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

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

"""

import traceback

import calendar

import logging

import itertools

from time import mktime

from datetime import timedelta, date

from tg import tmpl_context as c, request

from tg.i18n import ugettext as _

from webob.exc import HTTPBadRequest

from beaker.cache import cache_region, region_invalidate

from kallithea.lib.vcs.exceptions import ChangesetError, EmptyRepositoryError, \

    NodeDoesNotExistError

from kallithea.config.conf import ALL_READMES, ALL_EXTS, LANGUAGES_EXTENSIONS_MAP

from kallithea.model.db import Statistics, CacheInvalidation, User

from kallithea.lib.utils2 import safe_int, safe_str

from kallithea.lib.auth import LoginRequired, HasRepoPermissionLevelDecorator

from kallithea.lib.base import BaseRepoController, render, jsonify

from kallithea.lib.vcs.backends.base import EmptyChangeset

from kallithea.lib.markup_renderer import MarkupRenderer

from kallithea.lib.celerylib.tasks import get_commits_stats

from kallithea.lib.compat import json

from kallithea.lib.vcs.nodes import FileNode

from kallithea.lib.page import RepoPage

log = logging.getLogger(__name__)

README_FILES = [''.join([x[0][0], x[1][0]]) for x in

                    sorted(list(itertools.product(ALL_READMES, ALL_EXTS)),

                           key=lambda y:y[0][1] + y[1][1])]

class SummaryController(BaseRepoController):

    def __get_readme_data(self, db_repo):

        repo_name = db_repo.repo_name

        log.debug('Looking for README file')

        @cache_region('long_term', '_get_readme_from_cache')

        def _get_readme_from_cache(key, kind):

            readme_data = None

            readme_file = None

            try:

                # gets the landing revision! or tip if fails

                cs = db_repo.get_landing_changeset()

                if isinstance(cs, EmptyChangeset):

                    raise EmptyRepositoryError()

                renderer = MarkupRenderer()

                for f in README_FILES:

                    try:

                        readme = cs.get_node(f)

                        if not isinstance(readme, FileNode):

                            continue

                        readme_file = f

                        log.debug('Found README file `%s` rendering...',

                                  readme_file)

                        readme_data = renderer.render(readme.content,

                                                      filename=f)

                        break

                    except NodeDoesNotExistError:

                        continue

            except ChangesetError:

                log.error(traceback.format_exc())

                pass

            except EmptyRepositoryError:

                pass

            return readme_data, readme_file

        kind = 'README'

        valid = CacheInvalidation.test_and_set_valid(repo_name, kind)

        if not valid:

            region_invalidate(_get_readme_from_cache, None, '_get_readme_from_cache', repo_name, kind)

        return _get_readme_from_cache(repo_name, kind)

    @LoginRequired(allow_default_user=True)

    @HasRepoPermissionLevelDecorator('read')

    def index(self, repo_name):

        p = safe_int(request.GET.get('page'), 1)

        size = safe_int(request.GET.get('size'), 10)

        collection = c.db_repo_scm_instance

        c.cs_pagination = RepoPage(collection, page=p, items_per_page=size)

        page_revisions = [x.raw_id for x in list(c.cs_pagination)]

        c.cs_comments = c.db_repo.get_comments(page_revisions)

        c.cs_statuses = c.db_repo.statuses(page_revisions)

        if request.authuser.is_default_user:

            username = ''

        else:

            username = safe_str(request.authuser.username)

        if c.db_repo.enable_statistics:

            c.show_stats = True

        else:

            c.show_stats = False

        stats = Statistics.query() \

            .filter(Statistics.repository == c.db_repo) \

            .scalar()

        c.stats_percentage = 0

        if stats and stats.languages:

            c.no_data = False is c.db_repo.enable_statistics

            lang_stats_d = json.loads(stats.languages)

            lang_stats = ((x, {"count": y,

                               "desc": LANGUAGES_EXTENSIONS_MAP.get(x)})

                          for x, y in lang_stats_d.items())

            c.trending_languages = (

                sorted(lang_stats, reverse=True, key=lambda k: k[1])[:10]

            )

        else:

            c.no_data = True

            c.trending_languages = []

        c.enable_downloads = c.db_repo.enable_downloads

        c.readme_data, c.readme_file = \

            self.__get_readme_data(c.db_repo)

        return render('summary/summary.html')

    @LoginRequired()

    @HasRepoPermissionLevelDecorator('read')

    @jsonify

    def repo_size(self, repo_name):

        if request.is_xhr:

            return c.db_repo._repo_size()

        else:

            raise HTTPBadRequest()

    @LoginRequired(allow_default_user=True)

    @HasRepoPermissionLevelDecorator('read')

    def statistics(self, repo_name):

        if c.db_repo.enable_statistics:

            c.show_stats = True

            c.no_data_msg = _('No data ready yet')

        else:

            c.show_stats = False

            c.no_data_msg = _('Statistics are disabled for this repository')

        td = date.today() + timedelta(days=1)

        td_1m = td - timedelta(days=calendar.mdays[td.month])

        td_1y = td - timedelta(days=365)

        ts_min_m = mktime(td_1m.timetuple())

        ts_min_y = mktime(td_1y.timetuple())

        ts_max_y = mktime(td.timetuple())

        c.ts_min = ts_min_m

        c.ts_max = ts_max_y

        stats = Statistics.query() \

            .filter(Statistics.repository == c.db_repo) \

            .scalar()

        c.stats_percentage = 0

        if stats and stats.languages:

            c.no_data = False is c.db_repo.enable_statistics

            lang_stats_d = json.loads(stats.languages)

            c.commit_data = json.loads(stats.commit_activity)

            c.overview_data = json.loads(stats.commit_activity_combined)

            lang_stats = ((x, {"count": y,

                               "desc": LANGUAGES_EXTENSIONS_MAP.get(x)})

                          for x, y in lang_stats_d.items())

            c.trending_languages = (

                sorted(lang_stats, reverse=True, key=lambda k: k[1])[:10]

            )

            last_rev = stats.stat_on_revision + 1

            c.repo_last_rev = c.db_repo_scm_instance.count() \

                if c.db_repo_scm_instance.revisions else 0

            if last_rev == 0 or c.repo_last_rev == 0:

                pass

            else:

                c.stats_percentage = '%.2f' % ((float((last_rev)) /

                                                c.repo_last_rev) * 100)

        else:

            c.commit_data = {}

            c.overview_data = ([[ts_min_y, 0], [ts_max_y, 10]])

            c.trending_languages = {}

            c.no_data = True

        recurse_limit = 500  # don't recurse more than 500 times when parsing

        get_commits_stats(c.db_repo.repo_name, ts_min_y, ts_max_y, recurse_limit)

        return render('summary/statistics.html')

                            "get_group_%s" % _hash_key(group_name)

                          )

            )

        return q.scalar()

    @classmethod

    def get(cls, user_group_id, cache=False):

        user_group = cls.query()

        if cache:

            user_group = user_group.options(FromCache("sql_cache_short",

                                    "get_users_group_%s" % user_group_id))

        return user_group.get(user_group_id)

    def get_api_data(self, with_members=True):

        user_group = self

        data = dict(

            users_group_id=user_group.users_group_id,

            group_name=user_group.users_group_name,

            group_description=user_group.user_group_description,

            active=user_group.users_group_active,

            owner=user_group.owner.username,

        )

        if with_members:

            data['members'] = [

                ugm.user.get_api_data()

                for ugm in user_group.members

            ]

        return data

class UserGroupMember(Base, BaseDbModel):

    __tablename__ = 'users_groups_members'

    __table_args__ = (

        _table_args_default_dict,

    )

    users_group_member_id = Column(Integer(), primary_key=True)

    users_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)

    user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)

    user = relationship('User')

    users_group = relationship('UserGroup')

    def __init__(self, gr_id='', u_id=''):

        self.users_group_id = gr_id

        self.user_id = u_id

class RepositoryField(Base, BaseDbModel):

    __tablename__ = 'repositories_fields'

    __table_args__ = (

        UniqueConstraint('repository_id', 'field_key'),  # no-multi field

        _table_args_default_dict,

    )

    PREFIX = 'ex_'  # prefix used in form to not conflict with already existing fields

    repo_field_id = Column(Integer(), primary_key=True)

    repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)

    field_key = Column(String(250), nullable=False)

    field_label = Column(String(1024), nullable=False)

    field_value = Column(String(10000), nullable=False)

    field_desc = Column(String(1024), nullable=False)

    field_type = Column(String(255), nullable=False)

    created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)

    repository = relationship('Repository')

    @property

    def field_key_prefixed(self):

        return 'ex_%s' % self.field_key

    @classmethod

    def un_prefix_key(cls, key):

        if key.startswith(cls.PREFIX):

            return key[len(cls.PREFIX):]

        return key

    @classmethod

    def get_by_key_name(cls, key, repo):

        row = cls.query() \

                .filter(cls.repository == repo) \

                .filter(cls.field_key == key).scalar()

        return row

class Repository(Base, BaseDbModel):

    __tablename__ = 'repositories'

    __table_args__ = (

        Index('r_repo_name_idx', 'repo_name'),

        _table_args_default_dict,

    )

    DEFAULT_CLONE_URI = '{scheme}://{user}@{netloc}/{repo}'

    STATE_CREATED = u'repo_state_created'

    STATE_PENDING = u'repo_state_pending'

    STATE_ERROR = u'repo_state_error'

    repo_id = Column(Integer(), primary_key=True)

    repo_name = Column(Unicode(255), nullable=False, unique=True)

    repo_state = Column(String(255), nullable=False)

    clone_uri = Column(String(255), nullable=True) # FIXME: not nullable?

    repo_type = Column(String(255), nullable=False)

    owner_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)

    private = Column(Boolean(), nullable=False)

    enable_statistics = Column("statistics", Boolean(), nullable=False, default=True)

    enable_downloads = Column("downloads", Boolean(), nullable=False, default=True)

    description = Column(Unicode(10000), nullable=False)

    created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)

    updated_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)

    _landing_revision = Column("landing_revision", String(255), nullable=False)

    _changeset_cache = Column("changeset_cache", LargeBinary(), nullable=True) # JSON data # FIXME: not nullable?

    fork_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=True)

    group_id = Column(Integer(), ForeignKey('groups.group_id'), nullable=True)

    owner = relationship('User')

    fork = relationship('Repository', remote_side=repo_id)

    group = relationship('RepoGroup')

    repo_to_perm = relationship('UserRepoToPerm', cascade='all', order_by='UserRepoToPerm.repo_to_perm_id')

    users_group_to_perm = relationship('UserGroupRepoToPerm', cascade='all')

    stats = relationship('Statistics', cascade='all', uselist=False)

    followers = relationship('UserFollowing',

                             primaryjoin='UserFollowing.follows_repository_id==Repository.repo_id',

                             cascade='all')

    extra_fields = relationship('RepositoryField',

                                cascade="all, delete-orphan")

    logs = relationship('UserLog')

    comments = relationship('ChangesetComment', cascade="all, delete-orphan")

    pull_requests_org = relationship('PullRequest',

                    primaryjoin='PullRequest.org_repo_id==Repository.repo_id',

                    cascade="all, delete-orphan")

    pull_requests_other = relationship('PullRequest',

                    primaryjoin='PullRequest.other_repo_id==Repository.repo_id',

                    cascade="all, delete-orphan")

    def __unicode__(self):

        return u"<%s('%s:%s')>" % (self.__class__.__name__, self.repo_id,

                                   safe_unicode(self.repo_name))

    @hybrid_property

    def landing_rev(self):

        # always should return [rev_type, rev]

        if self._landing_revision:

            _rev_info = self._landing_revision.split(':')

            if len(_rev_info) < 2:

                _rev_info.insert(0, 'rev')

            return [_rev_info[0], _rev_info[1]]

        return [None, None]

    @landing_rev.setter

    def landing_rev(self, val):

        if ':' not in val:

            raise ValueError('value must be delimited with `:` and consist '

                             'of <rev_type>:<rev>, got %s instead' % val)

        self._landing_revision = val

    @hybrid_property

    def changeset_cache(self):

        try:

            cs_cache = json.loads(self._changeset_cache) # might raise on bad data

            cs_cache['raw_id'] # verify data, raise exception on error

            return cs_cache

        except (TypeError, KeyError, ValueError):

            return EmptyChangeset().__json__()

    @changeset_cache.setter

    def changeset_cache(self, val):

        try:

            self._changeset_cache = json.dumps(val)

        except Exception:

            log.error(traceback.format_exc())

    @classmethod

    def query(cls, sorted=False):

        """Add Repository-specific helpers for common query constructs.

        sorted: if True, apply the default ordering (name, case insensitive).

        """

        q = super(Repository, cls).query()

        if sorted:

            q = q.order_by(func.lower(Repository.repo_name))

        # into a valid system path

        p += self.repo_name.split(Repository.url_sep())

        return os.path.join(*map(safe_unicode, p))

    @property

    def cache_keys(self):

        """

        Returns associated cache keys for that repo

        """

        return CacheInvalidation.query() \

            .filter(CacheInvalidation.cache_args == self.repo_name) \

            .order_by(CacheInvalidation.cache_key) \

            .all()

    def get_new_name(self, repo_name):

        """

        returns new full repository name based on assigned group and new new

        :param group_name:

        """

        path_prefix = self.group.full_path_splitted if self.group else []

        return Repository.url_sep().join(path_prefix + [repo_name])

    @property

    def _ui(self):

        """

        Creates an db based ui object for this repository

        """

        from kallithea.lib.utils import make_ui

        return make_ui(clear_session=False)

    @classmethod

    def is_valid(cls, repo_name):

        """

        returns True if given repo name is a valid filesystem repository

        :param cls:

        :param repo_name:

        """

        from kallithea.lib.utils import is_valid_repo

        return is_valid_repo(repo_name, cls.base_path())

    def get_api_data(self, with_revision_names=False,

                           with_pullrequests=False):

        """

        Common function for generating repo api data.

        Optionally, also return tags, branches, bookmarks and PRs.

        """

        repo = self

        data = dict(

            repo_id=repo.repo_id,

            repo_name=repo.repo_name,

            repo_type=repo.repo_type,

            clone_uri=repo.clone_uri,

            private=repo.private,

            created_on=repo.created_on,

            description=repo.description,

            landing_rev=repo.landing_rev,

            owner=repo.owner.username,

            fork_of=repo.fork.repo_name if repo.fork else None,

            enable_statistics=repo.enable_statistics,

            enable_downloads=repo.enable_downloads,

            last_changeset=repo.changeset_cache,

        )

        if with_revision_names:

            scm_repo = repo.scm_instance_no_cache()

            data.update(dict(

                tags=scm_repo.tags,

                branches=scm_repo.branches,

                bookmarks=scm_repo.bookmarks,

            ))

        if with_pullrequests:

            data['pull_requests'] = repo.pull_requests_other

        rc_config = Setting.get_app_settings()

        repository_fields = str2bool(rc_config.get('repository_fields'))

        if repository_fields:

            for f in self.extra_fields:

                data[f.field_key_prefixed] = f.field_value

        return data

    @property

    def last_db_change(self):

        return self.updated_on

    @property

    def clone_uri_hidden(self):

        clone_uri = self.clone_uri

        if clone_uri:

            import urlobject

            url_obj = urlobject.URLObject(self.clone_uri)

            if url_obj.password:

                clone_uri = url_obj.with_password('*****')

        return clone_uri

        import kallithea.lib.helpers as h

        prefix_url = h.canonical_url('home')

        return get_clone_url(clone_uri_tmpl=clone_uri_tmpl,

                             prefix_url=prefix_url,

                             repo_name=self.repo_name,

                             repo_id=self.repo_id, **override)

    def set_state(self, state):

        self.repo_state = state

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

    # SCM PROPERTIES

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

    def get_changeset(self, rev=None):

        return get_changeset_safe(self.scm_instance, rev)

    def get_landing_changeset(self):

        """

        Returns landing changeset, or if that doesn't exist returns the tip

        """

        _rev_type, _rev = self.landing_rev

        cs = self.get_changeset(_rev)

        if isinstance(cs, EmptyChangeset):

            return self.get_changeset()

        return cs

    def update_changeset_cache(self, cs_cache=None):

        """

        Update cache of last changeset for repository, keys should be::

            short_id

            raw_id

            revision

            message

            date

            author

        :param cs_cache:

        """

        from kallithea.lib.vcs.backends.base import BaseChangeset

        if cs_cache is None:

            cs_cache = EmptyChangeset()

            # use no-cache version here

            scm_repo = self.scm_instance_no_cache()

            if scm_repo:

                cs_cache = scm_repo.get_changeset()

        if isinstance(cs_cache, BaseChangeset):

            cs_cache = cs_cache.__json__()

        if (not self.changeset_cache or cs_cache['raw_id'] != self.changeset_cache['raw_id']):

            _default = datetime.datetime.fromtimestamp(0)

            last_change = cs_cache.get('date') or _default

            log.debug('updated repo %s with new cs cache %s',

                      self.repo_name, cs_cache)

            self.updated_on = last_change

            self.changeset_cache = cs_cache

            Session().commit()

        else:

            log.debug('changeset_cache for %s already up to date with %s',

                      self.repo_name, cs_cache['raw_id'])

    @property

    def tip(self):

        return self.get_changeset('tip')

    @property

    def author(self):

        return self.tip.author

    @property

    def last_change(self):

        return self.scm_instance.last_change

    def get_comments(self, revisions=None):

        """

        Returns comments for this repository grouped by revisions

        :param revisions: filter query by revisions only

        """

        cmts = ChangesetComment.query() \

            .filter(ChangesetComment.repo == self)

        if revisions is not None:

            if not revisions:

                return {} # don't use sql 'in' on empty set

            cmts = cmts.filter(ChangesetComment.revision.in_(revisions))

        grouped = collections.defaultdict(list)

        for cmt in cmts.all():

            grouped[cmt.revision].append(cmt)

        return grouped

    def statuses(self, revisions):

        """