##// END OF EJS Templates
docs: Update notes regarding Pylons / Pyramid
docs: Update notes regarding Pylons / Pyramid

File last commit:

r413:f591e19f default
r739:3f65f91e default
Show More
whoosh_fallback_schema.py
75 lines | 2.8 KiB | text/x-python | PythonLexer
/ rhodecode / lib / index / whoosh_fallback_schema.py
# -*- coding: utf-8 -*-
# Copyright (C) 2012-2016 RhodeCode GmbH
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License, version 3
# (only), as published by the Free Software Foundation.
#
# 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 Affero General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
#
# This program is dual-licensed. If you wish to learn more about the
# RhodeCode Enterprise Edition, including its added features, Support services,
# and proprietary license terms, please see https://rhodecode.com/licenses/
"""
Whoosh fallback schema for RhodeCode in case rhodecode_tools defined one is
not available
"""
from __future__ import absolute_import
from whoosh.analysis import RegexTokenizer, LowercaseFilter
from whoosh.formats import Characters
from whoosh.fields import (
TEXT, ID, STORED, NUMERIC, BOOLEAN, Schema, FieldType, DATETIME)
# CUSTOM ANALYZER wordsplit + lowercase filter for case insensitive search
ANALYZER = RegexTokenizer(expression=r"\w+") | LowercaseFilter()
# FILE INDEX SCHEMA DEFINITION
FILE_INDEX_NAME = 'FILE_INDEX'
FILE_SCHEMA = Schema(
fileid=ID(unique=True), # Path
repository=ID(stored=True),
repository_id=NUMERIC(unique=True, stored=True), # Numeric id of repo
repo_name=TEXT(stored=True),
owner=TEXT(),
path=TEXT(stored=True),
content=FieldType(format=Characters(), analyzer=ANALYZER,
scorable=True, stored=True),
modtime=STORED(),
md5=STORED(),
extension=ID(stored=True),
commit_id=TEXT(stored=True),
size=NUMERIC(int, 64, signed=False, stored=True),
mimetype=TEXT(stored=True),
lines=NUMERIC(int, 64, signed=False, stored=True),
)
# COMMIT INDEX SCHEMA
COMMIT_INDEX_NAME = 'COMMIT_INDEX'
COMMIT_SCHEMA = Schema(
commit_id=ID(unique=True, stored=True),
repository=ID(unique=True, stored=True),
repository_id=NUMERIC(unique=True, stored=True),
commit_idx=NUMERIC(stored=True, sortable=True),
commit_idx_sort=ID(),
date=NUMERIC(int, 64, signed=False, stored=True, sortable=True),
owner=TEXT(stored=True),
author=TEXT(stored=True),
message=FieldType(format=Characters(), analyzer=ANALYZER,
scorable=True, stored=True),
parents=TEXT(stored=True),
added=TEXT(stored=True), # space separated names of added files
removed=TEXT(stored=True), # space separated names of removed files
changed=TEXT(stored=True), # space separated names of changed files
)