_parsed = diff_processor.prepare(inline_diff=False)

        limited_diff = False

        if isinstance(_parsed, LimitedDiffContainer):

            limited_diff = True

        for st in _parsed:

            st.update({'added': st['stats'][0],

                       'removed': st['stats'][1]})

            changes.append('\n %(operation)s %(filename)s '

                           '(%(added)s lines added, %(removed)s lines removed)'

                            % st)

        if limited_diff:

            changes = changes + ['\n ' +

                                 _('Changeset was too big and was cut off...')]

        return diff_processor, changes

    def __get_desc(self, cs):

        desc_msg = []

        desc_msg.append((_('%s committed on %s')

                         % (h.person(cs.author), h.fmt_date(cs.date))) + '<br/>')

        #branches, tags, bookmarks

        if cs.branch:

            desc_msg.append('branch: %s<br/>' % cs.branch)

        if h.is_hg(c.rhodecode_repo):

            for book in cs.bookmarks:

                desc_msg.append('bookmark: %s<br/>' % book)

        for tag in cs.tags:

            desc_msg.append('tag: %s<br/>' % tag)

        diff_processor, changes = self.__changes(cs)

        # rev link

        _url = url('changeset_home', repo_name=cs.repository.name,

                   revision=cs.raw_id, qualified=True)

        desc_msg.append('changeset: <a href="%s">%s</a>' % (_url, cs.raw_id[:8]))

        desc_msg.append('<pre>')

        desc_msg.append(cs.message)

        desc_msg.append('\n')

        desc_msg.extend(changes)

        if self.include_diff:

            desc_msg.append('\n\n')

            desc_msg.append(diff_processor.as_raw())

        desc_msg.append('</pre>')

        return map(safe_unicode, desc_msg)

    def atom(self, repo_name):

        """Produce an atom-1.0 feed via feedgenerator module"""

        @cache_region('long_term')

            feed = Atom1Feed(

                 title=self.title % repo_name,

                 link=url('summary_home', repo_name=repo_name,

                          qualified=True),

                 description=self.description % repo_name,

                 language=self.language,

                 ttl=self.ttl

            )

            for cs in reversed(list(c.rhodecode_repo[-self.feed_nr:])):

                feed.add_item(title=self._get_title(cs),

                              link=url('changeset_home', repo_name=repo_name,

                                       revision=cs.raw_id, qualified=True),

                              author_name=cs.author,

                              description=''.join(self.__get_desc(cs)),

                              pubdate=cs.date,

                              )

            response.content_type = feed.mime_type

            return feed.writeString('utf-8')

        if not valid:

    def rss(self, repo_name):

        """Produce an rss2 feed via feedgenerator module"""

        @cache_region('long_term')

            feed = Rss201rev2Feed(

                title=self.title % repo_name,

                link=url('summary_home', repo_name=repo_name,

                         qualified=True),

                description=self.description % repo_name,

                language=self.language,

                ttl=self.ttl

            )

            for cs in reversed(list(c.rhodecode_repo[-self.feed_nr:])):

                feed.add_item(title=self._get_title(cs),

                              link=url('changeset_home', repo_name=repo_name,

                                       revision=cs.raw_id, qualified=True),

                              author_name=cs.author,

                              description=''.join(self.__get_desc(cs)),

                              pubdate=cs.date,

                             )

            response.content_type = feed.mime_type

            return feed.writeString('utf-8')

        if not valid:

from rhodecode.lib.utils2 import safe_unicode, safe_str

from rhodecode.lib.auth import LoginRequired, HasRepoPermissionAnyDecorator,\

    NotAnonymous

from rhodecode.lib.base import BaseRepoController, render

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

from rhodecode.lib.markup_renderer import MarkupRenderer

from rhodecode.lib.celerylib import run_task

from rhodecode.lib.celerylib.tasks import get_commits_stats

from rhodecode.lib.helpers import RepoPage

from rhodecode.lib.compat import json, OrderedDict

from rhodecode.lib.vcs.nodes import FileNode

from rhodecode.controllers.changelog import _load_changelog_summary

log = logging.getLogger(__name__)

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

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

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

class SummaryController(BaseRepoController):

    def __before__(self):

        super(SummaryController, self).__before__()

    def _get_download_links(self, repo):

        download_l = []

        branches_group = ([], _("Branches"))

        tags_group = ([], _("Tags"))

        for name, chs in c.rhodecode_repo.branches.items():

            #chs = chs.split(':')[-1]

            branches_group[0].append((chs, name),)

        download_l.append(branches_group)

        for name, chs in c.rhodecode_repo.tags.items():

            #chs = chs.split(':')[-1]

            tags_group[0].append((chs, name),)

        download_l.append(tags_group)

        return download_l

    def __get_readme_data(self, db_repo):

        repo_name = db_repo.repo_name

        @cache_region('long_term')

            readme_data = None

            readme_file = None

            log.debug('Looking for README file')

            try:

                # get's 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, f)

                        break

                    except NodeDoesNotExistError:

                        continue

            except ChangesetError:

                log.error(traceback.format_exc())

                pass

            except EmptyRepositoryError:

                pass

            except Exception:

                log.error(traceback.format_exc())

            return readme_data, readme_file

        if not valid:

    @LoginRequired()

    @HasRepoPermissionAnyDecorator('repository.read', 'repository.write',

                                   'repository.admin')

    def index(self, repo_name):

        c.dbrepo = dbrepo = c.rhodecode_db_repo

        _load_changelog_summary()

        if self.rhodecode_user.username == 'default':

            # for default(anonymous) user we don't need to pass credentials

            username = ''

            password = ''

        else:

            username = str(self.rhodecode_user.username)

            password = '@'

        parsed_url = urlparse(url.current(qualified=True))

        default_clone_uri = '{scheme}://{user}{pass}{netloc}{path}'

        uri_tmpl = config.get('clone_uri', default_clone_uri)

        uri_tmpl = uri_tmpl.replace('{', '%(').replace('}', ')s')

        decoded_path = safe_unicode(urllib.unquote(parsed_url.path))

        uri_dict = {

           'user': urllib.quote(username),

           'pass': password,

           'scheme': parsed_url.scheme,

           'netloc': parsed_url.netloc,

           'path': urllib.quote(safe_str(decoded_path))

        }

        uri = (uri_tmpl % uri_dict)

        # generate another clone url by id

        uri_dict.update(

         {'path': decoded_path.replace(repo_name, '_%s' % c.dbrepo.repo_id)}

        )

        uri_id = uri_tmpl % uri_dict

        c.clone_repo_url = uri

        c.clone_repo_url_id = uri_id

        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())

    from rhodecode.model.scm import ScmModel

    sa = meta.Session()

    rm = RepoModel()

    user = User.get_first_admin()

    added = []

    ##creation defaults

    defs = RhodeCodeSetting.get_default_repo_settings(strip_prefix=True)

    enable_statistics = defs.get('repo_enable_statistics')

    enable_locking = defs.get('repo_enable_locking')

    enable_downloads = defs.get('repo_enable_downloads')

    private = defs.get('repo_private')

    for name, repo in initial_repo_list.items():

        group = map_groups(name)

        db_repo = rm.get_by_repo_name(name)

        # found repo that is on filesystem not in RhodeCode database

        if not db_repo:

            log.info('repository %s not found, creating now' % name)

            added.append(name)

            desc = (repo.description

                    if repo.description != 'unknown'

                    else '%s repository' % name)

            new_repo = rm.create_repo(

                repo_name=name,

                repo_type=repo.alias,

                description=desc,

                repos_group=getattr(group, 'group_id', None),

                owner=user,

                just_db=True,

                enable_locking=enable_locking,

                enable_downloads=enable_downloads,

                enable_statistics=enable_statistics,

                private=private

            )

            # we added that repo just now, and make sure it has githook

            # installed

            if new_repo.repo_type == 'git':

                ScmModel().install_git_hook(new_repo.scm_instance)

            new_repo.update_changeset_cache()

        elif install_git_hook:

            if db_repo.repo_type == 'git':

                ScmModel().install_git_hook(db_repo.scm_instance)

        # during starting install all cache keys for all repositories in the

        # system, this will register all repos and multiple instances

        cache_key = CacheInvalidation._get_cache_key(name)

        log.debug("Creating invalidation cache key for %s: %s", name, cache_key)

    sa.commit()

    removed = []

    if remove_obsolete:

        # remove from database those repositories that are not in the filesystem

        for repo in sa.query(Repository).all():

            if repo.repo_name not in initial_repo_list.keys():

                log.debug("Removing non-existing repository found in db `%s`" %

                          repo.repo_name)

                try:

                    removed.append(repo.repo_name)

                    RepoModel(sa).delete(repo, forks='detach', fs_remove=False)

                    sa.commit()

                except Exception:

                    #don't hold further removals on error

                    log.error(traceback.format_exc())

                    sa.rollback()

    return added, removed

# set cache regions for beaker so celery can utilise it

def add_cache(settings):

    cache_settings = {'regions': None}

    for key in settings.keys():

        for prefix in ['beaker.cache.', 'cache.']:

            if key.startswith(prefix):

                name = key.split(prefix)[1].strip()

                cache_settings[name] = settings[key].strip()

    if cache_settings['regions']:

        for region in cache_settings['regions'].split(','):

            region = region.strip()

            region_settings = {}

            for key, value in cache_settings.items():

                if key.startswith(region):

                    region_settings[key.split('.')[1]] = value

            region_settings['expire'] = int(region_settings.get('expire',

                                                                60))

            region_settings.setdefault('lock_dir',

                                       cache_settings.get('lock_dir'))

            region_settings.setdefault('data_dir',

                                       cache_settings.get('data_dir'))

            if 'type' not in region_settings:

                region_settings['type'] = cache_settings.get('type',

                                                             'memory')

            beaker.cache.cache_regions[region] = region_settings

        status_lbl = ChangesetStatus.get_status_lbl(stat)

        for pr in PullRequest.query().filter(PullRequest.org_repo == self).all():

            for rev in pr.revisions:

                pr_id = pr.pull_request_id

                pr_repo = pr.other_repo.repo_name

                grouped[rev] = [stat, status_lbl, pr_id, pr_repo]

        for stat in statuses.all():

            pr_id = pr_repo = None

            if stat.pull_request:

                pr_id = stat.pull_request.pull_request_id

                pr_repo = stat.pull_request.other_repo.repo_name

            grouped[stat.revision] = [str(stat.status), stat.status_lbl,

                                      pr_id, pr_repo]

        return grouped

    def _repo_size(self):

        from rhodecode.lib import helpers as h

        log.debug('calculating repository size...')

        return h.format_byte_size(self.scm_instance.size)

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

    # SCM CACHE INSTANCE

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

    def set_invalidate(self):

        """

        Mark caches of this repo as invalid.

        """

        CacheInvalidation.set_invalidate(self.repo_name)

    def scm_instance_no_cache(self):

        return self.__get_instance()

    @property

    def scm_instance(self):

        import rhodecode

        full_cache = str2bool(rhodecode.CONFIG.get('vcs_full_cache'))

        if full_cache:

            return self.scm_instance_cached()

        return self.__get_instance()

    def scm_instance_cached(self, valid_cache_keys=None):

        @cache_region('long_term')

        def _c(repo_name):

            return self.__get_instance()

        rn = self.repo_name

        if not valid:

            log.debug('Cache for %s invalidated, getting new object' % (rn))

            region_invalidate(_c, None, rn)

        else:

            log.debug('Getting obj for %s from cache' % (rn))

        return _c(rn)

    def __get_instance(self):

        repo_full_path = self.repo_full_path

        try:

            alias = get_scm(repo_full_path)[0]

            log.debug('Creating instance of %s repository from %s'

                      % (alias, repo_full_path))

            backend = get_backend(alias)

        except VCSError:

            log.error(traceback.format_exc())

            log.error('Perhaps this repository is in db and not in '

                      'filesystem run rescan repositories with '

                      '"destroy old data " option from admin panel')

            return

        if alias == 'hg':

            repo = backend(safe_str(repo_full_path), create=False,

                           baseui=self._ui)

            # skip hidden web repository

            if repo._get_hidden():

                return

        else:

            repo = backend(repo_full_path, create=False)

        return repo

class RepoGroup(Base, BaseModel):

    __tablename__ = 'groups'

    __table_args__ = (

        UniqueConstraint('group_name', 'group_parent_id'),

        CheckConstraint('group_id != group_parent_id'),

        {'extend_existing': True, 'mysql_engine': 'InnoDB',

         'mysql_charset': 'utf8'},

    )

    __mapper_args__ = {'order_by': 'group_name'}

    group_id = Column("group_id", Integer(), nullable=False, unique=True, default=None, primary_key=True)

    group_name = Column("group_name", String(255, convert_unicode=False, assert_unicode=None), nullable=False, unique=True, default=None)

    group_parent_id = Column("group_parent_id", Integer(), ForeignKey('groups.group_id'), nullable=True, unique=None, default=None)

    group_description = Column("group_description", String(10000, convert_unicode=False, assert_unicode=None), nullable=True, unique=None, default=None)

    def get_suffix(self):

        """

        get suffix that might have been used in _get_cache_key to

        generate self.cache_key. Only used for informational purposes

        in repo_edit.html.

        """

        # prefix, repo_name, suffix

        return self._cache_key_partition()[2]

    @classmethod

    def clear_cache(cls):

        """

        Delete all cache keys from database.

        Should only be run when all instances are down and all entries thus stale.

        """

        cls.query().delete()

        Session().commit()

    @classmethod

    def _get_cache_key(cls, key):

        """

        Wrapper for generating a unique cache key for this instance and "key".

        key must / will start with a repo_name which will be stored in .cache_args .

        """

        import rhodecode

        prefix = rhodecode.CONFIG.get('instance_id', '')

        return "%s%s" % (prefix, key)

    @classmethod

    def set_invalidate(cls, repo_name):

        """

        Mark all caches of a repo as invalid in the database.

        """

        inv_objs = Session().query(cls).filter(cls.cache_args == repo_name).all()

        try:

            for inv_obj in inv_objs:

                log.debug('marking %s key for invalidation based on repo_name=%s'

                          % (inv_obj, safe_str(repo_name)))

                inv_obj.cache_active = False

                Session().add(inv_obj)

            Session().commit()

        except Exception:

            log.error(traceback.format_exc())

            Session().rollback()

    @classmethod

        """

        Mark this cache key as active and currently cached.

        Return True if the existing cache registration still was valid.

        Return False to indicate that it had been invalidated and caches should be refreshed.

        """

        cache_key = cls._get_cache_key(key)

        if valid_cache_keys and cache_key in valid_cache_keys:

            return True

        try:

            inv_obj = cls.query().filter(cls.cache_key == cache_key).scalar()

            if not inv_obj:

                inv_obj = CacheInvalidation(cache_key, repo_name)

            was_valid = inv_obj.cache_active

            inv_obj.cache_active = True

            Session().add(inv_obj)

            Session().commit()

            return was_valid

        except Exception:

            log.error(traceback.format_exc())

            Session().rollback()

            return False

    @classmethod

    def get_valid_cache_keys(cls):

        """

        Return opaque object with information of which caches still are valid

        and can be used without checking for invalidation.

        """

        return set(inv_obj.cache_key for inv_obj in cls.query().filter(cls.cache_active).all())

class ChangesetComment(Base, BaseModel):

    __tablename__ = 'changeset_comments'

    __table_args__ = (

        Index('cc_revision_idx', 'revision'),

        {'extend_existing': True, 'mysql_engine': 'InnoDB',

         'mysql_charset': 'utf8'},

    )

    comment_id = Column('comment_id', Integer(), nullable=False, primary_key=True)

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

    revision = Column('revision', String(40), nullable=True)

    pull_request_id = Column("pull_request_id", Integer(), ForeignKey('pull_requests.pull_request_id'), nullable=True)

    line_no = Column('line_no', Unicode(10), nullable=True)

    hl_lines = Column('hl_lines', Unicode(512), nullable=True)

    f_path = Column('f_path', Unicode(1000), nullable=True)

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

    text = Column('text', UnicodeText(25000), nullable=False)

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

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

    author = relationship('User', lazy='joined')

    repo = relationship('Repository')

    status_change = relationship('ChangesetStatus', cascade="all, delete, delete-orphan")

    pull_request = relationship('PullRequest', lazy='joined')

    @classmethod