diff options
Diffstat (limited to 'service/pixelated/adapter/search/__init__.py')
| -rw-r--r-- | service/pixelated/adapter/search/__init__.py | 40 |
1 files changed, 23 insertions, 17 deletions
diff --git a/service/pixelated/adapter/search/__init__.py b/service/pixelated/adapter/search/__init__.py index b8d3e7ca..56ab2255 100644 --- a/service/pixelated/adapter/search/__init__.py +++ b/service/pixelated/adapter/search/__init__.py @@ -18,6 +18,8 @@ from pixelated.support.encrypted_file_storage import EncryptedFileStorage import os import re +import dateutil.parser +import time from pixelated.adapter.model.status import Status from pixelated.adapter.search.contacts import contacts_suggestions from whoosh.index import FileIndex @@ -27,7 +29,6 @@ from whoosh.qparser import MultifieldParser from whoosh.writing import AsyncWriter from whoosh import sorting from pixelated.support.functional import unique -from pixelated.support.date import milliseconds import traceback @@ -102,7 +103,6 @@ class SearchEngine(object): to=KEYWORD(stored=False, commas=True), cc=KEYWORD(stored=False, commas=True), bcc=KEYWORD(stored=False, commas=True), - bounced=KEYWORD(stored=False, commas=True), subject=TEXT(stored=False), date=NUMERIC(stored=False, sortable=True, bits=64, signed=False), body=TEXT(stored=False), @@ -121,32 +121,38 @@ class SearchEngine(object): def _index_mail(self, writer, mail): mdict = mail.as_dict() header = mdict['header'] - tags = mdict.get('tags', []) - tags.append(mail.mailbox_name.lower()) - bounced = mail.bounced if mail.bounced else [''] + tags = set(mdict.get('tags', {})) + tags.add(mail.mailbox_name.lower()) index_data = { - 'sender': self._unicode_header_field(header.get('from', '')), - 'subject': self._unicode_header_field(header.get('subject', '')), - 'date': milliseconds(header.get('date', '')), - 'to': u','.join([h.decode('utf-8') for h in header.get('to', [''])]), - 'cc': u','.join([h.decode('utf-8') for h in header.get('cc', [''])]), - 'bcc': u','.join([h.decode('utf-8') for h in header.get('bcc', [''])]), + 'sender': self._empty_string_to_none(header.get('from', '')), + 'subject': self._empty_string_to_none(header.get('subject', '')), + 'date': self._format_utc_integer(header.get('date', '')), + 'to': self._format_recipient(header, 'to'), + 'cc': self._format_recipient(header, 'cc'), + 'bcc': self._format_recipient(header, 'bcc'), 'tag': u','.join(unique(tags)), - 'bounced': u','.join(bounced), - 'body': unicode(mdict['textPlainBody']), + 'body': unicode(mdict['textPlainBody'] if 'textPlainBody' in mdict else mdict['body']), 'ident': unicode(mdict['ident']), 'flags': unicode(','.join(unique(mail.flags))), - 'raw': unicode(mail.raw.decode('utf-8')) + 'raw': unicode(mail.raw) } writer.update_document(**index_data) - def _unicode_header_field(self, field_value): + def _format_utc_integer(self, date): + timetuple = dateutil.parser.parse(date).utctimetuple() + return time.strftime('%s', timetuple) + + def _format_recipient(self, headers, name): + list = headers.get(name, ['']) + return u','.join(list) if list else u'' + + def _empty_string_to_none(self, field_value): if not field_value: return None - - return unicode(field_value.decode('utf-8')) + else: + return field_value def index_mails(self, mails, callback=None): try: |