This file was forked by the Kallithea project in July 2014.

Original author and date, and relevant copyright and licensing information is below:

:created_on: Oct 6, 2010

:author: marcink

:copyright: (c) 2013 RhodeCode GmbH, and others.

:license: GPLv3, see LICENSE.md for more details.

"""

import email.message

import email.utils

import os

import smtplib

import time

import traceback

from collections import OrderedDict

from operator import itemgetter

from time import mktime

import celery.utils.log

from tg import config

import kallithea

import kallithea.lib.helpers as h

from kallithea.lib import celerylib, conf, ext_json, hooks

from kallithea.lib.indexers.daemon import WhooshIndexingDaemon

from kallithea.lib.utils2 import asbool, ascii_bytes

from kallithea.lib.vcs.utils import author_email

from kallithea.model import db, repo, userlog

__all__ = ['whoosh_index', 'get_commits_stats', 'send_email']

log = celery.utils.log.get_task_logger(__name__)

@celerylib.task

@celerylib.locked_task

@celerylib.dbsession

def whoosh_index(repo_location, full_index):

    celerylib.get_session() # initialize database connection

    index_location = config['index_dir']

    WhooshIndexingDaemon(index_location=index_location,

                         repo_location=repo_location) \

                         .run(full_index=full_index)

@celerylib.task

@celerylib.dbsession

def get_commits_stats(repo_name, ts_min_y, ts_max_y, recurse_limit=100):

    DBS = celerylib.get_session()

    lockkey = celerylib.__get_lockkey('get_commits_stats', repo_name, ts_min_y,

                            ts_max_y)

    lockkey_path = config.get('cache_dir') or config['app_conf']['cache_dir']  # Backward compatibility for TurboGears < 2.4

    log.info('running task with lockkey %s', lockkey)

    try:

        lock = celerylib.DaemonLock(os.path.join(lockkey_path, lockkey))

        co_day_auth_aggr = {}

        commits_by_day_aggregate = {}

        db_repo = db.Repository.get_by_repo_name(repo_name)

        if db_repo is None:

            return True

        scm_repo = db_repo.scm_instance

        repo_size = scm_repo.count()

        # return if repo have no revisions

        if repo_size < 1:

            lock.release()

            return True

        skip_date_limit = True

        parse_limit = int(config.get('commit_parse_limit'))

        last_rev = None

        last_cs = None

        timegetter = itemgetter('time')

        dbrepo = DBS.query(db.Repository) \

            .filter(db.Repository.repo_name == repo_name).scalar()

        cur_stats = DBS.query(db.Statistics) \

            .filter(db.Statistics.repository == dbrepo).scalar()

        if cur_stats is not None:

            last_rev = cur_stats.stat_on_revision

        if last_rev == scm_repo.get_changeset().revision and repo_size > 1:

            # pass silently without any work if we're not on first revision or

            # current state of parsing revision(from db marker) is the

            # last revision

            lock.release()

            return True

        if cur_stats:

            commits_by_day_aggregate = OrderedDict(ext_json.loads(

                                        cur_stats.commit_activity_combined))

            co_day_auth_aggr = ext_json.loads(cur_stats.commit_activity)

        log.debug('starting parsing %s', parse_limit)

        last_rev = last_rev + 1 if last_rev and last_rev >= 0 else 0

        log.debug('Getting revisions from %s to %s',

             last_rev, last_rev + parse_limit

        )

        for cs in scm_repo[last_rev:last_rev + parse_limit]:

            log.debug('parsing %s', cs)

            last_cs = cs  # remember last parsed changeset

            tt = cs.date.timetuple()

            k = mktime(tt[:3] + (0, 0, 0, 0, 0, 0))

                try:

                    l = [timegetter(x) for x in

                    time_pos = l.index(k)

                except ValueError:

                    time_pos = None

                if time_pos is not None and time_pos >= 0:

                    datadict = \

                    datadict["commits"] += 1

                    datadict["added"] += len(cs.added)

                    datadict["changed"] += len(cs.changed)

                    datadict["removed"] += len(cs.removed)

                else:

                    if k >= ts_min_y and k <= ts_max_y or skip_date_limit:

                        datadict = {"time": k,

                                    "commits": 1,

                                    "added": len(cs.added),

                                    "changed": len(cs.changed),

                                    "removed": len(cs.removed),

                                   }

                            .append(datadict)

            else:

                if k >= ts_min_y and k <= ts_max_y or skip_date_limit:

                                        "data": [{"time": k,

                                                 "commits": 1,

                                                 "added": len(cs.added),

                                                 "changed": len(cs.changed),

                                                 "removed": len(cs.removed),

                                                 }],

                                        "schema": ["commits"],

                                        }

            # gather all data by day

            if k in commits_by_day_aggregate:

                commits_by_day_aggregate[k] += 1

            else:

                commits_by_day_aggregate[k] = 1

        overview_data = sorted(commits_by_day_aggregate.items(),

                               key=itemgetter(0))

        stats = cur_stats if cur_stats else db.Statistics()

        stats.commit_activity = ascii_bytes(ext_json.dumps(co_day_auth_aggr))

        stats.commit_activity_combined = ascii_bytes(ext_json.dumps(overview_data))

        log.debug('last revision %s', last_rev)

        leftovers = len(scm_repo.revisions[last_rev:])

        log.debug('revisions to parse %s', leftovers)

        if last_rev == 0 or leftovers < parse_limit:

            log.debug('getting code trending stats')

            stats.languages = ascii_bytes(ext_json.dumps(__get_codes_stats(repo_name)))

        try:

            stats.repository = dbrepo

            stats.stat_on_revision = last_cs.revision if last_cs else 0

            DBS.add(stats)

            DBS.commit()

        except:

            log.error(traceback.format_exc())

            DBS.rollback()

            lock.release()

            return False

        # final release

        lock.release()

        # execute another task if celery is enabled

        if len(scm_repo.revisions) > 1 and kallithea.CELERY_APP and recurse_limit > 0:

            get_commits_stats(repo_name, ts_min_y, ts_max_y, recurse_limit - 1)

        elif recurse_limit <= 0: