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

"""

    rhodecode.controllers.search

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

    Search controller for rhodecode

    :created_on: Aug 7, 2010

    :author: marcink

    :copyright: (C) 2010-2012 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, 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 logging

import traceback

from pylons.i18n.translation import _

from pylons import request, config, tmpl_context as c

from rhodecode.lib.auth import LoginRequired

from rhodecode.lib.base import BaseController, render

from webhelpers.paginate import Page

from webhelpers.util import update_params

from whoosh.index import open_dir, EmptyIndexError

from whoosh.qparser import QueryParser, QueryParserError

from whoosh.query import Phrase, Wildcard, Term, Prefix

log = logging.getLogger(__name__)

class SearchController(BaseController):

    @LoginRequired()

    def __before__(self):

        super(SearchController, self).__before__()

    def index(self, search_repo=None):

        c.repo_name = search_repo

        c.formated_results = []

        c.runtime = ''

        c.cur_query = request.GET.get('q', None)

        c.cur_type = request.GET.get('type', 'source')

        c.cur_search = search_type = {'content': 'content',

                                      'commit': 'content',

                                      'path': 'path',

                                      'repository': 'repository'}\

                                      .get(c.cur_type, 'content')

        if c.cur_query:

            cur_query = c.cur_query.lower()

        if c.cur_query:

            p = int(request.params.get('page', 1))

            highlight_items = set()

            try:

                idx = open_dir(config['app_conf']['index_dir'],

                               indexname=IDX_NAME)

                searcher = idx.searcher()

                qp = QueryParser(search_type, schema=SCHEMA)

                if c.repo_name:

                    cur_query = u'repository:%s %s' % (c.repo_name, cur_query)

                try:

                    query = qp.parse(unicode(cur_query))

                    # extract words for highlight

                    if isinstance(query, Phrase):

                        highlight_items.update(query.words)

                    elif isinstance(query, Prefix):

                        highlight_items.add(query.text)

                    else:

                        for i in query.all_terms():

                            if i[0] == 'content':

                                highlight_items.add(i[1])

                    matcher = query.matcher(searcher)

                    log.debug(query)

                    log.debug(highlight_items)

                    results = searcher.search(query)

                    res_ln = len(results)

                    c.runtime = '%s results (%.3f seconds)' % (

                        res_ln, results.runtime

                    )

                    def url_generator(**kw):

                        return update_params("?q=%s&type=%s" \

                                           % (c.cur_query, c.cur_search), **kw)

                    c.formated_results = Page(

                        page=p,

                        item_count=res_ln,

                        items_per_page=10,

                        url=url_generator

                    )

                except QueryParserError:

                    c.runtime = _('Invalid search query. Try quoting it.')

                searcher.close()

            except (EmptyIndexError, IOError):

                log.error(traceback.format_exc())

                log.error('Empty Index data')

                c.runtime = _('There is no index to search in. '

                              'Please run whoosh indexer')

            except (Exception):

                log.error(traceback.format_exc())

                c.runtime = _('An error occurred during this search operation')

        # Return a rendered template

        return render('/search/search.html')

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, HtmlFormatter, ContextFragmenter

from webhelpers.html.builder import escape

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

# 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=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)

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

        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,)\

                .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 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)

        self.search_type = search_type

        self.searcher = searcher

        self.matcher = matcher

        self.highlight_items = highlight_items

        self.fragment_size = 200

    @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])

        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])

    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)]

        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 != 'content':

            return ''

        hl = highlight(

            text=escape(content),

            terms=self.highlight_items,

            analyzer=ANALYZER,

            fragmenter=FRAGMENTER,

            formatter=FORMATTER,

            top=top

        )

        return hl