# may require a full re-index.

                        start_id = results[0]['raw_id']

                        last_rev = repo.get_changeset(revision=start_id).revision

                    # there are new changesets to index or a new repo to index

                    if last_rev == 0 or num_of_revs > last_rev + 1:

                        # delete the docs in the index for the previous

                        # last changeset(s)

                        for hit in results:

                            q = qp.parse(u"last:t AND %s AND raw_id:%s" %

                                            (repo_name, hit['raw_id']))

                            writer.delete_by_query(q)

                        # index from the previous last changeset + all new ones

                        indexed_total += self.index_changesets(writer,

                                                repo_name, repo, start_id)

                        writer_is_dirty = True

                log.debug('indexed %s changesets for repo %s' % (

                             indexed_total, repo_name)

                )

            finally:

                if writer_is_dirty:

                    log.debug('>> COMMITING CHANGES TO CHANGESET INDEX<<')

                    writer.commit(merge=True)

                else:

                    writer.cancel

    def update_file_index(self):

        log.debug((u'STARTING INCREMENTAL INDEXING UPDATE FOR EXTENSIONS %s '

                   'AND REPOS %s') % (INDEX_EXTENSIONS, self.repo_paths.keys()))

        idx = open_dir(self.index_location, indexname=self.indexname)

        # The set of all paths in the index

        indexed_paths = set()

        # The set of all paths we need to re-index

        to_index = set()

        writer = idx.writer()

        writer_is_dirty = False

        try:

            with idx.reader() as reader:

                # Loop over the stored fields in the index

                for fields in reader.all_stored_fields():

                    indexed_path = fields['path']

                    indexed_repo_path = fields['repository']

                    indexed_paths.add(indexed_path)

                    if not indexed_repo_path in self.filtered_repo_update_paths:

                        continue

                    continue

                ri_cnt = 0   # indexed

                riwc_cnt = 0  # indexed with content

                for path in self.get_paths(repo):

                    path = safe_unicode(path)

                    if path in to_index or path not in indexed_paths:

                        # This is either a file that's changed, or a new file

                        # that wasn't indexed before. So index it!

                        i, iwc = self.add_doc(writer, path, repo, repo_name)

                        writer_is_dirty = True

                        log.debug('re indexing %s' % path)

                        ri_cnt += i

                        ri_cnt_total += 1

                        riwc_cnt += iwc

                        riwc_cnt_total += iwc

                log.debug('added %s files %s with content for repo %s' % (

                             ri_cnt + riwc_cnt, riwc_cnt, repo.path)

                )

            log.debug('indexed %s files in total and %s with content' % (

                        ri_cnt_total, riwc_cnt_total)

            )

        finally:

            if writer_is_dirty:

                writer.commit(merge=True)

            else:

                writer.cancel()

    def build_indexes(self):

        if os.path.exists(self.index_location):

            log.debug('removing previous index')

            rmtree(self.index_location)

        if not os.path.exists(self.index_location):

            os.mkdir(self.index_location)

        chgset_idx = create_in(self.index_location, CHGSETS_SCHEMA,

                               indexname=CHGSET_IDX_NAME)

        chgset_idx_writer = chgset_idx.writer()

        file_idx = create_in(self.index_location, SCHEMA, indexname=IDX_NAME)

        file_idx_writer = file_idx.writer()

        log.debug('BUILDING INDEX FOR EXTENSIONS %s '

                  'AND REPOS %s' % (INDEX_EXTENSIONS, self.repo_paths.keys()))

        for repo_name, repo in self.repo_paths.items():

            # skip indexing if there aren't any revisions

            if len(repo) < 1:

                continue