Changeset 236 for trunk/email2trac.py.in

Timestamp:

12/05/08 06:03:24 (14 years ago)

Author:

bromine

Message:

email2trac.py.in:

• better handling of inline attachments and multiple body parts, closes #105

File:

• trunk/email2trac.py.in (modified) (14 diffs)

trunk/email2trac.py.in

@@ -155 +155 @@
                         os.umask(int(parameters['umask'], 8))
 
+                if parameters.has_key('quote_attachment_filenames'):
+                        self.QUOTE_ATTACHMENT_FILENAMES = int(parameters['quote_attachment_filenames'])
+                else:
+                        self.QUOTE_ATTACHMENT_FILENAMES = 1
+
                 if parameters.has_key('debug'):
                         self.DEBUG = int(parameters['debug'])
@@ -355 +360 @@
                 return str
 
-        def debug_attachments(self, message):
+        def debug_body(self, message_body, tempfile=False):
+                if tempfile:
+                        import tempfile
+                        body_file = tempfile.mktemp('.email2trac')
+                else:
+                        body_file = os.path.join(self.TMPDIR, 'body.txt')
+
+                print 'TD: writing body (%s)' % body_file
+                fx = open(body_file, 'wb')
+                if not message_body:
+                        message_body = '(None)'
+                fx.write(message_body)
+                fx.close()
+                try:
+                        os.chmod(body_file,S_IRWXU|S_IRWXG|S_IRWXO)
+                except OSError:
+                        pass
+
+        def debug_attachments(self, message_parts):
                 n = 0
-                for part in message.walk():
-                        if part.get_content_maintype() == 'multipart':      # multipart/* is just a container
-                                print 'TD: multipart container'
+                for part in message_parts:
+                        # Skip inline text parts
+                        if not isinstance(part, tuple):
                                 continue
+                                
+                        (filename, part) = part
 
                         n = n + 1
@@ -366 +391 @@
                         print 'TD: part%d: filename: %s' % (n, part.get_filename())
 
-                        if part.is_multipart():
-                                print 'TD: this part is multipart'
-                                payload = part.get_payload(decode=1)
-                                print 'TD: payload:', payload
-                        else:
-                                print 'TD: this part is not multipart'
-
-                        file = 'part%d' %n
-                        part_file = os.path.join(self.TMPDIR, file)
+                        part_file = os.path.join(self.TMPDIR, filename)
                         #part_file = '/var/tmp/part%d' % n
                         print 'TD: writing part%d (%s)' % (n,part_file)
@@ -643 +660 @@
                         self.update_ticket_fields(tkt, update_fields)
 
-                body_text = self.get_body_text(m)
+                message_parts = self.get_message_parts(m)
+                message_parts = self.unique_attachment_names(message_parts, tkt)
 
                 if self.EMAIL_HEADER:
-                        head = self.email_header_txt(m)
-                        body_text = u"%s\r\n%s" %(head, body_text)
+                        message_parts.insert(0, self.email_header_txt(m))
+
+                body_text = self.body_text(message_parts)
 
                 if body_text.strip():
@@ -655 +674 @@
 
                 if self.VERSION  == 0.9:
-                        str = self.attachments(m, tkt, True)
-                else:
-                        str = self.attachments(m, tkt)
+                        str = self.attachments(message_parts, tkt, True)
+                else:
+                        str = self.attachments(message_parts, tkt)
 
                 if self.notification and not spam:
@@ -755 +774 @@
                         head = self.email_header_txt(msg)
                         
-                body_text = self.get_body_text(msg)
-
-                tkt['description'] = '%s\r\n%s' \
-                        %(head, body_text)
+                message_parts = self.get_message_parts(msg)
+                message_parts = self.unique_attachment_names(message_parts)
+                
+                if self.EMAIL_HEADER > 0:
+                        message_parts.insert(0, self.email_header_txt(msg))
+                        
+                body_text = self.body_text(message_parts)
+
+                tkt['description'] = body_text
 
                 #when = int(time.time())
@@ -785 +809 @@
                                 %(head, mailto, body_text)
 
-                str =  self.attachments(msg, tkt)
+                str =  self.attachments(message_parts, tkt)
                 if str:
                         changed = True
@@ -815 +839 @@
 
                 if self.DEBUG > 1:        # save the entire e-mail message text
+                        message_parts = self.get_message_parts(m)
+                        message_parts = self.unique_attachment_names(message_parts)
                         self.save_email_for_debug(m, True)
-                        self.debug_attachments(m)
+                        body_text = self.body_text(message_parts)
+                        self.debug_body(body_text, True)
+                        self.debug_attachments(message_parts)
 
                 self.db = self.env.get_db_cnx()
@@ -946 +974 @@
 
 
-        def get_body_text(self, msg):
-                """
-                put the message text in the ticket description or in the changes field.
-                message text can be plain text or html or something else
-                """
-                has_description = 0
-                encoding = True
-                ubody_text = u'No plain text message'
+        def get_message_parts(self, msg):
+                """
+                parses the email message and returns a list of body parts and attachments
+                body parts are returned as strings, attachments are returned as tuples of (filename, Message object)
+                """
+                message_parts = []
+
                 for part in msg.walk():
-
                         # 'multipart/*' is a container for multipart messages
                         #
+                        if self.DEBUG:
+                                print 'TD: Message part: Content-Type: %s' % part.get_content_type()
+
                         if part.get_content_maintype() == 'multipart':
                                 continue
 
-                        if part.get_content_type() == 'text/plain':
+                        # Check if this is an inline part. It's inline if there is co Cont-Disp header, or if there is one and it says "inline"
+                        inline = self.inline_part(part)
+
+                        # Inline text parts are where the body is
+                        if part.get_content_type() == 'text/plain' and inline:
+                                if self.DEBUG:
+                                        print 'TD:               Inline body part'
+
                                 # Try to decode, if fails then do not decode
                                 #
@@ -998 +1034 @@
                                         ubody_text = 'ERROR: Could not find charset: %s, please install' %(charset)
 
-                        elif part.get_content_type() == 'text/html':
-                                ubody_text = '(see attachment for HTML mail message)'
-
+                                if self.VERBATIM_FORMAT:
+                                        message_parts.append('{{{\r\n%s\r\n}}}' %ubody_text)
+                                else:
+                                        message_parts.append('%s' %ubody_text)
                         else:
-                                ubody_text = '(see attachment for message)'
-
-                        has_description = 1
-                        break           # we have the description, so break
-
-                if not has_description:
-                        ubody_text = '(see attachment for message)'
-
-                # A patch so that the web-interface will not update the description
-                # field of a ticket
-                #
-                ubody_text = ('\r\n'.join(ubody_text.splitlines()))
-
-                #  If we can unicode it try to encode it for trac
-                #  else we a lot of garbage
-                #
-                #if encoding:
-                #       ubody_text = ubody_text.encode('utf-8')
-
-                if self.VERBATIM_FORMAT:
-                        ubody_text = '{{{\r\n%s\r\n}}}' %ubody_text
-                else:
-                        ubody_text = '%s' %ubody_text
-
-                return ubody_text
+                                if self.DEBUG:
+                                        print 'TD:               Filename: %s' % part.get_filename()
+
+                                message_parts.append((part.get_filename(), part))
+
+                return message_parts
+                
+        def unique_attachment_names(self, message_parts, tkt = None):
+                renamed_parts = []
+                attachment_names = set()
+                for part in message_parts:
+                        
+                        # If not an attachment, leave it alone
+                        if not isinstance(part, tuple):
+                                renamed_parts.append(part)
+                                continue
+                                
+                        (filename, part) = part
+                        # Decode the filename
+                        if filename:
+                                filename = self.email_to_unicode(filename)                      
+                        # If no name, use a default one
+                        else:
+                                filename = 'untitled-part'
+
+                                # Guess the extension from the content type
+                                ext = mimetypes.guess_extension(part.get_content_type())
+                                if not ext:
+                                        ext = '.bin'
+
+                                filename = '%s%s' % (filename, ext)
+
+                        # Discard relative paths in attachment names
+                        filename = filename.replace('\\', '/').replace(':', '/') 
+                        filename = os.path.basename(filename)
+
+                        # We try to normalize the filename to utf-8 NFC if we can.
+                        # Files uploaded from OS X might be in NFD.
+                        # Check python version and then try it
+                        #
+                        if sys.version_info[0] > 2 or (sys.version_info[0] == 2 and sys.version_info[1] >= 3):
+                                try:
+                                        filename = unicodedata.normalize('NFC', unicode(filename, 'utf-8')).encode('utf-8')  
+                                except TypeError:
+                                        pass
+                                        
+                        if self.QUOTE_ATTACHMENT_FILENAMES:
+                                filename = urllib.quote(filename)
+
+                        # Make the filename unique for this ticket
+                        num = 0
+                        unique_filename = filename
+                        filename, ext = os.path.splitext(filename)
+
+                        while unique_filename in attachment_names or self.attachment_exists(tkt, unique_filename):
+                                num += 1
+                                unique_filename = "%s-%s%s" % (filename, num, ext)
+                                
+                        if self.DEBUG:
+                                print 'TD: Attachment with filename %s will be saved as %s' % (filename, unique_filename)
+
+                        attachment_names.add(unique_filename)
+
+                        renamed_parts.append((filename, unique_filename, part))
+                
+                return renamed_parts
+                        
+        def inline_part(self, part):
+                return part.get_param('inline', None, 'Content-Disposition') == '' or not part.has_key('Content-Disposition')
+                
+                        
+        def attachment_exists(self, tkt, filename):
+                if tkt is None:
+                        return False
+                        
+                try:
+                        Attachment(self.env, 'ticket', tkt['id'], filename)
+                        return True
+                except ResourceNotFound:
+                        return False
+                        
+        def body_text(self, message_parts):
+                body_text = []
+                
+                for part in message_parts:
+                        # Plain text part, append it
+                        if not isinstance(part, tuple):
+                                body_text.extend(part.strip().splitlines())
+                                body_text.append("")
+                                continue
+                                
+                        (original, filename, part) = part
+                        inline = self.inline_part(part)
+                        
+                        if part.get_content_maintype() == 'image' and inline:
+                                body_text.append('[[Image(%s)]]' % filename)
+                                body_text.append("")
+                        else:
+                                body_text.append('[attachment:"%s"]' % filename)
+                                body_text.append("")
+                                
+                body_text = '\r\n'.join(body_text)
+                return body_text
 
         def notify(self, tkt , new=True, modtime=0):
@@ -1087 +1202 @@
                 return str
 
-        def attachments(self, message, ticket, update=False):
+        def attachments(self, message_parts, ticket, update=False):
                 '''
                 save any attachments as files in the ticket's directory
@@ -1099 +1214 @@
                 max_size = int(self.get_config('attachment', 'max_size'))
                 status   = ''
-
-                for part in message.walk():
-                        if part.get_content_maintype() == 'multipart':          # multipart/* is just a container
+                
+                for part in message_parts:
+                        # Skip body parts
+                        if not isinstance(part, tuple):
                                 continue
-
-                        if not first:                                                                           # first content is the message
-                                first = 1
-                                if part.get_content_type() == 'text/plain':             # if first is text, is was already put in the description
-                                        continue
-
-                        filename = part.get_filename() 
-                        if not filename:
-                                number = number + 1
-                                filename = 'part%04d' % number
-
-                                ext = mimetypes.guess_extension(part.get_content_type())
-                                if not ext:
-                                        ext = '.bin'
-
-                                filename = '%s%s' % (filename, ext)
-                        else:
-                                filename = self.email_to_unicode(filename)
-
-                        # From the trac code
-                        #
-                        filename = filename.replace('\\', '/').replace(':', '/') 
-                        filename = os.path.basename(filename)
-
-                        # We try to normalize the filename to utf-8 NFC if we can.
-                        # Files uploaded from OS X might be in NFD.
-                        # Check python version and then try it
-                        #
-                        if sys.version_info[0] > 2 or (sys.version_info[0] == 2 and sys.version_info[1] >= 3):
-                                try:
-                                        filename = unicodedata.normalize('NFC', unicode(filename, 'utf-8')).encode('utf-8')  
-                                except TypeError:
-                                        pass
-
-                        url_filename = urllib.quote(filename)
+                                
+                        (original, filename, part) = part
                         #
                         # Must be tuneables HvB
                         #
-                        path, fd =  util.create_unique_file(os.path.join(self.TMPDIR, url_filename))
+                        path, fd =  util.create_unique_file(os.path.join(self.TMPDIR, filename))
                         text = part.get_payload(decode=1)
                         if not text:
@@ -1157 +1240 @@
                         if (max_size != -1) and (file_size > max_size):
                                 status = '%s\nFile %s is larger then allowed attachment size (%d > %d)\n\n' \
-                                        %(status, filename, file_size, max_size)
+                                        %(status, original, file_size, max_size)
 
                                 os.unlink(path)
@@ -1176 +1259 @@
                                 att.description = self.email_to_unicode('Added by email2trac')
 
-                        att.insert(url_filename, fd, file_size)
+                        att.insert(filename, fd, file_size)
                         #except  util.TracError, detail:
                         #       print detail