when indexing changesets use the raw_id to locate the point from

which to start indexing rather than the revision which can be unreliable.

rhodecode/lib/indexers/__init__.py

@@ -1,265 +1,264
 # -*- coding: utf-8 -*-
 """
     rhodecode.lib.indexers.__init__
     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
     Whoosh indexing module for RhodeCode
 
     :created_on: Aug 17, 2010
     :author: marcink
     :copyright: (C) 2010-2012 Marcin Kuzminski <[e-mail]>
     :license: GPLv3, see COPYING for more details.
 """
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # the Free Software Foundation, either version 3 of the License, or
 # (at your option) any later version.
 #
 # This program is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 # GNU General Public License for more details.
 #
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 import os
 import sys
 import traceback
 import logging
 from os.path import dirname as dn, join as jn
 
 #to get the rhodecode import
 sys.path.append(dn(dn(dn(os.path.realpath(__file__)))))
 
 from string import strip
 from shutil import rmtree
 
 from whoosh.analysis import RegexTokenizer, LowercaseFilter, StopFilter
 from whoosh.fields import TEXT, ID, STORED, NUMERIC, BOOLEAN, Schema, FieldType
 from whoosh.index import create_in, open_dir
 from whoosh.formats import Characters
 from whoosh.highlight import highlight, HtmlFormatter, ContextFragmenter
 
 from webhelpers.html.builder import escape, literal
 from sqlalchemy import engine_from_config
 
 from rhodecode.model import init_model
 from rhodecode.model.scm import ScmModel
 from rhodecode.model.repo import RepoModel
 from rhodecode.config.environment import load_environment
 from rhodecode.lib.utils2 import LazyProperty
 from rhodecode.lib.utils import BasePasterCommand, Command, add_cache,\
     load_rcextensions
 
 log = logging.getLogger(__name__)
 
 # CUSTOM ANALYZER wordsplit + lowercase filter
 ANALYZER = RegexTokenizer(expression=r"\w+") | LowercaseFilter()
 
 #INDEX SCHEMA DEFINITION
 SCHEMA = Schema(
     fileid=ID(unique=True),
     owner=TEXT(),
     repository=TEXT(stored=True),
     path=TEXT(stored=True),
     content=FieldType(format=Characters(), analyzer=ANALYZER,
                       scorable=True, stored=True),
     modtime=STORED(),
     extension=TEXT(stored=True)
 )
 
 IDX_NAME = 'HG_INDEX'
 FORMATTER = HtmlFormatter('span', between='\n<span class="break">...</span>\n')
 FRAGMENTER = ContextFragmenter(200)
 
 CHGSETS_SCHEMA = Schema(
     raw_id=ID(unique=True, stored=True),
-    revision=NUMERIC(unique=True, stored=True),
     last=BOOLEAN(),
     owner=TEXT(),
     repository=ID(unique=True, stored=True),
     author=TEXT(stored=True),
     message=FieldType(format=Characters(), analyzer=ANALYZER,
                       scorable=True, stored=True),
     parents=TEXT(),
     added=TEXT(),
     removed=TEXT(),
     changed=TEXT(),
 )
 
 CHGSET_IDX_NAME = 'CHGSET_INDEX'
 
 class MakeIndex(BasePasterCommand):
 
     max_args = 1
     min_args = 1
 
     usage = "CONFIG_FILE"
     summary = "Creates index for full text search given configuration file"
     group_name = "RhodeCode"
     takes_config_file = -1
     parser = Command.standard_parser(verbose=True)
 
     def command(self):
         logging.config.fileConfig(self.path_to_ini_file)
         from pylons import config
         add_cache(config)
         engine = engine_from_config(config, 'sqlalchemy.db1.')
         init_model(engine)
         index_location = config['index_dir']
         repo_location = self.options.repo_location \
             if self.options.repo_location else RepoModel().repos_path
         repo_list = map(strip, self.options.repo_list.split(',')) \
             if self.options.repo_list else None
         repo_update_list = map(strip, self.options.repo_update_list.split(',')) \
             if self.options.repo_update_list else None
         load_rcextensions(config['here'])
         #======================================================================
         # WHOOSH DAEMON
         #======================================================================
         from rhodecode.lib.pidlock import LockHeld, DaemonLock
         from rhodecode.lib.indexers.daemon import WhooshIndexingDaemon
         try:
             l = DaemonLock(file_=jn(dn(dn(index_location)), 'make_index.lock'))
             WhooshIndexingDaemon(index_location=index_location,
                                  repo_location=repo_location,
                                  repo_list=repo_list,
                                  repo_update_list=repo_update_list)\
                 .run(full_index=self.options.full_index)
             l.release()
         except LockHeld:
             sys.exit(1)
 
     def update_parser(self):
         self.parser.add_option('--repo-location',
                           action='store',
                           dest='repo_location',
                           help="Specifies repositories location to index OPTIONAL",
                           )
         self.parser.add_option('--index-only',
                           action='store',
                           dest='repo_list',
                           help="Specifies a comma separated list of repositores "
                                 "to build index on. If not given all repositories "
                                 "are scanned for indexing. OPTIONAL",
                           )
         self.parser.add_option('--update-only',
                           action='store',
                           dest='repo_update_list',
                           help="Specifies a comma separated list of repositores "
                                 "to re-build index on. OPTIONAL",
                           )
         self.parser.add_option('-f',
                           action='store_true',
                           dest='full_index',
                           help="Specifies that index should be made full i.e"
                                 " destroy old and build from scratch",
                           default=False)
 
 
 class WhooshResultWrapper(object):
     def __init__(self, search_type, searcher, matcher, highlight_items,
                  repo_location):
         self.search_type = search_type
         self.searcher = searcher
         self.matcher = matcher
         self.highlight_items = highlight_items
         self.fragment_size = 200
         self.repo_location = repo_location
 
     @LazyProperty
     def doc_ids(self):
         docs_id = []
         while self.matcher.is_active():
             docnum = self.matcher.id()
             chunks = [offsets for offsets in self.get_chunks()]
             docs_id.append([docnum, chunks])
             self.matcher.next()
         return docs_id
 
     def __str__(self):
         return '<%s at %s>' % (self.__class__.__name__, len(self.doc_ids))
 
     def __repr__(self):
         return self.__str__()
 
     def __len__(self):
         return len(self.doc_ids)
 
     def __iter__(self):
         """
         Allows Iteration over results,and lazy generate content
 
         *Requires* implementation of ``__getitem__`` method.
         """
         for docid in self.doc_ids:
             yield self.get_full_content(docid)
 
     def __getitem__(self, key):
         """
         Slicing of resultWrapper
         """
         i, j = key.start, key.stop
 
         slices = []
         for docid in self.doc_ids[i:j]:
             slices.append(self.get_full_content(docid))
         return slices
 
     def get_full_content(self, docid):
         res = self.searcher.stored_fields(docid[0])
         log.debug('result: %s' % res)
         if self.search_type == 'content':
             full_repo_path = jn(self.repo_location, res['repository'])
             f_path = res['path'].split(full_repo_path)[-1]
             f_path = f_path.lstrip(os.sep)
             content_short = self.get_short_content(res, docid[1])
             res.update({'content_short': content_short,
                         'content_short_hl': self.highlight(content_short),
                         'f_path': f_path
                       })
         elif self.search_type == 'message':
             res.update({'message_hl': self.highlight(res['message'])})
 
         log.debug('result: %s' % res)
 
         return res
 
     def get_short_content(self, res, chunks):
 
         return ''.join([res['content'][chunk[0]:chunk[1]] for chunk in chunks])
 
     def get_chunks(self):
         """
         Smart function that implements chunking the content
         but not overlap chunks so it doesn't highlight the same
         close occurrences twice.
 
         :param matcher:
         :param size:
         """
         memory = [(0, 0)]
         if self.matcher.supports('positions'): 
             for span in self.matcher.spans():
                 start = span.startchar or 0
                 end = span.endchar or 0
                 start_offseted = max(0, start - self.fragment_size)
                 end_offseted = end + self.fragment_size
 
                 if start_offseted < memory[-1][1]:
                     start_offseted = memory[-1][1]
                 memory.append((start_offseted, end_offseted,))
                 yield (start_offseted, end_offseted,)
 
     def highlight(self, content, top=5):
         if self.search_type not in ['content', 'message']:
             return ''
         hl = highlight(
             text=content,
             terms=self.highlight_items,
             analyzer=ANALYZER,
             fragmenter=FRAGMENTER,
             formatter=FORMATTER,
             top=top
         )
         return hl

rhodecode/lib/indexers/daemon.py

@@ -1,373 +1,390
 # -*- coding: utf-8 -*-
 """
     rhodecode.lib.indexers.daemon
     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
     A daemon will read from task table and run tasks
 
     :created_on: Jan 26, 2010
     :author: marcink
     :copyright: (C) 2010-2012 Marcin Kuzminski <[e-mail]>
     :license: GPLv3, see COPYING for more details.
 """
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # the Free Software Foundation, either version 3 of the License, or
 # (at your option) any later version.
 #
 # This program is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 # GNU General Public License for more details.
 #
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 from __future__ import with_statement
 
 import os
 import sys
 import logging
 import traceback
 
 from shutil import rmtree
 from time import mktime
 
 from os.path import dirname as dn
 from os.path import join as jn
 
 #to get the rhodecode import
 project_path = dn(dn(dn(dn(os.path.realpath(__file__)))))
 sys.path.append(project_path)
 
 from rhodecode.config.conf import INDEX_EXTENSIONS
 from rhodecode.model.scm import ScmModel
 from rhodecode.lib.utils2 import safe_unicode
 from rhodecode.lib.indexers import SCHEMA, IDX_NAME, CHGSETS_SCHEMA, CHGSET_IDX_NAME
 
 from rhodecode.lib.vcs.exceptions import ChangesetError, RepositoryError, \
     NodeDoesNotExistError
 
 from whoosh.index import create_in, open_dir, exists_in
 from whoosh.query import *
 from whoosh.qparser import QueryParser
 
 log = logging.getLogger('whoosh_indexer')
 
 
 class WhooshIndexingDaemon(object):
     """
     Daemon for atomic indexing jobs
     """
 
     def __init__(self, indexname=IDX_NAME, index_location=None,
                  repo_location=None, sa=None, repo_list=None,
                  repo_update_list=None):
         self.indexname = indexname
 
         self.index_location = index_location
         if not index_location:
             raise Exception('You have to provide index location')
 
         self.repo_location = repo_location
         if not repo_location:
             raise Exception('You have to provide repositories location')
 
         self.repo_paths = ScmModel(sa).repo_scan(self.repo_location)
 
         #filter repo list
         if repo_list:
             self.filtered_repo_paths = {}
             for repo_name, repo in self.repo_paths.items():
                 if repo_name in repo_list:
                     self.filtered_repo_paths[repo_name] = repo
 
             self.repo_paths = self.filtered_repo_paths
 
         #filter update repo list
         self.filtered_repo_update_paths = {}
         if repo_update_list:
             self.filtered_repo_update_paths = {}
             for repo_name, repo in self.repo_paths.items():
                 if repo_name in repo_update_list:
                     self.filtered_repo_update_paths[repo_name] = repo
             self.repo_paths = self.filtered_repo_update_paths
 
         self.initial = True
         if not os.path.isdir(self.index_location):
             os.makedirs(self.index_location)
             log.info('Cannot run incremental index since it does not'
                      ' yet exist running full build')
         elif not exists_in(self.index_location, IDX_NAME):
             log.info('Running full index build as the file content'
                      ' index does not exist')
         elif not exists_in(self.index_location, CHGSET_IDX_NAME):
             log.info('Running full index build as the changeset'
                      ' index does not exist')
         else:
             self.initial = False
 
     def get_paths(self, repo):
         """
         recursive walk in root dir and return a set of all path in that dir
         based on repository walk function
         """
         index_paths_ = set()
         try:
             tip = repo.get_changeset('tip')
             for topnode, dirs, files in tip.walk('/'):
                 for f in files:
                     index_paths_.add(jn(repo.path, f.path))
 
         except RepositoryError, e:
             log.debug(traceback.format_exc())
             pass
         return index_paths_
 
     def get_node(self, repo, path):
         n_path = path[len(repo.path) + 1:]
         node = repo.get_changeset().get_node(n_path)
         return node
 
     def get_node_mtime(self, node):
         return mktime(node.last_changeset.date.timetuple())
 
     def add_doc(self, writer, path, repo, repo_name):
         """
         Adding doc to writer this function itself fetches data from
         the instance of vcs backend
         """
 
         node = self.get_node(repo, path)
         indexed = indexed_w_content = 0
         # we just index the content of chosen files, and skip binary files
         if node.extension in INDEX_EXTENSIONS and not node.is_binary:
             u_content = node.content
             if not isinstance(u_content, unicode):
                 log.warning('  >> %s Could not get this content as unicode '
                             'replacing with empty content' % path)
                 u_content = u''
             else:
                 log.debug('    >> %s [WITH CONTENT]' % path)
                 indexed_w_content += 1
 
         else:
             log.debug('    >> %s' % path)
             # just index file name without it's content
             u_content = u''
             indexed += 1
 
         p = safe_unicode(path)
         writer.add_document(
             fileid=p,
             owner=unicode(repo.contact),
             repository=safe_unicode(repo_name),
             path=p,
             content=u_content,
             modtime=self.get_node_mtime(node),
             extension=node.extension
         )
         return indexed, indexed_w_content
 
-    def index_changesets(self, writer, repo_name, repo, start_rev=0):
+    def index_changesets(self, writer, repo_name, repo, start_rev=None):
         """
         Add all changeset in the vcs repo starting at start_rev
         to the index writer
+
+        :param writer: the whoosh index writer to add to
+        :param repo_name: name of the repository from whence the
+          changeset originates including the repository group
+        :param repo: the vcs repository instance to index changesets for,
+          the presumption is the repo has changesets to index
+        :param start_rev=None: the full sha id to start indexing from
+          if start_rev is None then index from the first changeset in
+          the repo
         """
 
-        log.debug('indexing changesets in %s[%d:]' % (repo_name, start_rev))
+        if start_rev is None:
+            start_rev = repo[0].raw_id
+
+        log.debug('indexing changesets in %s starting at rev: %s' % (repo_name, start_rev))
 
         indexed=0
-        for cs in repo[start_rev:]:
+        for cs in repo.get_changesets(start=start_rev):
             writer.add_document(
                 raw_id=unicode(cs.raw_id),
                 owner=unicode(repo.contact),
                 repository=safe_unicode(repo_name),
                 author=cs.author,
                 message=cs.message,
-                revision=cs.revision,
                 last=cs.last,
                 added=u' '.join([node.path for node in cs.added]).lower(),
                 removed=u' '.join([node.path for node in cs.removed]).lower(),
                 changed=u' '.join([node.path for node in cs.changed]).lower(),
                 parents=u' '.join([cs.raw_id for cs in cs.parents]),
             )
             indexed += 1
 
         log.debug('indexed %d changesets for repo %s' % (indexed, repo_name))
 
     def index_files(self, file_idx_writer, repo_name, repo):
         i_cnt = iwc_cnt = 0
         log.debug('building index for [%s]' % repo.path)
         for idx_path in self.get_paths(repo):
             i, iwc = self.add_doc(file_idx_writer, idx_path, repo, repo_name)
             i_cnt += i
             iwc_cnt += iwc
 
         log.debug('added %s files %s with content for repo %s' % (i_cnt + iwc_cnt, iwc_cnt, repo.path))
 
     def update_changeset_index(self):
         idx = open_dir(self.index_location, indexname=CHGSET_IDX_NAME)
 
         with idx.searcher() as searcher:
             writer = idx.writer()
             writer_is_dirty = False
             try:
                 for repo_name, repo in self.repo_paths.items():
                     # skip indexing if there aren't any revs in the repo
-                    revs = repo.revisions
-                    if len(revs) < 1:
+                    num_of_revs = len(repo)
+                    if num_of_revs < 1:
                         continue
 
                     qp = QueryParser('repository', schema=CHGSETS_SCHEMA)
                     q = qp.parse(u"last:t AND %s" % repo_name)
 
-                    results = searcher.search(q, sortedby='revision')
+                    results = searcher.search(q)
 
+                    # default to scanning the entire repo
                     last_rev = 0
+                    start_id = None
+
                     if len(results) > 0:
-                        last_rev = results[0]['revision']
+                        # assuming that there is only one result, if not this
+                        # 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 len(revs) > last_rev + 1:
+                    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
-                        self.index_changesets(writer, repo_name, repo, last_rev)
+                        self.index_changesets(writer, repo_name, repo, start_id)
                         writer_is_dirty = True
 
             finally:
                 if writer_is_dirty:
                     log.debug('>> COMMITING CHANGES TO CHANGESET INDEX<<')
                     writer.commit(merge=True)
                     log.debug('>> COMMITTED CHANGES TO CHANGESET INDEX<<')
                 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
 
                     repo = self.repo_paths[indexed_repo_path]
 
                     try:
                         node = self.get_node(repo, indexed_path)
                         # Check if this file was changed since it was indexed
                         indexed_time = fields['modtime']
                         mtime = self.get_node_mtime(node)
                         if mtime > indexed_time:
                             # The file has changed, delete it and add it to the list of
                             # files to reindex
                             log.debug('adding to reindex list %s mtime: %s vs %s' % (
                                             indexed_path, mtime, indexed_time)
                             )
                             writer.delete_by_term('fileid', indexed_path)
                             writer_is_dirty = True
 
                             to_index.add(indexed_path)
                     except (ChangesetError, NodeDoesNotExistError):
                         # This file was deleted since it was indexed
                         log.debug('removing from index %s' % indexed_path)
                         writer.delete_by_term('path', indexed_path)
                         writer_is_dirty = True
 
             # Loop over the files in the filesystem
             # Assume we have a function that gathers the filenames of the
             # documents to be indexed
             ri_cnt_total = 0  # indexed
             riwc_cnt_total = 0  # indexed with content
             for repo_name, repo in self.repo_paths.items():
                 # skip indexing if there aren't any revisions
                 if len(repo) < 1:
                     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:
                 log.debug('>> COMMITING CHANGES <<')
                 writer.commit(merge=True)
                 log.debug('>>> FINISHED REBUILDING INDEX <<<')
             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
 
             self.index_files(file_idx_writer, repo_name, repo)
             self.index_changesets(chgset_idx_writer, repo_name, repo)
 
         log.debug('>> COMMITING CHANGES <<')
         file_idx_writer.commit(merge=True)
         chgset_idx_writer.commit(merge=True)
         log.debug('>>> FINISHED BUILDING INDEX <<<')
 
     def update_indexes(self):
         self.update_file_index()
         self.update_changeset_index()
 
     def run(self, full_index=False):
         """Run daemon"""
         if full_index or self.initial:
             self.build_indexes()
         else:
             self.update_indexes()

rhodecode/tests/functional/test_search.py

@@ -1,78 +1,78
 import os
 from rhodecode.tests import *
 from nose.plugins.skip import SkipTest
 
 
 class TestSearchController(TestController):
 
     def test_index(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'))
 
         self.assertTrue('class="small" id="q" name="q" type="text"' in
                         response.body)
         # Test response...
 
     def test_empty_search(self):
         if os.path.isdir(self.index_location):
             raise SkipTest('skipped due to existing index')
         else:
             self.log_user()
             response = self.app.get(url(controller='search', action='index'),
                                     {'q': HG_REPO})
             self.assertTrue('There is no index to search in. '
                             'Please run whoosh indexer' in response.body)
 
     def test_normal_search(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
                                 {'q': 'def repo'})
         response.mustcontain('10 results')
 
     def test_repo_search(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
                                 {'q': 'repository:%s def test' % HG_REPO})
 
         response.mustcontain('4 results')
 
     def test_search_last(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
                                 {'q': 'last:t', 'type': 'commit'})
 
         response.mustcontain('1 results')
 
     def test_search_commit_message(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
                                 {'q': 'bother to ask where to fetch repo during tests',
                                  'type': 'commit'})
 
         response.mustcontain('1 results')
         response.mustcontain('a00c1b6f5d7a6ae678fd553a8b81d92367f7ecf1')
 
     def test_search_commit_changed_file(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
                                 {'q': 'changed:tests/utils.py',
                                  'type': 'commit'})
 
         response.mustcontain('a00c1b6f5d7a6ae678fd553a8b81d92367f7ecf1')
 
     def test_search_commit_added_file(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
                                 {'q': 'added:README.rst',
                                  'type': 'commit'})
 
         response.mustcontain('1 results')
         response.mustcontain('3803844fdbd3b711175fc3da9bdacfcd6d29a6fb')
 
     def test_search_author(self):
         self.log_user()
         response = self.app.get(url(controller='search', action='index'),
-                                {'q': 'author:[e-mail] revision:0',
+                                {'q': 'author:[e-mail] raw_id:b986218ba1c9b0d6a259fac9b050b1724ed8e545',
                                  'type': 'commit'})
 
         response.mustcontain('1 results')