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

"""

    rhodecode.lib.celerylib.__init__

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

    celery libs for RhodeCode

    :created_on: Nov 27, 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, 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 socket

import traceback

import logging

from hashlib import md5

from decorator import decorator

from pylons import  config

from vcs.utils.lazy import LazyProperty

from rhodecode.lib import str2bool

from rhodecode.lib.pidlock import DaemonLock, LockHeld

from celery.messaging import establish_connection

log = logging.getLogger(__name__)

try:

    CELERY_ON = str2bool(config['app_conf'].get('use_celery'))

except KeyError:

    CELERY_ON = False

class ResultWrapper(object):

    def __init__(self, task):

        self.task = task

    @LazyProperty

    def result(self):

        return self.task

def run_task(task, *args, **kwargs):

    if CELERY_ON:

        try:

            t = task.apply_async(args=args, kwargs=kwargs)

            log.info('running task %s:%s', t.task_id, task)

            return t

        except socket.error, e:

            if  e.errno == 111:

                log.debug('Unable to connect to celeryd. Sync execution')

            else:

                log.error(traceback.format_exc())

        except KeyError, e:

                log.debug('Unable to connect to celeryd. Sync execution')

        except Exception, e:

            log.error(traceback.format_exc())

    log.debug('executing task %s in sync mode', task)

    return ResultWrapper(task(*args, **kwargs))

def __get_lockkey(func, *fargs, **fkwargs):

    params = list(fargs)

    params.extend(['%s-%s' % ar for ar in fkwargs.items()])

    func_name = str(func.__name__) if hasattr(func, '__name__') else str(func)

        md5(func_name + '-' + '-'.join(map(str, params))).hexdigest()

    return lockkey

def locked_task(func):

    def __wrapper(func, *fargs, **fkwargs):

        lockkey = __get_lockkey(func, *fargs, **fkwargs)

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

        try:

            ret = func(*fargs, **fkwargs)

            l.release()

            return ret

        except LockHeld:

            log.info('LockHeld')

            return 'Task with key %s already running' % lockkey

    return decorator(__wrapper, func)

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

"""

    rhodecode.lib.celerylib.tasks

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

    RhodeCode task modules, containing all task that suppose to be run

    by celery daemon

    :created_on: Oct 6, 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, 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 celery.decorators import task

import os

import traceback

import logging

from time import mktime

from operator import itemgetter

from string import lower

from pylons import config

from pylons.i18n.translation import _

from rhodecode.lib import LANGUAGES_EXTENSIONS_MAP

from rhodecode.lib.celerylib import run_task, locked_task, str2bool, \

    __get_lockkey, LockHeld, DaemonLock

from rhodecode.lib.helpers import person

from rhodecode.lib.smtp_mailer import SmtpMailer

from rhodecode.lib.utils import add_cache

from rhodecode.lib.odict import OrderedDict

from rhodecode.model import init_model

from rhodecode.model import meta

from rhodecode.model.db import RhodeCodeUi, Statistics, Repository

from vcs.backends import get_repo

from sqlalchemy import engine_from_config

add_cache(config)

try:

    import json

except ImportError:

    #python 2.5 compatibility

    import simplejson as json

__all__ = ['whoosh_index', 'get_commits_stats',

           'reset_user_password', 'send_email']

CELERY_ON = str2bool(config['app_conf'].get('use_celery'))

def get_session():

    if CELERY_ON:

        engine = engine_from_config(config, 'sqlalchemy.db1.')

        init_model(engine)

    sa = meta.Session()

    return sa

def get_repos_path():

    sa = get_session()

    q = sa.query(RhodeCodeUi).filter(RhodeCodeUi.ui_key == '/').one()

    return q.ui_value

@task(ignore_result=True)

@locked_task

def whoosh_index(repo_location, full_index):

    #log = whoosh_index.get_logger()

    from rhodecode.lib.indexers.daemon import WhooshIndexingDaemon

    index_location = config['index_dir']

    WhooshIndexingDaemon(index_location=index_location,

                         repo_location=repo_location, sa=get_session())\

                         .run(full_index=full_index)

@task(ignore_result=True)

def get_commits_stats(repo_name, ts_min_y, ts_max_y):

    try:

        log = get_commits_stats.get_logger()

    except:

        log = logging.getLogger(__name__)

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

                            ts_max_y)

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

    try:

        #for js data compatibilty cleans the key for person from '

        akc = lambda k: person(k).replace('"', "")

        co_day_auth_aggr = {}

        commits_by_day_aggregate = {}

        repos_path = get_repos_path()

        p = os.path.join(repos_path, repo_name)

        repo = get_repo(p)

        repo_size = len(repo.revisions)

        #return if repo have no revisions

        if repo_size < 1:

            lock.release()

            return True

        skip_date_limit = True

        parse_limit = int(config['app_conf'].get('commit_parse_limit'))

        last_rev = 0

        last_cs = None

        timegetter = itemgetter('time')

        sa = get_session()

        dbrepo = sa.query(Repository)\

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

        cur_stats = sa.query(Statistics)\

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

        if cur_stats is not None:

            last_rev = cur_stats.stat_on_revision

        if last_rev == 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(json.loads(

                                        cur_stats.commit_activity_combined))

            co_day_auth_aggr = json.loads(cur_stats.commit_activity)

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

        lmktime = mktime

        last_rev = last_rev + 1 if last_rev > 0 else last_rev

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

            last_cs = cs  # remember last parsed changeset

            k = lmktime([cs.date.timetuple()[0], cs.date.timetuple()[1],

                          cs.date.timetuple()[2], 0, 0, 0, 0, 0, 0])

            if akc(cs.author) in co_day_auth_aggr:

                try:

                    l = [timegetter(x) for x in

                         co_day_auth_aggr[akc(cs.author)]['data']]

                    time_pos = l.index(k)

                except ValueError:

                    time_pos = False

                if time_pos >= 0 and time_pos is not False:

                    datadict = \

                        co_day_auth_aggr[akc(cs.author)]['data'][time_pos]

                    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),

                                   }

                        co_day_auth_aggr[akc(cs.author)]['data']\

                            .append(datadict)

            else:

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

                    co_day_auth_aggr[akc(cs.author)] = {

                                        "label": akc(cs.author),

                                        "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))

        if not co_day_auth_aggr:

            co_day_auth_aggr[akc(repo.contact)] = {

                "label": akc(repo.contact),

                "data": [0, 1],

                "schema": ["commits"],

            }

        stats = cur_stats if cur_stats else Statistics()

        stats.commit_activity = json.dumps(co_day_auth_aggr)

        stats.commit_activity_combined = json.dumps(overview_data)

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

        leftovers = len(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 = json.dumps(__get_codes_stats(repo_name))

        try:

            stats.repository = dbrepo

            stats.stat_on_revision = last_cs.revision if last_cs else 0

            sa.add(stats)

            sa.commit()

        except:

            log.error(traceback.format_exc())

            sa.rollback()

            lock.release()

            return False

        #final release

        lock.release()

        #execute another task if celery is enabled

        if len(repo.revisions) > 1 and CELERY_ON:

            run_task(get_commits_stats, repo_name, ts_min_y, ts_max_y)

        return True

    except LockHeld:

        log.info('LockHeld')

        return 'Task with key %s already running' % lockkey

@task(ignore_result=True)

def reset_user_password(user_email):

    try:

        log = reset_user_password.get_logger()

    except:

        log = logging.getLogger(__name__)

    from rhodecode.lib import auth

    from rhodecode.model.db import User

    try:

        try:

            sa = get_session()

            user = sa.query(User).filter(User.email == user_email).scalar()

            new_passwd = auth.PasswordGenerator().gen_password(8,

                             auth.PasswordGenerator.ALPHABETS_BIG_SMALL)

            if user:

                user.password = auth.get_crypt_password(new_passwd)

                user.api_key = auth.generate_api_key(user.username)

                sa.add(user)

                sa.commit()

                log.info('change password for %s', user_email)

            if new_passwd is None:

                raise Exception('unable to generate new password')

        except:

            log.error(traceback.format_exc())

            sa.rollback()

        run_task(send_email, user_email,

                 "Your new rhodecode password",

                 'Your new rhodecode password:%s' % (new_passwd))

        log.info('send new password mail to %s', user_email)

    except:

        log.error('Failed to update user password')

        log.error(traceback.format_exc())

    return True

@task(ignore_result=True)

def send_email(recipients, subject, body):

    """

    Sends an email with defined parameters from the .ini files.

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

"""

    rhodecode.lib.indexers.__init__

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

    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, 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

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, Schema, FieldType

from whoosh.index import create_in, open_dir

from whoosh.formats import Characters

from whoosh.highlight import highlight, SimpleFragmenter, HtmlFormatter

from webhelpers.html.builder import escape

from sqlalchemy import engine_from_config

from vcs.utils.lazy import LazyProperty

from rhodecode.model import init_model

from rhodecode.model.scm import ScmModel

from rhodecode.config.environment import load_environment

from rhodecode.lib import LANGUAGES_EXTENSIONS_MAP

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

#EXTENSIONS WE WANT TO INDEX CONTENT OFF

INDEX_EXTENSIONS = LANGUAGES_EXTENSIONS_MAP.keys()

#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

        repo_list = map(strip, self.options.repo_list.split(',')) \

            if self.options.repo_list else None

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

        # WHOOSH DAEMON

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

        from rhodecode.lib.pidlock import LockHeld, DaemonLock

        from rhodecode.lib.indexers.daemon import WhooshIndexingDaemon

        try:

            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 __getitem__(self, key):

        """

        Slicing of resultWrapper

        """

        i, j = key.start, key.stop

        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])

    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(escape(content),

                 self.highlight_items,

                 analyzer=ANALYZER,

                 fragmenter=FRAGMENTER,

                 formatter=FORMATTER,

                 top=top)

        return hl

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

"""

    rhodecode.lib.utils

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

    Utilities library for RhodeCode

    :created_on: Apr 18, 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, 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 logging

import datetime

import traceback

import paste

import beaker

from paste.script.command import Command, BadCommand

from UserDict import DictMixin

from mercurial import ui, config, hg

from mercurial.error import RepoError

from webhelpers.text import collapse, remove_formatting, strip_tags

from vcs.backends.base import BaseChangeset

from vcs.utils.lazy import LazyProperty

from rhodecode.model import meta

from rhodecode.model.caching_query import FromCache

from rhodecode.model.db import Repository, User, RhodeCodeUi, UserLog, Group, \

    RhodeCodeSettings

from rhodecode.model.repo import RepoModel

from rhodecode.model.user import UserModel

log = logging.getLogger(__name__)

def recursive_replace(str, replace=' '):

    """Recursive replace of given sign to just one instance

    :param str: given string

    :param replace: char to find and replace multiple instances

    Examples::

    >>> recursive_replace("Mighty---Mighty-Bo--sstones",'-')

    'Mighty-Mighty-Bo-sstones'

    """

    if str.find(replace * 2) == -1:

        return str

    else:

        str = str.replace(replace * 2, replace)

        return recursive_replace(str, replace)

def repo_name_slug(value):

    """Return slug of name of repository

    This function is called on each creation/modification

    of repository to prevent bad names in repo

    """

    slug = remove_formatting(value)

    slug = strip_tags(slug)

    for c in """=[]\;'"<>,/~!@#$%^&*()+{}|: """:

        slug = slug.replace(c, '-')

    slug = recursive_replace(slug, '-')

    slug = collapse(slug, '-')

    return slug

def get_repo_slug(request):

    return request.environ['pylons.routes_dict'].get('repo_name')

def action_logger(user, action, repo, ipaddr='', sa=None):

    """

    Action logger for various actions made by users

    :param user: user that made this action, can be a unique username string or

        object containing user_id attribute

    :param action: action to log, should be on of predefined unique actions for

        easy translations

    :param repo: string name of repository or object containing repo_id,

        that action was made on

    :param ipaddr: optional ip address from what the action was made

    :param sa: optional sqlalchemy session

    """

    if not sa:

        sa = meta.Session()

    try:

        um = UserModel()

        if hasattr(user, 'user_id'):

            user_obj = user

        elif isinstance(user, basestring):

            user_obj = um.get_by_username(user, cache=False)

        else:

            raise Exception('You have to provide user object or username')

        rm = RepoModel()

        if hasattr(repo, 'repo_id'):

            repo_obj = rm.get(repo.repo_id, cache=False)

            repo_name = repo_obj.repo_name

        elif  isinstance(repo, basestring):

            repo_name = repo.lstrip('/')

            repo_obj = rm.get_by_repo_name(repo_name, cache=False)

        else:

            raise Exception('You have to provide repository to action logger')

        user_log = UserLog()

        user_log.user_id = user_obj.user_id

        user_log.action = action

        user_log.repository_id = repo_obj.repo_id

        user_log.repository_name = repo_name

        user_log.action_date = datetime.datetime.now()

        user_log.user_ip = ipaddr

        sa.add(user_log)

        sa.commit()

        log.info('Adding user %s, action %s on %s', user_obj, action, repo)

    except:

        log.error(traceback.format_exc())

        sa.rollback()

def get_repos(path, recursive=False):

    """

    Scans given path for repos and return (name,(type,path)) tuple

    :param path: path to scann for repositories

    :param recursive: recursive search and return names with subdirs in front

    """

    from vcs.utils.helpers import get_scm

    from vcs.exceptions import VCSError

    if path.endswith(os.sep):

        #remove ending slash for better results

        path = path[:-1]

    def _get_repos(p):

        if not os.access(p, os.W_OK):

            return

        for dirpath in os.listdir(p):

            if os.path.isfile(os.path.join(p, dirpath)):

                continue

            cur_path = os.path.join(p, dirpath)

            try:

                scm_info = get_scm(cur_path)

                yield scm_info[1].split(path)[-1].lstrip(os.sep), scm_info

            except VCSError:

                if not recursive:

                    continue

                #check if this dir containts other repos for recursive scan

                rec_path = os.path.join(p, dirpath)

                if os.path.isdir(rec_path):

                    for inner_scm in _get_repos(rec_path):

                        yield inner_scm

    return _get_repos(path)

def check_repo_fast(repo_name, base_path):

    """

    Check given path for existence of directory

    :param repo_name:

    :param base_path:

    :return False: if this directory is present

    """

    if os.path.isdir(os.path.join(base_path, repo_name)):

        return False

    return True

def check_repo(repo_name, base_path, verify=True):

    repo_path = os.path.join(base_path, repo_name)

    try:

        if not check_repo_fast(repo_name, base_path):

            return False

        r = hg.repository(ui.ui(), repo_path)

        if verify:

            hg.verify(r)

        #here we hnow that repo exists it was verified

        log.info('%s repo is already created', repo_name)

        return False

    except RepoError:

        #it means that there is no valid repo there...

        log.info('%s repo is free for creation', repo_name)

        return True

def ask_ok(prompt, retries=4, complaint='Yes or no, please!'):

    while True:

        ok = raw_input(prompt)

        if ok in ('y', 'ye', 'yes'):

            return True

        if ok in ('n', 'no', 'nop', 'nope'):

            return False

        retries = retries - 1

        meta.Session.remove()

    return baseui