Changeset - 072c0352dd36
[Not reviewed]
default
0 5 0
Mads Kiilerich (mads) - 5 years ago 2020-10-11 01:02:46
mads@kiilerich.com
Grafted from: 260bc0b57a56
imports: always use CONFIG as kallithea.CONFIG
5 files changed with 22 insertions and 25 deletions:
0 comments (0 inline, 0 general)
kallithea/controllers/feed.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.feed
 
~~~~~~~~~~~~~~~~~~~~~~~~~~
 

	
 
Feed 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: Apr 23, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 
"""
 

	
 

	
 
import logging
 

	
 
from beaker.cache import cache_region
 
from tg import response
 
from tg import tmpl_context as c
 
from tg.i18n import ugettext as _
 

	
 
from kallithea import CONFIG
 
import kallithea
 
from kallithea.lib import feeds
 
from kallithea.lib import helpers as h
 
from kallithea.lib.auth import HasRepoPermissionLevelDecorator, LoginRequired
 
from kallithea.lib.base import BaseRepoController
 
from kallithea.lib.diffs import DiffProcessor
 
from kallithea.lib.utils2 import asbool, safe_int, safe_str
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class FeedController(BaseRepoController):
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @HasRepoPermissionLevelDecorator('read')
 
    def _before(self, *args, **kwargs):
 
        super(FeedController, self)._before(*args, **kwargs)
 

	
 
    def _get_title(self, cs):
 
        return h.shorter(cs.message, 160)
 

	
 
    def __get_desc(self, cs):
 
        desc_msg = [(_('%s committed on %s')
 
                     % (h.person(cs.author), h.fmt_date(cs.date))) + '<br/>']
 
        # branches, tags, bookmarks
 
        for branch in cs.branches:
 
            desc_msg.append('branch: %s<br/>' % branch)
 
        for book in cs.bookmarks:
 
            desc_msg.append('bookmark: %s<br/>' % book)
 
        for tag in cs.tags:
 
            desc_msg.append('tag: %s<br/>' % tag)
 

	
 
        changes = []
 
        diff_limit = safe_int(CONFIG.get('rss_cut_off_limit', 32 * 1024))
 
        diff_limit = safe_int(kallithea.CONFIG.get('rss_cut_off_limit', 32 * 1024))
 
        raw_diff = cs.diff()
 
        diff_processor = DiffProcessor(raw_diff,
 
                                       diff_limit=diff_limit,
 
                                       inline_diff=False)
 

	
 
        for st in diff_processor.parsed:
 
            st.update({'added': st['stats']['added'],
 
                       'removed': st['stats']['deleted']})
 
            changes.append('\n %(operation)s %(filename)s '
 
                           '(%(added)s lines added, %(removed)s lines removed)'
 
                            % st)
 
        if diff_processor.limited_diff:
 
            changes = changes + ['\n ' +
 
                                 _('Changeset was too big and was cut off...')]
 

	
 
        # rev link
 
        _url = h.canonical_url('changeset_home', repo_name=c.db_repo.repo_name,
 
                   revision=cs.raw_id)
 
        desc_msg.append('changeset: <a href="%s">%s</a>' % (_url, cs.raw_id[:8]))
 

	
 
        desc_msg.append('<pre>')
 
        desc_msg.append(h.urlify_text(cs.message))
 
        desc_msg.append('\n')
 
        desc_msg.extend(changes)
 
        if asbool(CONFIG.get('rss_include_diff', False)):
 
        if asbool(kallithea.CONFIG.get('rss_include_diff', False)):
 
            desc_msg.append('\n\n')
 
            desc_msg.append(safe_str(raw_diff))
 
        desc_msg.append('</pre>')
 
        return desc_msg
 

	
 
    def _feed(self, repo_name, feeder):
 
        """Produce a simple feed"""
 

	
 
        @cache_region('long_term_file', '_get_feed_from_cache')
 
        def _get_feed_from_cache(*_cache_keys):  # parameters are not really used - only as caching key
 
            header = dict(
 
                title=_('%s %s feed') % (c.site_name, repo_name),
 
                link=h.canonical_url('summary_home', repo_name=repo_name),
 
                description=_('Changes on %s repository') % repo_name,
 
            )
 

	
 
            rss_items_per_page = safe_int(CONFIG.get('rss_items_per_page', 20))
 
            rss_items_per_page = safe_int(kallithea.CONFIG.get('rss_items_per_page', 20))
 
            entries=[]
 
            for cs in reversed(list(c.db_repo_scm_instance[-rss_items_per_page:])):
 
                entries.append(dict(
 
                    title=self._get_title(cs),
 
                    link=h.canonical_url('changeset_home', repo_name=repo_name, revision=cs.raw_id),
 
                    author_email=cs.author_email,
 
                    author_name=cs.author_name,
 
                    description=''.join(self.__get_desc(cs)),
 
                    pubdate=cs.date,
 
                ))
 
            return feeder.render(header, entries)
 

	
 
        response.content_type = feeder.content_type
 
        return _get_feed_from_cache(repo_name, feeder.__name__)
 

	
 
    def atom(self, repo_name):
 
        """Produce a simple atom-1.0 feed"""
 
        return self._feed(repo_name, feeds.AtomFeed)
 

	
 
    def rss(self, repo_name):
 
        """Produce a simple rss2 feed"""
 
        return self._feed(repo_name, feeds.RssFeed)
kallithea/controllers/files.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.files
 
~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

	
 
Files 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: Apr 21, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 
"""
 

	
 
import logging
 
import os
 
import posixpath
 
import shutil
 
import tempfile
 
import traceback
 
from collections import OrderedDict
 

	
 
from tg import request, response
 
from tg import tmpl_context as c
 
from tg.i18n import ugettext as _
 
from webob.exc import HTTPFound, HTTPNotFound
 

	
 
import kallithea
 
from kallithea.config.routing import url
 
from kallithea.lib import diffs
 
from kallithea.lib import helpers as h
 
from kallithea.lib.auth import HasRepoPermissionLevelDecorator, LoginRequired
 
from kallithea.lib.base import BaseRepoController, jsonify, render
 
from kallithea.lib.exceptions import NonRelativePathError
 
from kallithea.lib.utils import action_logger
 
from kallithea.lib.utils2 import asbool, convert_line_endings, detect_mode, safe_str
 
from kallithea.lib.vcs.backends.base import EmptyChangeset
 
from kallithea.lib.vcs.conf import settings
 
from kallithea.lib.vcs.exceptions import (ChangesetDoesNotExistError, ChangesetError, EmptyRepositoryError, ImproperArchiveTypeError, NodeAlreadyExistsError,
 
                                          NodeDoesNotExistError, NodeError, RepositoryError, VCSError)
 
from kallithea.lib.vcs.nodes import FileNode
 
from kallithea.model import db
 
from kallithea.model.repo import RepoModel
 
from kallithea.model.scm import ScmModel
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
class FilesController(BaseRepoController):
 

	
 
    def _before(self, *args, **kwargs):
 
        super(FilesController, self)._before(*args, **kwargs)
 

	
 
    def __get_cs(self, rev, silent_empty=False):
 
        """
 
        Safe way to get changeset if error occur it redirects to tip with
 
        proper message
 

	
 
        :param rev: revision to fetch
 
        :silent_empty: return None if repository is empty
 
        """
 

	
 
        try:
 
            return c.db_repo_scm_instance.get_changeset(rev)
 
        except EmptyRepositoryError as e:
 
            if silent_empty:
 
                return None
 
            url_ = url('files_add_home',
 
                       repo_name=c.repo_name,
 
                       revision=0, f_path='', anchor='edit')
 
            add_new = h.link_to(_('Click here to add new file'), url_, class_="alert-link")
 
            h.flash(_('There are no files yet.') + ' ' + add_new, category='warning')
 
            raise HTTPNotFound()
 
        except (ChangesetDoesNotExistError, LookupError):
 
            msg = _('Such revision does not exist for this repository')
 
            h.flash(msg, category='error')
 
            raise HTTPNotFound()
 
        except RepositoryError as e:
 
            h.flash(e, category='error')
 
            raise HTTPNotFound()
 

	
 
    def __get_filenode(self, cs, path):
 
        """
 
        Returns file_node or raise HTTP error.
 

	
 
        :param cs: given changeset
 
        :param path: path to lookup
 
        """
 

	
 
        try:
 
            file_node = cs.get_node(path)
 
            if file_node.is_dir():
 
                raise RepositoryError('given path is a directory')
 
        except ChangesetDoesNotExistError:
 
            msg = _('Such revision does not exist for this repository')
 
            h.flash(msg, category='error')
 
            raise HTTPNotFound()
 
        except RepositoryError as e:
 
            h.flash(e, category='error')
 
            raise HTTPNotFound()
 

	
 
        return file_node
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @HasRepoPermissionLevelDecorator('read')
 
    def index(self, repo_name, revision, f_path, annotate=False):
 
        # redirect to given revision from form if given
 
        post_revision = request.POST.get('at_rev', None)
 
        if post_revision:
 
            cs = self.__get_cs(post_revision) # FIXME - unused!
 

	
 
        c.revision = revision
 
        c.changeset = self.__get_cs(revision)
 
        c.branch = request.GET.get('branch', None)
 
        c.f_path = f_path
 
        c.annotate = annotate
 
        cur_rev = c.changeset.revision
 
        # used in files_source.html:
 
        c.cut_off_limit = self.cut_off_limit
 
        c.fulldiff = request.GET.get('fulldiff')
 

	
 
        # prev link
 
        try:
 
@@ -411,199 +412,198 @@ class FilesController(BaseRepoController
 
            content = convert_line_endings(r_post.get('content', ''), unix_mode)
 

	
 
            message = r_post.get('message') or c.default_message
 
            filename = r_post.get('filename')
 
            location = r_post.get('location', '')
 
            file_obj = r_post.get('upload_file', None)
 

	
 
            if file_obj is not None and hasattr(file_obj, 'filename'):
 
                filename = file_obj.filename
 
                content = file_obj.file
 

	
 
                if hasattr(content, 'file'):
 
                    # non posix systems store real file under file attr
 
                    content = content.file
 

	
 
            if not content:
 
                h.flash(_('No content'), category='warning')
 
                raise HTTPFound(location=url('changeset_home', repo_name=c.repo_name,
 
                                    revision='tip'))
 
            if not filename:
 
                h.flash(_('No filename'), category='warning')
 
                raise HTTPFound(location=url('changeset_home', repo_name=c.repo_name,
 
                                    revision='tip'))
 
            # strip all crap out of file, just leave the basename
 
            filename = os.path.basename(filename)
 
            node_path = posixpath.join(location, filename)
 
            author = request.authuser.full_contact
 

	
 
            try:
 
                nodes = {
 
                    node_path: {
 
                        'content': content
 
                    }
 
                }
 
                self.scm_model.create_nodes(
 
                    user=request.authuser.user_id,
 
                    ip_addr=request.ip_addr,
 
                    repo=c.db_repo,
 
                    message=message,
 
                    nodes=nodes,
 
                    parent_cs=c.cs,
 
                    author=author,
 
                )
 

	
 
                h.flash(_('Successfully committed to %s') % node_path,
 
                        category='success')
 
            except NonRelativePathError as e:
 
                h.flash(_('Location must be relative path and must not '
 
                          'contain .. in path'), category='warning')
 
                raise HTTPFound(location=url('changeset_home', repo_name=c.repo_name,
 
                                    revision='tip'))
 
            except (NodeError, NodeAlreadyExistsError) as e:
 
                h.flash(_(e), category='error')
 
            except Exception:
 
                log.error(traceback.format_exc())
 
                h.flash(_('Error occurred during commit'), category='error')
 
            raise HTTPFound(location=url('changeset_home',
 
                                repo_name=c.repo_name, revision='tip'))
 

	
 
        return render('files/files_add.html')
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @HasRepoPermissionLevelDecorator('read')
 
    def archivefile(self, repo_name, fname):
 
        fileformat = None
 
        revision = None
 
        ext = None
 
        subrepos = request.GET.get('subrepos') == 'true'
 

	
 
        for a_type, ext_data in settings.ARCHIVE_SPECS.items():
 
            archive_spec = fname.split(ext_data[1])
 
            if len(archive_spec) == 2 and archive_spec[1] == '':
 
                fileformat = a_type or ext_data[1]
 
                revision = archive_spec[0]
 
                ext = ext_data[1]
 

	
 
        try:
 
            dbrepo = RepoModel().get_by_repo_name(repo_name)
 
            if not dbrepo.enable_downloads:
 
                return _('Downloads disabled') # TODO: do something else?
 

	
 
            if c.db_repo_scm_instance.alias == 'hg':
 
                # patch and reset hooks section of UI config to not run any
 
                # hooks on fetching archives with subrepos
 
                for k, v in c.db_repo_scm_instance._repo.ui.configitems('hooks'):
 
                    c.db_repo_scm_instance._repo.ui.setconfig('hooks', k, None)
 

	
 
            cs = c.db_repo_scm_instance.get_changeset(revision)
 
            content_type = settings.ARCHIVE_SPECS[fileformat][0]
 
        except ChangesetDoesNotExistError:
 
            return _('Unknown revision %s') % revision
 
        except EmptyRepositoryError:
 
            return _('Empty repository')
 
        except (ImproperArchiveTypeError, KeyError):
 
            return _('Unknown archive type')
 

	
 
        from kallithea import CONFIG
 
        rev_name = cs.raw_id[:12]
 
        archive_name = '%s-%s%s' % (repo_name.replace('/', '_'), rev_name, ext)
 

	
 
        archive_path = None
 
        cached_archive_path = None
 
        archive_cache_dir = CONFIG.get('archive_cache_dir')
 
        archive_cache_dir = kallithea.CONFIG.get('archive_cache_dir')
 
        if archive_cache_dir and not subrepos: # TODO: subrepo caching?
 
            if not os.path.isdir(archive_cache_dir):
 
                os.makedirs(archive_cache_dir)
 
            cached_archive_path = os.path.join(archive_cache_dir, archive_name)
 
            if os.path.isfile(cached_archive_path):
 
                log.debug('Found cached archive in %s', cached_archive_path)
 
                archive_path = cached_archive_path
 
            else:
 
                log.debug('Archive %s is not yet cached', archive_name)
 

	
 
        if archive_path is None:
 
            # generate new archive
 
            fd, archive_path = tempfile.mkstemp()
 
            log.debug('Creating new temp archive in %s', archive_path)
 
            with os.fdopen(fd, 'wb') as stream:
 
                cs.fill_archive(stream=stream, kind=fileformat, subrepos=subrepos)
 
                # stream (and thus fd) has been closed by cs.fill_archive
 
            if cached_archive_path is not None:
 
                # we generated the archive - move it to cache
 
                log.debug('Storing new archive in %s', cached_archive_path)
 
                shutil.move(archive_path, cached_archive_path)
 
                archive_path = cached_archive_path
 

	
 
        def get_chunked_archive(archive_path):
 
            stream = open(archive_path, 'rb')
 
            while True:
 
                data = stream.read(16 * 1024)
 
                if not data:
 
                    break
 
                yield data
 
            stream.close()
 
            if archive_path != cached_archive_path:
 
                log.debug('Destroying temp archive %s', archive_path)
 
                os.remove(archive_path)
 

	
 
        action_logger(user=request.authuser,
 
                      action='user_downloaded_archive:%s' % (archive_name),
 
                      repo=repo_name, ipaddr=request.ip_addr, commit=True)
 

	
 
        response.content_disposition = str('attachment; filename=%s' % (archive_name))
 
        response.content_type = str(content_type)
 
        return get_chunked_archive(archive_path)
 

	
 
    @LoginRequired(allow_default_user=True)
 
    @HasRepoPermissionLevelDecorator('read')
 
    def diff(self, repo_name, f_path):
 
        ignore_whitespace_diff = h.get_ignore_whitespace_diff(request.GET)
 
        diff_context_size = h.get_diff_context_size(request.GET)
 
        diff2 = request.GET.get('diff2', '')
 
        diff1 = request.GET.get('diff1', '') or diff2
 
        c.action = request.GET.get('diff')
 
        c.no_changes = diff1 == diff2
 
        c.f_path = f_path
 
        c.big_diff = False
 
        fulldiff = request.GET.get('fulldiff')
 
        c.changes = OrderedDict()
 
        c.changes[diff2] = []
 

	
 
        # special case if we want a show rev only, it's impl here
 
        # to reduce JS and callbacks
 

	
 
        if request.GET.get('show_rev'):
 
            if asbool(request.GET.get('annotate', 'False')):
 
                _url = url('files_annotate_home', repo_name=c.repo_name,
 
                           revision=diff1, f_path=c.f_path)
 
            else:
 
                _url = url('files_home', repo_name=c.repo_name,
 
                           revision=diff1, f_path=c.f_path)
 

	
 
            raise HTTPFound(location=_url)
 
        try:
 
            if diff1 not in ['', None, 'None', '0' * 12, '0' * 40]:
 
                c.changeset_1 = c.db_repo_scm_instance.get_changeset(diff1)
 
                try:
 
                    node1 = c.changeset_1.get_node(f_path)
 
                    if node1.is_dir():
 
                        raise NodeError('%s path is a %s not a file'
 
                                        % (node1, type(node1)))
 
                except NodeDoesNotExistError:
 
                    c.changeset_1 = EmptyChangeset(cs=diff1,
 
                                                   revision=c.changeset_1.revision,
 
                                                   repo=c.db_repo_scm_instance)
 
                    node1 = FileNode(f_path, '', changeset=c.changeset_1)
 
            else:
 
                c.changeset_1 = EmptyChangeset(repo=c.db_repo_scm_instance)
 
                node1 = FileNode(f_path, '', changeset=c.changeset_1)
 

	
 
            if diff2 not in ['', None, 'None', '0' * 12, '0' * 40]:
 
                c.changeset_2 = c.db_repo_scm_instance.get_changeset(diff2)
 
                try:
 
                    node2 = c.changeset_2.get_node(f_path)
 
                    if node2.is_dir():
 
                        raise NodeError('%s path is a %s not a file'
 
                                        % (node2, type(node2)))
 
                except NodeDoesNotExistError:
 
                    c.changeset_2 = EmptyChangeset(cs=diff2,
kallithea/lib/helpers.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/>.
 
"""
 
Helper functions
 

	
 
Consists of functions to typically be used within templates, but also
 
available to Controllers. This module is available to both as 'h'.
 
"""
 
import hashlib
 
import json
 
import logging
 
import random
 
import re
 
import textwrap
 
import urllib.parse
 

	
 
from beaker.cache import cache_region
 
from pygments import highlight as code_highlight
 
from pygments.formatters.html import HtmlFormatter
 
from tg.i18n import ugettext as _
 
from webhelpers2.html import HTML, escape, literal
 
from webhelpers2.html.tags import NotGiven, Option, Options, _input, _make_safe_id_component, checkbox, end_form
 
from webhelpers2.html.tags import form as insecure_form
 
from webhelpers2.html.tags import hidden, link_to, password, radio
 
from webhelpers2.html.tags import select as webhelpers2_select
 
from webhelpers2.html.tags import submit, text, textarea
 
from webhelpers2.number import format_byte_size
 
from webhelpers2.text import chop_at, truncate, wrap_paragraphs
 

	
 
import kallithea
 
from kallithea.config.routing import url
 
from kallithea.lib.annotate import annotate_highlight
 
#==============================================================================
 
# PERMS
 
#==============================================================================
 
from kallithea.lib.auth import HasPermissionAny, HasRepoGroupPermissionLevel, HasRepoPermissionLevel
 
from kallithea.lib.markup_renderer import url_re
 
from kallithea.lib.pygmentsutils import get_custom_lexer
 
from kallithea.lib.utils2 import MENTIONS_REGEX, AttributeDict
 
from kallithea.lib.utils2 import age as _age
 
from kallithea.lib.utils2 import asbool, credentials_filter, safe_bytes, safe_int, safe_str, time_to_datetime
 
from kallithea.lib.vcs.backends.base import BaseChangeset, EmptyChangeset
 
from kallithea.lib.vcs.exceptions import ChangesetDoesNotExistError
 
#==============================================================================
 
# SCM FILTERS available via h.
 
#==============================================================================
 
from kallithea.lib.vcs.utils import author_email, author_name
 

	
 

	
 
# mute pyflakes "imported but unused"
 
assert Option
 
assert checkbox
 
assert end_form
 
assert password
 
assert radio
 
assert submit
 
assert text
 
assert textarea
 
assert format_byte_size
 
assert chop_at
 
assert wrap_paragraphs
 
assert HasPermissionAny
 
assert HasRepoGroupPermissionLevel
 
assert HasRepoPermissionLevel
 
assert time_to_datetime
 
assert EmptyChangeset
 

	
 

	
 
log = logging.getLogger(__name__)
 

	
 

	
 
def canonical_url(*args, **kargs):
 
    '''Like url(x, qualified=True), but returns url that not only is qualified
 
    but also canonical, as configured in canonical_url'''
 
    from kallithea import CONFIG
 
    try:
 
        parts = CONFIG.get('canonical_url', '').split('://', 1)
 
        parts = kallithea.CONFIG.get('canonical_url', '').split('://', 1)
 
        kargs['host'] = parts[1]
 
        kargs['protocol'] = parts[0]
 
    except IndexError:
 
        kargs['qualified'] = True
 
    return url(*args, **kargs)
 

	
 

	
 
def canonical_hostname():
 
    '''Return canonical hostname of system'''
 
    from kallithea import CONFIG
 
    try:
 
        parts = CONFIG.get('canonical_url', '').split('://', 1)
 
        parts = kallithea.CONFIG.get('canonical_url', '').split('://', 1)
 
        return parts[1].split('/', 1)[0]
 
    except IndexError:
 
        parts = url('home', qualified=True).split('://', 1)
 
        return parts[1].split('/', 1)[0]
 

	
 

	
 
def html_escape(s):
 
    """Return string with all html escaped.
 
    This is also safe for javascript in html but not necessarily correct.
 
    """
 
    return (s
 
        .replace('&', '&amp;')
 
        .replace(">", "&gt;")
 
        .replace("<", "&lt;")
 
        .replace('"', "&quot;")
 
        .replace("'", "&apos;") # Note: this is HTML5 not HTML4 and might not work in mails
 
        )
 

	
 
def js(value):
 
    """Convert Python value to the corresponding JavaScript representation.
 

	
 
    This is necessary to safely insert arbitrary values into HTML <script>
 
    sections e.g. using Mako template expression substitution.
 

	
 
    Note: Rather than using this function, it's preferable to avoid the
 
    insertion of values into HTML <script> sections altogether. Instead,
 
    data should (to the extent possible) be passed to JavaScript using
 
    data attributes or AJAX calls, eliminating the need for JS specific
 
    escaping.
 

	
 
    Note: This is not safe for use in attributes (e.g. onclick), because
 
    quotes are not escaped.
 

	
 
    Because the rules for parsing <script> varies between XHTML (where
 
    normal rules apply for any special characters) and HTML (where
 
    entities are not interpreted, but the literal string "</script>"
 
    is forbidden), the function ensures that the result never contains
 
    '&', '<' and '>', thus making it safe in both those contexts (but
 
    not in attributes).
 
    """
 
    return literal(
 
        ('(' + json.dumps(value) + ')')
 
        # In JSON, the following can only appear in string literals.
 
        .replace('&', r'\x26')
 
        .replace('<', r'\x3c')
 
        .replace('>', r'\x3e')
 
    )
 

	
 

	
 
def jshtml(val):
 
    """HTML escapes a string value, then converts the resulting string
 
    to its corresponding JavaScript representation (see `js`).
 

	
 
    This is used when a plain-text string (possibly containing special
 
    HTML characters) will be used by a script in an HTML context (e.g.
 
    element.innerHTML or jQuery's 'html' method).
 

	
 
    If in doubt, err on the side of using `jshtml` over `js`, since it's
 
    better to escape too much than too little.
 
    """
 
    return js(escape(val))
 

	
 

	
 
def shorter(s, size=20, firstline=False, postfix='...'):
 
    """Truncate s to size, including the postfix string if truncating.
 
    If firstline, truncate at newline.
 
    """
 
    if firstline:
 
        s = s.split('\n', 1)[0].rstrip()
 
    if len(s) > size:
 
        return s[:size - len(postfix)] + postfix
 
    return s
 

	
 

	
 
def reset(name, value, id=NotGiven, **attrs):
 
    """Create a reset button, similar to webhelpers2.html.tags.submit ."""
 
    return _input("reset", name, value, id, attrs)
 

	
 

	
 
def select(name, selected_values, options, id=NotGiven, **attrs):
 
    """Convenient wrapper of webhelpers2 to let it accept options as a tuple list"""
 
    if isinstance(options, list):
 
        option_list = options
 
        # Handle old value,label lists ... where value also can be value,label lists
 
        options = Options()
 
        for x in option_list:
 
            if isinstance(x, tuple) and len(x) == 2:
 
                value, label = x
 
            elif isinstance(x, str):
 
                value = label = x
 
            else:
 
                log.error('invalid select option %r', x)
 
                raise
 
            if isinstance(value, list):
 
                og = options.add_optgroup(label)
 
                for x in value:
 
@@ -467,195 +466,194 @@ def pygmentize_annotation(repo_name, fil
 
    return literal(markup_whitespace(annotate_highlight(filenode, url_func, **kwargs)))
 

	
 

	
 
class _Message(object):
 
    """A message returned by ``pop_flash_messages()``.
 

	
 
    Converting the message to a string returns the message text. Instances
 
    also have the following attributes:
 

	
 
    * ``category``: the category specified when the message was created.
 
    * ``message``: the html-safe message text.
 
    """
 

	
 
    def __init__(self, category, message):
 
        self.category = category
 
        self.message = message
 

	
 

	
 
def _session_flash_messages(append=None, clear=False):
 
    """Manage a message queue in tg.session: return the current message queue
 
    after appending the given message, and possibly clearing the queue."""
 
    key = 'flash'
 
    from tg import session
 
    if key in session:
 
        flash_messages = session[key]
 
    else:
 
        if append is None:  # common fast path - also used for clearing empty queue
 
            return []  # don't bother saving
 
        flash_messages = []
 
        session[key] = flash_messages
 
    if append is not None and append not in flash_messages:
 
        flash_messages.append(append)
 
    if clear:
 
        session.pop(key, None)
 
    session.save()
 
    return flash_messages
 

	
 

	
 
def flash(message, category, logf=None):
 
    """
 
    Show a message to the user _and_ log it through the specified function
 

	
 
    category: notice (default), warning, error, success
 
    logf: a custom log function - such as log.debug
 

	
 
    logf defaults to log.info, unless category equals 'success', in which
 
    case logf defaults to log.debug.
 
    """
 
    assert category in ('error', 'success', 'warning'), category
 
    if hasattr(message, '__html__'):
 
        # render to HTML for storing in cookie
 
        safe_message = str(message)
 
    else:
 
        # Apply str - the message might be an exception with __str__
 
        # Escape, so we can trust the result without further escaping, without any risk of injection
 
        safe_message = html_escape(str(message))
 
    if logf is None:
 
        logf = log.info
 
        if category == 'success':
 
            logf = log.debug
 

	
 
    logf('Flash %s: %s', category, safe_message)
 

	
 
    _session_flash_messages(append=(category, safe_message))
 

	
 

	
 
def pop_flash_messages():
 
    """Return all accumulated messages and delete them from the session.
 

	
 
    The return value is a list of ``Message`` objects.
 
    """
 
    return [_Message(category, message) for category, message in _session_flash_messages(clear=True)]
 

	
 

	
 
def age(x, y=False):
 
    return _age(x, y)
 

	
 
def capitalize(x):
 
    return x.capitalize()
 

	
 
email = author_email
 

	
 
def short_id(x):
 
    return x[:12]
 

	
 
def hide_credentials(x):
 
    return ''.join(credentials_filter(x))
 

	
 

	
 
def show_id(cs):
 
    """
 
    Configurable function that shows ID
 
    by default it's r123:fffeeefffeee
 

	
 
    :param cs: changeset instance
 
    """
 
    from kallithea import CONFIG
 
    def_len = safe_int(CONFIG.get('show_sha_length', 12))
 
    show_rev = asbool(CONFIG.get('show_revision_number', False))
 
    def_len = safe_int(kallithea.CONFIG.get('show_sha_length', 12))
 
    show_rev = asbool(kallithea.CONFIG.get('show_revision_number', False))
 

	
 
    raw_id = cs.raw_id[:def_len]
 
    if show_rev:
 
        return 'r%s:%s' % (cs.revision, raw_id)
 
    else:
 
        return raw_id
 

	
 

	
 
def fmt_date(date):
 
    if date:
 
        return date.strftime("%Y-%m-%d %H:%M:%S")
 
    return ""
 

	
 

	
 
def is_git(repository):
 
    if hasattr(repository, 'alias'):
 
        _type = repository.alias
 
    elif hasattr(repository, 'repo_type'):
 
        _type = repository.repo_type
 
    else:
 
        _type = repository
 
    return _type == 'git'
 

	
 

	
 
def is_hg(repository):
 
    if hasattr(repository, 'alias'):
 
        _type = repository.alias
 
    elif hasattr(repository, 'repo_type'):
 
        _type = repository.repo_type
 
    else:
 
        _type = repository
 
    return _type == 'hg'
 

	
 

	
 
@cache_region('long_term', 'user_attr_or_none')
 
def user_attr_or_none(author, show_attr):
 
    """Try to match email part of VCS committer string with a local user and return show_attr
 
    - or return None if user not found"""
 
    email = author_email(author)
 
    if email:
 
        from kallithea.model.db import User
 
        user = User.get_by_email(email)
 
        if user is not None:
 
            return getattr(user, show_attr)
 
    return None
 

	
 

	
 
def email_or_none(author):
 
    """Try to match email part of VCS committer string with a local user.
 
    Return primary email of user, email part of the specified author name, or None."""
 
    if not author:
 
        return None
 
    email = user_attr_or_none(author, 'email')
 
    if email is not None:
 
        return email # always use user's main email address - not necessarily the one used to find user
 

	
 
    # extract email from the commit string
 
    email = author_email(author)
 
    if email:
 
        return email
 

	
 
    # No valid email, not a valid user in the system, none!
 
    return None
 

	
 

	
 
def person(author, show_attr="username"):
 
    """Find the user identified by 'author', return one of the users attributes,
 
    default to the username attribute, None if there is no user"""
 
    from kallithea.model.db import User
 

	
 
    # if author is already an instance use it for extraction
 
    if isinstance(author, User):
 
        return getattr(author, show_attr)
 

	
 
    value = user_attr_or_none(author, show_attr)
 
    if value is not None:
 
        return value
 

	
 
    # Still nothing?  Just pass back the author name if any, else the email
 
    return author_name(author) or email(author)
 

	
 

	
 
def person_by_id(id_, show_attr="username"):
 
    from kallithea.model.db import User
 

	
 
    # maybe it's an ID ?
 
    if str(id_).isdigit() or isinstance(id_, int):
 
        id_ = int(id_)
 
        user = User.get(id_)
 
        if user is not None:
 
            return getattr(user, show_attr)
 
    return id_
 

	
 

	
 
def boolicon(value):
 
    """Returns boolean value of a value, represented as small html image of true/false
 
@@ -1119,211 +1117,210 @@ def urlify_text(s, repo_name=None, link_
 
    Parses given text message and make literal html with markup.
 
    The text will be truncated to the specified length.
 
    Hashes are turned into changeset links to specified repository.
 
    URLs links to what they say.
 
    Issues are linked to given issue-server.
 
    If link_ is provided, all text not already linking somewhere will link there.
 
    >>> urlify_text("Urlify http://example.com/ and 'https://example.com' *and* <b>markup/b>")
 
    literal('Urlify <a href="http://example.com/">http://example.com/</a> and &#39;<a href="https://example.com&apos">https://example.com&apos</a>; <b>*and*</b> &lt;b&gt;markup/b&gt;')
 
    """
 

	
 
    def _replace(match_obj):
 
        url = match_obj.group('url')
 
        if url is not None:
 
            return '<a href="%(url)s">%(url)s</a>' % {'url': url}
 
        mention = match_obj.group('mention')
 
        if mention is not None:
 
            return '<b>%s</b>' % mention
 
        hash_ = match_obj.group('hash')
 
        if hash_ is not None and repo_name is not None:
 
            from kallithea.config.routing import url  # doh, we need to re-import url to mock it later
 
            return '<a class="changeset_hash" href="%(url)s">%(hash)s</a>' % {
 
                 'url': url('changeset_home', repo_name=repo_name, revision=hash_),
 
                 'hash': hash_,
 
                }
 
        bold = match_obj.group('bold')
 
        if bold is not None:
 
            return '<b>*%s*</b>' % _urlify(bold[1:-1])
 
        if stylize:
 
            seen = match_obj.group('seen')
 
            if seen:
 
                return '<div class="label label-meta" data-tag="see">see =&gt; %s</div>' % seen
 
            license = match_obj.group('license')
 
            if license:
 
                return '<div class="label label-meta" data-tag="license"><a href="http://www.opensource.org/licenses/%s">%s</a></div>' % (license, license)
 
            tagtype = match_obj.group('tagtype')
 
            if tagtype:
 
                tagvalue = match_obj.group('tagvalue')
 
                return '<div class="label label-meta" data-tag="%s">%s =&gt; <a href="/%s">%s</a></div>' % (tagtype, tagtype, tagvalue, tagvalue)
 
            lang = match_obj.group('lang')
 
            if lang:
 
                return '<div class="label label-meta" data-tag="lang">%s</div>' % lang
 
            tag = match_obj.group('tag')
 
            if tag:
 
                return '<div class="label label-meta" data-tag="%s">%s</div>' % (tag, tag)
 
        return match_obj.group(0)
 

	
 
    def _urlify(s):
 
        """
 
        Extract urls from text and make html links out of them
 
        """
 
        return _URLIFY_RE.sub(_replace, s)
 

	
 
    if truncate is None:
 
        s = s.rstrip()
 
    else:
 
        s = truncatef(s, truncate, whole_word=True)
 
    s = html_escape(s)
 
    s = _urlify(s)
 
    if repo_name is not None:
 
        s = urlify_issues(s, repo_name)
 
    if link_ is not None:
 
        # make href around everything that isn't a href already
 
        s = linkify_others(s, link_)
 
    s = s.replace('\r\n', '<br/>').replace('\n', '<br/>')
 
    # Turn HTML5 into more valid HTML4 as required by some mail readers.
 
    # (This is not done in one step in html_escape, because character codes like
 
    # &#123; risk to be seen as an issue reference due to the presence of '#'.)
 
    s = s.replace("&apos;", "&#39;")
 
    return literal(s)
 

	
 

	
 
def linkify_others(t, l):
 
    """Add a default link to html with links.
 
    HTML doesn't allow nesting of links, so the outer link must be broken up
 
    in pieces and give space for other links.
 
    """
 
    urls = re.compile(r'(\<a.*?\<\/a\>)',)
 
    links = []
 
    for e in urls.split(t):
 
        if e.strip() and not urls.match(e):
 
            links.append('<a class="message-link" href="%s">%s</a>' % (l, e))
 
        else:
 
            links.append(e)
 

	
 
    return ''.join(links)
 

	
 

	
 
# Global variable that will hold the actual urlify_issues function body.
 
# Will be set on first use when the global configuration has been read.
 
_urlify_issues_f = None
 

	
 

	
 
def urlify_issues(newtext, repo_name):
 
    """Urlify issue references according to .ini configuration"""
 
    global _urlify_issues_f
 
    if _urlify_issues_f is None:
 
        from kallithea import CONFIG
 
        from kallithea.model.db import URL_SEP
 
        assert CONFIG['sqlalchemy.url'] # make sure config has been loaded
 
        assert kallithea.CONFIG['sqlalchemy.url'] # make sure config has been loaded
 

	
 
        # Build chain of urlify functions, starting with not doing any transformation
 
        def tmp_urlify_issues_f(s):
 
            return s
 

	
 
        issue_pat_re = re.compile(r'issue_pat(.*)')
 
        for k in CONFIG:
 
        for k in kallithea.CONFIG:
 
            # Find all issue_pat* settings that also have corresponding server_link and prefix configuration
 
            m = issue_pat_re.match(k)
 
            if m is None:
 
                continue
 
            suffix = m.group(1)
 
            issue_pat = CONFIG.get(k)
 
            issue_server_link = CONFIG.get('issue_server_link%s' % suffix)
 
            issue_sub = CONFIG.get('issue_sub%s' % suffix)
 
            issue_prefix = CONFIG.get('issue_prefix%s' % suffix)
 
            issue_pat = kallithea.CONFIG.get(k)
 
            issue_server_link = kallithea.CONFIG.get('issue_server_link%s' % suffix)
 
            issue_sub = kallithea.CONFIG.get('issue_sub%s' % suffix)
 
            issue_prefix = kallithea.CONFIG.get('issue_prefix%s' % suffix)
 
            if issue_prefix:
 
                log.error('found unsupported issue_prefix%s = %r - use issue_sub%s instead', suffix, issue_prefix, suffix)
 
            if not issue_pat:
 
                log.error('skipping incomplete issue pattern %r: it needs a regexp', k)
 
                continue
 
            if not issue_server_link:
 
                log.error('skipping incomplete issue pattern %r: it needs issue_server_link%s', k, suffix)
 
                continue
 
            if issue_sub is None: # issue_sub can be empty but should be present
 
                log.error('skipping incomplete issue pattern %r: it needs (a potentially empty) issue_sub%s', k, suffix)
 
                continue
 

	
 
            # Wrap tmp_urlify_issues_f with substitution of this pattern, while making sure all loop variables (and compiled regexpes) are bound
 
            try:
 
                issue_re = re.compile(issue_pat)
 
            except re.error as e:
 
                log.error('skipping invalid issue pattern %r: %r -> %r %r. Error: %s', k, issue_pat, issue_server_link, issue_sub, str(e))
 
                continue
 

	
 
            log.debug('issue pattern %r: %r -> %r %r', k, issue_pat, issue_server_link, issue_sub)
 

	
 
            def issues_replace(match_obj,
 
                               issue_server_link=issue_server_link, issue_sub=issue_sub):
 
                try:
 
                    issue_url = match_obj.expand(issue_server_link)
 
                except (IndexError, re.error) as e:
 
                    log.error('invalid issue_url setting %r -> %r %r. Error: %s', issue_pat, issue_server_link, issue_sub, str(e))
 
                    issue_url = issue_server_link
 
                issue_url = issue_url.replace('{repo}', repo_name)
 
                issue_url = issue_url.replace('{repo_name}', repo_name.split(URL_SEP)[-1])
 
                # if issue_sub is empty use the matched issue reference verbatim
 
                if not issue_sub:
 
                    issue_text = match_obj.group()
 
                else:
 
                    try:
 
                        issue_text = match_obj.expand(issue_sub)
 
                    except (IndexError, re.error) as e:
 
                        log.error('invalid issue_sub setting %r -> %r %r. Error: %s', issue_pat, issue_server_link, issue_sub, str(e))
 
                        issue_text = match_obj.group()
 

	
 
                return (
 
                    '<a class="issue-tracker-link" href="%(url)s">'
 
                    '%(text)s'
 
                    '</a>'
 
                    ) % {
 
                     'url': issue_url,
 
                     'text': issue_text,
 
                    }
 

	
 
            def tmp_urlify_issues_f(s, issue_re=issue_re, issues_replace=issues_replace, chain_f=tmp_urlify_issues_f):
 
                return issue_re.sub(issues_replace, chain_f(s))
 

	
 
        # Set tmp function globally - atomically
 
        _urlify_issues_f = tmp_urlify_issues_f
 

	
 
    return _urlify_issues_f(newtext)
 

	
 

	
 
def render_w_mentions(source, repo_name=None):
 
    """
 
    Render plain text with revision hashes and issue references urlified
 
    and with @mention highlighting.
 
    """
 
    s = safe_str(source)
 
    s = urlify_text(s, repo_name=repo_name)
 
    return literal('<div class="formatted-fixed">%s</div>' % s)
 

	
 

	
 
def short_ref(ref_type, ref_name):
 
    if ref_type == 'rev':
 
        return short_id(ref_name)
 
    return ref_name
 

	
 

	
 
def link_to_ref(repo_name, ref_type, ref_name, rev=None):
 
    """
 
    Return full markup for a href to changeset_home for a changeset.
 
    If ref_type is branch it will link to changelog.
 
    ref_name is shortened if ref_type is 'rev'.
 
    if rev is specified show it too, explicitly linking to that revision.
 
    """
 
    txt = short_ref(ref_type, ref_name)
 
    if ref_type == 'branch':
 
        u = url('changelog_home', repo_name=repo_name, branch=ref_name)
 
    else:
 
        u = url('changeset_home', repo_name=repo_name, revision=ref_name)
 
    l = link_to(repo_name + '#' + txt, u)
 
    if rev and ref_type != 'rev':
 
        l = literal('%s (%s)' % (l, link_to(short_id(rev), url('changeset_home', repo_name=repo_name, revision=rev))))
 
    return l
 

	
 

	
 
def changeset_status(repo, revision):
 
    from kallithea.model.changeset_status import ChangesetStatusModel
 
    return ChangesetStatusModel().get_status(repo, revision)
 

	
kallithea/tests/functional/test_search_indexing.py
Show inline comments
 
import mock
 

	
 
from kallithea import CONFIG
 
import kallithea
 
from kallithea.config.conf import INDEX_FILENAMES
 
from kallithea.model.meta import Session
 
from kallithea.model.repo import RepoModel
 
from kallithea.model.repo_group import RepoGroupModel
 
from kallithea.tests import base
 
from kallithea.tests.fixture import Fixture, create_test_index
 

	
 

	
 
fixture = Fixture()
 

	
 

	
 
def init_indexing_test(repo):
 
    prev = fixture.commit_change(repo.repo_name,
 
                                 filename='this_should_be_unique_filename.txt',
 
                                 content='this_should_be_unique_content\n',
 
                                 message='this_should_be_unique_commit_log',
 
                                 vcs_type='hg',
 
                                 newfile=True)
 

	
 
def init_stopword_test(repo):
 
    prev = fixture.commit_change(repo.repo_name,
 
                                 filename='this/is/it',
 
                                 content='def test\n',
 
                                 message='bother to ask where - in folder',
 
                                 vcs_type='hg',
 
                                 newfile=True)
 
    prev = fixture.commit_change(repo.repo_name,
 
                                 filename='join.us',
 
                                 content='def test\n',
 
                                 message='bother to ask where - top level',
 
                                 author='this is it <this-is-it@foo.bar.com>',
 
                                 vcs_type='hg',
 
                                 parent=prev,
 
                                 newfile=True)
 

	
 

	
 
repos = [
 
    # reponame,              init func or fork base, groupname
 
    ('indexing_test',       init_indexing_test,     None),
 
    ('indexing_test-fork',  'indexing_test',       None),
 
    ('group/indexing_test', 'indexing_test',       'group'),
 
    ('this-is-it',          'indexing_test',       None),
 
    ('indexing_test-foo',   'indexing_test',       None),
 
    ('stopword_test',       init_stopword_test,     None),
 
]
 

	
 

	
 
# map: name => id
 
repoids = {}
 
groupids = {}
 

	
 

	
 
def rebuild_index(full_index):
 
    with mock.patch('kallithea.lib.indexers.daemon.log.debug',
 
                    lambda *args, **kwargs: None):
 
        # The more revisions managed repositories have, the more
 
        # memory capturing "log.debug()" output in "indexers.daemon"
 
        # requires. This may cause unintentional failure of subsequent
 
        # tests, if ENOMEM at forking "git" prevents from rebuilding
 
        # index for search.
 
        # Therefore, "log.debug()" is disabled regardless of logging
 
        # level while rebuilding index.
 
        # (FYI, ENOMEM occurs at forking "git" with python 2.7.3,
 
        # Linux 3.2.78-1 x86_64, 3GB memory, and no ulimit
 
        # configuration for memory)
 
        create_test_index(base.TESTS_TMP_PATH, CONFIG, full_index=full_index)
 
        create_test_index(base.TESTS_TMP_PATH, kallithea.CONFIG, full_index=full_index)
 

	
 

	
 
class TestSearchControllerIndexing(base.TestController):
 
    @classmethod
 
    def setup_class(cls):
 
        for reponame, init_or_fork, groupname in repos:
 
            if groupname and groupname not in groupids:
 
                group = fixture.create_repo_group(groupname)
 
                groupids[groupname] = group.group_id
 
            if callable(init_or_fork):
 
                repo = fixture.create_repo(reponame,
 
                                           repo_group=groupname)
 
                init_or_fork(repo)
 
            else:
 
                repo = fixture.create_fork(init_or_fork, reponame,
 
                                           repo_group=groupname)
 
            repoids[reponame] = repo.repo_id
 

	
 
        # treat "it" as indexable filename
 
        filenames_mock = list(INDEX_FILENAMES)
 
        filenames_mock.append('it')
 
        with mock.patch('kallithea.lib.indexers.daemon.INDEX_FILENAMES',
 
                        filenames_mock):
 
            rebuild_index(full_index=False) # only for newly added repos
 

	
 
    @classmethod
 
    def teardown_class(cls):
 
        # delete in reversed order, to delete fork destination at first
 
        for reponame, init_or_fork, groupname in reversed(repos):
 
            RepoModel().delete(repoids[reponame])
 

	
 
        for reponame, init_or_fork, groupname in reversed(repos):
 
            if groupname in groupids:
 
                RepoGroupModel().delete(groupids.pop(groupname),
 
                                        force_delete=True)
 

	
 
        Session().commit()
 
        Session.remove()
 

	
 
        rebuild_index(full_index=True) # rebuild fully for subsequent tests
 

	
 
    @base.parametrize('reponame', [
 
        ('indexing_test'),
 
        ('indexing_test-fork'),
 
        ('group/indexing_test'),
 
        ('this-is-it'),
 
        ('*-fork'),
 
        ('group/*'),
 
    ])
 
    @base.parametrize('searchtype,query,hit', [
 
        ('content', 'this_should_be_unique_content', 1),
 
        ('commit', 'this_should_be_unique_commit_log', 1),
 
        ('path', 'this_should_be_unique_filename.txt', 1),
 
    ])
 
    def test_repository_tokenization(self, reponame, searchtype, query, hit):
 
        self.log_user()
 

	
 
        q = 'repository:%s %s' % (reponame, query)
 
        response = self.app.get(base.url(controller='search', action='index'),
 
                                {'q': q, 'type': searchtype})
 
        response.mustcontain('>%d results' % hit)
 

	
 
    @base.parametrize('reponame', [
 
        ('indexing_test'),
 
        ('indexing_test-fork'),
 
        ('group/indexing_test'),
 
        ('this-is-it'),
 
    ])
 
    @base.parametrize('searchtype,query,hit', [
 
        ('content', 'this_should_be_unique_content', 1),
 
        ('commit', 'this_should_be_unique_commit_log', 1),
 
        ('path', 'this_should_be_unique_filename.txt', 1),
 
    ])
 
    def test_searching_under_repository(self, reponame, searchtype, query, hit):
 
        self.log_user()
 

	
 
        response = self.app.get(base.url(controller='search', action='index',
 
                                    repo_name=reponame),
 
                                {'q': query, 'type': searchtype})
 
        response.mustcontain('>%d results' % hit)
 

	
 
    @base.parametrize('searchtype,query,hit', [
 
        ('content', 'path:this/is/it def test', 1),
 
        ('commit', 'added:this/is/it bother to ask where', 1),
 
        # this condition matches against files below, because
 
        # "path:" condition is also applied on "repository path".
 
        # - "this/is/it" in "stopword_test" repo
 
        # - "this_should_be_unique_filename.txt" in "this-is-it" repo
 
        ('path', 'this/is/it', 2),
 

	
 
        ('content', 'extension:us', 1),
 
        ('path', 'extension:us', 1),
 
    ])
 
    def test_filename_stopword(self, searchtype, query, hit):
 
        response = self.app.get(base.url(controller='search', action='index'),
 
                                {'q': query, 'type': searchtype})
kallithea/tests/other/test_vcs_operations.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/>.
 
"""
 
Test suite for vcs push/pull operations.
 

	
 
The tests need Git > 1.8.1.
 

	
 
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: Dec 30, 2010
 
:author: marcink
 
:copyright: (c) 2013 RhodeCode GmbH, and others.
 
:license: GPLv3, see LICENSE.md for more details.
 

	
 
"""
 

	
 
import json
 
import os
 
import re
 
import tempfile
 
import time
 
import urllib.request
 
from subprocess import PIPE, Popen
 
from tempfile import _RandomNameSequence
 

	
 
import pytest
 

	
 
from kallithea import CONFIG
 
import kallithea
 
from kallithea.lib.utils2 import ascii_bytes, safe_str
 
from kallithea.model.db import Repository, Ui, User, UserIpMap, UserLog
 
from kallithea.model.meta import Session
 
from kallithea.model.ssh_key import SshKeyModel
 
from kallithea.model.user import UserModel
 
from kallithea.tests import base
 
from kallithea.tests.fixture import Fixture
 

	
 

	
 
DEBUG = True
 
HOST = '127.0.0.1:4999'  # test host
 

	
 
fixture = Fixture()
 

	
 

	
 
# Parameterize different kinds of VCS testing - both the kind of VCS and the
 
# access method (HTTP/SSH)
 

	
 
# Mixin for using HTTP and SSH URLs
 
class HttpVcsTest(object):
 
    @staticmethod
 
    def repo_url_param(webserver, repo_name, **kwargs):
 
        return webserver.repo_url(repo_name, **kwargs)
 

	
 
class SshVcsTest(object):
 
    public_keys = {
 
        base.TEST_USER_REGULAR_LOGIN: 'ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAAAgQC6Ycnc2oUZHQnQwuqgZqTTdMDZD7ataf3JM7oG2Fw8JR6cdmz4QZLe5mfDwaFwG2pWHLRpVqzfrD/Pn3rIO++bgCJH5ydczrl1WScfryV1hYMJ/4EzLGM657J1/q5EI+b9SntKjf4ax+KP322L0TNQGbZUHLbfG2MwHMrYBQpHUQ== kallithea@localhost',
 
        base.TEST_USER_ADMIN_LOGIN: 'ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAAAgQC6Ycnc2oUZHQnQwuqgZqTTdMDZD7ataf3JM7oG2Fw8JR6cdmz4QZLe5mfDwaFwG2pWHLRpVqzfrD/Pn3rIO++bgCJH5ydczrl1WScfryV1hYMJ/4EzLGM657J1/q5EI+b9SntKjf4ax+KP322L0TNQGbZUHLbfG2MwHMrYBQpHUq== kallithea@localhost',
 
    }
 

	
 
    @classmethod
 
    def repo_url_param(cls, webserver, repo_name, username=base.TEST_USER_ADMIN_LOGIN, password=base.TEST_USER_ADMIN_PASS, client_ip=base.IP_ADDR):
 
        user = User.get_by_username(username)
 
        if user.ssh_keys:
 
            ssh_key = user.ssh_keys[0]
 
        else:
 
            sshkeymodel = SshKeyModel()
 
            ssh_key = sshkeymodel.create(user, 'test key', cls.public_keys[user.username])
 
            Session().commit()
 

	
 
        return cls._ssh_param(repo_name, user, ssh_key, client_ip)
 

	
 
# Mixins for using Mercurial and Git
 
class HgVcsTest(object):
 
    repo_type = 'hg'
 
    repo_name = base.HG_REPO
 

	
 
class GitVcsTest(object):
 
    repo_type = 'git'
 
    repo_name = base.GIT_REPO
 

	
 
# Combine mixins to give the combinations we want to parameterize tests with
 
class HgHttpVcsTest(HgVcsTest, HttpVcsTest):
 
    pass
 

	
 
class GitHttpVcsTest(GitVcsTest, HttpVcsTest):
 
    pass
 

	
 
class HgSshVcsTest(HgVcsTest, SshVcsTest):
 
    @staticmethod
 
    def _ssh_param(repo_name, user, ssh_key, client_ip):
 
        # Specify a custom ssh command on the command line
 
        return r"""--config ui.ssh="bash -c 'SSH_ORIGINAL_COMMAND=\"\$2\" SSH_CONNECTION=\"%s 1024 127.0.0.1 22\" kallithea-cli ssh-serve -c %s %s %s' --" ssh://someuser@somehost/%s""" % (
 
            client_ip,
 
            CONFIG['__file__'],
 
            kallithea.CONFIG['__file__'],
 
            user.user_id,
 
            ssh_key.user_ssh_key_id,
 
            repo_name)
 

	
 
class GitSshVcsTest(GitVcsTest, SshVcsTest):
 
    @staticmethod
 
    def _ssh_param(repo_name, user, ssh_key, client_ip):
 
        # Set a custom ssh command in the global environment
 
        os.environ['GIT_SSH_COMMAND'] = r"""bash -c 'SSH_ORIGINAL_COMMAND="$2" SSH_CONNECTION="%s 1024 127.0.0.1 22" kallithea-cli ssh-serve -c %s %s %s' --""" % (
 
            client_ip,
 
            CONFIG['__file__'],
 
            kallithea.CONFIG['__file__'],
 
            user.user_id,
 
            ssh_key.user_ssh_key_id)
 
        return "ssh://someuser@somehost/%s""" % repo_name
 

	
 
parametrize_vcs_test = base.parametrize('vt', [
 
    HgHttpVcsTest,
 
    GitHttpVcsTest,
 
    HgSshVcsTest,
 
    GitSshVcsTest,
 
])
 
parametrize_vcs_test_hg = base.parametrize('vt', [
 
    HgHttpVcsTest,
 
    HgSshVcsTest,
 
])
 
parametrize_vcs_test_http = base.parametrize('vt', [
 
    HgHttpVcsTest,
 
    GitHttpVcsTest,
 
])
 

	
 
class Command(object):
 

	
 
    def __init__(self, cwd):
 
        self.cwd = cwd
 

	
 
    def execute(self, *args, **environ):
 
        """
 
        Runs command on the system with given ``args`` using simple space
 
        join without safe quoting.
 
        """
 
        command = ' '.join(args)
 
        ignoreReturnCode = environ.pop('ignoreReturnCode', False)
 
        if DEBUG:
 
            print('*** CMD %s ***' % command)
 
        testenv = dict(os.environ)
 
        testenv['LANG'] = 'en_US.UTF-8'
 
        testenv['LANGUAGE'] = 'en_US:en'
 
        testenv['HGPLAIN'] = ''
 
        testenv['HGRCPATH'] = ''
 
        testenv.update(environ)
 
        p = Popen(command, shell=True, stdout=PIPE, stderr=PIPE, cwd=self.cwd, env=testenv)
 
        stdout, stderr = p.communicate()
 
        if DEBUG:
 
            if stdout:
 
                print('stdout:', stdout)
 
            if stderr:
 
                print('stderr:', stderr)
 
        if not ignoreReturnCode:
 
            assert p.returncode == 0
 
        return safe_str(stdout), safe_str(stderr)
 

	
 

	
 
def _get_tmp_dir(prefix='vcs_operations-', suffix=''):
 
    return tempfile.mkdtemp(dir=base.TESTS_TMP_PATH, prefix=prefix, suffix=suffix)
 

	
 

	
 
def _add_files(vcs, dest_dir, files_no=3):
 
    """
 
    Generate some files, add it to dest_dir repo and push back
 
    vcs is git or hg and defines what VCS we want to make those files for
 

	
 
    :param vcs:
 
    :param dest_dir:
 
    """
 
    added_file = '%ssetup.py' % next(_RandomNameSequence())
 
    open(os.path.join(dest_dir, added_file), 'a').close()
 
    Command(dest_dir).execute(vcs, 'add', added_file)
 

	
 
    email = 'me@example.com'
 
    if os.name == 'nt':
 
        author_str = 'User <%s>' % email
 
    else:
 
        author_str = 'User ǝɯɐᴎ <%s>' % email
 
    for i in range(files_no):
 
        cmd = """echo "added_line%s" >> %s""" % (i, added_file)
 
        Command(dest_dir).execute(cmd)
 
        if vcs == 'hg':
 
            cmd = """hg commit -m "committed new %s" -u "%s" "%s" """ % (
 
                i, author_str, added_file
 
            )
 
        elif vcs == 'git':
 
            cmd = """git commit -m "committed new %s" --author "%s" "%s" """ % (
 
                i, author_str, added_file
 
            )
 
        # git commit needs EMAIL on some machines
 
        Command(dest_dir).execute(cmd, EMAIL=email)
 

	
 
def _add_files_and_push(webserver, vt, dest_dir, clone_url, ignoreReturnCode=False, files_no=3):
 
    _add_files(vt.repo_type, dest_dir, files_no=files_no)
 
    # PUSH it back
 
    stdout = stderr = None
 
    if vt.repo_type == 'hg':
 
        stdout, stderr = Command(dest_dir).execute('hg push -f --verbose', clone_url, ignoreReturnCode=ignoreReturnCode)
 
    elif vt.repo_type == 'git':
 
        stdout, stderr = Command(dest_dir).execute('git push -f --verbose', clone_url, "master", ignoreReturnCode=ignoreReturnCode)
 

	
 
    return stdout, stderr
0 comments (0 inline, 0 general)