diff options
Diffstat (limited to 'app/models/incoming_message.rb')
| -rw-r--r-- | app/models/incoming_message.rb | 452 |
1 files changed, 91 insertions, 361 deletions
diff --git a/app/models/incoming_message.rb b/app/models/incoming_message.rb index a4519a17d..131970ba6 100644 --- a/app/models/incoming_message.rb +++ b/app/models/incoming_message.rb @@ -1,7 +1,5 @@ -# encoding: UTF-8 - # == Schema Information -# Schema version: 95 +# Schema version: 108 # # Table name: incoming_messages # @@ -10,11 +8,19 @@ # created_at :datetime not null # updated_at :datetime not null # raw_email_id :integer not null -# cached_attachment_text_clipped :text -# cached_main_body_text_folded :text -# cached_main_body_text_unfolded :text +# cached_attachment_text_clipped :text +# cached_main_body_text_folded :text +# cached_main_body_text_unfolded :text +# sent_at :time +# subject :text +# mail_from_domain :text +# valid_to_reply_to :boolean +# last_parsed :datetime +# mail_from :text # +# encoding: UTF-8 + # models/incoming_message.rb: # An (email) message from really anybody to be logged with a request. e.g. A # response from the public body. @@ -44,275 +50,6 @@ module TMail end end -# This is the type which is used to send data about attachments to the view -class FOIAttachment - attr_accessor :body - attr_accessor :content_type - attr_accessor :filename - attr_accessor :url_part_number - attr_accessor :within_rfc822_subject # we use the subject as the filename for email attachments - - # List of DSN codes taken from RFC 3463 - # http://tools.ietf.org/html/rfc3463 - DsnToMessage = { - 'X.1.0' => 'Other address status', - 'X.1.1' => 'Bad destination mailbox address', - 'X.1.2' => 'Bad destination system address', - 'X.1.3' => 'Bad destination mailbox address syntax', - 'X.1.4' => 'Destination mailbox address ambiguous', - 'X.1.5' => 'Destination mailbox address valid', - 'X.1.6' => 'Mailbox has moved', - 'X.1.7' => 'Bad sender\'s mailbox address syntax', - 'X.1.8' => 'Bad sender\'s system address', - 'X.2.0' => 'Other or undefined mailbox status', - 'X.2.1' => 'Mailbox disabled, not accepting messages', - 'X.2.2' => 'Mailbox full', - 'X.2.3' => 'Message length exceeds administrative limit.', - 'X.2.4' => 'Mailing list expansion problem', - 'X.3.0' => 'Other or undefined mail system status', - 'X.3.1' => 'Mail system full', - 'X.3.2' => 'System not accepting network messages', - 'X.3.3' => 'System not capable of selected features', - 'X.3.4' => 'Message too big for system', - 'X.4.0' => 'Other or undefined network or routing status', - 'X.4.1' => 'No answer from host', - 'X.4.2' => 'Bad connection', - 'X.4.3' => 'Routing server failure', - 'X.4.4' => 'Unable to route', - 'X.4.5' => 'Network congestion', - 'X.4.6' => 'Routing loop detected', - 'X.4.7' => 'Delivery time expired', - 'X.5.0' => 'Other or undefined protocol status', - 'X.5.1' => 'Invalid command', - 'X.5.2' => 'Syntax error', - 'X.5.3' => 'Too many recipients', - 'X.5.4' => 'Invalid command arguments', - 'X.5.5' => 'Wrong protocol version', - 'X.6.0' => 'Other or undefined media error', - 'X.6.1' => 'Media not supported', - 'X.6.2' => 'Conversion required and prohibited', - 'X.6.3' => 'Conversion required but not supported', - 'X.6.4' => 'Conversion with loss performed', - 'X.6.5' => 'Conversion failed', - 'X.7.0' => 'Other or undefined security status', - 'X.7.1' => 'Delivery not authorized, message refused', - 'X.7.2' => 'Mailing list expansion prohibited', - 'X.7.3' => 'Security conversion required but not possible', - 'X.7.4' => 'Security features not supported', - 'X.7.5' => 'Cryptographic failure', - 'X.7.6' => 'Cryptographic algorithm not supported', - 'X.7.7' => 'Message integrity failure' - } - - # Returns HTML, of extra comment to put by attachment - def extra_note - # For delivery status notification attachments, extract the status and - # look up what it means in the DSN table. - if @content_type == 'message/delivery-status' - if !@body.match(/Status:\s+([0-9]+\.([0-9]+\.[0-9]+))\s+/) - return "" - end - dsn = $1 - dsn_part = 'X.' + $2 - - dsn_message = "" - if DsnToMessage.include?(dsn_part) - dsn_message = " (" + DsnToMessage[dsn_part] + ")" - end - - return "<br><em>DSN: " + dsn + dsn_message + "</em>" - end - return "" - end - - # Called by controller so old filenames still work - def old_display_filename - filename = self._internal_display_filename - - # Convert weird spaces (e.g. \n) to normal ones - filename = filename.gsub(/\s/, " ") - # Remove slashes, they mess with URLs - filename = filename.gsub(/\//, "-") - - return filename - end - - # XXX changing this will break existing URLs, so have a care - maybe - # make another old_display_filename see above - def display_filename - filename = self._internal_display_filename - - # Sometimes filenames have e.g. %20 in - no point butchering that - # (without unescaping it, this would remove the % and leave 20s in there) - filename = CGI.unescape(filename) - - # Remove weird spaces - filename = filename.gsub(/\s+/, " ") - # Remove non-alphabetic characters - filename = filename.gsub(/[^A-Za-z0-9.]/, " ") - # Remove spaces near dots - filename = filename.gsub(/\s*\.\s*/, ".") - # Compress adjacent spaces down to a single one - filename = filename.gsub(/\s+/, " ") - filename = filename.strip - - return filename - end - - def _internal_display_filename - calc_ext = AlaveteliFileTypes.mimetype_to_extension(@content_type) - - if @filename - # Put right extension on if missing - if !filename.match(/\.#{calc_ext}$/) && calc_ext - filename + "." + calc_ext - else - filename - end - else - if !calc_ext - calc_ext = "bin" - end - if @within_rfc822_subject - @within_rfc822_subject + "." + calc_ext - else - "attachment." + calc_ext - end - end - end - - # Size to show next to the download link for the attachment - def display_size - s = self.body.size - - if s > 1024 * 1024 - return sprintf("%.1f", s.to_f / 1024 / 1024) + 'M' - else - return (s / 1024).to_s + 'K' - end - end - - # Whether this type can be shown in the Google Docs Viewer. - # The full list of supported types can be found at - # https://docs.google.com/support/bin/answer.py?hl=en&answer=1189935 - def has_google_docs_viewer? - return !! { - "application/pdf" => true, # .pdf - "image/tiff" => true, # .tiff - - "application/vnd.ms-word" => true, # .doc - "application/vnd.openxmlformats-officedocument.wordprocessingml.document" => true, # .docx - - "application/vnd.ms-powerpoint" => true, # .ppt - "application/vnd.openxmlformats-officedocument.presentationml.presentation" => true, # .pptx - - "application/vnd.ms-excel" => true, # .xls - "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet" => true, # .xlsx - - } [self.content_type] - end - - # Whether this type has a "View as HTML" - def has_body_as_html? - return ( - !!{ - "text/plain" => true, - "application/rtf" => true, - }[self.content_type] or - self.has_google_docs_viewer? - ) - end - - # Name of type of attachment type - only valid for things that has_body_as_html? - def name_of_content_type - return { - "text/plain" => "Text file", - 'application/rtf' => "RTF file", - - 'application/pdf' => "PDF file", - 'image/tiff' => "TIFF image", - - 'application/vnd.ms-word' => "Word document", - 'application/vnd.openxmlformats-officedocument.wordprocessingml.document' => "Word document", - - 'application/vnd.ms-powerpoint' => "PowerPoint presentation", - 'application/vnd.openxmlformats-officedocument.presentationml.presentation' => "PowerPoint presentation", - - 'application/vnd.ms-excel' => "Excel spreadsheet", - 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet' => "Excel spreadsheet", - }[self.content_type] - end - - # For "View as HTML" of attachment - def body_as_html(dir) - html = nil - wrapper_id = "wrapper" - - # simple cases, can never fail - if self.content_type == 'text/plain' - text = self.body.strip - text = CGI.escapeHTML(text) - text = MySociety::Format.make_clickable(text) - html = text.gsub(/\n/, '<br>') - return '<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" - "http://www.w3.org/TR/html4/loose.dtd"><html><head><title></title></head><body>' + html + "</body></html>", wrapper_id - end - - # the extractions will also produce image files, which go in the - # current directory, so change to the directory the function caller - # wants everything in - Dir.chdir(dir) do - tempfile = Tempfile.new('foiextract', '.') - tempfile.print self.body - tempfile.flush - - if self.content_type == 'application/pdf' - IO.popen("#{`which pdftohtml`.chomp} -nodrm -zoom 1.0 -stdout -enc UTF-8 -noframes " + tempfile.path + "", "r") do |child| - html = child.read() - end - elsif self.content_type == 'application/rtf' - IO.popen("/usr/bin/unrtf --html " + tempfile.path + "", "r") do |child| - html = child.read() - end - elsif self.has_google_docs_viewer? - html = '' # force error and using Google docs viewer - else - raise "No HTML conversion available for type " + self.content_type - end - - tempfile.close - tempfile.delete - end - - # We need to look at: - # a) Any error code - # b) The output size, as pdftohtml does not return an error code upon error. - # c) For cases when there is no text in the body of the HTML, or - # images, so nothing will be rendered. This is to detect some bug in - # pdftohtml, which sometimes makes it return just <hr>s and no other - # content. - html.match(/(\<body[^>]*\>.*)/mi) - body = $1.to_s - body_without_tags = body.gsub(/\s+/,"").gsub(/\<[^\>]*\>/, "") - contains_images = html.match(/<img/mi) ? true : false - if !$?.success? || html.size == 0 || (body_without_tags.size == 0 && !contains_images) - ret = "<html><head></head><body>"; - if self.has_google_docs_viewer? - wrapper_id = "wrapper_google_embed" - ret = ret + "<iframe src='http://docs.google.com/viewer?url=<attachment-url-here>&embedded=true' width='100%' height='100%' style='border: none;'></iframe>"; - else - ret = ret + "<p>Sorry, we were unable to convert this file to HTML. Please use the download link at the top right.</p>" - end - ret = ret + "</body></html>" - return ret, wrapper_id - end - - return html, wrapper_id - end - -end - - class IncomingMessage < ActiveRecord::Base belongs_to :info_request validates_presence_of :info_request @@ -380,7 +117,7 @@ class IncomingMessage < ActiveRecord::Base if !self.mail['return-path'].nil? && self.mail['return-path'].addr == "<>" return false end - if !self.mail['auto-submitted'].nil? && !self.mail['auto-submitted'].keys.empty? + if !self.mail['auto-submitted'].nil? return false end return true @@ -390,22 +127,27 @@ class IncomingMessage < ActiveRecord::Base # The following fields may be absent; we treat them as cached # values in case we want to regenerate them (due to mail # parsing bugs, etc). + if self.raw_email.nil? + raise "Incoming message id=#{id} has no raw_email" + end if (!force.nil? || self.last_parsed.nil?) - self.extract_attachments! - self.sent_at = self.mail.date || self.created_at - self.subject = self.mail.subject - # XXX can probably remove from_name_if_present (which is a - # monkey patch) by just calling .from_addrs[0].name here - # instead? - self.mail_from = self.mail.from_name_if_present - begin - self.mail_from_domain = PublicBody.extract_domain_from_email(self.mail.from_addrs[0].spec) - rescue NoMethodError - self.mail_from_domain = "" + ActiveRecord::Base.transaction do + self.extract_attachments! + self.sent_at = self.mail.date || self.created_at + self.subject = self.mail.subject + # XXX can probably remove from_name_if_present (which is a + # monkey patch) by just calling .from_addrs[0].name here + # instead? + self.mail_from = self.mail.from_name_if_present + begin + self.mail_from_domain = PublicBody.extract_domain_from_email(self.mail.from_addrs[0].spec) + rescue NoMethodError + self.mail_from_domain = "" + end + self.valid_to_reply_to = self._calculate_valid_to_reply_to + self.last_parsed = Time.now + self.save! end - self.valid_to_reply_to = self._calculate_valid_to_reply_to - self.last_parsed = Time.now - self.save! end end @@ -527,11 +269,7 @@ class IncomingMessage < ActiveRecord::Base # Special cases for some content types if content_type == 'application/pdf' uncompressed_text = nil - IO.popen("#{`which pdftk`.chomp} - output - uncompress", "r+") do |child| - child.write(text) - child.close_write() - uncompressed_text = child.read() - end + uncompressed_text = AlaveteliExternalCommand.run("pdftk", "-", "output", "-", "uncompress", :stdin_string => text) # if we managed to uncompress the PDF... if !uncompressed_text.nil? && !uncompressed_text.empty? # then censor stuff (making a copy so can compare again in a bit) @@ -542,15 +280,11 @@ class IncomingMessage < ActiveRecord::Base # then use the altered file (recompressed) recompressed_text = nil if MySociety::Config.get('USE_GHOSTSCRIPT_COMPRESSION') == true - command = "gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen -dNOPAUSE -dQUIET -dBATCH -sOutputFile=- -" + command = ["gs", "-sDEVICE=pdfwrite", "-dCompatibilityLevel=1.4", "-dPDFSETTINGS=/screen", "-dNOPAUSE", "-dQUIET", "-dBATCH", "-sOutputFile=-", "-"] else - command = "#{`which pdftk`.chomp} - output - compress" - end - IO.popen(command, "r+") do |child| - child.write(censored_uncompressed_text) - child.close_write() - recompressed_text = child.read() + command = ["pdftk", "-", "output", "-", "compress"] end + recompressed_text = AlaveteliExternalCommand.run(*(command + [{:stdin_string=>censored_uncompressed_text}])) if recompressed_text.nil? || recompressed_text.empty? # buggy versions of pdftk sometimes fail on # compression, I don't see it's a disaster in @@ -586,8 +320,8 @@ class IncomingMessage < ActiveRecord::Base emails = ascii_chars.scan(MySociety::Validate.email_find_regexp) # Convert back to UCS-2, making a mask at the same time emails.map! {|email| [ - Iconv.conv('ucs-2', 'ascii', email[0]), - Iconv.conv('ucs-2', 'ascii', email[0].gsub(/[^@.]/, 'x')) + Iconv.conv('ucs-2le', 'ascii', email[0]), + Iconv.conv('ucs-2le', 'ascii', email[0].gsub(/[^@.]/, 'x')) ] } # Now search and replace the UCS-2 email with the UCS-2 mask for email, mask in emails @@ -792,7 +526,7 @@ class IncomingMessage < ActiveRecord::Base # it into conflict with ensure_parts_counted which it has to be # called both before and after. It will fail with cases of # attachments of attachments etc. - + charset = curr_mail.charset # save this, because overwriting content_type also resets charset # Don't allow nil content_types if curr_mail.content_type.nil? curr_mail.content_type = 'application/octet-stream' @@ -822,7 +556,6 @@ class IncomingMessage < ActiveRecord::Base curr_mail.content_type = 'application/octet-stream' end end - # If the part is an attachment of email if curr_mail.content_type == 'message/rfc822' || curr_mail.content_type == 'application/vnd.ms-outlook' || curr_mail.content_type == 'application/ms-tnef' ensure_parts_counted # fills in rfc822_attachment variable @@ -832,6 +565,8 @@ class IncomingMessage < ActiveRecord::Base curr_mail.within_rfc822_attachment = within_rfc822_attachment leaves_found += [curr_mail] end + # restore original charset + curr_mail.charset = charset end return leaves_found end @@ -887,64 +622,58 @@ class IncomingMessage < ActiveRecord::Base end # Returns body text from main text part of email, converted to UTF-8 def get_main_body_text_internal + parse_raw_email! main_part = get_main_body_text_part return _convert_part_body_to_text(main_part) end + # Given a main text part, converts it to text def _convert_part_body_to_text(part) if part.nil? text = "[ Email has no body, please see attachments ]" - text_charset = "utf-8" + source_charset = "utf-8" else - text = part.body - text_charset = part.charset + text = part.body # by default, TMail converts to UTF8 in this call + source_charset = part.charset if part.content_type == 'text/html' # e.g. http://www.whatdotheyknow.com/request/35/response/177 - # XXX This is a bit of a hack as it is calling a convert to text routine. - # Could instead call a sanitize HTML one. - text = self.class._get_attachment_text_internal_one_file(part.content_type, text) - end - end - - # Charset conversion, turn everything into UTF-8 - if not text_charset.nil? - begin - # XXX specially convert unicode pound signs, was needed here - # http://www.whatdotheyknow.com/request/88/response/352 - text = text.gsub("£", Iconv.conv(text_charset, 'utf-8', '£')) - # Try proper conversion - text = Iconv.conv('utf-8', text_charset, text) - rescue Iconv::IllegalSequence, Iconv::InvalidEncoding - # Clearly specified charset was nonsense - text_charset = nil + # XXX This is a bit of a hack as it is calling a + # convert to text routine. Could instead call a + # sanitize HTML one. + + # If the text isn't UTF8, it means TMail had a problem + # converting it (invalid characters, etc), and we + # should instead tell elinks to respect the source + # charset + use_charset = "utf-8" + begin + text = Iconv.conv('utf-8', 'utf-8', text) + rescue Iconv::IllegalSequence + use_charset = source_charset + end + text = self.class._get_attachment_text_internal_one_file(part.content_type, text, use_charset) end end - if text_charset.nil? - # No specified charset, so guess - - # Could use rchardet here, but it had trouble with - # http://www.whatdotheyknow.com/request/107/response/144 - # So I gave up - most likely in UK we'll only get windows-1252 anyway. + # If TMail can't convert text, it just returns it, so we sanitise it. + begin + # Test if it's good UTF-8 + text = Iconv.conv('utf-8', 'utf-8', text) + rescue Iconv::IllegalSequence + # Text looks like unlabelled nonsense, + # strip out anything that isn't UTF-8 begin - # See if it is good UTF-8 anyway - text = Iconv.conv('utf-8', 'utf-8', text) - rescue Iconv::IllegalSequence - begin - # Or is it good windows-1252, most likely - text = Iconv.conv('utf-8', 'windows-1252', text) - rescue Iconv::IllegalSequence - # Text looks like unlabelled nonsense, strip out anything that isn't UTF-8 - text = Iconv.conv('utf-8//IGNORE', 'utf-8', text) + - _("\n\n[ {{site_name}} note: The above text was badly encoded, and has had strange characters removed. ]", - :site_name => MySociety::Config.get('SITE_NAME', 'Alaveteli')) + text = Iconv.conv('utf-8//IGNORE', source_charset, text) + + _("\n\n[ {{site_name}} note: The above text was badly encoded, and has had strange characters removed. ]", + :site_name => MySociety::Config.get('SITE_NAME', 'Alaveteli')) + rescue Iconv::InvalidEncoding, Iconv::IllegalSequence + if source_charset != "utf-8" + source_charset = "utf-8" + retry end end end - # An assertion that we have ended up with UTF-8 XXX can remove as this should - # always be fine if code above is - Iconv.conv('utf-8', 'utf-8', text) # Fix DOS style linefeeds to Unix style ones (or other later regexps won't work) # Needed for e.g. http://www.whatdotheyknow.com/request/60/response/98 @@ -1004,9 +733,7 @@ class IncomingMessage < ActiveRecord::Base tempfile = Tempfile.new('foiuu') tempfile.print uu tempfile.flush - IO.popen("/usr/bin/uudecode " + tempfile.path + " -o -", "r") do |child| - content = child.read() - end + content = AlaveteliExternalCommand.run("uudecode", "-o", "/dev/stdout", tempfile.path) tempfile.close # Make attachment type from it, working out filename and mime type filename = uu.match(/^begin\s+[0-9]+\s+(.*)$/)[1] @@ -1192,7 +919,9 @@ class IncomingMessage < ActiveRecord::Base return self.cached_attachment_text_clipped end - def IncomingMessage._get_attachment_text_internal_one_file(content_type, body) + def IncomingMessage._get_attachment_text_internal_one_file(content_type, body, charset = 'utf-8') + # note re. charset: TMail always tries to convert email bodies + # to UTF8 by default, so normally it should already be that. text = '' # XXX - tell all these command line tools to return utf-8 if content_type == 'text/plain' @@ -1202,22 +931,23 @@ class IncomingMessage < ActiveRecord::Base tempfile.print body tempfile.flush if content_type == 'application/vnd.ms-word' - AlaveteliExternalCommand.run(`which wvText`.chomp, tempfile.path, tempfile.path + ".txt") + AlaveteliExternalCommand.run("wvText", tempfile.path, tempfile.path + ".txt") # Try catdoc if we get into trouble (e.g. for InfoRequestEvent 2701) if not File.exists?(tempfile.path + ".txt") - AlaveteliExternalCommand.run(`which catdoc`.chomp, tempfile.path, :append_to => text) + AlaveteliExternalCommand.run("catdoc", tempfile.path, :append_to => text) else text += File.read(tempfile.path + ".txt") + "\n\n" File.unlink(tempfile.path + ".txt") end elsif content_type == 'application/rtf' # catdoc on RTF prodcues less comments and extra bumf than --text option to unrtf - AlaveteliExternalCommand.run(`which catdoc`.chomp, tempfile.path, :append_to => text) + AlaveteliExternalCommand.run("catdoc", tempfile.path, :append_to => text) elsif content_type == 'text/html' - # lynx wordwraps links in its output, which then don't get formatted properly - # by Alaveteli. We use elinks instead, which doesn't do that. - AlaveteliExternalCommand.run(`which elinks`.chomp, "-eval", "'set document.codepage.assume = \"utf-8\"'", "-dump-charset", "utf-8", "-force-html", "-dump", - tempfile.path, :append_to => text) + # lynx wordwraps links in its output, which then don't + # get formatted properly by Alaveteli. We use elinks + # instead, which doesn't do that. + AlaveteliExternalCommand.run("elinks", "-eval", "set document.codepage.assume = \"#{charset}\"", "-eval", "set document.codepage.force_assumed = 1", "-dump-charset", "utf-8", "-force-html", "-dump", + tempfile.path, :append_to => text, :env => {"LANG" => "C"}) elsif content_type == 'application/vnd.ms-excel' # Bit crazy using /usr/bin/strings - but xls2csv, xlhtml and # py_xls2txt only extract text from cells, not from floating @@ -1227,9 +957,9 @@ class IncomingMessage < ActiveRecord::Base elsif content_type == 'application/vnd.ms-powerpoint' # ppthtml seems to catch more text, but only outputs HTML when # we want text, so just use catppt for now - AlaveteliExternalCommand.run(`which catppt`.chomp, tempfile.path, :append_to => text) + AlaveteliExternalCommand.run("catppt", tempfile.path, :append_to => text) elsif content_type == 'application/pdf' - AlaveteliExternalCommand.run(`which pdftotext`.chomp, tempfile.path, "-", :append_to => text) + AlaveteliExternalCommand.run("pdftotext", tempfile.path, "-", :append_to => text) elsif content_type == 'application/vnd.openxmlformats-officedocument.wordprocessingml.document' # This is Microsoft's XML office document format. # Just pull out the main XML file, and strip it of text. @@ -1283,7 +1013,7 @@ class IncomingMessage < ActiveRecord::Base text = '' attachments = self.get_attachments_for_display for attachment in attachments - text += IncomingMessage._get_attachment_text_internal_one_file(attachment.content_type, attachment.body) + text += IncomingMessage._get_attachment_text_internal_one_file(attachment.content_type, attachment.body, attachment.charset) end # Remove any bad characters text = Iconv.conv('utf-8//IGNORE', 'utf-8', text) @@ -1376,7 +1106,7 @@ class IncomingMessage < ActiveRecord::Base if !self.mail['return-path'].nil? && self.mail['return-path'].addr == "<>" return false end - if !self.mail['auto-submitted'].nil? && !self.mail['auto-submitted'].keys.empty? + if !self.mail['auto-submitted'].nil? return false end return true |