Changeset - 1112e440b921
[Not reviewed]
default
0 13 0
Mads Kiilerich - 6 years ago 2019-12-28 19:59:15
mads@kiilerich.com
Grafted from: 736eece1b4ed
py3: add safe_str where we really need it to get a str - probably from bytes
13 files changed with 45 insertions and 44 deletions:
0 comments (0 inline, 0 general)
kallithea/config/app_cfg.py
Show inline comments
 
# -*- coding: utf-8 -*-
 
# 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/>.
 
"""
 
Global configuration file for TurboGears2 specific settings in Kallithea.
 

	
 
This file complements the .ini file.
 
"""
 

	
 
import logging
 
import os
 
import platform
 
import sys
 

	
 
import alembic.config
 
import mercurial
 
import tg
 
from alembic.migration import MigrationContext
 
from alembic.script.base import ScriptDirectory
 
from sqlalchemy import create_engine
 
from tg.configuration import AppConfig
 
from tg.support.converters import asbool
 

	
 
import kallithea.lib.locale
 
import kallithea.model.base
 
import kallithea.model.meta
 
from kallithea.lib.middleware.https_fixup import HttpsFixup
 
from kallithea.lib.middleware.permanent_repo_url import PermanentRepoUrl
 
from kallithea.lib.middleware.simplegit import SimpleGit
 
from kallithea.lib.middleware.simplehg import SimpleHg
 
from kallithea.lib.middleware.wrapper import RequestWrapper
 
from kallithea.lib.utils import check_git_version, load_rcextensions, make_ui, set_app_settings, set_indexer_config, set_vcs_config
 
from kallithea.lib.utils2 import str2bool
 
from kallithea.lib.utils2 import safe_str, str2bool
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class KallitheaAppConfig(AppConfig):
 
    # Note: AppConfig has a misleading name, as it's not the application
 
    # configuration, but the application configurator. The AppConfig values are
 
    # used as a template to create the actual configuration, which might
 
    # overwrite or extend the one provided by the configurator template.
 

	
 
    # To make it clear, AppConfig creates the config and sets into it the same
 
    # values that AppConfig itself has. Then the values from the config file and
 
    # gearbox options are loaded and merged into the configuration. Then an
 
    # after_init_config(conf) method of AppConfig is called for any change that
 
    # might depend on options provided by configuration files.
 

	
 
    def __init__(self):
 
        super(KallitheaAppConfig, self).__init__()
 

	
 
        self['package'] = kallithea
 

	
 
        self['prefer_toscawidgets2'] = False
 
        self['use_toscawidgets'] = False
 

	
 
        self['renderers'] = []
 

	
 
        # Enable json in expose
 
        self['renderers'].append('json')
 

	
 
        # Configure template rendering
 
        self['renderers'].append('mako')
 
        self['default_renderer'] = 'mako'
 
        self['use_dotted_templatenames'] = False
 

	
 
        # Configure Sessions, store data as JSON to avoid pickle security issues
 
        self['session.enabled'] = True
 
        self['session.data_serializer'] = 'json'
 

	
 
        # Configure the base SQLALchemy Setup
 
        self['use_sqlalchemy'] = True
 
        self['model'] = kallithea.model.base
 
        self['DBSession'] = kallithea.model.meta.Session
 

	
 
        # Configure App without an authentication backend.
 
        self['auth_backend'] = None
 

	
 
        # Use custom error page for these errors. By default, Turbogears2 does not add
 
        # 400 in this list.
 
        # Explicitly listing all is considered more robust than appending to defaults,
 
        # in light of possible future framework changes.
 
        self['errorpage.status_codes'] = [400, 401, 403, 404]
 

	
 
        # Disable transaction manager -- currently Kallithea takes care of transactions itself
 
        self['tm.enabled'] = False
 

	
 
        # Set the i18n source language so TG doesn't search beyond 'en' in Accept-Language.
 
        # Don't force the default here if configuration force something else.
 
        if not self.get('i18n.lang'):
 
            self['i18n.lang'] = 'en'
 

	
 

	
 
base_config = KallitheaAppConfig()
 

	
 
# DebugBar, a debug toolbar for TurboGears2.
 
# (https://github.com/TurboGears/tgext.debugbar)
 
# To enable it, install 'tgext.debugbar' and 'kajiki', and run Kallithea with
 
# 'debug = true' (not in production!)
 
# See the Kallithea documentation for more information.
 
try:
 
    from tgext.debugbar import enable_debugbar
 
    import kajiki # only to check its existence
 
except ImportError:
 
    pass
 
else:
 
    base_config['renderers'].append('kajiki')
 
    enable_debugbar(base_config)
 

	
 

	
 
def setup_configuration(app):
 
    config = app.config
 

	
 
    if not kallithea.lib.locale.current_locale_is_valid():
 
        log.error("Terminating ...")
 
        sys.exit(1)
 

	
 
    # Mercurial sets encoding at module import time, so we have to monkey patch it
 
    hgencoding = config.get('hgencoding')
 
    if hgencoding:
 
        mercurial.encoding.encoding = hgencoding
 

	
 
    if config.get('ignore_alembic_revision', False):
 
        log.warn('database alembic revision checking is disabled')
 
    else:
 
        dbconf = config['sqlalchemy.url']
 
        alembic_cfg = alembic.config.Config()
 
        alembic_cfg.set_main_option('script_location', 'kallithea:alembic')
 
        alembic_cfg.set_main_option('sqlalchemy.url', dbconf)
 
        script_dir = ScriptDirectory.from_config(alembic_cfg)
 
        available_heads = sorted(script_dir.get_heads())
 

	
 
        engine = create_engine(dbconf)
 
        with engine.connect() as conn:
 
            context = MigrationContext.configure(conn)
 
            current_heads = sorted(str(s) for s in context.get_current_heads())
 
        if current_heads != available_heads:
 
            log.error('Failed to run Kallithea:\n\n'
 
                      'The database version does not match the Kallithea version.\n'
 
                      'Please read the documentation on how to upgrade or downgrade the database.\n'
 
                      'Current database version id(s): %s\n'
 
                      'Expected database version id(s): %s\n'
 
                      'If you are a developer and you know what you are doing, you can add `ignore_alembic_revision = True` '
 
                      'to your .ini file to skip the check.\n' % (' '.join(current_heads), ' '.join(available_heads)))
 
            sys.exit(1)
 

	
 
    # store some globals into kallithea
 
    kallithea.CELERY_ON = str2bool(config.get('use_celery'))
 
    kallithea.CELERY_EAGER = str2bool(config.get('celery.always.eager'))
 
    kallithea.CONFIG = config
 

	
 
    load_rcextensions(root_path=config['here'])
 

	
 
    repos_path = make_ui().configitems(b'paths')[0][1]
 
    repos_path = safe_str(make_ui().configitems(b'paths')[0][1])
 
    config['base_path'] = repos_path
 
    set_app_settings(config)
 

	
 
    instance_id = kallithea.CONFIG.get('instance_id', '*')
 
    if instance_id == '*':
 
        instance_id = '%s-%s' % (platform.uname()[1], os.getpid())
 
        kallithea.CONFIG['instance_id'] = instance_id
 

	
 
    # update kallithea.CONFIG with the meanwhile changed 'config'
 
    kallithea.CONFIG.update(config)
 

	
 
    # configure vcs and indexer libraries (they are supposed to be independent
 
    # as much as possible and thus avoid importing tg.config or
 
    # kallithea.CONFIG).
 
    set_vcs_config(kallithea.CONFIG)
 
    set_indexer_config(kallithea.CONFIG)
 

	
 
    check_git_version()
 

	
 
    kallithea.model.meta.Session.remove()
 

	
 

	
 
tg.hooks.register('configure_new_app', setup_configuration)
 

	
 

	
 
def setup_application(app):
 
    config = app.config
 

	
 
    # we want our low level middleware to get to the request ASAP. We don't
 
    # need any stack middleware in them - especially no StatusCodeRedirect buffering
 
    app = SimpleHg(app, config)
 
    app = SimpleGit(app, config)
 

	
 
    # Enable https redirects based on HTTP_X_URL_SCHEME set by proxy
 
    if any(asbool(config.get(x)) for x in ['https_fixup', 'force_https', 'use_htsts']):
 
        app = HttpsFixup(app, config)
 

	
 
    app = PermanentRepoUrl(app, config)
 

	
 
    # Optional and undocumented wrapper - gives more verbose request/response logging, but has a slight overhead
 
    if str2bool(config.get('use_wsgi_wrapper')):
 
        app = RequestWrapper(app, config)
 

	
 
    return app
 

	
 

	
 
tg.hooks.register('before_config', setup_application)
kallithea/controllers/home.py
Show inline comments
 
# -*- coding: utf-8 -*-
 
# 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/>.
 
"""
 
kallithea.controllers.home
 
~~~~~~~~~~~~~~~~~~~~~~~~~~
 

	
 
Home controller for Kallithea
 

	
 
This file was forked by the Kallithea project in July 2014.
 
Original author and date, and relevant copyright and licensing information is below:
 
:created_on: Feb 18, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 

	
 
"""
 

	
 
import logging
 

	
 
from sqlalchemy import or_
 
from tg import request
 
from tg import tmpl_context as c
 
from tg.i18n import ugettext as _
 
from webob.exc import HTTPBadRequest
 

	
 
from kallithea.lib import helpers as h
 
from kallithea.lib.auth import HasRepoPermissionLevelDecorator, LoginRequired
 
from kallithea.lib.base import BaseController, jsonify, render
 
from kallithea.lib.utils2 import safe_str
 
from kallithea.model.db import RepoGroup, Repository, User, UserGroup
 
from kallithea.model.repo import RepoModel
 
from kallithea.model.scm import UserGroupList
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class HomeController(BaseController):
 

	
 
    def about(self):
 
        return render('/about.html')
 

	
 
    @LoginRequired(allow_default_user=True)
 
    def index(self):
 
        c.group = None
 

	
 
        repo_groups_list = self.scm_model.get_repo_groups()
 
        repos_list = Repository.query(sorted=True).filter_by(group=None).all()
 

	
 
        c.data = RepoModel().get_repos_as_dict(repos_list,
 
                                               repo_groups_list=repo_groups_list,
 
                                               short_name=True)
 

	
 
        return render('/index.html')
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @jsonify
 
    def repo_switcher_data(self):
 
        if request.is_xhr:
 
            all_repos = Repository.query(sorted=True).all()
 
            repo_iter = self.scm_model.get_repos(all_repos)
 
            all_groups = RepoGroup.query(sorted=True).all()
 
            repo_groups_iter = self.scm_model.get_repo_groups(all_groups)
 

	
 
            res = [{
 
                    'text': _('Groups'),
 
                    'children': [
 
                       {'id': obj.group_name,
 
                        'text': obj.group_name,
 
                        'type': 'group',
 
                        'obj': {}}
 
                       for obj in repo_groups_iter
 
                    ],
 
                   },
 
                   {
 
                    'text': _('Repositories'),
 
                    'children': [
 
                       {'id': obj.repo_name,
 
                        'text': obj.repo_name,
 
                        'type': 'repo',
 
                        'obj': obj.get_dict()}
 
                       for obj in repo_iter
 
                    ],
 
                   }]
 

	
 
            for res_dict in res:
 
                for child in (res_dict['children']):
 
                    child['obj'].pop('_changeset_cache', None)  # bytes cannot be encoded in json ... but this value isn't relevant on client side at all ...
 

	
 
            data = {
 
                'more': False,
 
                'results': res,
 
            }
 
            return data
 

	
 
        else:
 
            raise HTTPBadRequest()
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @HasRepoPermissionLevelDecorator('read')
 
    @jsonify
 
    def repo_refs_data(self, repo_name):
 
        repo = Repository.get_by_repo_name(repo_name).scm_instance
 
        res = []
 
        _branches = repo.branches.items()
 
        if _branches:
 
            res.append({
 
                'text': _('Branch'),
 
                'children': [{'id': rev, 'text': name, 'type': 'branch'} for name, rev in _branches]
 
                'children': [{'id': safe_str(rev), 'text': safe_str(name), 'type': 'branch'} for name, rev in _branches]
 
            })
 
        _closed_branches = repo.closed_branches.items()
 
        if _closed_branches:
 
            res.append({
 
                'text': _('Closed Branches'),
 
                'children': [{'id': rev, 'text': name, 'type': 'closed-branch'} for name, rev in _closed_branches]
 
                'children': [{'id': safe_str(rev), 'text': safe_str(name), 'type': 'closed-branch'} for name, rev in _closed_branches]
 
            })
 
        _tags = repo.tags.items()
 
        if _tags:
 
            res.append({
 
                'text': _('Tag'),
 
                'children': [{'id': rev, 'text': name, 'type': 'tag'} for name, rev in _tags]
 
                'children': [{'id': safe_str(rev), 'text': safe_str(name), 'type': 'tag'} for name, rev in _tags]
 
            })
 
        _bookmarks = repo.bookmarks.items()
 
        if _bookmarks:
 
            res.append({
 
                'text': _('Bookmark'),
 
                'children': [{'id': rev, 'text': name, 'type': 'book'} for name, rev in _bookmarks]
 
                'children': [{'id': safe_str(rev), 'text': safe_str(name), 'type': 'book'} for name, rev in _bookmarks]
 
            })
 
        data = {
 
            'more': False,
 
            'results': res
 
        }
 
        return data
 

	
 
    @LoginRequired()
 
    @jsonify
 
    def users_and_groups_data(self):
 
        """
 
        Returns 'results' with a list of users and user groups.
 

	
 
        You can either use the 'key' GET parameter to get a user by providing
 
        the exact user key or you can use the 'query' parameter to
 
        search for users by user key, first name and last name.
 
        'types' defaults to just 'users' but can be set to 'users,groups' to
 
        get both users and groups.
 
        No more than 500 results (of each kind) will be returned.
 
        """
 
        types = request.GET.get('types', 'users').split(',')
 
        key = request.GET.get('key', '')
 
        query = request.GET.get('query', '')
 
        results = []
 
        if 'users' in types:
 
            user_list = []
 
            if key:
 
                u = User.get_by_username(key)
 
                if u:
 
                    user_list = [u]
 
            elif query:
 
                user_list = User.query() \
 
                    .filter(User.is_default_user == False) \
 
                    .filter(User.active == True) \
 
                    .filter(or_(
 
                        User.username.ilike("%%" + query + "%%"),
 
                        User.name.ilike("%%" + query + "%%"),
 
                        User.lastname.ilike("%%" + query + "%%"),
 
                    )) \
 
                    .order_by(User.username) \
 
                    .limit(500) \
 
                    .all()
 
            for u in user_list:
 
                results.append({
 
                    'type': 'user',
 
                    'id': u.user_id,
 
                    'nname': u.username,
 
                    'fname': u.name,
 
                    'lname': u.lastname,
 
                    'gravatar_lnk': h.gravatar_url(u.email, size=28, default='default'),
 
                    'gravatar_size': 14,
 
                })
 
        if 'groups' in types:
 
            grp_list = []
 
            if key:
 
                grp = UserGroup.get_by_group_name(key)
 
                if grp:
 
                    grp_list = [grp]
 
            elif query:
 
                grp_list = UserGroup.query() \
 
                    .filter(UserGroup.users_group_name.ilike("%%" + query + "%%")) \
 
                    .filter(UserGroup.users_group_active == True) \
 
                    .order_by(UserGroup.users_group_name) \
 
                    .limit(500) \
 
                    .all()
 
            for g in UserGroupList(grp_list, perm_level='read'):
 
                results.append({
 
                    'type': 'group',
 
                    'id': g.users_group_id,
 
                    'grname': g.users_group_name,
 
                })
 
        return dict(results=results)
kallithea/lib/base.py
Show inline comments
 
@@ -79,193 +79,193 @@ def _filter_proxy(ip):
 
    return ip
 

	
 

	
 
def _get_ip_addr(environ):
 
    proxy_key = 'HTTP_X_REAL_IP'
 
    proxy_key2 = 'HTTP_X_FORWARDED_FOR'
 
    def_key = 'REMOTE_ADDR'
 

	
 
    ip = environ.get(proxy_key)
 
    if ip:
 
        return _filter_proxy(ip)
 

	
 
    ip = environ.get(proxy_key2)
 
    if ip:
 
        return _filter_proxy(ip)
 

	
 
    ip = environ.get(def_key, '0.0.0.0')
 
    return _filter_proxy(ip)
 

	
 

	
 
def get_path_info(environ):
 
    """Return unicode PATH_INFO from environ ... using tg.original_request if available.
 
    """
 
    org_req = environ.get('tg.original_request')
 
    if org_req is not None:
 
        environ = org_req.environ
 
    return safe_str(environ['PATH_INFO'])
 

	
 

	
 
def log_in_user(user, remember, is_external_auth, ip_addr):
 
    """
 
    Log a `User` in and update session and cookies. If `remember` is True,
 
    the session cookie is set to expire in a year; otherwise, it expires at
 
    the end of the browser session.
 

	
 
    Returns populated `AuthUser` object.
 
    """
 
    # It should not be possible to explicitly log in as the default user.
 
    assert not user.is_default_user, user
 

	
 
    auth_user = AuthUser.make(dbuser=user, is_external_auth=is_external_auth, ip_addr=ip_addr)
 
    if auth_user is None:
 
        return None
 

	
 
    user.update_lastlogin()
 
    meta.Session().commit()
 

	
 
    # Start new session to prevent session fixation attacks.
 
    session.invalidate()
 
    session['authuser'] = cookie = auth_user.to_cookie()
 

	
 
    # If they want to be remembered, update the cookie.
 
    # NOTE: Assumes that beaker defaults to browser session cookie.
 
    if remember:
 
        t = datetime.datetime.now() + datetime.timedelta(days=365)
 
        session._set_cookie_expires(t)
 

	
 
    session.save()
 

	
 
    log.info('user %s is now authenticated and stored in '
 
             'session, session attrs %s', user.username, cookie)
 

	
 
    # dumps session attrs back to cookie
 
    session._update_cookie_out()
 

	
 
    return auth_user
 

	
 

	
 
class BasicAuth(paste.auth.basic.AuthBasicAuthenticator):
 

	
 
    def __init__(self, realm, authfunc, auth_http_code=None):
 
        self.realm = realm
 
        self.authfunc = authfunc
 
        self._rc_auth_http_code = auth_http_code
 

	
 
    def build_authentication(self, environ):
 
        head = paste.httpheaders.WWW_AUTHENTICATE.tuples('Basic realm="%s"' % self.realm)
 
        # Consume the whole body before sending a response
 
        try:
 
            request_body_size = int(environ.get('CONTENT_LENGTH', 0))
 
        except (ValueError):
 
            request_body_size = 0
 
        environ['wsgi.input'].read(request_body_size)
 
        if self._rc_auth_http_code and self._rc_auth_http_code == '403':
 
            # return 403 if alternative http return code is specified in
 
            # Kallithea config
 
            return paste.httpexceptions.HTTPForbidden(headers=head)
 
        return paste.httpexceptions.HTTPUnauthorized(headers=head)
 

	
 
    def authenticate(self, environ):
 
        authorization = paste.httpheaders.AUTHORIZATION(environ)
 
        if not authorization:
 
            return self.build_authentication(environ)
 
        (authmeth, auth) = authorization.split(' ', 1)
 
        if 'basic' != authmeth.lower():
 
            return self.build_authentication(environ)
 
        auth = base64.b64decode(auth.strip())
 
        auth = safe_str(base64.b64decode(auth.strip()))
 
        _parts = auth.split(':', 1)
 
        if len(_parts) == 2:
 
            username, password = _parts
 
            if self.authfunc(username, password, environ) is not None:
 
                return username
 
        return self.build_authentication(environ)
 

	
 
    __call__ = authenticate
 

	
 

	
 
class BaseVCSController(object):
 
    """Base controller for handling Mercurial/Git protocol requests
 
    (coming from a VCS client, and not a browser).
 
    """
 

	
 
    scm_alias = None # 'hg' / 'git'
 

	
 
    def __init__(self, application, config):
 
        self.application = application
 
        self.config = config
 
        # base path of repo locations
 
        self.basepath = self.config['base_path']
 
        # authenticate this VCS request using the authentication modules
 
        self.authenticate = BasicAuth('', auth_modules.authenticate,
 
                                      config.get('auth_ret_code'))
 

	
 
    @classmethod
 
    def parse_request(cls, environ):
 
        """If request is parsed as a request for this VCS, return a namespace with the parsed request.
 
        If the request is unknown, return None.
 
        """
 
        raise NotImplementedError()
 

	
 
    def _authorize(self, environ, action, repo_name, ip_addr):
 
        """Authenticate and authorize user.
 

	
 
        Since we're dealing with a VCS client and not a browser, we only
 
        support HTTP basic authentication, either directly via raw header
 
        inspection, or by using container authentication to delegate the
 
        authentication to the web server.
 

	
 
        Returns (user, None) on successful authentication and authorization.
 
        Returns (None, wsgi_app) to send the wsgi_app response to the client.
 
        """
 
        # Use anonymous access if allowed for action on repo.
 
        default_user = User.get_default_user(cache=True)
 
        default_authuser = AuthUser.make(dbuser=default_user, ip_addr=ip_addr)
 
        if default_authuser is None:
 
            log.debug('No anonymous access at all') # move on to proper user auth
 
        else:
 
            if self._check_permission(action, default_authuser, repo_name):
 
                return default_authuser, None
 
            log.debug('Not authorized to access this repository as anonymous user')
 

	
 
        username = None
 
        #==============================================================
 
        # DEFAULT PERM FAILED OR ANONYMOUS ACCESS IS DISABLED SO WE
 
        # NEED TO AUTHENTICATE AND ASK FOR AUTH USER PERMISSIONS
 
        #==============================================================
 

	
 
        # try to auth based on environ, container auth methods
 
        log.debug('Running PRE-AUTH for container based authentication')
 
        pre_auth = auth_modules.authenticate('', '', environ)
 
        if pre_auth is not None and pre_auth.get('username'):
 
            username = pre_auth['username']
 
        log.debug('PRE-AUTH got %s as username', username)
 

	
 
        # If not authenticated by the container, running basic auth
 
        if not username:
 
            self.authenticate.realm = self.config['realm']
 
            result = self.authenticate(environ)
 
            if isinstance(result, str):
 
                paste.httpheaders.AUTH_TYPE.update(environ, 'basic')
 
                paste.httpheaders.REMOTE_USER.update(environ, result)
 
                username = result
 
            else:
 
                return None, result.wsgi_application
 

	
 
        #==============================================================
 
        # CHECK PERMISSIONS FOR THIS REQUEST USING GIVEN USERNAME
 
        #==============================================================
 
        try:
 
            user = User.get_by_username_or_email(username)
 
        except Exception:
 
            log.error(traceback.format_exc())
 
            return None, webob.exc.HTTPInternalServerError()
 

	
 
        authuser = AuthUser.make(dbuser=user, ip_addr=ip_addr)
 
        if authuser is None:
 
            return None, webob.exc.HTTPForbidden()
 
        if not self._check_permission(action, authuser, repo_name):
 
            return None, webob.exc.HTTPForbidden()
 

	
 
        return user, None
 

	
 
    def _handle_request(self, environ, start_response):
kallithea/lib/diffs.py
Show inline comments
 
@@ -445,196 +445,196 @@ class DiffProcessor(object):
 
                    pass
 

	
 
        return _files
 

	
 
    def stat(self):
 
        """
 
        Returns tuple of added, and removed lines for this instance
 
        """
 
        return self.adds, self.removes
 

	
 

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

	
 

	
 
def _escaper(string):
 
    """
 
    Do HTML escaping/markup
 
    """
 

	
 
    def substitute(m):
 
        groups = m.groups()
 
        if groups[0]:
 
            return '&amp;'
 
        if groups[1]:
 
            return '&lt;'
 
        if groups[2]:
 
            return '&gt;'
 
        if groups[3]:
 
            return '<u>\t</u>'
 
        if groups[4]:
 
            return '<u class="cr"></u>'
 
        if groups[5]:
 
            return ' <i></i>'
 
        assert False
 

	
 
    return _escape_re.sub(substitute, safe_str(string))
 

	
 

	
 
_git_header_re = re.compile(br"""
 
    ^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|$))?
 
    (?:^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(br"""
 
    ^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)
 

	
 

	
 
_header_next_check = re.compile(br'''(?!@)(?!literal )(?!delta )''')
 

	
 

	
 
def _get_header(vcs, diff_chunk):
 
    """
 
    Parses a Git diff for a single file (header and chunks) and returns a tuple with:
 

	
 
    1. A dict with meta info:
 

	
 
        a_path, b_path, similarity_index, rename_from, rename_to,
 
        old_mode, new_mode, new_file_mode, deleted_file_mode,
 
        a_blob_id, b_blob_id, b_mode, a_file, b_file
 

	
 
    2. An iterator yielding lines with simple HTML markup.
 
    """
 
    match = None
 
    if vcs == 'git':
 
        match = _git_header_re.match(diff_chunk)
 
    elif vcs == 'hg':
 
        match = _hg_header_re.match(diff_chunk)
 
    if match is None:
 
        raise Exception('diff not recognized as valid %s diff' % vcs)
 
    meta_info = match.groupdict()
 
    meta_info = {k: None if v is None else safe_str(v) for k, v in match.groupdict().items()}
 
    rest = diff_chunk[match.end():]
 
    if rest and _header_next_check.match(rest):
 
        raise Exception('cannot parse %s diff header: %r followed by %r' % (vcs, diff_chunk[:match.end()], rest[:1000]))
 
        raise Exception('cannot parse %s diff header: %r followed by %r' % (vcs, safe_str(bytes(diff_chunk[:match.end()])), safe_str(bytes(rest[:1000]))))
 
    diff_lines = (_escaper(m.group(0)) for m in re.finditer(br'.*\n|.+$', rest)) # don't split on \r as str.splitlines do
 
    return meta_info, diff_lines
 

	
 

	
 
_chunk_re = re.compile(r'^@@ -(\d+)(?:,(\d+))? \+(\d+)(?:,(\d+))? @@(.*)')
 
_newline_marker = re.compile(r'^\\ No newline at end of file')
 

	
 

	
 
def _parse_lines(diff_lines):
 
    """
 
    Given an iterator of diff body lines, parse them and return a dict per
 
    line and added/removed totals.
 
    """
 
    added = deleted = 0
 
    old_line = old_end = new_line = new_end = None
 

	
 
    chunks = []
 
    try:
 
        line = next(diff_lines)
 

	
 
        while True:
 
            lines = []
 
            chunks.append(lines)
 

	
 
            match = _chunk_re.match(line)
 

	
 
            if not match:
 
                raise Exception('error parsing diff @@ line %r' % line)
 

	
 
            gr = match.groups()
 
            (old_line, old_end,
 
             new_line, new_end) = [int(x or 1) for x in gr[:-1]]
 
            old_line -= 1
 
            new_line -= 1
 

	
 
            context = len(gr) == 5
 
            old_end += old_line
 
            new_end += new_line
 

	
 
            if context:
 
                # skip context only if it's first line
 
                if int(gr[0]) > 1:
 
                    lines.append({
 
                        'old_lineno': '...',
 
                        'new_lineno': '...',
 
                        'action':     'context',
 
                        'line':       line,
 
                    })
 

	
 
            line = next(diff_lines)
 

	
 
            while old_line < old_end or new_line < new_end:
 
                if not line:
 
                    raise Exception('error parsing diff - empty line at -%s+%s' % (old_line, new_line))
 

	
 
                affects_old = affects_new = False
 

	
 
                command = line[0]
 
                if command == '+':
 
                    affects_new = True
 
                    action = 'add'
 
                    added += 1
 
                elif command == '-':
 
                    affects_old = True
 
                    action = 'del'
 
                    deleted += 1
 
                elif command == ' ':
 
                    affects_old = affects_new = True
 
                    action = 'unmod'
 
                else:
 
                    raise Exception('error parsing diff - unknown command in line %r at -%s+%s' % (line, old_line, new_line))
 

	
 
                if not _newline_marker.match(line):
 
                    old_line += affects_old
 
                    new_line += affects_new
 
                    lines.append({
 
                        'old_lineno':   affects_old and old_line or '',
 
                        'new_lineno':   affects_new and new_line or '',
 
                        'action':       action,
 
                        'line':         line[1:],
 
                    })
 

	
 
                line = next(diff_lines)
 

	
 
                if _newline_marker.match(line):
 
                    # we need to append to lines, since this is not
 
                    # counted in the line specs of diff
 
                    lines.append({
 
                        'old_lineno':   '...',
 
                        'new_lineno':   '...',
 
                        'action':       'context',
 
                        'line':         line,
 
                    })
 
                    line = next(diff_lines)
 
            if old_line > old_end:
 
                raise Exception('error parsing diff - more than %s "-" lines at -%s+%s' % (old_end, old_line, new_line))
kallithea/lib/hooks.py
Show inline comments
 
# -*- coding: utf-8 -*-
 
# 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/>.
 
"""
 
kallithea.lib.hooks
 
~~~~~~~~~~~~~~~~~~~
 

	
 
Hooks run by Kallithea
 

	
 
This file was forked by the Kallithea project in July 2014.
 
Original author and date, and relevant copyright and licensing information is below:
 
:created_on: Aug 6, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 
"""
 

	
 
import os
 
import sys
 
import time
 

	
 
import mercurial.scmutil
 

	
 
from kallithea.lib import helpers as h
 
from kallithea.lib.exceptions import UserCreationError
 
from kallithea.lib.utils import action_logger, make_ui
 
from kallithea.lib.utils2 import HookEnvironmentError, ascii_str, get_hook_environment, safe_bytes
 
from kallithea.lib.utils2 import HookEnvironmentError, ascii_str, get_hook_environment, safe_bytes, safe_str
 
from kallithea.lib.vcs.backends.base import EmptyChangeset
 
from kallithea.model.db import Repository, User
 

	
 

	
 
def _get_scm_size(alias, root_path):
 
    if not alias.startswith('.'):
 
        alias += '.'
 

	
 
    size_scm, size_root = 0, 0
 
    for path, dirs, files in os.walk(root_path):
 
        if path.find(alias) != -1:
 
            for f in files:
 
                try:
 
                    size_scm += os.path.getsize(os.path.join(path, f))
 
                except OSError:
 
                    pass
 
        else:
 
            for f in files:
 
                try:
 
                    size_root += os.path.getsize(os.path.join(path, f))
 
                except OSError:
 
                    pass
 

	
 
    size_scm_f = h.format_byte_size(size_scm)
 
    size_root_f = h.format_byte_size(size_root)
 
    size_total_f = h.format_byte_size(size_root + size_scm)
 

	
 
    return size_scm_f, size_root_f, size_total_f
 

	
 

	
 
def repo_size(ui, repo, hooktype=None, **kwargs):
 
    """Show size of Mercurial repository, to be called after push."""
 
    size_hg_f, size_root_f, size_total_f = _get_scm_size('.hg', repo.root)
 
    size_hg_f, size_root_f, size_total_f = _get_scm_size('.hg', safe_str(repo.root))
 

	
 
    last_cs = repo[len(repo) - 1]
 

	
 
    msg = ('Repository size .hg: %s Checkout: %s Total: %s\n'
 
           'Last revision is now r%s:%s\n') % (
 
        size_hg_f, size_root_f, size_total_f, last_cs.rev(), ascii_str(last_cs.hex())[:12]
 
    )
 
    ui.status(safe_bytes(msg))
 

	
 

	
 
def log_pull_action(ui, repo, **kwargs):
 
    """Logs user last pull action
 

	
 
    Called as Mercurial hook outgoing.pull_logger or from Kallithea before invoking Git.
 

	
 
    Does *not* use the action from the hook environment but is always 'pull'.
 
    """
 
    ex = get_hook_environment()
 

	
 
    user = User.get_by_username(ex.username)
 
    action = 'pull'
 
    action_logger(user, action, ex.repository, ex.ip, commit=True)
 
    # extension hook call
 
    from kallithea import EXTENSIONS
 
    callback = getattr(EXTENSIONS, 'PULL_HOOK', None)
 
    if callable(callback):
 
        kw = {}
 
        kw.update(ex)
 
        callback(**kw)
 

	
 
    return 0
 

	
 

	
 
def log_push_action(ui, repo, node, node_last, **kwargs):
 
    """
 
    Entry point for Mercurial hook changegroup.push_logger.
 

	
 
    The pushed changesets is given by the revset 'node:node_last'.
 

	
 
    Note: This hook is not only logging, but also the side effect invalidating
 
    cahes! The function should perhaps be renamed.
 
    """
 
    revs = [ascii_str(repo[r].hex()) for r in mercurial.scmutil.revrange(repo, [b'%s:%s' % (node, node_last)])]
 
    process_pushed_raw_ids(revs)
 
    return 0
 

	
 

	
 
def process_pushed_raw_ids(revs):
 
    """
 
    Register that changes have been added to the repo - log the action *and* invalidate caches.
 

	
 
    Called from  Mercurial changegroup.push_logger calling hook log_push_action,
 
    or from the Git post-receive hook calling handle_git_post_receive ...
 
    or from scm _handle_push.
 
    """
 
    ex = get_hook_environment()
 

	
 
    action = '%s:%s' % (ex.action, ','.join(revs))
 
    action_logger(ex.username, action, ex.repository, ex.ip, commit=True)
 

	
 
    from kallithea.model.scm import ScmModel
 
    ScmModel().mark_for_invalidation(ex.repository)
 

	
 
    # extension hook call
 
    from kallithea import EXTENSIONS
 
    callback = getattr(EXTENSIONS, 'PUSH_HOOK', None)
 
    if callable(callback):
 
        kw = {'pushed_revs': revs}
 
        kw.update(ex)
 
        callback(**kw)
 

	
 

	
 
def log_create_repository(repository_dict, created_by, **kwargs):
 
    """
 
    Post create repository Hook.
 

	
 
    :param repository: dict dump of repository object
 
    :param created_by: username who created repository
 

	
 
    available keys of repository_dict:
 

	
 
     'repo_type',
 
     'description',
 
     'private',
 
     'created_on',
 
     'enable_downloads',
 
     'repo_id',
 
     'owner_id',
 
     'enable_statistics',
 
     'clone_uri',
 
     'fork_id',
 
     'group_id',
 
     'repo_name'
 

	
 
    """
 
    from kallithea import EXTENSIONS
kallithea/lib/utils.py
Show inline comments
 
# -*- coding: utf-8 -*-
 
# 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/>.
 
"""
 
kallithea.lib.utils
 
~~~~~~~~~~~~~~~~~~~
 

	
 
Utilities library for Kallithea
 

	
 
This file was forked by the Kallithea project in July 2014.
 
Original author and date, and relevant copyright and licensing information is below:
 
:created_on: Apr 18, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 
"""
 

	
 
import datetime
 
import logging
 
import os
 
import re
 
import sys
 
import traceback
 
from distutils.version import StrictVersion
 

	
 
import beaker.cache
 
import mercurial.config
 
import mercurial.ui
 
from tg.i18n import ugettext as _
 

	
 
import kallithea.config.conf
 
from kallithea.lib.exceptions import HgsubversionImportError
 
from kallithea.lib.utils2 import ascii_bytes, aslist, get_current_authuser, safe_bytes
 
from kallithea.lib.utils2 import ascii_bytes, aslist, get_current_authuser, safe_bytes, safe_str
 
from kallithea.lib.vcs.backends.git.repository import GitRepository
 
from kallithea.lib.vcs.backends.hg.repository import MercurialRepository
 
from kallithea.lib.vcs.conf import settings
 
from kallithea.lib.vcs.exceptions import RepositoryError, VCSError
 
from kallithea.lib.vcs.utils.fakemod import create_module
 
from kallithea.lib.vcs.utils.helpers import get_scm
 
from kallithea.model import meta
 
from kallithea.model.db import RepoGroup, Repository, Setting, Ui, User, UserGroup, UserLog
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 
REMOVED_REPO_PAT = re.compile(r'rm__\d{8}_\d{6}_\d{6}_.*')
 

	
 

	
 
#==============================================================================
 
# PERM DECORATOR HELPERS FOR EXTRACTING NAMES FOR PERM CHECKS
 
#==============================================================================
 
def get_repo_slug(request):
 
    _repo = request.environ['pylons.routes_dict'].get('repo_name')
 
    if _repo:
 
        _repo = _repo.rstrip('/')
 
    return _repo
 

	
 

	
 
def get_repo_group_slug(request):
 
    _group = request.environ['pylons.routes_dict'].get('group_name')
 
    if _group:
 
        _group = _group.rstrip('/')
 
    return _group
 

	
 

	
 
def get_user_group_slug(request):
 
    _group = request.environ['pylons.routes_dict'].get('id')
 
    _group = UserGroup.get(_group)
 
    if _group:
 
        return _group.users_group_name
 
    return None
 

	
 

	
 
def _get_permanent_id(s):
 
    """Helper for decoding stable URLs with repo ID. For a string like '_123'
 
    return 123.
 
    """
 
    by_id_match = re.match(r'^_(\d+)$', s)
 
    if by_id_match is None:
 
        return None
 
    return int(by_id_match.group(1))
 

	
 

	
 
def fix_repo_id_name(path):
 
    """
 
    Rewrite repo_name for _<ID> permanent URLs.
 

	
 
    Given a path, if the first path element is like _<ID>, return the path with
 
    this part expanded to the corresponding full repo name, else return the
 
    provided path.
 
    """
 
    first, rest = path, ''
 
    if '/' in path:
 
        first, rest_ = path.split('/', 1)
 
        rest = '/' + rest_
 
    repo_id = _get_permanent_id(first)
 
    if repo_id is not None:
 
        repo = Repository.get(repo_id)
 
        if repo is not None:
 
            return repo.repo_name + rest
 
    return path
 

	
 

	
 
def action_logger(user, action, repo, ipaddr='', commit=False):
 
    """
 
    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
 

	
 
    """
 

	
 
    # if we don't get explicit IP address try to get one from registered user
 
    # in tmpl context var
 
    if not ipaddr:
 
        ipaddr = getattr(get_current_authuser(), 'ip_addr', '')
 

	
 
    if getattr(user, 'user_id', None):
 
        user_obj = User.get(user.user_id)
 
    elif isinstance(user, str):
 
        user_obj = User.get_by_username(user)
 
    else:
 
        raise Exception('You have to provide a user object or a username')
 

	
 
@@ -493,154 +493,154 @@ def repo2db_mapper(initial_repo_dict, re
 
                repo_name=name,
 
                repo_type=repo.alias,
 
                description=desc,
 
                repo_group=getattr(group, 'group_id', None),
 
                owner=user,
 
                enable_downloads=enable_downloads,
 
                enable_statistics=enable_statistics,
 
                private=private,
 
                state=Repository.STATE_CREATED
 
            )
 
            sa.commit()
 
            # we added that repo just now, and make sure it has githook
 
            # installed, and updated server info
 
            if new_repo.repo_type == 'git':
 
                git_repo = new_repo.scm_instance
 
                ScmModel().install_git_hooks(git_repo)
 
                # update repository server-info
 
                log.debug('Running update server info')
 
                git_repo._update_server_info()
 
            new_repo.update_changeset_cache()
 
        elif install_git_hooks:
 
            if db_repo.repo_type == 'git':
 
                ScmModel().install_git_hooks(db_repo.scm_instance, force_create=overwrite_git_hooks)
 

	
 
    removed = []
 
    # 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_dict:
 
            if remove_obsolete:
 
                log.debug("Removing non-existing repository found in db `%s`",
 
                          repo.repo_name)
 
                try:
 
                    RepoModel().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()
 
            removed.append(repo.repo_name)
 
    return added, removed
 

	
 

	
 
def load_rcextensions(root_path):
 
    path = os.path.join(root_path, 'rcextensions', '__init__.py')
 
    if os.path.isfile(path):
 
        rcext = create_module('rc', path)
 
        EXT = kallithea.EXTENSIONS = rcext
 
        log.debug('Found rcextensions now loading %s...', rcext)
 

	
 
        # Additional mappings that are not present in the pygments lexers
 
        kallithea.config.conf.LANGUAGES_EXTENSIONS_MAP.update(getattr(EXT, 'EXTRA_MAPPINGS', {}))
 

	
 
        # OVERRIDE OUR EXTENSIONS FROM RC-EXTENSIONS (if present)
 

	
 
        if getattr(EXT, 'INDEX_EXTENSIONS', []):
 
            log.debug('settings custom INDEX_EXTENSIONS')
 
            kallithea.config.conf.INDEX_EXTENSIONS = getattr(EXT, 'INDEX_EXTENSIONS', [])
 

	
 
        # ADDITIONAL MAPPINGS
 
        log.debug('adding extra into INDEX_EXTENSIONS')
 
        kallithea.config.conf.INDEX_EXTENSIONS.extend(getattr(EXT, 'EXTRA_INDEX_EXTENSIONS', []))
 

	
 
        # auto check if the module is not missing any data, set to default if is
 
        # this will help autoupdate new feature of rcext module
 
        #from kallithea.config import rcextensions
 
        #for k in dir(rcextensions):
 
        #    if not k.startswith('_') and not hasattr(EXT, k):
 
        #        setattr(EXT, k, getattr(rcextensions, k))
 

	
 

	
 
#==============================================================================
 
# MISC
 
#==============================================================================
 

	
 
git_req_ver = StrictVersion('1.7.4')
 

	
 
def check_git_version():
 
    """
 
    Checks what version of git is installed on the system, and raise a system exit
 
    if it's too old for Kallithea to work properly.
 
    """
 
    if 'git' not in kallithea.BACKENDS:
 
        return None
 

	
 
    if not settings.GIT_EXECUTABLE_PATH:
 
        log.warning('No git executable configured - check "git_path" in the ini file.')
 
        return None
 

	
 
    try:
 
        stdout, stderr = GitRepository._run_git_command(['--version'])
 
    except RepositoryError as e:
 
        # message will already have been logged as error
 
        log.warning('No working git executable found - check "git_path" in the ini file.')
 
        return None
 

	
 
    if stderr:
 
        log.warning('Error/stderr from "%s --version":\n%s', settings.GIT_EXECUTABLE_PATH, stderr)
 
        log.warning('Error/stderr from "%s --version":\n%s', settings.GIT_EXECUTABLE_PATH, safe_str(stderr))
 

	
 
    if not stdout:
 
        log.warning('No working git executable found - check "git_path" in the ini file.')
 
        return None
 

	
 
    output = stdout.strip()
 
    output = safe_str(stdout).strip()
 
    m = re.search(r"\d+.\d+.\d+", output)
 
    if m:
 
        ver = StrictVersion(m.group(0))
 
        log.debug('Git executable: "%s", version %s (parsed from: "%s")',
 
                  settings.GIT_EXECUTABLE_PATH, ver, output)
 
        if ver < git_req_ver:
 
            log.error('Kallithea detected %s version %s, which is too old '
 
                      'for the system to function properly. '
 
                      'Please upgrade to version %s or later. '
 
                      'If you strictly need Mercurial repositories, you can '
 
                      'clear the "git_path" setting in the ini file.',
 
                      settings.GIT_EXECUTABLE_PATH, ver, git_req_ver)
 
            log.error("Terminating ...")
 
            sys.exit(1)
 
    else:
 
        ver = StrictVersion('0.0.0')
 
        log.warning('Error finding version number in "%s --version" stdout:\n%s',
 
                    settings.GIT_EXECUTABLE_PATH, output)
 

	
 
    return ver
 

	
 

	
 
#===============================================================================
 
# CACHE RELATED METHODS
 
#===============================================================================
 

	
 
def conditional_cache(region, prefix, condition, func):
 
    """
 

	
 
    Conditional caching function use like::
 
        def _c(arg):
 
            #heavy computation function
 
            return data
 

	
 
        # depending from condition the compute is wrapped in cache or not
 
        compute = conditional_cache('short_term', 'cache_desc', condition=True, func=func)
 
        return compute(arg)
 

	
 
    :param region: name of cache region
 
    :param prefix: cache region prefix
 
    :param condition: condition for cache to be triggered, and return data cached
 
    :param func: wrapped heavy function to compute
 

	
 
    """
 
    wrapped = func
 
    if condition:
 
        log.debug('conditional_cache: True, wrapping call of '
 
                  'func: %s into %s region cache' % (region, func))
 
        wrapped = beaker.cache._cache_decorate((prefix,), None, None, region)(func)
 

	
 
    return wrapped
kallithea/lib/vcs/backends/git/changeset.py
Show inline comments
 
@@ -3,248 +3,247 @@ from io import BytesIO
 
from itertools import chain
 
from subprocess import PIPE, Popen
 

	
 
from dulwich import objects
 
from dulwich.config import ConfigFile
 

	
 
from kallithea.lib.vcs.backends.base import BaseChangeset, EmptyChangeset
 
from kallithea.lib.vcs.conf import settings
 
from kallithea.lib.vcs.exceptions import ChangesetDoesNotExistError, ChangesetError, ImproperArchiveTypeError, NodeDoesNotExistError, RepositoryError, VCSError
 
from kallithea.lib.vcs.nodes import (
 
    AddedFileNodesGenerator, ChangedFileNodesGenerator, DirNode, FileNode, NodeKind, RemovedFileNodesGenerator, RootNode, SubModuleNode)
 
from kallithea.lib.vcs.utils import ascii_bytes, ascii_str, date_fromtimestamp, safe_int, safe_str
 
from kallithea.lib.vcs.utils.lazy import LazyProperty
 

	
 

	
 
class GitChangeset(BaseChangeset):
 
    """
 
    Represents state of the repository at a revision.
 
    """
 

	
 
    def __init__(self, repository, revision):
 
        self._stat_modes = {}
 
        self.repository = repository
 
        try:
 
            commit = self.repository._repo[ascii_bytes(revision)]
 
            if isinstance(commit, objects.Tag):
 
                revision = safe_str(commit.object[1])
 
                commit = self.repository._repo.get_object(commit.object[1])
 
        except KeyError:
 
            raise RepositoryError("Cannot get object with id %s" % revision)
 
        self.raw_id = ascii_str(commit.id)
 
        self.short_id = self.raw_id[:12]
 
        self._commit = commit  # a Dulwich Commmit with .id
 
        self._tree_id = commit.tree
 
        self._committer_property = 'committer'
 
        self._author_property = 'author'
 
        self._date_property = 'commit_time'
 
        self._date_tz_property = 'commit_timezone'
 
        self.revision = repository.revisions.index(self.raw_id)
 

	
 
        self.nodes = {}
 
        self._paths = {}
 

	
 
    @LazyProperty
 
    def bookmarks(self):
 
        return ()
 

	
 
    @LazyProperty
 
    def message(self):
 
        return safe_str(self._commit.message)
 

	
 
    @LazyProperty
 
    def committer(self):
 
        return safe_str(getattr(self._commit, self._committer_property))
 

	
 
    @LazyProperty
 
    def author(self):
 
        return safe_str(getattr(self._commit, self._author_property))
 

	
 
    @LazyProperty
 
    def date(self):
 
        return date_fromtimestamp(getattr(self._commit, self._date_property),
 
                                  getattr(self._commit, self._date_tz_property))
 

	
 
    @LazyProperty
 
    def _timestamp(self):
 
        return getattr(self._commit, self._date_property)
 

	
 
    @LazyProperty
 
    def status(self):
 
        """
 
        Returns modified, added, removed, deleted files for current changeset
 
        """
 
        return self.changed, self.added, self.removed
 

	
 
    @LazyProperty
 
    def tags(self):
 
        _tags = []
 
        for tname, tsha in self.repository.tags.items():
 
            if tsha == self.raw_id:
 
                _tags.append(tname)
 
        return _tags
 

	
 
    @LazyProperty
 
    def branch(self):
 
        # Note: This function will return one branch name for the changeset -
 
        # that might not make sense in Git where branches() is a better match
 
        # for the basic model
 
        heads = self.repository._heads(reverse=False)
 
        ref = heads.get(self._commit.id)
 
        if ref:
 
            return safe_str(ref)
 

	
 
    @LazyProperty
 
    def branches(self):
 
        heads = self.repository._heads(reverse=True)
 
        return [b for b in heads if heads[b] == self._commit.id] # FIXME: Inefficient ... and returning None!
 
        return [safe_str(b) for b in heads if heads[b] == self._commit.id] # FIXME: Inefficient ... and returning None!
 

	
 
    def _fix_path(self, path):
 
        """
 
        Paths are stored without trailing slash so we need to get rid off it if
 
        needed.
 
        """
 
        if path.endswith('/'):
 
            path = path.rstrip('/')
 
        return path
 

	
 
    def _get_id_for_path(self, path):
 
        # FIXME: Please, spare a couple of minutes and make those codes cleaner;
 
        if path not in self._paths:
 
            path = path.strip('/')
 
            # set root tree
 
            tree = self.repository._repo[self._tree_id]
 
            if path == '':
 
                self._paths[''] = tree.id
 
                return tree.id
 
            splitted = path.split('/')
 
            dirs, name = splitted[:-1], splitted[-1]
 
            curdir = ''
 

	
 
            # initially extract things from root dir
 
            for item, stat, id in tree.items():
 
                name = safe_str(item)
 
                if curdir:
 
                    name = '/'.join((curdir, item))
 
                else:
 
                    name = item
 
                    name = '/'.join((curdir, name))
 
                self._paths[name] = id
 
                self._stat_modes[name] = stat
 

	
 
            for dir in dirs:
 
                if curdir:
 
                    curdir = '/'.join((curdir, dir))
 
                else:
 
                    curdir = dir
 
                dir_id = None
 
                for item, stat, id in tree.items():
 
                    if dir == item:
 
                    name = safe_str(item)
 
                    if dir == name:
 
                        dir_id = id
 
                if dir_id:
 
                    # Update tree
 
                    tree = self.repository._repo[dir_id]
 
                    if not isinstance(tree, objects.Tree):
 
                        raise ChangesetError('%s is not a directory' % curdir)
 
                else:
 
                    raise ChangesetError('%s have not been found' % curdir)
 

	
 
                # cache all items from the given traversed tree
 
                for item, stat, id in tree.items():
 
                    name = safe_str(item)
 
                    if curdir:
 
                        name = '/'.join((curdir, item))
 
                    else:
 
                        name = item
 
                        name = '/'.join((curdir, name))
 
                    self._paths[name] = id
 
                    self._stat_modes[name] = stat
 
            if path not in self._paths:
 
                raise NodeDoesNotExistError("There is no file nor directory "
 
                    "at the given path '%s' at revision %s"
 
                    % (path, self.short_id))
 
        return self._paths[path]
 

	
 
    def _get_kind(self, path):
 
        obj = self.repository._repo[self._get_id_for_path(path)]
 
        if isinstance(obj, objects.Blob):
 
            return NodeKind.FILE
 
        elif isinstance(obj, objects.Tree):
 
            return NodeKind.DIR
 

	
 
    def _get_filectx(self, path):
 
        path = self._fix_path(path)
 
        if self._get_kind(path) != NodeKind.FILE:
 
            raise ChangesetError("File does not exist for revision %s at "
 
                " '%s'" % (self.raw_id, path))
 
        return path
 

	
 
    def _get_file_nodes(self):
 
        return chain(*(t[2] for t in self.walk()))
 

	
 
    @LazyProperty
 
    def parents(self):
 
        """
 
        Returns list of parents changesets.
 
        """
 
        return [self.repository.get_changeset(ascii_str(parent_id))
 
                for parent_id in self._commit.parents]
 

	
 
    @LazyProperty
 
    def children(self):
 
        """
 
        Returns list of children changesets.
 
        """
 
        rev_filter = settings.GIT_REV_FILTER
 
        so = self.repository.run_git_command(
 
            ['rev-list', rev_filter, '--children']
 
        )
 
        return [
 
            self.repository.get_changeset(cs)
 
            for parts in (l.split(' ') for l in so.splitlines())
 
            if parts[0] == self.raw_id
 
            for cs in parts[1:]
 
        ]
 

	
 
    def next(self, branch=None):
 
        if branch and self.branch != branch:
 
            raise VCSError('Branch option used on changeset not belonging '
 
                           'to that branch')
 

	
 
        cs = self
 
        while True:
 
            try:
 
                next_ = cs.revision + 1
 
                next_rev = cs.repository.revisions[next_]
 
            except IndexError:
 
                raise ChangesetDoesNotExistError
 
            cs = cs.repository.get_changeset(next_rev)
 

	
 
            if not branch or branch == cs.branch:
 
                return cs
 

	
 
    def prev(self, branch=None):
 
        if branch and self.branch != branch:
 
            raise VCSError('Branch option used on changeset not belonging '
 
                           'to that branch')
 

	
 
        cs = self
 
        while True:
 
            try:
 
                prev_ = cs.revision - 1
 
                if prev_ < 0:
 
                    raise IndexError
 
                prev_rev = cs.repository.revisions[prev_]
 
            except IndexError:
 
                raise ChangesetDoesNotExistError
 
            cs = cs.repository.get_changeset(prev_rev)
 

	
 
            if not branch or branch == cs.branch:
 
                return cs
 

	
 
    def diff(self, ignore_whitespace=True, context=3):
 
        # Only used to feed diffstat
 
        rev1 = self.parents[0] if self.parents else self.repository.EMPTY_CHANGESET
 
        rev2 = self
 
        return b''.join(self.repository.get_diff(rev1, rev2,
 
                                    ignore_whitespace=ignore_whitespace,
 
                                    context=context))
 

	
 
    def get_file_mode(self, path):
 
        """
 
        Returns stat mode of the file at the given ``path``.
 
@@ -311,242 +310,241 @@ class GitChangeset(BaseChangeset):
 
        return [self.repository.get_changeset(ascii_str(x.commit.id.decode))
 
                for x in walker]
 

	
 
    def get_file_annotate(self, path):
 
        """
 
        Returns a generator of four element tuples with
 
            lineno, sha, changeset lazy loader and line
 
        """
 
        # TODO: This function now uses os underlying 'git' command which is
 
        # generally not good. Should be replaced with algorithm iterating
 
        # commits.
 
        cmd = ['blame', '-l', '--root', '-r', self.raw_id, '--', path]
 
        # -l     ==> outputs long shas (and we need all 40 characters)
 
        # --root ==> doesn't put '^' character for boundaries
 
        # -r sha ==> blames for the given revision
 
        so = self.repository.run_git_command(cmd)
 

	
 
        for i, blame_line in enumerate(so.split('\n')[:-1]):
 
            sha, line = re.split(r' ', blame_line, 1)
 
            yield (i + 1, sha, lambda sha=sha: self.repository.get_changeset(sha), line)
 

	
 
    def fill_archive(self, stream=None, kind='tgz', prefix=None,
 
                     subrepos=False):
 
        """
 
        Fills up given stream.
 

	
 
        :param stream: file like object.
 
        :param kind: one of following: ``zip``, ``tgz`` or ``tbz2``.
 
            Default: ``tgz``.
 
        :param prefix: name of root directory in archive.
 
            Default is repository name and changeset's raw_id joined with dash
 
            (``repo-tip.<KIND>``).
 
        :param subrepos: include subrepos in this archive.
 

	
 
        :raise ImproperArchiveTypeError: If given kind is wrong.
 
        :raise VcsError: If given stream is None
 
        """
 
        allowed_kinds = settings.ARCHIVE_SPECS
 
        if kind not in allowed_kinds:
 
            raise ImproperArchiveTypeError('Archive kind not supported use one'
 
                'of %s' % ' '.join(allowed_kinds))
 

	
 
        if stream is None:
 
            raise VCSError('You need to pass in a valid stream for filling'
 
                           ' with archival data')
 

	
 
        if prefix is None:
 
            prefix = '%s-%s' % (self.repository.name, self.short_id)
 
        elif prefix.startswith('/'):
 
            raise VCSError("Prefix cannot start with leading slash")
 
        elif prefix.strip() == '':
 
            raise VCSError("Prefix cannot be empty")
 

	
 
        if kind == 'zip':
 
            frmt = 'zip'
 
        else:
 
            frmt = 'tar'
 
        _git_path = settings.GIT_EXECUTABLE_PATH
 
        cmd = '%s archive --format=%s --prefix=%s/ %s' % (_git_path,
 
                                                frmt, prefix, self.raw_id)
 
        if kind == 'tgz':
 
            cmd += ' | gzip -9'
 
        elif kind == 'tbz2':
 
            cmd += ' | bzip2 -9'
 

	
 
        if stream is None:
 
            raise VCSError('You need to pass in a valid stream for filling'
 
                           ' with archival data')
 
        popen = Popen(cmd, stdout=PIPE, stderr=PIPE, shell=True,
 
                      cwd=self.repository.path)
 

	
 
        buffer_size = 1024 * 8
 
        chunk = popen.stdout.read(buffer_size)
 
        while chunk:
 
            stream.write(chunk)
 
            chunk = popen.stdout.read(buffer_size)
 
        # Make sure all descriptors would be read
 
        popen.communicate()
 

	
 
    def get_nodes(self, path):
 
        """
 
        Returns combined ``DirNode`` and ``FileNode`` objects list representing
 
        state of changeset at the given ``path``. If node at the given ``path``
 
        is not instance of ``DirNode``, ChangesetError would be raised.
 
        """
 

	
 
        if self._get_kind(path) != NodeKind.DIR:
 
            raise ChangesetError("Directory does not exist for revision %s at "
 
                " '%s'" % (self.revision, path))
 
        path = self._fix_path(path)
 
        id = self._get_id_for_path(path)
 
        tree = self.repository._repo[id]
 
        dirnodes = []
 
        filenodes = []
 
        als = self.repository.alias
 
        for name, stat, id in tree.items():
 
            obj_path = safe_str(name)
 
            if path != '':
 
                obj_path = '/'.join((path, name))
 
            else:
 
                obj_path = name
 
                obj_path = '/'.join((path, obj_path))
 
            if objects.S_ISGITLINK(stat):
 
                root_tree = self.repository._repo[self._tree_id]
 
                cf = ConfigFile.from_file(BytesIO(self.repository._repo.get_object(root_tree[b'.gitmodules'][1]).data))
 
                url = ascii_str(cf.get(('submodule', obj_path), 'url'))
 
                dirnodes.append(SubModuleNode(obj_path, url=url, changeset=ascii_str(id),
 
                                              alias=als))
 
                continue
 

	
 
            obj = self.repository._repo.get_object(id)
 
            if obj_path not in self._stat_modes:
 
                self._stat_modes[obj_path] = stat
 
            if isinstance(obj, objects.Tree):
 
                dirnodes.append(DirNode(obj_path, changeset=self))
 
            elif isinstance(obj, objects.Blob):
 
                filenodes.append(FileNode(obj_path, changeset=self, mode=stat))
 
            else:
 
                raise ChangesetError("Requested object should be Tree "
 
                                     "or Blob, is %r" % type(obj))
 
        nodes = dirnodes + filenodes
 
        for node in nodes:
 
            if node.path not in self.nodes:
 
                self.nodes[node.path] = node
 
        nodes.sort()
 
        return nodes
 

	
 
    def get_node(self, path):
 
        """
 
        Returns ``Node`` object from the given ``path``. If there is no node at
 
        the given ``path``, ``ChangesetError`` would be raised.
 
        """
 
        path = self._fix_path(path)
 
        if path not in self.nodes:
 
            try:
 
                id_ = self._get_id_for_path(path)
 
            except ChangesetError:
 
                raise NodeDoesNotExistError("Cannot find one of parents' "
 
                    "directories for a given path: %s" % path)
 

	
 
            _GL = lambda m: m and objects.S_ISGITLINK(m)
 
            if _GL(self._stat_modes.get(path)):
 
                tree = self.repository._repo[self._tree_id]
 
                cf = ConfigFile.from_file(BytesIO(self.repository._repo.get_object(tree[b'.gitmodules'][1]).data))
 
                url = ascii_str(cf.get(('submodule', path), 'url'))
 
                node = SubModuleNode(path, url=url, changeset=ascii_str(id_),
 
                                     alias=self.repository.alias)
 
            else:
 
                obj = self.repository._repo.get_object(id_)
 

	
 
                if isinstance(obj, objects.Tree):
 
                    if path == '':
 
                        node = RootNode(changeset=self)
 
                    else:
 
                        node = DirNode(path, changeset=self)
 
                    node._tree = obj
 
                elif isinstance(obj, objects.Blob):
 
                    node = FileNode(path, changeset=self)
 
                    node._blob = obj
 
                else:
 
                    raise NodeDoesNotExistError("There is no file nor directory "
 
                        "at the given path: '%s' at revision %s"
 
                        % (path, self.short_id))
 
            # cache node
 
            self.nodes[path] = node
 
        return self.nodes[path]
 

	
 
    @LazyProperty
 
    def affected_files(self):
 
        """
 
        Gets a fast accessible file changes for given changeset
 
        """
 
        added, modified, deleted = self._changes_cache
 
        return list(added.union(modified).union(deleted))
 

	
 
    @LazyProperty
 
    def _changes_cache(self):
 
        added = set()
 
        modified = set()
 
        deleted = set()
 
        _r = self.repository._repo
 

	
 
        parents = self.parents
 
        if not self.parents:
 
            parents = [EmptyChangeset()]
 
        for parent in parents:
 
            if isinstance(parent, EmptyChangeset):
 
                oid = None
 
            else:
 
                oid = _r[parent._commit.id].tree
 
            changes = _r.object_store.tree_changes(oid, _r[self._commit.id].tree)
 
            for (oldpath, newpath), (_, _), (_, _) in changes:
 
                if newpath and oldpath:
 
                    modified.add(newpath)
 
                    modified.add(safe_str(newpath))
 
                elif newpath and not oldpath:
 
                    added.add(newpath)
 
                    added.add(safe_str(newpath))
 
                elif not newpath and oldpath:
 
                    deleted.add(oldpath)
 
                    deleted.add(safe_str(oldpath))
 
        return added, modified, deleted
 

	
 
    def _get_paths_for_status(self, status):
 
        """
 
        Returns sorted list of paths for given ``status``.
 

	
 
        :param status: one of: *added*, *modified* or *deleted*
 
        """
 
        added, modified, deleted = self._changes_cache
 
        return sorted({
 
            'added': list(added),
 
            'modified': list(modified),
 
            'deleted': list(deleted)}[status]
 
        )
 

	
 
    @LazyProperty
 
    def added(self):
 
        """
 
        Returns list of added ``FileNode`` objects.
 
        """
 
        if not self.parents:
 
            return list(self._get_file_nodes())
 
        return AddedFileNodesGenerator([n for n in
 
                                self._get_paths_for_status('added')], self)
 

	
 
    @LazyProperty
 
    def changed(self):
 
        """
 
        Returns list of modified ``FileNode`` objects.
 
        """
 
        if not self.parents:
 
            return []
 
        return ChangedFileNodesGenerator([n for n in
 
                                self._get_paths_for_status('modified')], self)
 

	
 
    @LazyProperty
 
    def removed(self):
 
        """
 
        Returns list of removed ``FileNode`` objects.
 
        """
 
        if not self.parents:
 
            return []
 
        return RemovedFileNodesGenerator([n for n in
 
                                self._get_paths_for_status('deleted')], self)
 

	
 
    extra = {}
kallithea/lib/vcs/backends/git/repository.py
Show inline comments
 
@@ -266,252 +266,252 @@ class GitRepository(BaseRepository):
 

	
 
        if revision in (None, '', 'tip', 'HEAD', 'head', -1):
 
            revision = -1
 

	
 
        if isinstance(revision, int):
 
            try:
 
                return self.revisions[revision]
 
            except IndexError:
 
                msg = "Revision %r does not exist for %s" % (revision, self.name)
 
                raise ChangesetDoesNotExistError(msg)
 

	
 
        if isinstance(revision, (str, unicode)):
 
            if revision.isdigit() and (len(revision) < 12 or len(revision) == revision.count('0')):
 
                try:
 
                    return self.revisions[int(revision)]
 
                except IndexError:
 
                    msg = "Revision %r does not exist for %s" % (revision, self)
 
                    raise ChangesetDoesNotExistError(msg)
 

	
 
            # get by branch/tag name
 
            _ref_revision = self._parsed_refs.get(safe_bytes(revision))
 
            if _ref_revision:  # and _ref_revision[1] in [b'H', b'RH', b'T']:
 
                return ascii_str(_ref_revision[0])
 

	
 
            if revision in self.revisions:
 
                return revision
 

	
 
            # maybe it's a tag ? we don't have them in self.revisions
 
            if revision in self.tags.values():
 
                return revision
 

	
 
            if SHA_PATTERN.match(revision):
 
                msg = "Revision %r does not exist for %s" % (revision, self.name)
 
                raise ChangesetDoesNotExistError(msg)
 

	
 
        raise ChangesetDoesNotExistError("Given revision %r not recognized" % revision)
 

	
 
    def get_ref_revision(self, ref_type, ref_name):
 
        """
 
        Returns ``GitChangeset`` object representing repository's
 
        changeset at the given ``revision``.
 
        """
 
        return self._get_revision(ref_name)
 

	
 
    def _get_archives(self, archive_name='tip'):
 

	
 
        for i in [('zip', '.zip'), ('gz', '.tar.gz'), ('bz2', '.tar.bz2')]:
 
            yield {"type": i[0], "extension": i[1], "node": archive_name}
 

	
 
    def _get_url(self, url):
 
        """
 
        Returns normalized url. If schema is not given, would fall to
 
        filesystem (``file:///``) schema.
 
        """
 
        if url != 'default' and '://' not in url:
 
            url = ':///'.join(('file', url))
 
        return url
 

	
 
    @LazyProperty
 
    def name(self):
 
        return os.path.basename(self.path)
 

	
 
    @LazyProperty
 
    def last_change(self):
 
        """
 
        Returns last change made on this repository as datetime object
 
        """
 
        return date_fromtimestamp(self._get_mtime(), makedate()[1])
 

	
 
    def _get_mtime(self):
 
        try:
 
            return time.mktime(self.get_changeset().date.timetuple())
 
        except RepositoryError:
 
            idx_loc = '' if self.bare else '.git'
 
            # fallback to filesystem
 
            in_path = os.path.join(self.path, idx_loc, "index")
 
            he_path = os.path.join(self.path, idx_loc, "HEAD")
 
            if os.path.exists(in_path):
 
                return os.stat(in_path).st_mtime
 
            else:
 
                return os.stat(he_path).st_mtime
 

	
 
    @LazyProperty
 
    def description(self):
 
        return safe_str(self._repo.get_description() or b'unknown')
 

	
 
    @LazyProperty
 
    def contact(self):
 
        undefined_contact = u'Unknown'
 
        return undefined_contact
 

	
 
    @property
 
    def branches(self):
 
        if not self.revisions:
 
            return {}
 
        sortkey = lambda ctx: ctx[0]
 
        _branches = [(key, ascii_str(sha))
 
        _branches = [(safe_str(key), ascii_str(sha))
 
                     for key, (sha, type_) in self._parsed_refs.items() if type_ == b'H']
 
        return OrderedDict(sorted(_branches, key=sortkey, reverse=False))
 

	
 
    @LazyProperty
 
    def closed_branches(self):
 
        return {}
 

	
 
    @LazyProperty
 
    def tags(self):
 
        return self._get_tags()
 

	
 
    def _get_tags(self):
 
        if not self.revisions:
 
            return {}
 

	
 
        sortkey = lambda ctx: ctx[0]
 
        _tags = [(key, ascii_str(sha))
 
        _tags = [(safe_str(key), ascii_str(sha))
 
                 for key, (sha, type_) in self._parsed_refs.items() if type_ == b'T']
 
        return OrderedDict(sorted(_tags, key=sortkey, reverse=True))
 

	
 
    def tag(self, name, user, revision=None, message=None, date=None,
 
            **kwargs):
 
        """
 
        Creates and returns a tag for the given ``revision``.
 

	
 
        :param name: name for new tag
 
        :param user: full username, i.e.: "Joe Doe <joe.doe@example.com>"
 
        :param revision: changeset id for which new tag would be created
 
        :param message: message of the tag's commit
 
        :param date: date of tag's commit
 

	
 
        :raises TagAlreadyExistError: if tag with same name already exists
 
        """
 
        if name in self.tags:
 
            raise TagAlreadyExistError("Tag %s already exists" % name)
 
        changeset = self.get_changeset(revision)
 
        message = message or "Added tag %s for commit %s" % (name,
 
            changeset.raw_id)
 
        self._repo.refs[b"refs/tags/%s" % safe_bytes(name)] = changeset._commit.id
 

	
 
        self._parsed_refs = self._get_parsed_refs()
 
        self.tags = self._get_tags()
 
        return changeset
 

	
 
    def remove_tag(self, name, user, message=None, date=None):
 
        """
 
        Removes tag with the given ``name``.
 

	
 
        :param name: name of the tag to be removed
 
        :param user: full username, i.e.: "Joe Doe <joe.doe@example.com>"
 
        :param message: message of the tag's removal commit
 
        :param date: date of tag's removal commit
 

	
 
        :raises TagDoesNotExistError: if tag with given name does not exists
 
        """
 
        if name not in self.tags:
 
            raise TagDoesNotExistError("Tag %s does not exist" % name)
 
        # self._repo.refs is a DiskRefsContainer, and .path gives the full absolute path of '.git'
 
        tagpath = os.path.join(self._repo.refs.path, 'refs', 'tags', name)
 
        tagpath = os.path.join(safe_str(self._repo.refs.path), 'refs', 'tags', name)
 
        try:
 
            os.remove(tagpath)
 
            self._parsed_refs = self._get_parsed_refs()
 
            self.tags = self._get_tags()
 
        except OSError as e:
 
            raise RepositoryError(e.strerror)
 

	
 
    @LazyProperty
 
    def bookmarks(self):
 
        """
 
        Gets bookmarks for this repository
 
        """
 
        return {}
 

	
 
    @LazyProperty
 
    def _parsed_refs(self):
 
        return self._get_parsed_refs()
 

	
 
    def _get_parsed_refs(self):
 
        """Return refs as a dict, like:
 
        { b'v0.2.0': [b'599ba911aa24d2981225f3966eb659dfae9e9f30', b'T'] }
 
        """
 
        _repo = self._repo
 
        refs = _repo.get_refs()
 
        keys = [(b'refs/heads/', b'H'),
 
                (b'refs/remotes/origin/', b'RH'),
 
                (b'refs/tags/', b'T')]
 
        _refs = {}
 
        for ref, sha in refs.items():
 
            for k, type_ in keys:
 
                if ref.startswith(k):
 
                    _key = ref[len(k):]
 
                    if type_ == b'T':
 
                        obj = _repo.get_object(sha)
 
                        if isinstance(obj, Tag):
 
                            sha = _repo.get_object(sha).object[1]
 
                    _refs[_key] = [sha, type_]
 
                    break
 
        return _refs
 

	
 
    def _heads(self, reverse=False):
 
        refs = self._repo.get_refs()
 
        heads = {}
 

	
 
        for key, val in refs.items():
 
            for ref_key in [b'refs/heads/', b'refs/remotes/origin/']:
 
                if key.startswith(ref_key):
 
                    n = key[len(ref_key):]
 
                    if n not in [b'HEAD']:
 
                        heads[n] = val
 

	
 
        return heads if reverse else dict((y, x) for x, y in heads.items())
 

	
 
    def get_changeset(self, revision=None):
 
        """
 
        Returns ``GitChangeset`` object representing commit from git repository
 
        at the given revision or head (most recent commit) if None given.
 
        """
 
        if isinstance(revision, GitChangeset):
 
            return revision
 
        return GitChangeset(repository=self, revision=self._get_revision(revision))
 

	
 
    def get_changesets(self, start=None, end=None, start_date=None,
 
           end_date=None, branch_name=None, reverse=False, max_revisions=None):
 
        """
 
        Returns iterator of ``GitChangeset`` objects from start to end (both
 
        are inclusive), in ascending date order (unless ``reverse`` is set).
 

	
 
        :param start: changeset ID, as str; first returned changeset
 
        :param end: changeset ID, as str; last returned changeset
 
        :param start_date: if specified, changesets with commit date less than
 
          ``start_date`` would be filtered out from returned set
 
        :param end_date: if specified, changesets with commit date greater than
 
          ``end_date`` would be filtered out from returned set
 
        :param branch_name: if specified, changesets not reachable from given
 
          branch would be filtered out from returned set
 
        :param reverse: if ``True``, returned generator would be reversed
 
          (meaning that returned changesets would have descending date order)
 

	
 
        :raise BranchDoesNotExistError: If given ``branch_name`` does not
 
            exist.
 
        :raise ChangesetDoesNotExistError: If changeset for given ``start`` or
 
          ``end`` could not be found.
 

	
 
        """
 
        if branch_name and branch_name not in self.branches:
 
            raise BranchDoesNotExistError("Branch '%s' not found"
 
                                          % branch_name)
 
        # actually we should check now if it's not an empty repo to not spaw
 
        # subprocess commands
 
        if self._empty:
 
            raise EmptyRepositoryError("There are no changesets yet")
 

	
 
        # %H at format means (full) commit hash, initial hashes are retrieved
 
        # in ascending date order
 
        cmd = ['log', '--date-order', '--reverse', '--pretty=format:%H']
 
@@ -619,118 +619,119 @@ class GitRepository(BaseRepository):
 
        if rev1 == self.EMPTY_CHANGESET:
 
            parts = stdout.split(b'\ndiff ', 1)
 
            if len(parts) > 1:
 
                stdout = b'diff ' + parts[1]
 
        return stdout
 

	
 
    @LazyProperty
 
    def in_memory_changeset(self):
 
        """
 
        Returns ``GitInMemoryChangeset`` object for this repository.
 
        """
 
        return GitInMemoryChangeset(self)
 

	
 
    def clone(self, url, update_after_clone=True, bare=False):
 
        """
 
        Tries to clone changes from external location.
 

	
 
        :param update_after_clone: If set to ``False``, git won't checkout
 
          working directory
 
        :param bare: If set to ``True``, repository would be cloned into
 
          *bare* git repository (no working directory at all).
 
        """
 
        url = self._get_url(url)
 
        cmd = ['clone', '-q']
 
        if bare:
 
            cmd.append('--bare')
 
        elif not update_after_clone:
 
            cmd.append('--no-checkout')
 
        cmd += ['--', url, self.path]
 
        # If error occurs run_git_command raises RepositoryError already
 
        self.run_git_command(cmd)
 

	
 
    def pull(self, url):
 
        """
 
        Tries to pull changes from external location.
 
        """
 
        url = self._get_url(url)
 
        cmd = ['pull', '--ff-only', url]
 
        # If error occurs run_git_command raises RepositoryError already
 
        self.run_git_command(cmd)
 

	
 
    def fetch(self, url):
 
        """
 
        Tries to pull changes from external location.
 
        """
 
        url = self._get_url(url)
 
        so = self.run_git_command(['ls-remote', '-h', url])
 
        cmd = ['fetch', url, '--']
 
        for line in (x for x in so.splitlines()):
 
            sha, ref = line.split('\t')
 
            cmd.append('+%s:%s' % (ref, ref))
 
        self.run_git_command(cmd)
 

	
 
    def _update_server_info(self):
 
        """
 
        runs gits update-server-info command in this repo instance
 
        """
 
        from dulwich.server import update_server_info
 
        try:
 
            update_server_info(self._repo)
 
        except OSError as e:
 
            if e.errno not in [errno.ENOENT, errno.EROFS]:
 
                raise
 
            # Workaround for dulwich crashing on for example its own dulwich/tests/data/repos/simple_merge.git/info/refs.lock
 
            log.error('Ignoring %s running update-server-info: %s', type(e).__name__, e)
 

	
 
    @LazyProperty
 
    def workdir(self):
 
        """
 
        Returns ``Workdir`` instance for this repository.
 
        """
 
        return GitWorkdir(self)
 

	
 
    def get_config_value(self, section, name, config_file=None):
 
        """
 
        Returns configuration value for a given [``section``] and ``name``.
 

	
 
        :param section: Section we want to retrieve value from
 
        :param name: Name of configuration we want to retrieve
 
        :param config_file: A path to file which should be used to retrieve
 
          configuration from (might also be a list of file paths)
 
        """
 
        if config_file is None:
 
            config_file = []
 
        elif isinstance(config_file, str):
 
            config_file = [config_file]
 

	
 
        def gen_configs():
 
            for path in config_file + self._config_files:
 
                try:
 
                    yield ConfigFile.from_path(path)
 
                except (IOError, OSError, ValueError):
 
                    continue
 

	
 
        for config in gen_configs():
 
            try:
 
                return config.get(section, name)
 
                value = config.get(section, name)
 
            except KeyError:
 
                continue
 
            return None if value is None else safe_str(value)
 
        return None
 

	
 
    def get_user_name(self, config_file=None):
 
        """
 
        Returns user's name from global configuration file.
 

	
 
        :param config_file: A path to file which should be used to retrieve
 
          configuration from (might also be a list of file paths)
 
        """
 
        return self.get_config_value('user', 'name', config_file)
 

	
 
    def get_user_email(self, config_file=None):
 
        """
 
        Returns user's email from global configuration file.
 

	
 
        :param config_file: A path to file which should be used to retrieve
 
          configuration from (might also be a list of file paths)
 
        """
 
        return self.get_config_value('user', 'email', config_file)
kallithea/lib/vcs/backends/git/workdir.py
Show inline comments
 
import re
 

	
 
from kallithea.lib.utils2 import ascii_str
 
from kallithea.lib.utils2 import ascii_str, safe_str
 
from kallithea.lib.vcs.backends.base import BaseWorkdir
 
from kallithea.lib.vcs.exceptions import BranchDoesNotExistError, RepositoryError
 

	
 

	
 
class GitWorkdir(BaseWorkdir):
 

	
 
    def get_branch(self):
 
        headpath = self.repository._repo.refs.refpath(b'HEAD')
 
        try:
 
            content = open(headpath).read()
 
            content = safe_str(open(headpath, 'rb').read())
 
            match = re.match(r'^ref: refs/heads/(?P<branch>.+)\n$', content)
 
            if match:
 
                return match.groupdict()['branch']
 
            else:
 
                raise RepositoryError("Couldn't compute workdir's branch")
 
        except IOError:
 
            # Try naive way...
 
            raise RepositoryError("Couldn't compute workdir's branch")
 

	
 
    def get_changeset(self):
 
        wk_dir_id = ascii_str(self.repository._repo.refs.as_dict().get(b'HEAD'))
 
        return self.repository.get_changeset(wk_dir_id)
 

	
 
    def checkout_branch(self, branch=None):
 
        if branch is None:
 
            branch = self.repository.DEFAULT_BRANCH_NAME
 
        if branch not in self.repository.branches:
 
            raise BranchDoesNotExistError
 
        self.repository.run_git_command(['checkout', branch])
kallithea/lib/vcs/backends/hg/changeset.py
Show inline comments
 
@@ -28,193 +28,193 @@ class MercurialChangeset(BaseChangeset):
 
        self.revision = self._ctx._rev
 
        self.nodes = {}
 

	
 
    @LazyProperty
 
    def tags(self):
 
        return [safe_str(tag) for tag in self._ctx.tags()]
 

	
 
    @LazyProperty
 
    def branch(self):
 
        return safe_str(self._ctx.branch())
 

	
 
    @LazyProperty
 
    def branches(self):
 
        return [safe_str(self._ctx.branch())]
 

	
 
    @LazyProperty
 
    def closesbranch(self):
 
        return self._ctx.closesbranch()
 

	
 
    @LazyProperty
 
    def obsolete(self):
 
        return self._ctx.obsolete()
 

	
 
    @LazyProperty
 
    def bumped(self):
 
        return self._ctx.phasedivergent()
 

	
 
    @LazyProperty
 
    def divergent(self):
 
        return self._ctx.contentdivergent()
 

	
 
    @LazyProperty
 
    def extinct(self):
 
        return self._ctx.extinct()
 

	
 
    @LazyProperty
 
    def unstable(self):
 
        return self._ctx.orphan()
 

	
 
    @LazyProperty
 
    def phase(self):
 
        if(self._ctx.phase() == 1):
 
            return 'Draft'
 
        elif(self._ctx.phase() == 2):
 
            return 'Secret'
 
        else:
 
            return ''
 

	
 
    @LazyProperty
 
    def successors(self):
 
        successors = mercurial.obsutil.successorssets(self._ctx._repo, self._ctx.node(), closest=True)
 
        if successors:
 
            # flatten the list here handles both divergent (len > 1)
 
            # and the usual case (len = 1)
 
            successors = [mercurial.node.hex(n)[:12] for sub in successors for n in sub if n != self._ctx.node()]
 

	
 
        return successors
 

	
 
    @LazyProperty
 
    def predecessors(self):
 
        return [mercurial.node.hex(n)[:12] for n in mercurial.obsutil.closestpredecessors(self._ctx._repo, self._ctx.node())]
 

	
 
    @LazyProperty
 
    def bookmarks(self):
 
        return [safe_str(bookmark) for bookmark in self._ctx.bookmarks()]
 

	
 
    @LazyProperty
 
    def message(self):
 
        return safe_str(self._ctx.description())
 

	
 
    @LazyProperty
 
    def committer(self):
 
        return safe_str(self.author)
 

	
 
    @LazyProperty
 
    def author(self):
 
        return safe_str(self._ctx.user())
 

	
 
    @LazyProperty
 
    def date(self):
 
        return date_fromtimestamp(*self._ctx.date())
 

	
 
    @LazyProperty
 
    def _timestamp(self):
 
        return self._ctx.date()[0]
 

	
 
    @LazyProperty
 
    def status(self):
 
        """
 
        Returns modified, added, removed, deleted files for current changeset
 
        """
 
        return self.repository._repo.status(self._ctx.p1().node(),
 
                                            self._ctx.node())
 

	
 
    @LazyProperty
 
    def _file_paths(self):
 
        return list(self._ctx)
 
        return list(safe_str(f) for f in self._ctx)
 

	
 
    @LazyProperty
 
    def _dir_paths(self):
 
        p = list(set(get_dirs_for_path(*self._file_paths)))
 
        p.insert(0, '')
 
        return p
 

	
 
    @LazyProperty
 
    def _paths(self):
 
        return self._dir_paths + self._file_paths
 

	
 
    @LazyProperty
 
    def short_id(self):
 
        return self.raw_id[:12]
 

	
 
    @LazyProperty
 
    def parents(self):
 
        """
 
        Returns list of parents changesets.
 
        """
 
        return [self.repository.get_changeset(parent.rev())
 
                for parent in self._ctx.parents() if parent.rev() >= 0]
 

	
 
    @LazyProperty
 
    def children(self):
 
        """
 
        Returns list of children changesets.
 
        """
 
        return [self.repository.get_changeset(child.rev())
 
                for child in self._ctx.children() if child.rev() >= 0]
 

	
 
    def next(self, branch=None):
 
        if branch and self.branch != branch:
 
            raise VCSError('Branch option used on changeset not belonging '
 
                           'to that branch')
 

	
 
        cs = self
 
        while True:
 
            try:
 
                next_ = cs.repository.revisions.index(cs.raw_id) + 1
 
                next_rev = cs.repository.revisions[next_]
 
            except IndexError:
 
                raise ChangesetDoesNotExistError
 
            cs = cs.repository.get_changeset(next_rev)
 

	
 
            if not branch or branch == cs.branch:
 
                return cs
 

	
 
    def prev(self, branch=None):
 
        if branch and self.branch != branch:
 
            raise VCSError('Branch option used on changeset not belonging '
 
                           'to that branch')
 

	
 
        cs = self
 
        while True:
 
            try:
 
                prev_ = cs.repository.revisions.index(cs.raw_id) - 1
 
                if prev_ < 0:
 
                    raise IndexError
 
                prev_rev = cs.repository.revisions[prev_]
 
            except IndexError:
 
                raise ChangesetDoesNotExistError
 
            cs = cs.repository.get_changeset(prev_rev)
 

	
 
            if not branch or branch == cs.branch:
 
                return cs
 

	
 
    def diff(self):
 
        # Only used to feed diffstat
 
        return b''.join(self._ctx.diff())
 

	
 
    def _fix_path(self, path):
 
        """
 
        Paths are stored without trailing slash so we need to get rid off it if
 
        needed. Also mercurial keeps filenodes as str so we need to decode
 
        from unicode to str
 
        """
 
        if path.endswith('/'):
 
            path = path.rstrip('/')
 

	
 
        return path
 

	
 
    def _get_kind(self, path):
 
        path = self._fix_path(path)
 
        if path in self._file_paths:
 
            return NodeKind.FILE
 
        elif path in self._dir_paths:
 
            return NodeKind.DIR
 
        else:
 
            raise ChangesetError("Node does not exist at the given path '%s'"
 
                % (path))
 

	
 
    def _get_filectx(self, path):
 
        path = self._fix_path(path)
 
        if self._get_kind(path) != NodeKind.FILE:
 
            raise ChangesetError("File does not exist for revision %s at "
 
@@ -293,115 +293,115 @@ class MercurialChangeset(BaseChangeset):
 
        :param stream: file like object.
 
        :param kind: one of following: ``zip``, ``tgz`` or ``tbz2``.
 
            Default: ``tgz``.
 
        :param prefix: name of root directory in archive.
 
            Default is repository name and changeset's raw_id joined with dash
 
            (``repo-tip.<KIND>``).
 
        :param subrepos: include subrepos in this archive.
 

	
 
        :raise ImproperArchiveTypeError: If given kind is wrong.
 
        :raise VcsError: If given stream is None
 
        """
 
        allowed_kinds = settings.ARCHIVE_SPECS
 
        if kind not in allowed_kinds:
 
            raise ImproperArchiveTypeError('Archive kind not supported use one'
 
                'of %s' % ' '.join(allowed_kinds))
 

	
 
        if stream is None:
 
            raise VCSError('You need to pass in a valid stream for filling'
 
                           ' with archival data')
 

	
 
        if prefix is None:
 
            prefix = '%s-%s' % (self.repository.name, self.short_id)
 
        elif prefix.startswith('/'):
 
            raise VCSError("Prefix cannot start with leading slash")
 
        elif prefix.strip() == '':
 
            raise VCSError("Prefix cannot be empty")
 

	
 
        mercurial.archival.archive(self.repository._repo, stream, ascii_bytes(self.raw_id),
 
                         safe_bytes(kind), prefix=safe_bytes(prefix), subrepos=subrepos)
 

	
 
    def get_nodes(self, path):
 
        """
 
        Returns combined ``DirNode`` and ``FileNode`` objects list representing
 
        state of changeset at the given ``path``. If node at the given ``path``
 
        is not instance of ``DirNode``, ChangesetError would be raised.
 
        """
 

	
 
        if self._get_kind(path) != NodeKind.DIR:
 
            raise ChangesetError("Directory does not exist for revision %s at "
 
                " '%s'" % (self.revision, path))
 
        path = self._fix_path(path)
 

	
 
        filenodes = [FileNode(f, changeset=self) for f in self._file_paths
 
            if os.path.dirname(f) == path]
 
        dirs = path == '' and '' or [d for d in self._dir_paths
 
            if d and posixpath.dirname(d) == path]
 
        dirnodes = [DirNode(d, changeset=self) for d in dirs
 
            if os.path.dirname(d) == path]
 

	
 
        als = self.repository.alias
 
        for k, vals in self._extract_submodules().items():
 
            #vals = url,rev,type
 
            loc = vals[0]
 
            cs = vals[1]
 
            dirnodes.append(SubModuleNode(k, url=loc, changeset=cs,
 
                                          alias=als))
 
        nodes = dirnodes + filenodes
 
        for node in nodes:
 
            self.nodes[node.path] = node
 
        nodes.sort()
 
        return nodes
 

	
 
    def get_node(self, path):
 
        """
 
        Returns ``Node`` object from the given ``path``. If there is no node at
 
        the given ``path``, ``ChangesetError`` would be raised.
 
        """
 
        path = self._fix_path(path)
 
        if path not in self.nodes:
 
            if path in self._file_paths:
 
                node = FileNode(path, changeset=self)
 
            elif path in self._dir_paths or path in self._dir_paths:
 
                if path == '':
 
                    node = RootNode(changeset=self)
 
                else:
 
                    node = DirNode(path, changeset=self)
 
            else:
 
                raise NodeDoesNotExistError("There is no file nor directory "
 
                    "at the given path: '%s' at revision %s"
 
                    % (path, self.short_id))
 
            # cache node
 
            self.nodes[path] = node
 
        return self.nodes[path]
 

	
 
    @LazyProperty
 
    def affected_files(self):
 
        """
 
        Gets a fast accessible file changes for given changeset
 
        """
 
        return self._ctx.files()
 

	
 
    @property
 
    def added(self):
 
        """
 
        Returns list of added ``FileNode`` objects.
 
        """
 
        return AddedFileNodesGenerator([n for n in self.status.added], self)
 
        return AddedFileNodesGenerator([safe_str(n) for n in self.status.added], self)
 

	
 
    @property
 
    def changed(self):
 
        """
 
        Returns list of modified ``FileNode`` objects.
 
        """
 
        return ChangedFileNodesGenerator([n for n in self.status.modified], self)
 
        return ChangedFileNodesGenerator([safe_str(n) for n in self.status.modified], self)
 

	
 
    @property
 
    def removed(self):
 
        """
 
        Returns list of removed ``FileNode`` objects.
 
        """
 
        return RemovedFileNodesGenerator([n for n in self.status.removed], self)
 
        return RemovedFileNodesGenerator([safe_str(n) for n in self.status.removed], self)
 

	
 
    @LazyProperty
 
    def extra(self):
 
        return self._ctx.extra()
kallithea/lib/vcs/backends/hg/inmemory.py
Show inline comments
 
import datetime
 

	
 
import mercurial.context
 
import mercurial.node
 

	
 
from kallithea.lib.vcs.backends.base import BaseInMemoryChangeset
 
from kallithea.lib.vcs.exceptions import RepositoryError
 
from kallithea.lib.vcs.utils import ascii_str, safe_bytes
 
from kallithea.lib.vcs.utils import ascii_str, safe_bytes, safe_str
 

	
 

	
 
class MercurialInMemoryChangeset(BaseInMemoryChangeset):
 

	
 
    def commit(self, message, author, parents=None, branch=None, date=None,
 
               **kwargs):
 
        """
 
        Performs in-memory commit (doesn't check workdir in any way) and
 
        returns newly created ``Changeset``. Updates repository's
 
        ``revisions``.
 

	
 
        :param message: message of the commit
 
        :param author: full username, i.e. "Joe Doe <joe.doe@example.com>"
 
        :param parents: single parent or sequence of parents from which commit
 
          would be derived
 
        :param date: ``datetime.datetime`` instance. Defaults to
 
          ``datetime.datetime.now()``.
 
        :param branch: branch name, as string. If none given, default backend's
 
          branch would be used.
 

	
 
        :raises ``CommitError``: if any error occurs while committing
 
        """
 
        self.check_integrity(parents)
 

	
 
        from .repository import MercurialRepository
 
        if not isinstance(message, unicode) or not isinstance(author, unicode):
 
            raise RepositoryError('Given message and author needs to be '
 
                                  'an <unicode> instance got %r & %r instead'
 
                                  % (type(message), type(author)))
 

	
 
        if branch is None:
 
            branch = MercurialRepository.DEFAULT_BRANCH_NAME
 
        kwargs[b'branch'] = safe_bytes(branch)
 

	
 
        def filectxfn(_repo, memctx, bytes_path):
 
            """
 
            Callback from Mercurial, returning ctx to commit for the given
 
            path.
 
            """
 
            path = bytes_path  # will be different for py3
 
            path = safe_str(bytes_path)
 

	
 
            # check if this path is removed
 
            if path in (node.path for node in self.removed):
 
                return None
 

	
 
            # check if this path is added
 
            for node in self.added:
 
                if node.path == path:
 
                    return mercurial.context.memfilectx(_repo, memctx, path=bytes_path,
 
                        data=node.content,
 
                        islink=False,
 
                        isexec=node.is_executable,
 
                        copysource=False)
 

	
 
            # or changed
 
            for node in self.changed:
 
                if node.path == path:
 
                    return mercurial.context.memfilectx(_repo, memctx, path=bytes_path,
 
                        data=node.content,
 
                        islink=False,
 
                        isexec=node.is_executable,
 
                        copysource=False)
 

	
 
            raise RepositoryError("Given path haven't been marked as added, "
 
                                  "changed or removed (%s)" % path)
 

	
 
        parents = [None, None]
 
        for i, parent in enumerate(self.parents):
 
            if parent is not None:
 
                parents[i] = parent._ctx.node()
 

	
 
        if date and isinstance(date, datetime.datetime):
 
            date = safe_bytes(date.strftime('%a, %d %b %Y %H:%M:%S'))
 

	
 
        commit_ctx = mercurial.context.memctx(
 
            repo=self.repository._repo,
 
            parents=parents,
 
            text=b'',
 
            files=[safe_bytes(x) for x in self.get_paths()],
 
            filectxfn=filectxfn,
 
            user=safe_bytes(author),
 
            date=date,
 
            extra=kwargs)
 

	
 
        # injecting given _repo params
 
        commit_ctx._text = safe_bytes(message)
 
        commit_ctx._user = safe_bytes(author)
 
        commit_ctx._date = date
 

	
 
        # TODO: Catch exceptions!
 
        n = self.repository._repo.commitctx(commit_ctx)
 
        # Returns mercurial node
 
        self._commit_ctx = commit_ctx  # For reference
 
        # Update vcs repository object & recreate mercurial _repo
 
        # new_ctx = self.repository._repo[node]
 
        # new_tip = ascii_str(self.repository.get_changeset(new_ctx.hex()))
 
        self.repository.revisions.append(ascii_str(mercurial.node.hex(n)))
 
        self._repo = self.repository._get_repo(create=False)
 
        self.repository.branches = self.repository._get_branches()
 
        tip = self.repository.get_changeset()
 
        self.reset()
 
        return tip
kallithea/lib/vcs/backends/hg/repository.py
Show inline comments
 
@@ -387,230 +387,231 @@ class MercurialRepository(BaseRepository
 
    @LazyProperty
 
    def in_memory_changeset(self):
 
        return MercurialInMemoryChangeset(self)
 

	
 
    @LazyProperty
 
    def description(self):
 
        _desc = self._repo.ui.config(b'web', b'description', None, untrusted=True)
 
        return safe_str(_desc or b'unknown')
 

	
 
    @LazyProperty
 
    def contact(self):
 
        return safe_str(mercurial.hgweb.common.get_contact(self._repo.ui.config)
 
                            or b'Unknown')
 

	
 
    @LazyProperty
 
    def last_change(self):
 
        """
 
        Returns last change made on this repository as datetime object
 
        """
 
        return date_fromtimestamp(self._get_mtime(), makedate()[1])
 

	
 
    def _get_mtime(self):
 
        try:
 
            return time.mktime(self.get_changeset().date.timetuple())
 
        except RepositoryError:
 
            # fallback to filesystem
 
            cl_path = os.path.join(self.path, '.hg', "00changelog.i")
 
            st_path = os.path.join(self.path, '.hg', "store")
 
            if os.path.exists(cl_path):
 
                return os.stat(cl_path).st_mtime
 
            else:
 
                return os.stat(st_path).st_mtime
 

	
 
    def _get_revision(self, revision):
 
        """
 
        Given any revision identifier, returns a 40 char string with revision hash.
 

	
 
        :param revision: str or int or None
 
        """
 
        if self._empty:
 
            raise EmptyRepositoryError("There are no changesets yet")
 

	
 
        if revision in [-1, None]:
 
            revision = b'tip'
 
        elif isinstance(revision, unicode):
 
            revision = safe_bytes(revision)
 

	
 
        try:
 
            if isinstance(revision, int):
 
                return ascii_str(self._repo[revision].hex())
 
            return ascii_str(mercurial.scmutil.revsymbol(self._repo, revision).hex())
 
        except (IndexError, ValueError, mercurial.error.RepoLookupError, TypeError):
 
            msg = "Revision %r does not exist for %s" % (safe_str(revision), self.name)
 
            raise ChangesetDoesNotExistError(msg)
 
        except (LookupError, ):
 
            msg = "Ambiguous identifier `%s` for %s" % (safe_str(revision), self.name)
 
            raise ChangesetDoesNotExistError(msg)
 

	
 
    def get_ref_revision(self, ref_type, ref_name):
 
        """
 
        Returns revision number for the given reference.
 
        """
 
        if ref_type == 'rev' and not ref_name.strip('0'):
 
            return self.EMPTY_CHANGESET
 
        # lookup up the exact node id
 
        _revset_predicates = {
 
                'branch': 'branch',
 
                'book': 'bookmark',
 
                'tag': 'tag',
 
                'rev': 'id',
 
            }
 
        # avoid expensive branch(x) iteration over whole repo
 
        rev_spec = "%%s & %s(%%s)" % _revset_predicates[ref_type]
 
        try:
 
            revs = self._repo.revs(rev_spec, ref_name, ref_name)
 
        except LookupError:
 
            msg = "Ambiguous identifier %s:%s for %s" % (ref_type, ref_name, self.name)
 
            raise ChangesetDoesNotExistError(msg)
 
        except mercurial.error.RepoLookupError:
 
            msg = "Revision %s:%s does not exist for %s" % (ref_type, ref_name, self.name)
 
            raise ChangesetDoesNotExistError(msg)
 
        if revs:
 
            revision = revs.last()
 
        else:
 
            # TODO: just report 'not found'?
 
            revision = ref_name
 

	
 
        return self._get_revision(revision)
 

	
 
    def _get_archives(self, archive_name='tip'):
 
        allowed = self.baseui.configlist(b"web", b"allow_archive",
 
                                         untrusted=True)
 
        for name, ext in [(b'zip', '.zip'), (b'gz', '.tar.gz'), (b'bz2', '.tar.bz2')]:
 
            if name in allowed or self._repo.ui.configbool(b"web",
 
                                                           b"allow" + name,
 
                                                           untrusted=True):
 
                yield {"type": name, "extension": ext, "node": archive_name}
 
                yield {"type": safe_str(name), "extension": ext, "node": archive_name}
 

	
 
    def _get_url(self, url):
 
        """
 
        Returns normalized url. If schema is not given, fall back to
 
        filesystem (``file:///``) schema.
 
        """
 
        if url != 'default' and '://' not in url:
 
            url = "file:" + urllib.request.pathname2url(url)
 
        return url
 

	
 
    def get_changeset(self, revision=None):
 
        """
 
        Returns ``MercurialChangeset`` object representing repository's
 
        changeset at the given ``revision``.
 
        """
 
        return MercurialChangeset(repository=self, revision=self._get_revision(revision))
 

	
 
    def get_changesets(self, start=None, end=None, start_date=None,
 
                       end_date=None, branch_name=None, reverse=False, max_revisions=None):
 
        """
 
        Returns iterator of ``MercurialChangeset`` objects from start to end
 
        (both are inclusive)
 

	
 
        :param start: None, str, int or mercurial lookup format
 
        :param end:  None, str, int or mercurial lookup format
 
        :param start_date:
 
        :param end_date:
 
        :param branch_name:
 
        :param reversed: return changesets in reversed order
 
        """
 
        start_raw_id = self._get_revision(start)
 
        start_pos = None if start is None else self.revisions.index(start_raw_id)
 
        end_raw_id = self._get_revision(end)
 
        end_pos = None if end is None else self.revisions.index(end_raw_id)
 

	
 
        if start_pos is not None and end_pos is not None and start_pos > end_pos:
 
            raise RepositoryError("Start revision '%s' cannot be "
 
                                  "after end revision '%s'" % (start, end))
 

	
 
        if branch_name and branch_name not in self.allbranches:
 
            msg = "Branch %r not found in %s" % (branch_name, self.name)
 
            raise BranchDoesNotExistError(msg)
 
        if end_pos is not None:
 
            end_pos += 1
 
        # filter branches
 
        filter_ = []
 
        if branch_name:
 
            filter_.append(b'branch("%s")' % safe_bytes(branch_name))
 
        if start_date:
 
            filter_.append(b'date(">%s")' % safe_bytes(str(start_date)))
 
        if end_date:
 
            filter_.append(b'date("<%s")' % safe_bytes(str(end_date)))
 
        if filter_ or max_revisions:
 
            if filter_:
 
                revspec = b' and '.join(filter_)
 
            else:
 
                revspec = b'all()'
 
            if max_revisions:
 
                revspec = b'limit(%s, %d)' % (revspec, max_revisions)
 
            revisions = mercurial.scmutil.revrange(self._repo, [revspec])
 
        else:
 
            revisions = self.revisions
 

	
 
        # this is very much a hack to turn this into a list; a better solution
 
        # would be to get rid of this function entirely and use revsets
 
        revs = list(revisions)[start_pos:end_pos]
 
        if reverse:
 
            revs.reverse()
 

	
 
        return CollectionGenerator(self, revs)
 

	
 
    def pull(self, url):
 
        """
 
        Tries to pull changes from external location.
 
        """
 
        other = mercurial.hg.peer(self._repo, {}, safe_bytes(self._get_url(url)))
 
        try:
 
            mercurial.exchange.pull(self._repo, other, heads=None, force=None)
 
        except mercurial.error.Abort as err:
 
            # Propagate error but with vcs's type
 
            raise RepositoryError(str(err))
 

	
 
    @LazyProperty
 
    def workdir(self):
 
        """
 
        Returns ``Workdir`` instance for this repository.
 
        """
 
        return MercurialWorkdir(self)
 

	
 
    def get_config_value(self, section, name=None, config_file=None):
 
        """
 
        Returns configuration value for a given [``section``] and ``name``.
 

	
 
        :param section: Section we want to retrieve value from
 
        :param name: Name of configuration we want to retrieve
 
        :param config_file: A path to file which should be used to retrieve
 
          configuration from (might also be a list of file paths)
 
        """
 
        if config_file is None:
 
            config_file = []
 
        elif isinstance(config_file, str):
 
            config_file = [config_file]
 

	
 
        config = self._repo.ui
 
        if config_file:
 
            config = mercurial.ui.ui()
 
            for path in config_file:
 
                config.readconfig(safe_bytes(path))
 
        return config.config(safe_bytes(section), safe_bytes(name))
 
        value = config.config(safe_bytes(section), safe_bytes(name))
 
        return value if value is None else safe_str(value)
 

	
 
    def get_user_name(self, config_file=None):
 
        """
 
        Returns user's name from global configuration file.
 

	
 
        :param config_file: A path to file which should be used to retrieve
 
          configuration from (might also be a list of file paths)
 
        """
 
        username = self.get_config_value('ui', 'username', config_file=config_file)
 
        if username:
 
            return author_name(username)
 
        return None
 

	
 
    def get_user_email(self, config_file=None):
 
        """
 
        Returns user's email from global configuration file.
 

	
 
        :param config_file: A path to file which should be used to retrieve
 
          configuration from (might also be a list of file paths)
 
        """
 
        username = self.get_config_value('ui', 'username', config_file=config_file)
 
        if username:
 
            return author_email(username)
 
        return None
kallithea/lib/vcs/backends/hg/workdir.py
Show inline comments
 
import mercurial.merge
 

	
 
from kallithea.lib.vcs.backends.base import BaseWorkdir
 
from kallithea.lib.vcs.exceptions import BranchDoesNotExistError
 
from kallithea.lib.vcs.utils import ascii_bytes, ascii_str
 
from kallithea.lib.vcs.utils import ascii_bytes, ascii_str, safe_str
 

	
 

	
 
class MercurialWorkdir(BaseWorkdir):
 

	
 
    def get_branch(self):
 
        return self.repository._repo.dirstate.branch()
 
        return safe_str(self.repository._repo.dirstate.branch())
 

	
 
    def get_changeset(self):
 
        wk_dir_id = ascii_str(self.repository._repo[None].parents()[0].hex())
 
        return self.repository.get_changeset(wk_dir_id)
 

	
 
    def checkout_branch(self, branch=None):
 
        if branch is None:
 
            branch = self.repository.DEFAULT_BRANCH_NAME
 
        if branch not in self.repository.branches:
 
            raise BranchDoesNotExistError
 

	
 
        raw_id = self.repository.branches[branch]
 
        mercurial.merge.update(self.repository._repo, ascii_bytes(raw_id), False, False, None)
0 comments (0 inline, 0 general)