Whoosh indexing module for RhodeCode

    :created_on: Aug 17, 2010

    :author: marcink

    :copyright: (C) 2009-2010 Marcin Kuzminski <marcin@python-works.com>    

    :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; version 2

# of the License or (at your opinion) any later version of the license.

# 

# 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, write to the Free Software

# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,

# MA  02110-1301, USA.

import os

import sys

import traceback

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 rhodecode.model import init_model

from rhodecode.model.scm import ScmModel

from rhodecode.config.environment import load_environment

from rhodecode.lib.utils import BasePasterCommand, Command, add_cache

from shutil import rmtree

from webhelpers.html.builder import escape

from vcs.utils.lazy import LazyProperty

from sqlalchemy import engine_from_config

from whoosh.analysis import RegexTokenizer, LowercaseFilter, StopFilter

from whoosh.fields import TEXT, ID, STORED, Schema, FieldType

from whoosh.index import create_in, open_dir

from whoosh.formats import Characters

from whoosh.highlight import highlight, SimpleFragmenter, HtmlFormatter

#EXTENSIONS WE WANT TO INDEX CONTENT OFF

INDEX_EXTENSIONS = ['action', 'adp', 'ashx', 'asmx', 'aspx', 'asx', 'axd', 'c',

                    'cfg', 'cfm', 'cpp', 'cs', 'css', 'diff', 'do', 'el', 'erl',

                    'h', 'htm', 'html', 'ini', 'java', 'js', 'jsp', 'jspx', 'lisp',

                    'lua', 'm', 'mako', 'ml', 'pas', 'patch', 'php', 'php3',

                    'php4', 'phtml', 'pm', 'py', 'rb', 'rst', 's', 'sh', 'sql',

                    'tpl', 'txt', 'vim', 'wss', 'xhtml', 'xml', 'xsl', 'xslt',

                    'yaws']

#CUSTOM ANALYZER wordsplit + lowercase filter

ANALYZER = RegexTokenizer(expression=r"\w+") | LowercaseFilter()

#INDEX SCHEMA DEFINITION

SCHEMA = Schema(owner=TEXT(),

                repository=TEXT(stored=True),

                path=TEXT(stored=True),

                content=FieldType(format=Characters(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 = SimpleFragmenter(200)

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):

        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

        #======================================================================

        # WHOOSH DAEMON

        #======================================================================

        from rhodecode.lib.pidlock import LockHeld, DaemonLock

        from rhodecode.lib.indexers.daemon import WhooshIndexingDaemon

        try:

            l = DaemonLock()

            WhooshIndexingDaemon(index_location=index_location,

                                 repo_location=repo_location,

                                 repo_list=repo_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 REQUIRED",

                          )

        self.parser.add_option('--index-only',

                          action='store',

                          dest='repo_list',

                          help="Specifies a comma separated list of repositores "

                                "to 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 ResultWrapper(object):

    def __init__(self, search_type, searcher, matcher, highlight_items):

        self.search_type = search_type

        self.searcher = searcher

        self.matcher = matcher

        self.highlight_items = highlight_items

        self.fragment_size = 200 / 2

    @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 __getslice__(self, i, j):

        """

        Slicing of resultWrapper

        """

        slice = []

        for docid in self.doc_ids[i:j]:

            slice.append(self.get_full_content(docid))

        return slice

    def get_full_content(self, docid):

        res = self.searcher.stored_fields(docid[0])

        f_path = res['path'][res['path'].find(res['repository']) \

                             + len(res['repository']):].lstrip('/')

        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})

        return res

    def get_short_content(self, res, chunks):

        return ''.join([res['content'][chunk[0]:chunk[1]] for chunk in chunks])

# -*- coding: utf-8 -*-

"""

    rhodecode.lib.indexers.daemon

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    A deamon will read from task table and run tasks

    :created_on: Jan 26, 2010

    :author: marcink

    :copyright: (C) 2009-2011 Marcin Kuzminski <marcin@python-works.com>

    :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; version 2

# of the License or (at your opinion) any later version of the license.

#

# 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, write to the Free Software

# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,

# MA  02110-1301, USA.

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.model.scm import ScmModel

from rhodecode.lib import safe_unicode

from rhodecode.lib.indexers import INDEX_EXTENSIONS, SCHEMA, IDX_NAME

from vcs.exceptions import ChangesetError, RepositoryError

from whoosh.index import create_in, open_dir

log = logging.getLogger('whooshIndexer')

# create logger

log.setLevel(logging.DEBUG)

log.propagate = False

# create console handler and set level to debug

ch = logging.StreamHandler()

ch.setLevel(logging.DEBUG)

# create formatter

# add formatter to ch

ch.setFormatter(formatter)

# add ch to logger

log.addHandler(ch)

class WhooshIndexingDaemon(object):

    """

    Deamon for atomic jobs

    """

    def __init__(self, indexname='HG_INDEX', index_location=None,

                 repo_location=None, sa=None, repo_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)

        if repo_list:

            filtered_repo_paths = {}

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

                if repo_name in repo_list:

                    filtered_repo_paths[repo_name] = repo

            self.repo_paths = filtered_repo_paths

        self.initial = False

        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')

            self.initial = True

    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))

                for dir in dirs:

                    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)

        #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)

        else:

            log.debug('    >> %s' % path)

            #just index file name without it's content

            u_content = u''

        writer.add_document(owner=unicode(repo.contact),

                        repository=safe_unicode(repo_name),

                        path=safe_unicode(path),

                        content=u_content,

                        modtime=self.get_node_mtime(node),

                        extension=node.extension)