utils: ported to python3 and new app

rhodecode/lib/utils.py

@@ -1,798 +1,809 b''
 
 # Copyright (C) 2010-2020 RhodeCode GmbH
 #
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU Affero General Public License, version 3
 # (only), as published by the Free Software Foundation.
 #
 # 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 Affero General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 #
 # This program is dual-licensed. If you wish to learn more about the
 # RhodeCode Enterprise Edition, including its added features, Support services,
 # and proprietary license terms, please see https://rhodecode.com/licenses/
 
 """
 Utilities library for RhodeCode
 """
 
 import datetime
 import decorator
-import json
 import logging
 import os
 import re
 import sys
 import shutil
 import socket
 import tempfile
 import traceback
 import tarfile
 import warnings
-import hashlib
 from os.path import join as jn
 
 import paste
 import pkg_resources
-from webhelpers2.text import collapse, remove_formatting
+from webhelpers2.text import collapse, strip_tags, convert_accented_entities, convert_misc_entities
+
 from mako import exceptions
-from pyramid.threadlocal import get_current_registry
 
+from rhodecode.lib.hash_utils import sha256_safe, md5, sha1
+from rhodecode.lib.str_utils import safe_bytes, safe_str
 from rhodecode.lib.vcs.backends.base import Config
 from rhodecode.lib.vcs.exceptions import VCSError
 from rhodecode.lib.vcs.utils.helpers import get_scm, get_scm_backend
-from rhodecode.lib.utils2 import (
-    safe_str, safe_unicode, get_current_rhodecode_user, md5, sha1)
+from rhodecode.lib.ext_json import sjson as json
 from rhodecode.model import meta
 from rhodecode.model.db import (
     Repository, User, RhodeCodeUi, UserLog, RepoGroup, UserGroup)
 from rhodecode.model.meta import Session
 
 
 log = logging.getLogger(__name__)
 
 REMOVED_REPO_PAT = re.compile(r'rm__\d{8}_\d{6}_\d{6}__.*')
 
 # String which contains characters that are not allowed in slug names for
 # repositories or repository groups. It is properly escaped to use it in
 # regular expressions.
-SLUG_BAD_CHARS = re.escape('`?=[]\;\'"<>,/~!@#$%^&*()+{}|:')
+SLUG_BAD_CHARS = re.escape(r'`?=[]\;\'"<>,/~!@#$%^&*()+{}|:')
 
 # Regex that matches forbidden characters in repo/group slugs.
-SLUG_BAD_CHAR_RE = re.compile('[{}\x00-\x08\x0b-\x0c\x0e-\x1f]'.format(SLUG_BAD_CHARS))
+SLUG_BAD_CHAR_RE = re.compile(r'[{}\x00-\x08\x0b-\x0c\x0e-\x1f]'.format(SLUG_BAD_CHARS))
 
 # Regex that matches allowed characters in repo/group slugs.
-SLUG_GOOD_CHAR_RE = re.compile('[^{}]'.format(SLUG_BAD_CHARS))
+SLUG_GOOD_CHAR_RE = re.compile(r'[^{}]'.format(SLUG_BAD_CHARS))
 
 # Regex that matches whole repo/group slugs.
-SLUG_RE = re.compile('[^{}]+'.format(SLUG_BAD_CHARS))
+SLUG_RE = re.compile(r'[^{}]+'.format(SLUG_BAD_CHARS))
 
 _license_cache = None
 
 
 def repo_name_slug(value):
     """
     Return slug of name of repository
     This function is called on each creation/modification
     of repository to prevent bad names in repo
     """
+
     replacement_char = '-'
 
-    slug = remove_formatting(value)
+    slug = strip_tags(value)
+    slug = convert_accented_entities(slug)
+    slug = convert_misc_entities(slug)
+
     slug = SLUG_BAD_CHAR_RE.sub('', slug)
-    slug = re.sub('[\s]+', '-', slug)
+    slug = re.sub(r'[\s]+', '-', slug)
     slug = collapse(slug, replacement_char)
+
     return slug
 
 
 #==============================================================================
 # PERM DECORATOR HELPERS FOR EXTRACTING NAMES FOR PERM CHECKS
 #==============================================================================
 def get_repo_slug(request):
     _repo = ''
 
-    if hasattr(request, 'db_repo'):
+    if hasattr(request, 'db_repo_name'):
         # if our requests has set db reference use it for name, this
         # translates the example.com/_<id> into proper repo names
-        _repo = request.db_repo.repo_name
+        _repo = request.db_repo_name
     elif getattr(request, 'matchdict', None):
         # pyramid
         _repo = request.matchdict.get('repo_name')
 
     if _repo:
         _repo = _repo.rstrip('/')
     return _repo
 
 
 def get_repo_group_slug(request):
     _group = ''
     if hasattr(request, 'db_repo_group'):
         # if our requests has set db reference use it for name, this
         # translates the example.com/_<id> into proper repo group names
         _group = request.db_repo_group.group_name
     elif getattr(request, 'matchdict', None):
         # pyramid
         _group = request.matchdict.get('repo_group_name')
 
     if _group:
         _group = _group.rstrip('/')
     return _group
 
 
 def get_user_group_slug(request):
     _user_group = ''
 
     if hasattr(request, 'db_user_group'):
         _user_group = request.db_user_group.users_group_name
     elif getattr(request, 'matchdict', None):
         # pyramid
         _user_group = request.matchdict.get('user_group_id')
         _user_group_name = request.matchdict.get('user_group_name')
         try:
             if _user_group:
                 _user_group = UserGroup.get(_user_group)
             elif _user_group_name:
                 _user_group = UserGroup.get_by_group_name(_user_group_name)
 
             if _user_group:
                 _user_group = _user_group.users_group_name
         except Exception:
             log.exception('Failed to get user group by id and name')
             # catch all failures here
             return None
 
     return _user_group
 
 
 def get_filesystem_repos(path, recursive=False, skip_removed_repos=True):
     """
     Scans given path for repos and return (name,(type,path)) tuple
 
     :param path: path to scan for repositories
     :param recursive: recursive search and return names with subdirs in front
     """
 
     # remove ending slash for better results
     path = path.rstrip(os.sep)
     log.debug('now scanning in %s location recursive:%s...', path, recursive)
 
     def _get_repos(p):
-        dirpaths = _get_dirpaths(p)
+        dirpaths = get_dirpaths(p)
         if not _is_dir_writable(p):
             log.warning('repo path without write access: %s', p)
 
         for dirpath in dirpaths:
             if os.path.isfile(os.path.join(p, dirpath)):
                 continue
             cur_path = os.path.join(p, dirpath)
 
             # skip removed repos
             if skip_removed_repos and REMOVED_REPO_PAT.match(dirpath):
                 continue
 
             #skip .<somethin> dirs
             if dirpath.startswith('.'):
                 continue
 
             try:
                 scm_info = get_scm(cur_path)
                 yield scm_info[1].split(path, 1)[-1].lstrip(os.sep), scm_info
             except VCSError:
                 if not recursive:
                     continue
                 #check if this dir containts other repos for recursive scan
                 rec_path = os.path.join(p, dirpath)
                 if os.path.isdir(rec_path):
                     for inner_scm in _get_repos(rec_path):
                         yield inner_scm
 
     return _get_repos(path)
 
 
-def _get_dirpaths(p):
+def get_dirpaths(p: str) -> list:
     try:
         # OS-independable way of checking if we have at least read-only
         # access or not.
         dirpaths = os.listdir(p)
     except OSError:
         log.warning('ignoring repo path without read access: %s', p)
         return []
 
     # os.listpath has a tweak: If a unicode is passed into it, then it tries to
     # decode paths and suddenly returns unicode objects itself. The items it
     # cannot decode are returned as strings and cause issues.
     #
     # Those paths are ignored here until a solid solution for path handling has
     # been built.
     expected_type = type(p)
 
     def _has_correct_type(item):
         if type(item) is not expected_type:
             log.error(
-                "Ignoring path %s since it cannot be decoded into unicode.",
+                "Ignoring path %s since it cannot be decoded into str.",
                 # Using "repr" to make sure that we see the byte value in case
                 # of support.
                 repr(item))
             return False
         return True
 
     dirpaths = [item for item in dirpaths if _has_correct_type(item)]
 
     return dirpaths
 
 
 def _is_dir_writable(path):
     """
     Probe if `path` is writable.
 
     Due to trouble on Cygwin / Windows, this is actually probing if it is
     possible to create a file inside of `path`, stat does not produce reliable
     results in this case.
     """
     try:
         with tempfile.TemporaryFile(dir=path):
             pass
     except OSError:
         return False
     return True
 
 
 def is_valid_repo(repo_name, base_path, expect_scm=None, explicit_scm=None, config=None):
     """
     Returns True if given path is a valid repository False otherwise.
     If expect_scm param is given also, compare if given scm is the same
     as expected from scm parameter. If explicit_scm is given don't try to
     detect the scm, just use the given one to check if repo is valid
 
     :param repo_name:
     :param base_path:
     :param expect_scm:
     :param explicit_scm:
     :param config:
 
     :return True: if given path is a valid repository
     """
     full_path = os.path.join(safe_str(base_path), safe_str(repo_name))
     log.debug('Checking if `%s` is a valid path for repository. '
               'Explicit type: %s', repo_name, explicit_scm)
 
     try:
         if explicit_scm:
             detected_scms = [get_scm_backend(explicit_scm)(
                 full_path, config=config).alias]
         else:
             detected_scms = get_scm(full_path)
 
         if expect_scm:
             return detected_scms[0] == expect_scm
         log.debug('path: %s is an vcs object:%s', full_path, detected_scms)
         return True
     except VCSError:
         log.debug('path: %s is not a valid repo !', full_path)
         return False
 
 
 def is_valid_repo_group(repo_group_name, base_path, skip_path_check=False):
     """
     Returns True if given path is a repository group, False otherwise
 
     :param repo_name:
     :param base_path:
     """
     full_path = os.path.join(safe_str(base_path), safe_str(repo_group_name))
     log.debug('Checking if `%s` is a valid path for repository group',
               repo_group_name)
 
     # check if it's not a repo
     if is_valid_repo(repo_group_name, base_path):
         log.debug('Repo called %s exist, it is not a valid repo group', repo_group_name)
         return False
 
     try:
         # we need to check bare git repos at higher level
         # since we might match branches/hooks/info/objects or possible
         # other things inside bare git repo
         maybe_repo = os.path.dirname(full_path)
         if maybe_repo == base_path:
             # skip root level repo check, we know root location CANNOT BE a repo group
             return False
 
         scm_ = get_scm(maybe_repo)
         log.debug('path: %s is a vcs object:%s, not valid repo group', full_path, scm_)
         return False
     except VCSError:
         pass
 
     # check if it's a valid path
     if skip_path_check or os.path.isdir(full_path):
         log.debug('path: %s is a valid repo group !', full_path)
         return True
 
     log.debug('path: %s is not a valid repo group !', full_path)
     return False
 
 
 def ask_ok(prompt, retries=4, complaint='[y]es or [n]o please!'):
     while True:
         ok = eval(input(prompt))
         if ok.lower() in ('y', 'ye', 'yes'):
             return True
         if ok.lower() in ('n', 'no', 'nop', 'nope'):
             return False
         retries = retries - 1
         if retries < 0:
             raise IOError
         print(complaint)
 
 # propagated from mercurial documentation
 ui_sections = [
     'alias', 'auth',
     'decode/encode', 'defaults',
     'diff', 'email',
     'extensions', 'format',
     'merge-patterns', 'merge-tools',
     'hooks', 'http_proxy',
     'smtp', 'patch',
     'paths', 'profiling',
     'server', 'trusted',
     'ui', 'web', ]
 
 
 def config_data_from_db(clear_session=True, repo=None):
     """
     Read the configuration data from the database and return configuration
     tuples.
     """
     from rhodecode.model.settings import VcsSettingsModel
 
     config = []
 
     sa = meta.Session()
     settings_model = VcsSettingsModel(repo=repo, sa=sa)
 
     ui_settings = settings_model.get_ui_settings()
 
     ui_data = []
     for setting in ui_settings:
         if setting.active:
             ui_data.append((setting.section, setting.key, setting.value))
             config.append((
                 safe_str(setting.section), safe_str(setting.key),
                 safe_str(setting.value)))
         if setting.key == 'push_ssl':
             # force set push_ssl requirement to False, rhodecode
             # handles that
             config.append((
                 safe_str(setting.section), safe_str(setting.key), False))
     log.debug(
         'settings ui from db@repo[%s]: %s',
         repo,
-        ','.join(map(lambda s: '[{}] {}={}'.format(*s), ui_data)))
+        ','.join(['[{}] {}={}'.format(*s) for s in ui_data]))
     if clear_session:
         meta.Session.remove()
 
     # TODO: mikhail: probably it makes no sense to re-read hooks information.
     # It's already there and activated/deactivated
     skip_entries = []
     enabled_hook_classes = get_enabled_hook_classes(ui_settings)
     if 'pull' not in enabled_hook_classes:
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PRE_PULL))
     if 'push' not in enabled_hook_classes:
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PRE_PUSH))
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PRETX_PUSH))
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PUSH_KEY))
 
     config = [entry for entry in config if entry[:2] not in skip_entries]
 
     return config
 
 
 def make_db_config(clear_session=True, repo=None):
     """
     Create a :class:`Config` instance based on the values in the database.
     """
     config = Config()
     config_data = config_data_from_db(clear_session=clear_session, repo=repo)
     for section, option, value in config_data:
         config.set(section, option, value)
     return config
 
 
 def get_enabled_hook_classes(ui_settings):
     """
     Return the enabled hook classes.
 
     :param ui_settings: List of ui_settings as returned
         by :meth:`VcsSettingsModel.get_ui_settings`
 
     :return: a list with the enabled hook classes. The order is not guaranteed.
     :rtype: list
     """
     enabled_hooks = []
     active_hook_keys = [
         key for section, key, value, active in ui_settings
         if section == 'hooks' and active]
 
     hook_names = {
         RhodeCodeUi.HOOK_PUSH: 'push',
         RhodeCodeUi.HOOK_PULL: 'pull',
         RhodeCodeUi.HOOK_REPO_SIZE: 'repo_size'
     }
 
     for key in active_hook_keys:
         hook = hook_names.get(key)
         if hook:
             enabled_hooks.append(hook)
 
     return enabled_hooks
 
 
 def set_rhodecode_config(config):
     """
     Updates pyramid config with new settings from database
 
     :param config:
     """
     from rhodecode.model.settings import SettingsModel
     app_settings = SettingsModel().get_all_settings()
 
-    for k, v in app_settings.items():
+    for k, v in list(app_settings.items()):
         config[k] = v
 
 
 def get_rhodecode_realm():
     """
     Return the rhodecode realm from database.
     """
     from rhodecode.model.settings import SettingsModel
     realm = SettingsModel().get_setting_by_name('realm')
     return safe_str(realm.app_settings_value)
 
 
 def get_rhodecode_base_path():
     """
     Returns the base path. The base path is the filesystem path which points
     to the repository store.
     """
-    from rhodecode.model.settings import SettingsModel
-    paths_ui = SettingsModel().get_ui_by_section_and_key('paths', '/')
-    return safe_str(paths_ui.ui_value)
+
+    import rhodecode
+    return rhodecode.CONFIG['default_base_path']
 
 
 def map_groups(path):
     """
     Given a full path to a repository, create all nested groups that this
     repo is inside. This function creates parent-child relationships between
     groups and creates default perms for all new groups.
 
     :param paths: full path to repository
     """
     from rhodecode.model.repo_group import RepoGroupModel
     sa = meta.Session()
     groups = path.split(Repository.NAME_SEP)
     parent = None
     group = None
 
     # last element is repo in nested groups structure
     groups = groups[:-1]
     rgm = RepoGroupModel(sa)
     owner = User.get_first_super_admin()
     for lvl, group_name in enumerate(groups):
         group_name = '/'.join(groups[:lvl] + [group_name])
         group = RepoGroup.get_by_group_name(group_name)
         desc = '%s group' % group_name
 
         # skip folders that are now removed repos
         if REMOVED_REPO_PAT.match(group_name):
             break
 
         if group is None:
             log.debug('creating group level: %s group_name: %s',
                        lvl, group_name)
             group = RepoGroup(group_name, parent)
             group.group_description = desc
             group.user = owner
             sa.add(group)
             perm_obj = rgm._create_default_perms(group)
             sa.add(perm_obj)
             sa.flush()
 
         parent = group
     return group
 
 
 def repo2db_mapper(initial_repo_list, remove_obsolete=False):
     """
     maps all repos given in initial_repo_list, non existing repositories
     are created, if remove_obsolete is True it also checks for db entries
     that are not in initial_repo_list and removes them.
 
     :param initial_repo_list: list of repositories found by scanning methods
     :param remove_obsolete: check for obsolete entries in database
     """
     from rhodecode.model.repo import RepoModel
     from rhodecode.model.repo_group import RepoGroupModel
     from rhodecode.model.settings import SettingsModel
 
     sa = meta.Session()
     repo_model = RepoModel()
     user = User.get_first_super_admin()
     added = []
 
     # creation defaults
     defs = SettingsModel().get_default_repo_settings(strip_prefix=True)
     enable_statistics = defs.get('repo_enable_statistics')
     enable_locking = defs.get('repo_enable_locking')
     enable_downloads = defs.get('repo_enable_downloads')
     private = defs.get('repo_private')
 
-    for name, repo in initial_repo_list.items():
+    for name, repo in list(initial_repo_list.items()):
         group = map_groups(name)
-        unicode_name = safe_unicode(name)
-        db_repo = repo_model.get_by_repo_name(unicode_name)
+        str_name = safe_str(name)
+        db_repo = repo_model.get_by_repo_name(str_name)
         # found repo that is on filesystem not in RhodeCode database
         if not db_repo:
             log.info('repository %s not found, creating now', name)
             added.append(name)
             desc = (repo.description
                     if repo.description != 'unknown'
                     else '%s repository' % name)
 
             db_repo = repo_model._create_repo(
                 repo_name=name,
                 repo_type=repo.alias,
                 description=desc,
                 repo_group=getattr(group, 'group_id', None),
                 owner=user,
                 enable_locking=enable_locking,
                 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 we updated server info
             if db_repo.repo_type == 'git':
                 git_repo = db_repo.scm_instance()
                 # update repository server-info
                 log.debug('Running update server info')
                 git_repo._update_server_info()
 
             db_repo.update_commit_cache()
 
         config = db_repo._config
         config.set('extensions', 'largefiles', '')
         repo = db_repo.scm_instance(config=config)
         repo.install_hooks()
 
     removed = []
     if remove_obsolete:
         # remove from database those repositories that are not in the filesystem
         for repo in sa.query(Repository).all():
-            if repo.repo_name not in initial_repo_list.keys():
+            if repo.repo_name not in list(initial_repo_list.keys()):
                 log.debug("Removing non-existing repository found in db `%s`",
                           repo.repo_name)
                 try:
                     RepoModel(sa).delete(repo, forks='detach', fs_remove=False)
                     sa.commit()
                     removed.append(repo.repo_name)
                 except Exception:
                     # don't hold further removals on error
                     log.error(traceback.format_exc())
                     sa.rollback()
 
         def splitter(full_repo_name):
             _parts = full_repo_name.rsplit(RepoGroup.url_sep(), 1)
             gr_name = None
             if len(_parts) == 2:
                 gr_name = _parts[0]
             return gr_name
 
         initial_repo_group_list = [splitter(x) for x in
-                                   initial_repo_list.keys() if splitter(x)]
+                                   list(initial_repo_list.keys()) if splitter(x)]
 
         # remove from database those repository groups that are not in the
         # filesystem due to parent child relationships we need to delete them
         # in a specific order of most nested first
         all_groups = [x.group_name for x in sa.query(RepoGroup).all()]
-        nested_sort = lambda gr: len(gr.split('/'))
+        def nested_sort(gr):
+            return len(gr.split('/'))
         for group_name in sorted(all_groups, key=nested_sort, reverse=True):
             if group_name not in initial_repo_group_list:
                 repo_group = RepoGroup.get_by_group_name(group_name)
                 if (repo_group.children.all() or
                     not RepoGroupModel().check_exist_filesystem(
                         group_name=group_name, exc_on_failure=False)):
                     continue
 
                 log.info(
                     'Removing non-existing repository group found in db `%s`',
                     group_name)
                 try:
                     RepoGroupModel(sa).delete(group_name, fs_remove=False)
                     sa.commit()
                     removed.append(group_name)
                 except Exception:
                     # don't hold further removals on error
                     log.exception(
                         'Unable to remove repository group `%s`',
                         group_name)
                     sa.rollback()
                     raise
 
     return added, removed
 
 
 def load_rcextensions(root_path):
     import rhodecode
     from rhodecode.config import conf
 
     path = os.path.join(root_path)
     sys.path.append(path)
 
     try:
         rcextensions = __import__('rcextensions')
     except ImportError:
         if os.path.isdir(os.path.join(path, 'rcextensions')):
-            log.warn('Unable to load rcextensions from %s', path)
+            log.warning('Unable to load rcextensions from %s', path)
         rcextensions = None
 
     if rcextensions:
         log.info('Loaded rcextensions from %s...', rcextensions)
         rhodecode.EXTENSIONS = rcextensions
 
         # Additional mappings that are not present in the pygments lexers
         conf.LANGUAGES_EXTENSIONS_MAP.update(
             getattr(rhodecode.EXTENSIONS, 'EXTRA_MAPPINGS', {}))
 
 
 def get_custom_lexer(extension):
     """
     returns a custom lexer if it is defined in rcextensions module, or None
     if there's no custom lexer defined
     """
     import rhodecode
     from pygments import lexers
 
     # custom override made by RhodeCode
     if extension in ['mako']:
         return lexers.get_lexer_by_name('html+mako')
 
     # check if we didn't define this extension as other lexer
     extensions = rhodecode.EXTENSIONS and getattr(rhodecode.EXTENSIONS, 'EXTRA_LEXERS', None)
     if extensions and extension in rhodecode.EXTENSIONS.EXTRA_LEXERS:
         _lexer_name = rhodecode.EXTENSIONS.EXTRA_LEXERS[extension]
         return lexers.get_lexer_by_name(_lexer_name)
 
 
 #==============================================================================
 # TEST FUNCTIONS AND CREATORS
 #==============================================================================
 def create_test_index(repo_location, config):
     """
     Makes default test index.
     """
-    import rc_testdata
-
+    try:
+        import rc_testdata
+    except ImportError:
+        raise ImportError('Failed to import rc_testdata, '
+                          'please make sure this package is installed from requirements_test.txt')
     rc_testdata.extract_search_index(
         'vcs_search_index', os.path.dirname(config['search.location']))
 
 
 def create_test_directory(test_path):
     """
     Create test directory if it doesn't exist.
     """
     if not os.path.isdir(test_path):
         log.debug('Creating testdir %s', test_path)
         os.makedirs(test_path)
 
 
 def create_test_database(test_path, config):
     """
     Makes a fresh database.
     """
     from rhodecode.lib.db_manage import DbManage
+    from rhodecode.lib.utils2 import get_encryption_key
 
     # PART ONE create db
     dbconf = config['sqlalchemy.db1.url']
+    enc_key = get_encryption_key(config)
+
     log.debug('making test db %s', dbconf)
 
     dbmanage = DbManage(log_sql=False, dbconf=dbconf, root=config['here'],
-                        tests=True, cli_args={'force_ask': True})
+                        tests=True, cli_args={'force_ask': True}, enc_key=enc_key)
     dbmanage.create_tables(override=True)
     dbmanage.set_db_version()
     # for tests dynamically set new root paths based on generated content
     dbmanage.create_settings(dbmanage.config_prompt(test_path))
     dbmanage.create_default_user()
     dbmanage.create_test_admin_and_users()
     dbmanage.create_permissions()
     dbmanage.populate_default_permissions()
     Session().commit()
 
 
 def create_test_repositories(test_path, config):
     """
     Creates test repositories in the temporary directory. Repositories are
     extracted from archives within the rc_testdata package.
     """
     import rc_testdata
     from rhodecode.tests import HG_REPO, GIT_REPO, SVN_REPO
 
     log.debug('making test vcs repositories')
 
     idx_path = config['search.location']
     data_path = config['cache_dir']
 
     # clean index and data
     if idx_path and os.path.exists(idx_path):
         log.debug('remove %s', idx_path)
         shutil.rmtree(idx_path)
 
     if data_path and os.path.exists(data_path):
         log.debug('remove %s', data_path)
         shutil.rmtree(data_path)
 
     rc_testdata.extract_hg_dump('vcs_test_hg', jn(test_path, HG_REPO))
     rc_testdata.extract_git_dump('vcs_test_git', jn(test_path, GIT_REPO))
 
     # Note: Subversion is in the process of being integrated with the system,
     # until we have a properly packed version of the test svn repository, this
     # tries to copy over the repo from a package "rc_testdata"
     svn_repo_path = rc_testdata.get_svn_repo_archive()
     with tarfile.open(svn_repo_path) as tar:
         tar.extractall(jn(test_path, SVN_REPO))
 
 
 def password_changed(auth_user, session):
     # Never report password change in case of default user or anonymous user.
     if auth_user.username == User.DEFAULT_USER or auth_user.user_id is None:
         return False
 
-    password_hash = md5(auth_user.password) if auth_user.password else None
+    password_hash = md5(safe_bytes(auth_user.password)) if auth_user.password else None
     rhodecode_user = session.get('rhodecode_user', {})
     session_password_hash = rhodecode_user.get('password', '')
     return password_hash != session_password_hash
 
 
 def read_opensource_licenses():
     global _license_cache
 
     if not _license_cache:
         licenses = pkg_resources.resource_string(
             'rhodecode', 'config/licenses.json')
         _license_cache = json.loads(licenses)
 
     return _license_cache
 
 
 def generate_platform_uuid():
     """
     Generates platform UUID based on it's name
     """
     import platform
 
     try:
         uuid_list = [platform.platform()]
-        return hashlib.sha256(':'.join(uuid_list)).hexdigest()
+        return sha256_safe(':'.join(uuid_list))
     except Exception as e:
         log.error('Failed to generate host uuid: %s', e)
         return 'UNDEFINED'
 
 
 def send_test_email(recipients, email_body='TEST EMAIL'):
     """
     Simple code for generating test emails.
     Usage::
 
         from rhodecode.lib import utils
         utils.send_test_email()
     """
     from rhodecode.lib.celerylib import tasks, run_task
 
     email_body = email_body_plaintext = email_body
     subject = 'SUBJECT FROM: {}'.format(socket.gethostname())
     tasks.send_email(recipients, subject, email_body_plaintext, email_body)

rhodecode/lib/utils2.py

@@ -1,1047 +1,989 b''
 
 
 # Copyright (C) 2011-2020 RhodeCode GmbH
 #
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU Affero General Public License, version 3
 # (only), as published by the Free Software Foundation.
 #
 # 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 Affero General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 #
 # This program is dual-licensed. If you wish to learn more about the
 # RhodeCode Enterprise Edition, including its added features, Support services,
 # and proprietary license terms, please see https://rhodecode.com/licenses/
 
 
 """
 Some simple helper functions
 """
 
 import collections
 import datetime
 import dateutil.relativedelta
 import logging
 import re
 import sys
 import time
 import urllib.request
 import urllib.parse
 import urllib.error
 import urlobject
 import uuid
 import getpass
 import socket
 import errno
 import random
-from functools import update_wrapper, partial, wraps
+import functools
 from contextlib import closing
 
 import pygments.lexers
 import sqlalchemy
+import sqlalchemy.event
 import sqlalchemy.engine.url
 import sqlalchemy.exc
 import sqlalchemy.sql
 import webob
-import pyramid.threadlocal
 from pyramid.settings import asbool
 
 import rhodecode
 from rhodecode.translation import _, _pluralize
 from rhodecode.lib.str_utils import safe_str, safe_int, safe_bytes
 from rhodecode.lib.hash_utils import md5, md5_safe, sha1, sha1_safe
-from rhodecode.lib.type_utils import aslist, str2bool
-from functools import reduce
-
-#TODO: there's no longer safe_unicode, we mock it now, but should remove it
-safe_unicode = safe_str
+from rhodecode.lib.type_utils import aslist, str2bool, StrictAttributeDict, AttributeDict
 
 
 def __get_lem(extra_mapping=None):
     """
     Get language extension map based on what's inside pygments lexers
     """
     d = collections.defaultdict(lambda: [])
 
     def __clean(s):
         s = s.lstrip('*')
         s = s.lstrip('.')
 
         if s.find('[') != -1:
             exts = []
             start, stop = s.find('['), s.find(']')
 
             for suffix in s[start + 1:stop]:
                 exts.append(s[:s.find('[')] + suffix)
             return [e.lower() for e in exts]
         else:
             return [s.lower()]
 
     for lx, t in sorted(pygments.lexers.LEXERS.items()):
         m = list(map(__clean, t[-2]))
         if m:
-            m = reduce(lambda x, y: x + y, m)
+            m = functools.reduce(lambda x, y: x + y, m)
             for ext in m:
                 desc = lx.replace('Lexer', '')
                 d[ext].append(desc)
 
     data = dict(d)
 
     extra_mapping = extra_mapping or {}
     if extra_mapping:
-        for k, v in extra_mapping.items():
+        for k, v in list(extra_mapping.items()):
             if k not in data:
                 # register new mapping2lexer
                 data[k] = [v]
 
     return data
 
 
-def convert_line_endings(line, mode):
+def convert_line_endings(line: str, mode) -> str:
     """
     Converts a given line  "line end" accordingly to given mode
 
     Available modes are::
         0 - Unix
         1 - Mac
         2 - DOS
 
     :param line: given line to convert
     :param mode: mode to convert to
-    :rtype: str
     :return: converted line according to mode
     """
     if mode == 0:
         line = line.replace('\r\n', '\n')
         line = line.replace('\r', '\n')
     elif mode == 1:
         line = line.replace('\r\n', '\r')
         line = line.replace('\n', '\r')
     elif mode == 2:
         line = re.sub('\r(?!\n)|(?<!\r)\n', '\r\n', line)
     return line
 
 
-def detect_mode(line, default):
+def detect_mode(line: str, default) -> int:
     """
     Detects line break for given line, if line break couldn't be found
     given default value is returned
 
     :param line: str line
     :param default: default
-    :rtype: int
     :return: value of line end on of 0 - Unix, 1 - Mac, 2 - DOS
     """
     if line.endswith('\r\n'):
         return 2
     elif line.endswith('\n'):
         return 0
     elif line.endswith('\r'):
         return 1
     else:
         return default
 
 
 def remove_suffix(s, suffix):
     if s.endswith(suffix):
         s = s[:-1 * len(suffix)]
     return s
 
 
 def remove_prefix(s, prefix):
     if s.startswith(prefix):
         s = s[len(prefix):]
     return s
 
 
-def find_calling_context(ignore_modules=None):
+def find_calling_context(ignore_modules=None, depth=4, output_writer=None, indent=True):
     """
     Look through the calling stack and return the frame which called
     this function and is part of core module ( ie. rhodecode.* )
 
     :param ignore_modules: list of modules to ignore eg. ['rhodecode.lib']
+    :param depth:
+    :param output_writer:
+    :param indent:
 
     usage::
+
         from rhodecode.lib.utils2 import find_calling_context
 
         calling_context = find_calling_context(ignore_modules=[
             'rhodecode.lib.caching_query',
             'rhodecode.model.settings',
         ])
 
-        if calling_context:
-            cc_str = 'call context %s:%s' % (
-                calling_context.f_code.co_filename,
-                calling_context.f_lineno,
-            )
-            print(cc_str)
     """
+    import inspect
+    if not output_writer:
+        try:
+            from rich import print as pprint
+        except ImportError:
+            pprint = print
+        output_writer = pprint
 
-    ignore_modules = ignore_modules or []
+    frame = inspect.currentframe()
+    cc = []
+    try:
+        for i in range(depth):  # current frame + 3 callers
+            frame = frame.f_back
+            if not frame:
+                break
 
-    f = sys._getframe(2)
-    while f.f_back is not None:
-        name = f.f_globals.get('__name__')
-        if name and name.startswith(__name__.split('.')[0]):
+            info = inspect.getframeinfo(frame)
+            name = frame.f_globals.get('__name__')
             if name not in ignore_modules:
-                return f
-        f = f.f_back
-    return None
+                cc.insert(0, f'CALL_CONTEXT:{i}: file {info.filename}:{info.lineno} -> {info.function}')
+    finally:
+        # Avoids a reference cycle
+        del frame
+
+    output_writer('* INFO: This code was called from: *')
+    for cnt, frm_info in enumerate(cc):
+        if not indent:
+            cnt = 1
+        output_writer(' ' * cnt + frm_info)
 
 
 def ping_connection(connection, branch):
     if branch:
         # "branch" refers to a sub-connection of a connection,
         # we don't want to bother pinging on these.
         return
 
     # turn off "close with result".  This flag is only used with
     # "connectionless" execution, otherwise will be False in any case
     save_should_close_with_result = connection.should_close_with_result
     connection.should_close_with_result = False
 
     try:
         # run a SELECT 1.   use a core select() so that
         # the SELECT of a scalar value without a table is
         # appropriately formatted for the backend
         connection.scalar(sqlalchemy.sql.select([1]))
     except sqlalchemy.exc.DBAPIError as err:
         # catch SQLAlchemy's DBAPIError, which is a wrapper
         # for the DBAPI's exception.  It includes a .connection_invalidated
         # attribute which specifies if this connection is a "disconnect"
         # condition, which is based on inspection of the original exception
         # by the dialect in use.
         if err.connection_invalidated:
             # run the same SELECT again - the connection will re-validate
             # itself and establish a new connection.  The disconnect detection
             # here also causes the whole connection pool to be invalidated
             # so that all stale connections are discarded.
             connection.scalar(sqlalchemy.sql.select([1]))
         else:
             raise
     finally:
         # restore "close with result"
         connection.should_close_with_result = save_should_close_with_result
 
 
 def engine_from_config(configuration, prefix='sqlalchemy.', **kwargs):
     """Custom engine_from_config functions."""
     log = logging.getLogger('sqlalchemy.engine')
     use_ping_connection = asbool(configuration.pop('sqlalchemy.db1.ping_connection', None))
     debug = asbool(configuration.pop('sqlalchemy.db1.debug_query', None))
 
     engine = sqlalchemy.engine_from_config(configuration, prefix, **kwargs)
 
     def color_sql(sql):
         color_seq = '\033[1;33m'  # This is yellow: code 33
         normal = '\x1b[0m'
         return ''.join([color_seq, sql, normal])
 
     if use_ping_connection:
         log.debug('Adding ping_connection on the engine config.')
         sqlalchemy.event.listen(engine, "engine_connect", ping_connection)
 
     if debug:
         # attach events only for debug configuration
         def before_cursor_execute(conn, cursor, statement,
                                   parameters, context, executemany):
             setattr(conn, 'query_start_time', time.time())
             log.info(color_sql(">>>>> STARTING QUERY >>>>>"))
-            calling_context = find_calling_context(ignore_modules=[
+            find_calling_context(ignore_modules=[
                 'rhodecode.lib.caching_query',
                 'rhodecode.model.settings',
-            ])
-            if calling_context:
-                log.info(color_sql('call context %s:%s' % (
-                    calling_context.f_code.co_filename,
-                    calling_context.f_lineno,
-                )))
+            ], output_writer=log.info)
 
         def after_cursor_execute(conn, cursor, statement,
                                  parameters, context, executemany):
             delattr(conn, 'query_start_time')
 
         sqlalchemy.event.listen(engine, "before_cursor_execute", before_cursor_execute)
         sqlalchemy.event.listen(engine, "after_cursor_execute", after_cursor_execute)
 
     return engine
 
 
-def get_encryption_key(config):
+def get_encryption_key(config) -> bytes:
     secret = config.get('rhodecode.encrypted_values.secret')
     default = config['beaker.session.secret']
-    return secret or default
+    enc_key = secret or default
+
+    return safe_bytes(enc_key)
 
 
 def age(prevdate, now=None, show_short_version=False, show_suffix=True, short_format=False):
     """
     Turns a datetime into an age string.
     If show_short_version is True, this generates a shorter string with
     an approximate age; ex. '1 day ago', rather than '1 day and 23 hours ago'.
 
     * IMPORTANT*
     Code of this function is written in special way so it's easier to
     backport it to javascript. If you mean to update it, please also update
     `jquery.timeago-extension.js` file
 
     :param prevdate: datetime object
     :param now: get current time, if not define we use
         `datetime.datetime.now()`
     :param show_short_version: if it should approximate the date and
         return a shorter string
     :param show_suffix:
     :param short_format: show short format, eg 2D instead of 2 days
     :rtype: unicode
     :returns: unicode words describing age
     """
 
     def _get_relative_delta(now, prevdate):
         base = dateutil.relativedelta.relativedelta(now, prevdate)
         return {
             'year': base.years,
             'month': base.months,
             'day': base.days,
             'hour': base.hours,
             'minute': base.minutes,
             'second': base.seconds,
         }
 
     def _is_leap_year(year):
         return year % 4 == 0 and (year % 100 != 0 or year % 400 == 0)
 
     def get_month(prevdate):
         return prevdate.month
 
     def get_year(prevdate):
         return prevdate.year
 
     now = now or datetime.datetime.now()
     order = ['year', 'month', 'day', 'hour', 'minute', 'second']
     deltas = {}
     future = False
 
     if prevdate > now:
         now_old = now
         now = prevdate
         prevdate = now_old
         future = True
     if future:
         prevdate = prevdate.replace(microsecond=0)
     # Get date parts deltas
     for part in order:
         rel_delta = _get_relative_delta(now, prevdate)
         deltas[part] = rel_delta[part]
 
     # Fix negative offsets (there is 1 second between 10:59:59 and 11:00:00,
     # not 1 hour, -59 minutes and -59 seconds)
     offsets = [[5, 60], [4, 60], [3, 24]]
     for element in offsets:  # seconds, minutes, hours
         num = element[0]
         length = element[1]
 
         part = order[num]
         carry_part = order[num - 1]
 
         if deltas[part] < 0:
             deltas[part] += length
             deltas[carry_part] -= 1
 
     # Same thing for days except that the increment depends on the (variable)
     # number of days in the month
     month_lengths = [31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31]
     if deltas['day'] < 0:
         if get_month(prevdate) == 2 and _is_leap_year(get_year(prevdate)):
             deltas['day'] += 29
         else:
             deltas['day'] += month_lengths[get_month(prevdate) - 1]
 
         deltas['month'] -= 1
 
     if deltas['month'] < 0:
         deltas['month'] += 12
         deltas['year'] -= 1
 
     # Format the result
     if short_format:
         fmt_funcs = {
             'year': lambda d: '%dy' % d,
             'month': lambda d: '%dm' % d,
             'day': lambda d: '%dd' % d,
             'hour': lambda d: '%dh' % d,
             'minute': lambda d: '%dmin' % d,
             'second': lambda d: '%dsec' % d,
         }
     else:
         fmt_funcs = {
             'year': lambda d: _pluralize('${num} year', '${num} years', d, mapping={'num': d}).interpolate(),
             'month': lambda d: _pluralize('${num} month', '${num} months', d, mapping={'num': d}).interpolate(),
             'day': lambda d: _pluralize('${num} day', '${num} days', d, mapping={'num': d}).interpolate(),
             'hour': lambda d: _pluralize('${num} hour', '${num} hours', d, mapping={'num': d}).interpolate(),
             'minute': lambda d: _pluralize('${num} minute', '${num} minutes', d, mapping={'num': d}).interpolate(),
             'second': lambda d: _pluralize('${num} second', '${num} seconds', d, mapping={'num': d}).interpolate(),
         }
 
     i = 0
     for part in order:
         value = deltas[part]
         if value != 0:
 
             if i < 5:
                 sub_part = order[i + 1]
                 sub_value = deltas[sub_part]
             else:
                 sub_value = 0
 
             if sub_value == 0 or show_short_version:
                 _val = fmt_funcs[part](value)
                 if future:
                     if show_suffix:
                         return _('in ${ago}', mapping={'ago': _val})
                     else:
                         return _(_val)
 
                 else:
                     if show_suffix:
                         return _('${ago} ago', mapping={'ago': _val})
                     else:
                         return _(_val)
 
             val = fmt_funcs[part](value)
             val_detail = fmt_funcs[sub_part](sub_value)
             mapping = {'val': val, 'detail': val_detail}
 
             if short_format:
                 datetime_tmpl = _('${val}, ${detail}', mapping=mapping)
                 if show_suffix:
                     datetime_tmpl = _('${val}, ${detail} ago', mapping=mapping)
                     if future:
                         datetime_tmpl = _('in ${val}, ${detail}', mapping=mapping)
             else:
                 datetime_tmpl = _('${val} and ${detail}', mapping=mapping)
                 if show_suffix:
                     datetime_tmpl = _('${val} and ${detail} ago', mapping=mapping)
                     if future:
                         datetime_tmpl = _('in ${val} and ${detail}', mapping=mapping)
 
             return datetime_tmpl
         i += 1
     return _('just now')
 
 
 def age_from_seconds(seconds):
     seconds = safe_int(seconds) or 0
     prevdate = time_to_datetime(time.time() + seconds)
     return age(prevdate, show_suffix=False, show_short_version=True)
 
 
 def cleaned_uri(uri):
     """
     Quotes '[' and ']' from uri if there is only one of them.
         according to RFC3986 we cannot use such chars in uri
     :param uri:
     :return: uri without this chars
     """
     return urllib.parse.quote(uri, safe='@$:/')
 
 
 def credentials_filter(uri):
     """
     Returns a url with removed credentials
 
     :param uri:
     """
     import urlobject
     if isinstance(uri, rhodecode.lib.encrypt.InvalidDecryptedValue):
         return 'InvalidDecryptionKey'
 
     url_obj = urlobject.URLObject(cleaned_uri(uri))
     url_obj = url_obj.without_password().without_username()
 
     return url_obj
 
 
 def get_host_info(request):
     """
     Generate host info, to obtain full url e.g https://server.com
     use this
     `{scheme}://{netloc}`
     """
     if not request:
         return {}
 
     qualified_home_url = request.route_url('home')
     parsed_url = urlobject.URLObject(qualified_home_url)
-    decoded_path = safe_unicode(urllib.parse.unquote(parsed_url.path.rstrip('/')))
+    decoded_path = safe_str(urllib.parse.unquote(parsed_url.path.rstrip('/')))
 
     return {
         'scheme': parsed_url.scheme,
         'netloc': parsed_url.netloc+decoded_path,
         'hostname': parsed_url.hostname,
     }
 
 
 def get_clone_url(request, uri_tmpl, repo_name, repo_id, repo_type, **override):
     qualified_home_url = request.route_url('home')
     parsed_url = urlobject.URLObject(qualified_home_url)
-    decoded_path = safe_unicode(urllib.parse.unquote(parsed_url.path.rstrip('/')))
+    decoded_path = safe_str(urllib.parse.unquote(parsed_url.path.rstrip('/')))
 
     args = {
         'scheme': parsed_url.scheme,
         'user': '',
         'sys_user': getpass.getuser(),
         # path if we use proxy-prefix
         'netloc': parsed_url.netloc+decoded_path,
         'hostname': parsed_url.hostname,
         'prefix': decoded_path,
         'repo': repo_name,
         'repoid': str(repo_id),
         'repo_type': repo_type
     }
     args.update(override)
     args['user'] = urllib.parse.quote(safe_str(args['user']))
 
-    for k, v in args.items():
-        uri_tmpl = uri_tmpl.replace('{%s}' % k, v)
+    for k, v in list(args.items()):
+        tmpl_key = '{%s}' % k
+        uri_tmpl = uri_tmpl.replace(tmpl_key, v)
 
     # special case for SVN clone url
     if repo_type == 'svn':
         uri_tmpl = uri_tmpl.replace('ssh://', 'svn+ssh://')
 
     # remove leading @ sign if it's present. Case of empty user
     url_obj = urlobject.URLObject(uri_tmpl)
     url = url_obj.with_netloc(url_obj.netloc.lstrip('@'))
 
-    return safe_unicode(url)
+    return safe_str(url)
 
 
 def get_commit_safe(repo, commit_id=None, commit_idx=None, pre_load=None,
                     maybe_unreachable=False, reference_obj=None):
     """
     Safe version of get_commit if this commit doesn't exists for a
     repository it returns a Dummy one instead
 
     :param repo: repository instance
     :param commit_id: commit id as str
     :param commit_idx: numeric commit index
     :param pre_load: optional list of commit attributes to load
     :param maybe_unreachable: translate unreachable commits on git repos
     :param reference_obj: explicitly search via a reference obj in git. E.g "branch:123" would mean branch "123"
     """
     # TODO(skreft): remove these circular imports
     from rhodecode.lib.vcs.backends.base import BaseRepository, EmptyCommit
     from rhodecode.lib.vcs.exceptions import RepositoryError
     if not isinstance(repo, BaseRepository):
         raise Exception('You must pass an Repository '
                         'object as first argument got %s', type(repo))
 
     try:
         commit = repo.get_commit(
             commit_id=commit_id, commit_idx=commit_idx, pre_load=pre_load,
             maybe_unreachable=maybe_unreachable, reference_obj=reference_obj)
     except (RepositoryError, LookupError):
         commit = EmptyCommit()
     return commit
 
 
 def datetime_to_time(dt):
     if dt:
         return time.mktime(dt.timetuple())
 
 
 def time_to_datetime(tm):
     if tm:
         if isinstance(tm, str):
             try:
                 tm = float(tm)
             except ValueError:
                 return
         return datetime.datetime.fromtimestamp(tm)
 
 
 def time_to_utcdatetime(tm):
     if tm:
         if isinstance(tm, str):
             try:
                 tm = float(tm)
             except ValueError:
                 return
         return datetime.datetime.utcfromtimestamp(tm)
 
 
 MENTIONS_REGEX = re.compile(
     # ^@ or @ without any special chars in front
     r'(?:^@|[^a-zA-Z0-9\-\_\.]@)'
     # main body starts with letter, then can be . - _
     r'([a-zA-Z0-9]{1}[a-zA-Z0-9\-\_\.]+)',
     re.VERBOSE | re.MULTILINE)
 
 
 def extract_mentioned_users(s):
     """
     Returns unique usernames from given string s that have @mention
 
     :param s: string to get mentions
     """
     usrs = set()
     for username in MENTIONS_REGEX.findall(s):
         usrs.add(username)
 
     return sorted(list(usrs), key=lambda k: k.lower())
 
 
-class AttributeDictBase(dict):
-    def __getstate__(self):
-        odict = self.__dict__  # get attribute dictionary
-        return odict
-
-    def __setstate__(self, dict):
-        self.__dict__ = dict
-
-    __setattr__ = dict.__setitem__
-    __delattr__ = dict.__delitem__
-
-
-class StrictAttributeDict(AttributeDictBase):
-    """
-    Strict Version of Attribute dict which raises an Attribute error when
-    requested attribute is not set
-    """
-    def __getattr__(self, attr):
-        try:
-            return self[attr]
-        except KeyError:
-            raise AttributeError('%s object has no attribute %s' % (
-                self.__class__, attr))
-
-
-class AttributeDict(AttributeDictBase):
-    def __getattr__(self, attr):
-        return self.get(attr, None)
-
-
 def fix_PATH(os_=None):
     """
     Get current active python path, and append it to PATH variable to fix
     issues of subprocess calls and different python versions
     """
     if os_ is None:
         import os
     else:
         os = os_
 
     cur_path = os.path.split(sys.executable)[0]
+    os_path = os.environ['PATH']
     if not os.environ['PATH'].startswith(cur_path):
-        os.environ['PATH'] = '%s:%s' % (cur_path, os.environ['PATH'])
+        os.environ['PATH'] = f'{cur_path}:{os_path}'
 
 
 def obfuscate_url_pw(engine):
     _url = engine or ''
     try:
         _url = sqlalchemy.engine.url.make_url(engine)
-        if _url.password:
-            _url.password = 'XXXXX'
     except Exception:
         pass
-    return str(_url)
+    return repr(_url)
 
 
 def get_server_url(environ):
     req = webob.Request(environ)
     return req.host_url + req.script_name
 
 
 def unique_id(hexlen=32):
     alphabet = "23456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghjklmnpqrstuvwxyz"
     return suuid(truncate_to=hexlen, alphabet=alphabet)
 
 
 def suuid(url=None, truncate_to=22, alphabet=None):
     """
     Generate and return a short URL safe UUID.
 
     If the url parameter is provided, set the namespace to the provided
     URL and generate a UUID.
 
     :param url to get the uuid for
     :truncate_to: truncate the basic 22 UUID to shorter version
 
     The IDs won't be universally unique any longer, but the probability of
     a collision will still be very low.
     """
     # Define our alphabet.
     _ALPHABET = alphabet or "23456789ABCDEFGHJKLMNPQRSTUVWXYZ"
 
     # If no URL is given, generate a random UUID.
     if url is None:
         unique_id = uuid.uuid4().int
     else:
         unique_id = uuid.uuid3(uuid.NAMESPACE_URL, url).int
 
     alphabet_length = len(_ALPHABET)
     output = []
     while unique_id > 0:
         digit = unique_id % alphabet_length
         output.append(_ALPHABET[digit])
         unique_id = int(unique_id / alphabet_length)
     return "".join(output)[:truncate_to]
 
 
 def get_current_rhodecode_user(request=None):
     """
     Gets rhodecode user from request
     """
+    import pyramid.threadlocal
     pyramid_request = request or pyramid.threadlocal.get_current_request()
 
     # web case
     if pyramid_request and hasattr(pyramid_request, 'user'):
         return pyramid_request.user
 
     # api case
     if pyramid_request and hasattr(pyramid_request, 'rpc_user'):
         return pyramid_request.rpc_user
 
     return None
 
 
 def action_logger_generic(action, namespace=''):
     """
     A generic logger for actions useful to the system overview, tries to find
     an acting user for the context of the call otherwise reports unknown user
 
     :param action: logging message eg 'comment 5 deleted'
     :param type: string
 
     :param namespace: namespace of the logging message eg. 'repo.comments'
     :param type: string
 
     """
 
     logger_name = 'rhodecode.actions'
 
     if namespace:
         logger_name += '.' + namespace
 
     log = logging.getLogger(logger_name)
 
     # get a user if we can
     user = get_current_rhodecode_user()
 
     logfunc = log.info
 
     if not user:
         user = '<unknown user>'
         logfunc = log.warning
 
     logfunc('Logging action by {}: {}'.format(user, action))
 
 
 def escape_split(text, sep=',', maxsplit=-1):
     r"""
     Allows for escaping of the separator: e.g. arg='foo\, bar'
 
     It should be noted that the way bash et. al. do command line parsing, those
     single quotes are required.
     """
     escaped_sep = r'\%s' % sep
 
     if escaped_sep not in text:
         return text.split(sep, maxsplit)
 
     before, _mid, after = text.partition(escaped_sep)
     startlist = before.split(sep, maxsplit)  # a regular split is fine here
     unfinished = startlist[-1]
     startlist = startlist[:-1]
 
     # recurse because there may be more escaped separators
     endlist = escape_split(after, sep, maxsplit)
 
     # finish building the escaped value. we use endlist[0] becaue the first
     # part of the string sent in recursion is the rest of the escaped value.
     unfinished += sep + endlist[0]
 
     return startlist + [unfinished] + endlist[1:]  # put together all the parts
 
 
 class OptionalAttr(object):
     """
     Special Optional Option that defines other attribute. Example::
 
         def test(apiuser, userid=Optional(OAttr('apiuser')):
             user = Optional.extract(userid)
             # calls
 
     """
 
     def __init__(self, attr_name):
         self.attr_name = attr_name
 
     def __repr__(self):
         return '<OptionalAttr:%s>' % self.attr_name
 
     def __call__(self):
         return self
 
 
 # alias
 OAttr = OptionalAttr
 
 
 class Optional(object):
     """
     Defines an optional parameter::
 
         param = param.getval() if isinstance(param, Optional) else param
         param = param() if isinstance(param, Optional) else param
 
     is equivalent of::
 
         param = Optional.extract(param)
 
     """
 
     def __init__(self, type_):
         self.type_ = type_
 
     def __repr__(self):
         return '<Optional:%s>' % self.type_.__repr__()
 
     def __call__(self):
         return self.getval()
 
     def getval(self):
         """
         returns value from this Optional instance
         """
         if isinstance(self.type_, OAttr):
             # use params name
             return self.type_.attr_name
         return self.type_
 
     @classmethod
     def extract(cls, val):
         """
         Extracts value from Optional() instance
 
         :param val:
         :return: original value if it's not Optional instance else
             value of instance
         """
         if isinstance(val, cls):
             return val.getval()
         return val
 
 
 def glob2re(pat):
-    """
-    Translate a shell PATTERN to a regular expression.
-
-    There is no way to quote meta-characters.
-    """
-
-    i, n = 0, len(pat)
-    res = ''
-    while i < n:
-        c = pat[i]
-        i = i+1
-        if c == '*':
-            #res = res + '.*'
-            res = res + '[^/]*'
-        elif c == '?':
-            #res = res + '.'
-            res = res + '[^/]'
-        elif c == '[':
-            j = i
-            if j < n and pat[j] == '!':
-                j = j+1
-            if j < n and pat[j] == ']':
-                j = j+1
-            while j < n and pat[j] != ']':
-                j = j+1
-            if j >= n:
-                res = res + '\\['
-            else:
-                stuff = pat[i:j].replace('\\','\\\\')
-                i = j+1
-                if stuff[0] == '!':
-                    stuff = '^' + stuff[1:]
-                elif stuff[0] == '^':
-                    stuff = '\\' + stuff
-                res = '%s[%s]' % (res, stuff)
-        else:
-            res = res + re.escape(c)
-    return res + '\Z(?ms)'
+    import fnmatch
+    return fnmatch.translate(pat)
 
 
 def parse_byte_string(size_str):
     match = re.match(r'(\d+)(MB|KB)', size_str, re.IGNORECASE)
     if not match:
-        raise ValueError('Given size:%s is invalid, please make sure '
-                         'to use format of <num>(MB|KB)' % size_str)
+        raise ValueError(f'Given size:{size_str} is invalid, please make sure '
+                         f'to use format of <num>(MB|KB)')
 
     _parts = match.groups()
     num, type_ = _parts
     return int(num) * {'mb': 1024*1024, 'kb': 1024}[type_.lower()]
 
 
 class CachedProperty(object):
     """
     Lazy Attributes. With option to invalidate the cache by running a method
 
     >>> class Foo(object):
     ...
     ...    @CachedProperty
     ...    def heavy_func(self):
     ...        return 'super-calculation'
     ...
     ... foo = Foo()
     ... foo.heavy_func() # first computation
     ... foo.heavy_func() # fetch from cache
     ... foo._invalidate_prop_cache('heavy_func')
 
     # at this point calling foo.heavy_func() will be re-computed
     """
 
     def __init__(self, func, func_name=None):
 
         if func_name is None:
             func_name = func.__name__
         self.data = (func, func_name)
-        update_wrapper(self, func)
+        functools.update_wrapper(self, func)
 
     def __get__(self, inst, class_):
         if inst is None:
             return self
 
         func, func_name = self.data
         value = func(inst)
         inst.__dict__[func_name] = value
         if '_invalidate_prop_cache' not in inst.__dict__:
-            inst.__dict__['_invalidate_prop_cache'] = partial(
+            inst.__dict__['_invalidate_prop_cache'] = functools.partial(
                 self._invalidate_prop_cache, inst)
         return value
 
     def _invalidate_prop_cache(self, inst, name):
         inst.__dict__.pop(name, None)
 
 
 def retry(func=None, exception=Exception, n_tries=5, delay=5, backoff=1, logger=True):
     """
     Retry decorator with exponential backoff.
 
     Parameters
     ----------
     func : typing.Callable, optional
         Callable on which the decorator is applied, by default None
     exception : Exception or tuple of Exceptions, optional
         Exception(s) that invoke retry, by default Exception
     n_tries : int, optional
         Number of tries before giving up, by default 5
     delay : int, optional
         Initial delay between retries in seconds, by default 5
     backoff : int, optional
         Backoff multiplier e.g. value of 2 will double the delay, by default 1
     logger : bool, optional
         Option to log or print, by default False
 
     Returns
     -------
     typing.Callable
         Decorated callable that calls itself when exception(s) occur.
 
     Examples
     --------
     >>> import random
     >>> @retry(exception=Exception, n_tries=3)
     ... def test_random(text):
     ...    x = random.random()
     ...    if x < 0.5:
     ...        raise Exception("Fail")
     ...    else:
     ...        print("Success: ", text)
     >>> test_random("It works!")
     """
 
     if func is None:
-        return partial(
+        return functools.partial(
             retry,
             exception=exception,
             n_tries=n_tries,
             delay=delay,
             backoff=backoff,
             logger=logger,
         )
 
-    @wraps(func)
+    @functools.wraps(func)
     def wrapper(*args, **kwargs):
         _n_tries, n_delay = n_tries, delay
         log = logging.getLogger('rhodecode.retry')
 
         while _n_tries > 1:
             try:
                 return func(*args, **kwargs)
             except exception as e:
                 e_details = repr(e)
                 msg = "Exception on calling func {func}: {e}, " \
                       "Retrying in {n_delay} seconds..."\
                     .format(func=func, e=e_details, n_delay=n_delay)
                 if logger:
                     log.warning(msg)
                 else:
                     print(msg)
                 time.sleep(n_delay)
                 _n_tries -= 1
                 n_delay *= backoff
 
         return func(*args, **kwargs)
 
     return wrapper
 
 
 def user_agent_normalizer(user_agent_raw, safe=True):
     log = logging.getLogger('rhodecode.user_agent_normalizer')
     ua = (user_agent_raw or '').strip().lower()
     ua = ua.replace('"', '')
 
     try:
         if 'mercurial/proto-1.0' in ua:
             ua = ua.replace('mercurial/proto-1.0', '')
             ua = ua.replace('(', '').replace(')', '').strip()
             ua = ua.replace('mercurial ', 'mercurial/')
         elif ua.startswith('git'):
             parts = ua.split(' ')
             if parts:
                 ua = parts[0]
-            ua = re.sub('\.windows\.\d', '', ua).strip()
+            ua = re.sub(r'\.windows\.\d', '', ua).strip()
 
         return ua
     except Exception:
         log.exception('Failed to parse scm user-agent')
         if not safe:
             raise
 
     return ua
 
 
 def get_available_port(min_port=40000, max_port=55555, use_range=False):
     hostname = ''
     for _ in range(min_port, max_port):
         pick_port = 0
         if use_range:
             pick_port = random.randint(min_port, max_port)
 
         with closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
             try:
                 s.bind((hostname, pick_port))
                 s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
                 return s.getsockname()[1]
             except OSError:
                 continue
             except socket.error as e:
                 if e.args[0] in [errno.EADDRINUSE, errno.ECONNREFUSED]:
                     continue
                 raise