7 + # to avoid re-encoding non-ascii params. + import formatparam_27 me = os.path.basename(sys.argv[0]) @@ -17,78 +22,90 @@ Broytman mimedecode.py version %s, %s def usage(code=0, errormsg=''): version(0) sys.stdout.write("""\ - Usage: %s [-h|--help] [-V|--version] [-cCDP] [-H|--host=hostname] [-f charset] [-d header1[,h2,...]|*[,-h1,...]] [-p header1[,h2,h3,...]:param1[,p2,p3,...]] [-r header1[,h2,...]|*[,-h1,...]] [-R header1[,h2,h3,...]:param1[,p2,p3,...]] [--set-header header:value] [--set-param header:param=value] [-Bbeit mask] [--save-headers|body|message mask] [-O dest_dir] [-o output_file] [input_file [output_file]] +Usage: %s [-h|--help] [-V|--version] [-cCDP] [-H|--host=hostname] [-f charset] [-d header1[,h2,...]|*[,-h1,...]] [-p header1[,h2,h3,...]:param1[,p2,p3,...]] [-r header1[,h2,...]|*[,-h1,...]] [-R header1[,h2,h3,...]:param1[,p2,p3,...]] [--set-header header:value] [--set-param header:param=value] [-Bbeit mask] [--save-headers|body|message mask] [-O dest_dir] [-o output_file] [input_file [output_file]] """ % me) if errormsg: - sys.stderr.write(errormsg + '\n') + sys.stderr.write(errormsg + os.linesep) sys.exit(code) def output_headers(msg): unix_from = msg.get_unixfrom() if unix_from: - output(unix_from + '\n') + output(unix_from) + output(os.linesep) for key, value in msg.items(): - output("%s: %s\n" % (key, value)) - output("\n") # End of headers + output(key) + output(": ") + value = value.split(';', 1) + output(value[0]) + if len(value) == 2: + output(";") + output(_decode_header(value[1], strip=False)) + output(os.linesep) + output(os.linesep) # End of headers def recode_if_needed(s, charset): - if charset and charset.lower() <> g.default_encoding: - s = unicode(s, charset, "replace").encode(g.default_encoding, "replace") + if bytes is str: # Python2 + if isinstance(s, bytes) and \ + charset and charset.lower() != g.default_encoding: + s = s.decode(charset, "replace").\ + encode(g.default_encoding, "replace") + else: # Python3 + if isinstance(s, bytes): + s = s.decode(charset, "replace") return s -def _decode_header(s): +def _decode_header(s, strip=True): """Return a decoded string according to RFC 2047. NOTE: This is almost the same as email.Utils.decode. """ - import email.Header + import email.header - L = email.Header.decode_header(s) + L = email.header.decode_header(s) if not isinstance(L, list): # s wasn't decoded return s rtn = [] for atom, charset in L: - if charset is None: - rtn.append(atom) - else: - rtn.append(recode_if_needed(atom, charset)) - rtn.append(' ') - del rtn[-1] # remove the last space + atom = recode_if_needed(atom, charset or g.default_encoding) + if strip: + atom = atom.strip() + rtn.append(atom) # Now that we've decoded everything, we just need to join all the parts # together into the final string. - return ''.join(rtn) + return ' '.join(rtn) def decode_header(msg, header): "Decode mail header (if exists) and put it back, if it was encoded" - if msg.has_key(header): + if header in msg: value = msg[header] new_value = _decode_header(value) - if new_value <> value: # do not bother to touch msg if not changed + if new_value != value: # do not bother to touch msg if not changed set_header(msg, header, new_value) def decode_header_param(msg, header, param): "Decode mail header's parameter (if exists) and put it back, if it was encoded" - if msg.has_key(header): + if header in msg: value = msg.get_param(param, header=header) if value: if isinstance(value, tuple): new_value = recode_if_needed(value[2], value[0]) else: new_value = _decode_header(value) - if new_value <> value: # do not bother to touch msg if not changed + if new_value != value: # do not bother to touch msg if not changed msg.set_param(param, new_value, header) def _get_exceptions(list): - return [x[1:].lower() for x in list[1:] if x[0]=='-'] + return [x[1:].lower() for x in list[1:] if x[0] == '-'] def _decode_headers_params(msg, header, decode_all_params, param_list): if decode_all_params: @@ -111,11 +128,11 @@ def _remove_headers_params(msg, header, remove_all_params, param_list): msg.del_param(param, header) else: value = msg[header] - if value is None: # No such header + if value is None: # No such header return - if ';' not in value: # There are no parameters + if ';' not in value: # There are no parameters return - del msg[header] # Delete all such headers + del msg[header] # Delete all such headers # Get the value without parameters and set it back msg[header] = value.split(';')[0].strip() else: @@ -127,53 +144,53 @@ def decode_headers(msg): for header_list in g.remove_headers: header_list = header_list.split(',') - if header_list[0] == '*': # Remove all headers except listed + if header_list[0] == '*': # Remove all headers except listed header_list = _get_exceptions(header_list) for header in msg.keys(): if header.lower() not in header_list: del msg[header] - else: # Remove listed headers + else: # Remove listed headers for header in header_list: del msg[header] for header_list, param_list in g.remove_headers_params: header_list = header_list.split(',') param_list = param_list.split(',') - remove_all_params = param_list[0] == '*' # Remove all params except listed + remove_all_params = param_list[0] == '*' # Remove all params except listed if remove_all_params: param_list = _get_exceptions(param_list) - if header_list[0] == '*': # Remove for all headers except listed + if header_list[0] == '*': # Remove for all headers except listed header_list = _get_exceptions(header_list) for header in msg.keys(): if header.lower() not in header_list: _remove_headers_params(msg, header, remove_all_params, param_list) - else: # Decode for listed headers + else: # Decode for listed headers for header in header_list: _remove_headers_params(msg, header, remove_all_params, param_list) for header_list in g.decode_headers: header_list = header_list.split(',') - if header_list[0] == '*': # Decode all headers except listed + if header_list[0] == '*': # Decode all headers except listed header_list = _get_exceptions(header_list) for header in msg.keys(): if header.lower() not in header_list: decode_header(msg, header) - else: # Decode listed headers + else: # Decode listed headers for header in header_list: decode_header(msg, header) for header_list, param_list in g.decode_header_params: header_list = header_list.split(',') param_list = param_list.split(',') - decode_all_params = param_list[0] == '*' # Decode all params except listed + decode_all_params = param_list[0] == '*' # Decode all params except listed if decode_all_params: param_list = _get_exceptions(param_list) - if header_list[0] == '*': # Decode for all headers except listed + if header_list[0] == '*': # Decode for all headers except listed header_list = _get_exceptions(header_list) for header in msg.keys(): if header.lower() not in header_list: _decode_headers_params(msg, header, decode_all_params, param_list) - else: # Decode for listed headers + else: # Decode for listed headers for header in header_list: _decode_headers_params(msg, header, decode_all_params, param_list) @@ -181,7 +198,7 @@ def decode_headers(msg): def set_header(msg, header, value): "Replace header" - if msg.has_key(header): + if header in msg: msg.replace_header(header, value) else: msg[header] = value @@ -194,7 +211,7 @@ def set_content_type(msg, newtype, charset=None): msg.set_param("charset", charset, "Content-Type") -caps = None # Globally stored mailcap database; initialized only if needed +caps = None # Globally stored mailcap database; initialized only if needed def decode_body(msg, s): "Decode body to plain text using first copiousoutput filter from mailcap" @@ -206,13 +223,17 @@ def decode_body(msg, s): caps = mailcap.getcaps() content_type = msg.get_content_type() + if content_type.startswith('text/'): + charset = msg.get_content_charset() + else: + charset = None filename = tempfile.mktemp() command = None entries = mailcap.lookup(caps, content_type, "view") for entry in entries: - if entry.has_key('copiousoutput'): - if entry.has_key('test'): + if 'copiousoutput' in entry: + if 'test' in entry: test = mailcap.subst(entry['test'], content_type, filename) if test and os.system(test) != 0: continue @@ -222,18 +243,29 @@ def decode_body(msg, s): if not command: return s - file = open(filename, 'w') - file.write(s) - file.close() + outfile = open(filename, 'wb') + if charset and bytes is not str and isinstance(s, bytes): # Python3 + s = s.decode(charset, "replace") + if not isinstance(s, bytes): + s = s.encode(g.default_encoding, "replace") + outfile.write(s) + outfile.close() - pipe = os.popen(command, 'r') - s = pipe.read() - pipe.close() + pipe = subprocess.Popen(command, shell=True, stdout=subprocess.PIPE) + new_s = pipe.stdout.read() + pipe.stdout.close() + if pipe.wait() == 0: # result=0, Ok + s = new_s + if bytes is not str and isinstance(s, bytes): # Python3 + s = s.decode(g.default_encoding, "replace") + if charset and not isinstance(s, bytes): + s = s.encode(charset, "replace") + set_content_type(msg, "text/plain") + msg["X-MIME-Autoconverted"] = "from %s to text/plain by %s id %s" % (content_type, g.host_name, command.split()[0]) + else: + msg["X-MIME-Autoconverted"] = "failed conversion from %s to text/plain by %s id %s" % (content_type, g.host_name, command.split()[0]) os.remove(filename) - set_content_type(msg, "text/plain") - msg["X-MIME-Autoconverted"] = "from %s to text/plain by %s id %s" % (content_type, g.host_name, command.split()[0]) - return s @@ -241,7 +273,7 @@ def recode_charset(msg, s): "Recode charset of the message to the default charset" save_charset = charset = msg.get_content_charset() - if charset and charset.lower() <> g.default_encoding: + if charset and charset.lower() != g.default_encoding: s = recode_if_needed(s, charset) content_type = msg.get_content_type() set_content_type(msg, content_type, g.default_encoding) @@ -282,7 +314,7 @@ def _save_message(msg, outstring, save_headers=False, save_body=False): fname = msg.get_param(param, header=header) if fname: if isinstance(fname, tuple): - fname = fname[2] # Do not recode if it isn't recoded yet + fname = fname[2] # Do not recode if it isn't recoded yet try: for forbidden in chr(0), '/', '\\': if forbidden in fname: @@ -302,7 +334,11 @@ def _save_message(msg, outstring, save_headers=False, save_body=False): global output save_output = output outfile = open_output_file(fname) - output = outfile.write + def _output_bytes(s): + if not isinstance(s, bytes): + s = s.encode(g.default_encoding, "replace") + outfile.write(s) + output = _output_bytes if save_headers: output_headers(msg) if save_body: @@ -333,11 +369,13 @@ def decode_part(msg): elif content_type in g.binary_mask: left_binary = True break + elif content_type in g.fully_ignore_mask: + return encoding = msg["Content-Transfer-Encoding"] if left_binary or encoding in (None, '', '7bit', '8bit', 'binary'): outstring = msg.get_payload() - else: # Decode from transfer ecoding to text or binary form + else: # Decode from transfer ecoding to text or binary form outstring = msg.get_payload(decode=1) set_header(msg, "Content-Transfer-Encoding", "8bit") msg["X-MIME-Autoconverted"] = "from %s to 8bit by %s id %s" % (encoding, g.host_name, me) @@ -347,13 +385,13 @@ def decode_part(msg): outstring = totext(msg, outstring) break elif content_type in g.binary_mask or \ - content_type in g.decoded_binary_mask: + content_type in g.decoded_binary_mask: output_headers(msg) output(outstring) break elif content_type in g.ignore_mask: output_headers(msg) - output("\nMessage body of type %s skipped.\n" % ctype) + output("%sMessage body of type %s skipped.%s" % (os.linesep, ctype, os.linesep)) break elif content_type in g.error_mask: break @@ -364,42 +402,90 @@ def decode_part(msg): for content_type in masks: if content_type in g.save_headers_mask: _save_message(msg, outstring, save_headers=True, save_body=False) - elif content_type in g.save_body_mask: + if content_type in g.save_body_mask: _save_message(msg, outstring, save_headers=False, save_body=True) - elif content_type in g.save_message_mask: + if content_type in g.save_message_mask: _save_message(msg, outstring, save_headers=True, save_body=True) for content_type in masks: if content_type in g.error_mask: - raise ValueError, "content type %s prohibited" % ctype + raise ValueError("content type %s prohibited" % ctype) def decode_multipart(msg): "Decode multipart" decode_headers(msg) + boundary = msg.get_boundary() + + masks = [] + ctype = msg.get_content_type() + if ctype: + masks.append(ctype) + mtype = ctype.split('/')[0] + masks.append(mtype + '/*') + masks.append('*/*') + + for content_type in masks: + if content_type in g.fully_ignore_mask: + return + elif content_type in g.ignore_mask: + output_headers(msg) + output("%sMessage body of type %s skipped.%s" % (os.linesep, ctype, os.linesep)) + if boundary: + output("%s--%s--%s" % (os.linesep, boundary, os.linesep)) + return + + for content_type in masks: + if content_type in g.save_body_mask or \ + content_type in g.save_message_mask: + _out_l = [] + first_subpart = True + for subpart in msg.get_payload(): + if first_subpart: + first_subpart = False + else: + _out_l.append(os.linesep) + _out_l.append("--%s%s" % (boundary, os.linesep)) + _out_l.append(subpart.as_string()) + _out_l.append("%s--%s--%s" % (os.linesep, boundary, os.linesep)) + outstring = ''.join(_out_l) + break + else: + outstring = None + + for content_type in masks: + if content_type in g.save_headers_mask: + _save_message(msg, outstring, save_headers=True, save_body=False) + if content_type in g.save_body_mask: + _save_message(msg, outstring, save_headers=False, save_body=True) + if content_type in g.save_message_mask: + _save_message(msg, outstring, save_headers=True, save_body=True) + + for content_type in masks: + if content_type in g.error_mask: + raise ValueError("content type %s prohibited" % ctype) + output_headers(msg) - if msg.preamble: # Preserve the first part, it is probably not a RFC822-message - output(msg.preamble) # Usually it is just a few lines of text (MIME warning) + if msg.preamble: # Preserve the first part, it is probably not a RFC822-message + output(msg.preamble) # Usually it is just a few lines of text (MIME warning) if msg.preamble is not None: - output("\n") + output(os.linesep) first_subpart = True - boundary = msg.get_boundary() - for subpart in msg.get_payload(): if boundary: if first_subpart: first_subpart = False else: - output("\n") - output("--%s\n" % boundary) + output(os.linesep) + output("--%s%s" % (boundary, os.linesep)) # Recursively decode all parts of the subpart decode_message(subpart) if boundary: - output("\n--%s--\n" % boundary) + output("%s--%s--%s" % (os.linesep, boundary, os.linesep)) if msg.epilogue: output(msg.epilogue) @@ -410,9 +496,9 @@ def decode_message(msg): if msg.is_multipart(): decode_multipart(msg) - elif len(msg): # Simple one-part message (there are headers) - decode it + elif len(msg): # Simple one-part message (there are headers) - decode it decode_part(msg) - else: # Not a message, just text - copy it literally + else: # Not a message, just text - copy it literally output(msg.as_string()) @@ -423,7 +509,7 @@ def open_output_file(filename): if create: os.makedirs(full_dir) try: - return open(fullpath, 'w') + return open(fullpath, 'wb') except: if create: os.removedirs(full_dir) @@ -431,7 +517,7 @@ def open_output_file(filename): class GlobalOptions: from m_lib.defenc import default_encoding - recode_charset = 1 # recode charset of message body + recode_charset = 1 # recode charset of message body host_name = None @@ -455,10 +541,11 @@ class GlobalOptions: # A list of header/parameter/value triples to set set_header_param = [] - totext_mask = [] # A list of content-types to decode - binary_mask = [] # A list of content-types to pass through - decoded_binary_mask = [] # A list of content-types to pass through (content-transfer-decoded) - ignore_mask = [] # Ignore (skip, do not decode and do not include into output) + totext_mask = [] # A list of content-types to decode + binary_mask = [] # A list of content-types to pass through + decoded_binary_mask = [] # A list of content-types to pass through (content-transfer-decoded) + ignore_mask = [] # Ignore (do not decode and do not include into output) but output a warning instead of the body + fully_ignore_mask = [] # Completely ignore - no headers, no body, no warning error_mask = [] # Raise error if encounter one of these save_counter = 0 @@ -477,8 +564,9 @@ def get_opts(): from getopt import getopt, GetoptError try: - options, arguments = getopt(sys.argv[1:], - 'hVcCDPH:f:d:p:r:R:b:B:e:i:t:O:o:', + options, arguments = getopt( + sys.argv[1:], + 'hVcCDPH:f:d:p:r:R:b:B:e:I:i:t:O:o:', ['help', 'version', 'host=', 'save-headers=', 'save-body=', 'save-message=', 'set-header=', 'set-param=']) @@ -527,6 +615,8 @@ def get_opts(): g.binary_mask.append(value) elif option == '-b': g.decoded_binary_mask.append(value) + elif option == '-I': + g.fully_ignore_mask.append(value) elif option == '-i': g.ignore_mask.append(value) elif option == '-e': @@ -594,7 +684,14 @@ if __name__ == "__main__": g.host_name = socket.gethostname() g.outfile = outfile - output = outfile.write + if hasattr(outfile, 'buffer'): + def output_bytes(s): + if not isinstance(s, bytes): + s = s.encode(g.default_encoding, "replace") + outfile.buffer.write(s) + output = output_bytes + else: + output = outfile.write import email msg = email.message_from_file(infile)