:created_on: Created on Nov 17, 2012

:author: marcink

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

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

"""

import logging

from kallithea import EXTERN_TYPE_INTERNAL

from kallithea.lib import auth_modules

from kallithea.lib.compat import formatted_json, hybrid_property

from kallithea.model.db import User

log = logging.getLogger(__name__)

class KallitheaAuthPlugin(auth_modules.KallitheaAuthPluginBase):

    def __init__(self):

        pass

    @hybrid_property

    def name(self):

        return EXTERN_TYPE_INTERNAL

    def settings(self):

        return []

    def user_activation_state(self):

        def_user_perms = User.get_default_user().AuthUser.permissions['global']

        return 'hg.register.auto_activate' in def_user_perms

    def accepts(self, user, accepts_empty=True):

        """

        Custom accepts for this auth that doesn't accept empty users. We

        know that user exists in database.

        """

        return super(KallitheaAuthPlugin, self).accepts(user,

                                                        accepts_empty=False)

    def auth(self, userobj, username, password, settings, **kwargs):

        if not userobj:

            log.debug('userobj was:%s skipping', userobj)

            return None

        if userobj.extern_type != self.name:

            log.warning("userobj:%s extern_type mismatch got:`%s` expected:`%s`",

                     userobj, userobj.extern_type, self.name)

            return None

        user_data = {

            "username": userobj.username,

            "firstname": userobj.firstname,

            "lastname": userobj.lastname,

            "groups": [],

            "email": userobj.email,

            "admin": userobj.admin,

            "active": userobj.active,

            "active_from_extern": userobj.active,

            "extern_name": userobj.user_id,

        }

        log.debug(formatted_json(user_data))

        if userobj.active:

            from kallithea.lib import auth

            password_match = auth.KallitheaCrypto.hash_check(password, userobj.password)

            if userobj.username == User.DEFAULT_USER and userobj.active:

                log.info('user %s authenticated correctly as anonymous user',

                         username)

                return user_data

            elif userobj.username == username and password_match:

                log.info('user %s authenticated correctly', user_data['username'])

                return user_data

            log.error("user %s had a bad password", username)

            return None

        else:

            log.warning('user %s tried auth but is disabled', username)

            return None

    def get_managed_fields(self):

        # Note: 'username' should only be editable (at least for user) if self registration is enabled

        return []

# Cache to store PAM authenticated users

_auth_cache = dict()

_pam_lock = threading.Lock()

class KallitheaAuthPlugin(auth_modules.KallitheaExternalAuthPlugin):

    # PAM authnetication can be slow. Repository operations involve a lot of

    # auth calls. Little caching helps speedup push/pull operations significantly

    AUTH_CACHE_TTL = 4

    def __init__(self):

        global _auth_cache

        ts = time.time()

        cleared_cache = dict(

            [(k, v) for (k, v) in _auth_cache.items() if

             (v + KallitheaAuthPlugin.AUTH_CACHE_TTL > ts)])

        _auth_cache = cleared_cache

    @hybrid_property

    def name(self):

        return "pam"

    def settings(self):

        settings = [

            {

                "name": "service",

                "validator": self.validators.UnicodeString(strip=True),

                "type": "string",

                "description": "PAM service name to use for authentication",

                "default": "login",

                "formname": "PAM service name"

            },

            {

                "name": "gecos",

                "validator": self.validators.UnicodeString(strip=True),

                "type": "string",

                "description": "Regex for extracting user name/email etc "

                               "from Unix userinfo",

                "default": "(?P<last_name>.+),\s*(?P<first_name>\w+)",

                "formname": "Gecos Regex"

            }

        ]

        return settings

    def use_fake_password(self):

        return True

    def auth(self, userobj, username, password, settings, **kwargs):

        if username not in _auth_cache:

            # Need lock here, as PAM authentication is not thread safe

            _pam_lock.acquire()

            try:

                auth_result = pam.authenticate(username, password,

                                               settings["service"])

                # cache result only if we properly authenticated

                if auth_result:

                    _auth_cache[username] = time.time()

            finally:

                _pam_lock.release()

            if not auth_result:

                log.error("PAM was unable to authenticate user: %s", username)

                return None

        else:

            log.debug("Using cached auth for user: %s", username)

        # old attrs fetched from Kallithea database

        admin = getattr(userobj, 'admin', False)

        active = getattr(userobj, 'active', True)

        email = getattr(userobj, 'email', '') or "%s@%s" % (username, socket.gethostname())

        firstname = getattr(userobj, 'firstname', '')

        lastname = getattr(userobj, 'lastname', '')

        user_data = {

            'username': username,

            'firstname': firstname,

            'lastname': lastname,

            'groups': [g.gr_name for g in grp.getgrall() if username in g.gr_mem],

            'email': email,

            'admin': admin,

            'active': active,

            "active_from_extern": None,

            'extern_name': username,

        }

        try:

            user_pw_data = pwd.getpwnam(username)

            regex = settings["gecos"]

            match = re.search(regex, user_pw_data.pw_gecos)

            if match:

                user_data["firstname"] = match.group('first_name')

                user_data["lastname"] = match.group('last_name')

        except Exception:

            log.warning("Cannot extract additional info for PAM user %s", username)

            pass

            raise VCSError("Given object should be FileNode object, not %s"

                % filenode.__class__)

    repo = filenode_new.changeset.repository

    old_raw_id = getattr(filenode_old.changeset, 'raw_id', repo.EMPTY_CHANGESET)

    new_raw_id = getattr(filenode_new.changeset, 'raw_id', repo.EMPTY_CHANGESET)

    vcs_gitdiff = repo.get_diff(old_raw_id, new_raw_id, filenode_new.path,

                                ignore_whitespace, context)

    return vcs_gitdiff

NEW_FILENODE = 1

DEL_FILENODE = 2

MOD_FILENODE = 3

RENAMED_FILENODE = 4

COPIED_FILENODE = 5

CHMOD_FILENODE = 6

BIN_FILENODE = 7

class DiffLimitExceeded(Exception):

    pass

class LimitedDiffContainer(object):

    def __init__(self, diff_limit, cur_diff_size, diff):

        self.diff = diff

        self.diff_limit = diff_limit

        self.cur_diff_size = cur_diff_size

    def __iter__(self):

        for l in self.diff:

            yield l

class DiffProcessor(object):

    """

    Give it a unified or git diff and it returns a list of the files that were

    mentioned in the diff together with a dict of meta information that

    can be used to render it in a HTML template.

    """

    _chunk_re = re.compile(r'^@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@(.*)')

    _newline_marker = re.compile(r'^\\ No newline at end of file')

    _git_header_re = re.compile(r"""

        # has already been split on this:

        # ^diff[ ]--git

            [ ]a/(?P<a_path>.+?)[ ]b/(?P<b_path>.+?)\n

        (?:^similarity[ ]index[ ](?P<similarity_index>\d+)%\n

           ^rename[ ]from[ ](?P<rename_from>.+)\n

           ^rename[ ]to[ ](?P<rename_to>.+)(?:\n|$))?

        (?:^new[ ]file[ ]mode[ ](?P<new_file_mode>.+)(?:\n|$))?

        (?:^deleted[ ]file[ ]mode[ ](?P<deleted_file_mode>.+)(?:\n|$))?

        (?:^index[ ](?P<a_blob_id>[0-9A-Fa-f]+)

            \.\.(?P<b_blob_id>[0-9A-Fa-f]+)[ ]?(?P<b_mode>.+)?(?:\n|$))?

        (?:^(?P<bin_patch>GIT[ ]binary[ ]patch)(?:\n|$))?

        (?:^---[ ](a/(?P<a_file>.+?)|/dev/null)\t?(?:\n|$))?

        (?:^\+\+\+[ ](b/(?P<b_file>.+?)|/dev/null)\t?(?:\n|$))?

    """, re.VERBOSE | re.MULTILINE)

    _hg_header_re = re.compile(r"""

        # has already been split on this:

        # ^diff[ ]--git

            [ ]a/(?P<a_path>.+?)[ ]b/(?P<b_path>.+?)\n

        (?:^old[ ]mode[ ](?P<old_mode>\d+)\n

           ^new[ ]mode[ ](?P<new_mode>\d+)(?:\n|$))?

        (?:^similarity[ ]index[ ](?P<similarity_index>\d+)%(?:\n|$))?

        (?:^rename[ ]from[ ](?P<rename_from>.+)\n

           ^rename[ ]to[ ](?P<rename_to>.+)(?:\n|$))?

        (?:^copy[ ]from[ ](?P<copy_from>.+)\n

           ^copy[ ]to[ ](?P<copy_to>.+)(?:\n|$))?

        (?:^new[ ]file[ ]mode[ ](?P<new_file_mode>.+)(?:\n|$))?

        (?:^deleted[ ]file[ ]mode[ ](?P<deleted_file_mode>.+)(?:\n|$))?

        (?:^index[ ](?P<a_blob_id>[0-9A-Fa-f]+)

            \.\.(?P<b_blob_id>[0-9A-Fa-f]+)[ ]?(?P<b_mode>.+)?(?:\n|$))?

        (?:^(?P<bin_patch>GIT[ ]binary[ ]patch)(?:\n|$))?

        (?:^---[ ](a/(?P<a_file>.+?)|/dev/null)\t?(?:\n|$))?

        (?:^\+\+\+[ ](b/(?P<b_file>.+?)|/dev/null)\t?(?:\n|$))?

    """, re.VERBOSE | re.MULTILINE)

    # Used for inline highlighter word split, must match the substitutions in _escaper

    _token_re = re.compile(r'()(&amp;|&lt;|&gt;|<u>\t</u>|<u class="cr"></u>| <i></i>|\W+?)')

    _escape_re = re.compile(r'(&)|(<)|(>)|(\t)|(\r)|(?<=.)( \n| $)')

    def __init__(self, diff, vcs='hg', format='gitdiff', diff_limit=None):

        """

        :param diff:   a text in diff format

        :param vcs: type of version control hg or git

        :param format: format of diff passed, `udiff` or `gitdiff`

        :param diff_limit: define the size of diff that is considered "big"

            based on that parameter cut off will be triggered, set to None

            to show full diff

        """

        if not isinstance(diff, basestring):

            raise Exception('Diff must be a basestring got %s instead' % type(diff))

        self._diff = diff

        self._format = format

        based on repository walk function

        """

        index_paths_ = set()

        try:

            cs = self._get_index_changeset(repo)

            for _topnode, _dirs, files in cs.walk('/'):

                for f in files:

                    index_paths_.add(jn(safe_str(repo.path), safe_str(f.path)))

        except RepositoryError:

            log.debug(traceback.format_exc())

            pass

        return index_paths_

    def get_node(self, repo, path, index_rev=None):

        """

        gets a filenode based on given full path. It operates on string for

        hg git compatibility.

        :param repo: scm repo instance

        :param path: full path including root location

        :return: FileNode

        """

        # FIXME: paths should be normalized ... or even better: don't include repo.path

        path = safe_str(path)

        repo_path = safe_str(repo.path)

        assert path.startswith(repo_path)

        assert path[len(repo_path)] in (os.path.sep, os.path.altsep)

        node_path = path[len(repo_path) + 1:]

        cs = self._get_index_changeset(repo, index_rev=index_rev)

        node = cs.get_node(node_path)

        return node

    def is_indexable_node(self, node):

        """

        Just index the content of chosen files, skipping binary files

        """

        return (node.extension in INDEX_EXTENSIONS or node.name in INDEX_FILENAMES) and \

               not node.is_binary

    def get_node_mtime(self, node):

        return mktime(node.last_changeset.date.timetuple())

    def add_doc(self, writer, path, repo, repo_name, index_rev=None):

        """

        Adding doc to writer this function itself fetches data from

        the instance of vcs backend

        """

        indexed = indexed_w_content = 0

        if self.is_indexable_node(node):

            u_content = node.content

            if not isinstance(u_content, unicode):

                log.warning('  >> %s Could not get this content as unicode '

                            'replacing with empty content' % path)

                u_content = u''

            else:

                log.debug('    >> %s [WITH CONTENT]', path)

                indexed_w_content += 1

        else:

            log.debug('    >> %s', path)

            # just index file name without it's content

            u_content = u''

            indexed += 1

        p = safe_unicode(path)

        writer.add_document(

            fileid=p,

            owner=unicode(repo.contact),

            repository=safe_unicode(repo_name),

            path=p,

            content=u_content,

            modtime=self.get_node_mtime(node),

            extension=node.extension

        )

        return indexed, indexed_w_content

    def index_changesets(self, writer, repo_name, repo, start_rev=None):

        """

        Add all changeset in the vcs repo starting at start_rev

        to the index writer

        :param writer: the whoosh index writer to add to

        :param repo_name: name of the repository from whence the

          changeset originates including the repository group

        :param repo: the vcs repository instance to index changesets for,

          the presumption is the repo has changesets to index

        :param start_rev=None: the full sha id to start indexing from

          if start_rev is None then index from the first changeset in

          the repo

        """

        if start_rev is None:

            start_rev = repo[0].raw_id

        log.debug('indexing changesets in %s starting at rev: %s',

        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 kallithea

        prefix = kallithea.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()

        log.debug('for repo %s got %s invalidation objects',

                  safe_str(repo_name), inv_objs)

        for inv_obj in inv_objs:

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

                      inv_obj, safe_str(repo_name))

            Session().delete(inv_obj)

        Session().commit()

    @classmethod

    def test_and_set_valid(cls, repo_name, kind, valid_cache_keys=None):

        """

        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.

        """

        key = (repo_name + '_' + kind) if kind else repo_name

        cache_key = cls._get_cache_key(key)

        if valid_cache_keys and cache_key in valid_cache_keys:

            return True

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

        if not inv_obj:

            inv_obj = CacheInvalidation(cache_key, repo_name)

        if inv_obj.cache_active:

            return True

        inv_obj.cache_active = True

        Session().add(inv_obj)

        Session().commit()

        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'),

        Index('cc_pull_request_id_idx', 'pull_request_id'),

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

         'mysql_charset': 'utf8', 'sqlite_autoincrement': True},

    )

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

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

    revision = Column(String(40))

    pull_request_id = Column(Integer(), ForeignKey('pull_requests.pull_request_id'))

    line_no = Column(Unicode(10))

    f_path = Column(Unicode(1000))

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

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

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

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

    author = relationship('User')

    repo = relationship('Repository')

    # status_change is frequently used directly in templates - make it a lazy

    # join to avoid fetching each related ChangesetStatus on demand.

    # There will only be one ChangesetStatus referencing each comment so the join will not explode.

    status_change = relationship('ChangesetStatus',

                                 cascade="all, delete-orphan", lazy='joined')

    pull_request = relationship('PullRequest')

    @classmethod

    def get_users(cls, revision=None, pull_request_id=None):

        """

        Returns user associated with this ChangesetComment. ie those

        who actually commented

        :param cls:

        :param revision:

        """

            if (a.text > b.text) {

                return 1;

            }

            if (a.text < b.text) {

                return -1;

            }

            return 0;

        });

    }

    return results;

};

var prefixFirstSort = function(results, container, query) {

    if (query.term) {

        return results.sort(function (a, b) {

            // if parent node, no sorting

            if (a.children != undefined || b.children != undefined) {

                return 0;

            }

            // Put prefix matches before matches in the line

            var aPos = a.text.toLowerCase().indexOf(query.term.toLowerCase()),

                bPos = b.text.toLowerCase().indexOf(query.term.toLowerCase());

            if (aPos === 0 && bPos !== 0) {

                return -1;

            }

            if (bPos === 0 && aPos !== 0) {

                return 1;

            }

            // Default sorting

            if (a.text > b.text) {

                return 1;

            }

            if (a.text < b.text) {

                return -1;

            }

            return 0;

        });

    }

    return results;

};

// global hooks after DOM is loaded

$(document).ready(function(){

    $('.diff-collapse-button').click(function(e) {

        var $button = $(e.currentTarget);

        if($target.hasClass('hidden')){

            $target.removeClass('hidden');

            $button.html("↑ {0} ↑".format(_TM['Collapse Diff']));

        }

        else if(!$target.hasClass('hidden')){

            $target.addClass('hidden');

            $button.html("↓ {0} ↓".format(_TM['Expand Diff']));

        }

    });

});

diff --git a/work-horus.xls b/file.xls

similarity index 100%

rename from work-horus.xls

rename to file.xls

         {'added': 0,

          'deleted': 0,

          'binary': True,

          'ops': {DEL_FILENODE: 'deleted file',

                  BIN_FILENODE: 'binary diff not shown'}}),

        ('index.html', 'M',

         {'added': 3,

          'deleted': 2,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

        ('js/global.js', 'D',

         {'added': 0,

          'deleted': 75,

          'binary': False,

          'ops': {DEL_FILENODE: 'deleted file'}}),

        ('js/jquery/hashgrid.js', 'A',

         {'added': 340,

          'deleted': 0,

          'binary': False,

          'ops': {NEW_FILENODE: 'new file 100755'}}),

        ('less/docs.less', 'M',

         {'added': 34,

          'deleted': 0,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

        ('less/scaffolding.less', 'M',

         {'added': 1,

          'deleted': 3,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

        ('readme.markdown', 'M',

         {'added': 1,

          'deleted': 10,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

    ],

    'git_diff_chmod.diff': [

        ('work-horus.xls', 'M',

         {'added': 0,

          'deleted': 0,

          'binary': True,

          'ops': {CHMOD_FILENODE: 'modified file chmod 100644 => 100755'}})

    ],

    'git_diff_rename_file.diff': [

        ('file.xls', 'R',

         {'added': 0,

          'deleted': 0,

          'binary': True,

    ],

    'git_diff_mod_single_binary_file.diff': [

        ('US Warszawa.jpg', 'M',

         {'added': 0,

          'deleted': 0,

          'binary': True,

          'ops': {MOD_FILENODE: 'modified file',

                  BIN_FILENODE: 'binary diff not shown'}})

    ],

    'git_diff_binary_and_normal.diff': [

        ('img/baseline-10px.png', 'A',

         {'added': 0,

          'deleted': 0,

          'binary': True,

          'ops': {NEW_FILENODE: 'new file 100644',

                  BIN_FILENODE: 'binary diff not shown'}}),

        ('img/baseline-20px.png', 'D',

         {'added': 0,

          'deleted': 0,

          'binary': True,

          'ops': {DEL_FILENODE: 'deleted file',

                  BIN_FILENODE: 'binary diff not shown'}}),

        ('index.html', 'M',

         {'added': 3,

          'deleted': 2,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

        ('js/global.js', 'D',

         {'added': 0,

          'deleted': 75,

          'binary': False,

          'ops': {DEL_FILENODE: 'deleted file'}}),

        ('js/jquery/hashgrid.js', 'A',

         {'added': 340,

          'deleted': 0,

          'binary': False,

          'ops': {NEW_FILENODE: 'new file 100755'}}),

        ('less/docs.less', 'M',

         {'added': 34,

          'deleted': 0,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

        ('less/scaffolding.less', 'M',

         {'added': 1,

          'deleted': 3,

          'binary': False,

          'ops': {MOD_FILENODE: 'modified file'}}),

        ('readme.markdown', 'M',