audit-logs: use stricter limit on how much data the commits key can hold.

- we're limited by a smaller column size
- helps avoid mysql problems
- found after sqlalchemy update

rhodecode/lib/hooks_base.py

@@ -1,425 +1,425 b''
 # -*- coding: utf-8 -*-
 
 # Copyright (C) 2013-2017 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/
 
 
 """
 Set of hooks run by RhodeCode Enterprise
 """
 
 import os
 import collections
 import logging
 
 import rhodecode
 from rhodecode import events
 from rhodecode.lib import helpers as h
 from rhodecode.lib import audit_logger
 from rhodecode.lib.utils2 import safe_str
 from rhodecode.lib.exceptions import HTTPLockedRC, UserCreationError
 from rhodecode.model.db import Repository, User
 
 log = logging.getLogger(__name__)
 
 
 HookResponse = collections.namedtuple('HookResponse', ('status', 'output'))
 
 
 def is_shadow_repo(extras):
     """
     Returns ``True`` if this is an action executed against a shadow repository.
     """
     return extras['is_shadow_repo']
 
 
 def _get_scm_size(alias, root_path):
 
     if not alias.startswith('.'):
         alias += '.'
 
     size_scm, size_root = 0, 0
     for path, unused_dirs, files in os.walk(safe_str(root_path)):
         if path.find(alias) != -1:
             for f in files:
                 try:
                     size_scm += os.path.getsize(os.path.join(path, f))
                 except OSError:
                     pass
         else:
             for f in files:
                 try:
                     size_root += os.path.getsize(os.path.join(path, f))
                 except OSError:
                     pass
 
     size_scm_f = h.format_byte_size_binary(size_scm)
     size_root_f = h.format_byte_size_binary(size_root)
     size_total_f = h.format_byte_size_binary(size_root + size_scm)
 
     return size_scm_f, size_root_f, size_total_f
 
 
 # actual hooks called by Mercurial internally, and GIT by our Python Hooks
 def repo_size(extras):
     """Present size of repository after push."""
     repo = Repository.get_by_repo_name(extras.repository)
     vcs_part = safe_str(u'.%s' % repo.repo_type)
     size_vcs, size_root, size_total = _get_scm_size(vcs_part,
                                                     repo.repo_full_path)
     msg = ('Repository `%s` size summary %s:%s repo:%s total:%s\n'
            % (repo.repo_name, vcs_part, size_vcs, size_root, size_total))
     return HookResponse(0, msg)
 
 
 def pre_push(extras):
     """
     Hook executed before pushing code.
 
     It bans pushing when the repository is locked.
     """
 
     usr = User.get_by_username(extras.username)
     output = ''
     if extras.locked_by[0] and usr.user_id != int(extras.locked_by[0]):
         locked_by = User.get(extras.locked_by[0]).username
         reason = extras.locked_by[2]
         # this exception is interpreted in git/hg middlewares and based
         # on that proper return code is server to client
         _http_ret = HTTPLockedRC(
             _locked_by_explanation(extras.repository, locked_by, reason))
         if str(_http_ret.code).startswith('2'):
             # 2xx Codes don't raise exceptions
             output = _http_ret.title
         else:
             raise _http_ret
 
     # Propagate to external components. This is done after checking the
     # lock, for consistent behavior.
     if not is_shadow_repo(extras):
         pre_push_extension(repo_store_path=Repository.base_path(), **extras)
         events.trigger(events.RepoPrePushEvent(
             repo_name=extras.repository, extras=extras))
 
     return HookResponse(0, output)
 
 
 def pre_pull(extras):
     """
     Hook executed before pulling the code.
 
     It bans pulling when the repository is locked.
     """
 
     output = ''
     if extras.locked_by[0]:
         locked_by = User.get(extras.locked_by[0]).username
         reason = extras.locked_by[2]
         # this exception is interpreted in git/hg middlewares and based
         # on that proper return code is server to client
         _http_ret = HTTPLockedRC(
             _locked_by_explanation(extras.repository, locked_by, reason))
         if str(_http_ret.code).startswith('2'):
             # 2xx Codes don't raise exceptions
             output = _http_ret.title
         else:
             raise _http_ret
 
     # Propagate to external components. This is done after checking the
     # lock, for consistent behavior.
     if not is_shadow_repo(extras):
         pre_pull_extension(**extras)
         events.trigger(events.RepoPrePullEvent(
             repo_name=extras.repository, extras=extras))
 
     return HookResponse(0, output)
 
 
 def post_pull(extras):
     """Hook executed after client pulls the code."""
 
     audit_user = audit_logger.UserWrap(
         username=extras.username,
         ip_addr=extras.ip)
     repo = audit_logger.RepoWrap(repo_name=extras.repository)
     audit_logger.store(
         'user.pull', action_data={
             'user_agent': extras.user_agent},
         user=audit_user, repo=repo, commit=True)
 
     # Propagate to external components.
     if not is_shadow_repo(extras):
         post_pull_extension(**extras)
         events.trigger(events.RepoPullEvent(
             repo_name=extras.repository, extras=extras))
 
     output = ''
     # make lock is a tri state False, True, None. We only make lock on True
     if extras.make_lock is True and not is_shadow_repo(extras):
         user = User.get_by_username(extras.username)
         Repository.lock(Repository.get_by_repo_name(extras.repository),
                         user.user_id,
                         lock_reason=Repository.LOCK_PULL)
         msg = 'Made lock on repo `%s`' % (extras.repository,)
         output += msg
 
     if extras.locked_by[0]:
         locked_by = User.get(extras.locked_by[0]).username
         reason = extras.locked_by[2]
         _http_ret = HTTPLockedRC(
             _locked_by_explanation(extras.repository, locked_by, reason))
         if str(_http_ret.code).startswith('2'):
             # 2xx Codes don't raise exceptions
             output += _http_ret.title
 
     return HookResponse(0, output)
 
 
 def post_push(extras):
     """Hook executed after user pushes to the repository."""
     commit_ids = extras.commit_ids
 
     # log the push call
     audit_user = audit_logger.UserWrap(
         username=extras.username, ip_addr=extras.ip)
     repo = audit_logger.RepoWrap(repo_name=extras.repository)
     audit_logger.store(
         'user.push', action_data={
             'user_agent': extras.user_agent,
-            'commit_ids': commit_ids[:10000]},
+            'commit_ids': commit_ids[:400]},
         user=audit_user, repo=repo, commit=True)
 
     # Propagate to external components.
     if not is_shadow_repo(extras):
         post_push_extension(
             repo_store_path=Repository.base_path(),
             pushed_revs=commit_ids,
             **extras)
         events.trigger(events.RepoPushEvent(
             repo_name=extras.repository,
             pushed_commit_ids=commit_ids,
             extras=extras))
 
     output = ''
     # make lock is a tri state False, True, None. We only release lock on False
     if extras.make_lock is False and not is_shadow_repo(extras):
         Repository.unlock(Repository.get_by_repo_name(extras.repository))
         msg = 'Released lock on repo `%s`\n' % extras.repository
         output += msg
 
     if extras.locked_by[0]:
         locked_by = User.get(extras.locked_by[0]).username
         reason = extras.locked_by[2]
         _http_ret = HTTPLockedRC(
             _locked_by_explanation(extras.repository, locked_by, reason))
         # TODO: johbo: if not?
         if str(_http_ret.code).startswith('2'):
             # 2xx Codes don't raise exceptions
             output += _http_ret.title
 
     if extras.new_refs:
         tmpl = \
             extras.server_url + '/' + \
             extras.repository + \
             "/pull-request/new?{ref_type}={ref_name}"
         for branch_name in extras.new_refs['branches']:
             output += 'RhodeCode: open pull request link: {}\n'.format(
                 tmpl.format(ref_type='branch', ref_name=branch_name))
 
         for book_name in extras.new_refs['bookmarks']:
             output += 'RhodeCode: open pull request link: {}\n'.format(
                 tmpl.format(ref_type='bookmark', ref_name=book_name))
 
     output += 'RhodeCode: push completed\n'
     return HookResponse(0, output)
 
 
 def _locked_by_explanation(repo_name, user_name, reason):
     message = (
         'Repository `%s` locked by user `%s`. Reason:`%s`'
         % (repo_name, user_name, reason))
     return message
 
 
 def check_allowed_create_user(user_dict, created_by, **kwargs):
     # pre create hooks
     if pre_create_user.is_active():
         allowed, reason = pre_create_user(created_by=created_by, **user_dict)
         if not allowed:
             raise UserCreationError(reason)
 
 
 class ExtensionCallback(object):
     """
     Forwards a given call to rcextensions, sanitizes keyword arguments.
 
     Does check if there is an extension active for that hook. If it is
     there, it will forward all `kwargs_keys` keyword arguments to the
     extension callback.
     """
 
     def __init__(self, hook_name, kwargs_keys):
         self._hook_name = hook_name
         self._kwargs_keys = set(kwargs_keys)
 
     def __call__(self, *args, **kwargs):
         log.debug('Calling extension callback for %s', self._hook_name)
 
         kwargs_to_pass = dict((key, kwargs[key]) for key in self._kwargs_keys)
         # backward compat for removed api_key for old hooks. THis was it works
         # with older rcextensions that require api_key present
         if self._hook_name in ['CREATE_USER_HOOK', 'DELETE_USER_HOOK']:
             kwargs_to_pass['api_key'] = '_DEPRECATED_'
 
         callback = self._get_callback()
         if callback:
             return callback(**kwargs_to_pass)
         else:
             log.debug('extensions callback not found skipping...')
 
     def is_active(self):
         return hasattr(rhodecode.EXTENSIONS, self._hook_name)
 
     def _get_callback(self):
         return getattr(rhodecode.EXTENSIONS, self._hook_name, None)
 
 
 pre_pull_extension = ExtensionCallback(
     hook_name='PRE_PULL_HOOK',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository'))
 
 
 post_pull_extension = ExtensionCallback(
     hook_name='PULL_HOOK',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository'))
 
 
 pre_push_extension = ExtensionCallback(
     hook_name='PRE_PUSH_HOOK',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'repo_store_path', 'commit_ids'))
 
 
 post_push_extension = ExtensionCallback(
     hook_name='PUSH_HOOK',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'repo_store_path', 'pushed_revs'))
 
 
 pre_create_user = ExtensionCallback(
     hook_name='PRE_CREATE_USER_HOOK',
     kwargs_keys=(
         'username', 'password', 'email', 'firstname', 'lastname', 'active',
         'admin', 'created_by'))
 
 
 log_create_pull_request = ExtensionCallback(
     hook_name='CREATE_PULL_REQUEST',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'pull_request_id', 'url', 'title', 'description',
         'status', 'created_on', 'updated_on', 'commit_ids', 'review_status',
         'mergeable', 'source', 'target', 'author', 'reviewers'))
 
 
 log_merge_pull_request = ExtensionCallback(
     hook_name='MERGE_PULL_REQUEST',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'pull_request_id', 'url', 'title', 'description',
         'status', 'created_on', 'updated_on', 'commit_ids', 'review_status',
         'mergeable', 'source', 'target', 'author', 'reviewers'))
 
 
 log_close_pull_request = ExtensionCallback(
     hook_name='CLOSE_PULL_REQUEST',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'pull_request_id', 'url', 'title', 'description',
         'status', 'created_on', 'updated_on', 'commit_ids', 'review_status',
         'mergeable', 'source', 'target', 'author', 'reviewers'))
 
 
 log_review_pull_request = ExtensionCallback(
     hook_name='REVIEW_PULL_REQUEST',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'pull_request_id', 'url', 'title', 'description',
         'status', 'created_on', 'updated_on', 'commit_ids', 'review_status',
         'mergeable', 'source', 'target', 'author', 'reviewers'))
 
 
 log_update_pull_request = ExtensionCallback(
     hook_name='UPDATE_PULL_REQUEST',
     kwargs_keys=(
         'server_url', 'config', 'scm', 'username', 'ip', 'action',
         'repository', 'pull_request_id', 'url', 'title', 'description',
         'status', 'created_on', 'updated_on', 'commit_ids', 'review_status',
         'mergeable', 'source', 'target', 'author', 'reviewers'))
 
 
 log_create_user = ExtensionCallback(
     hook_name='CREATE_USER_HOOK',
     kwargs_keys=(
         'username', 'full_name_or_username', 'full_contact', 'user_id',
         'name', 'firstname', 'short_contact', 'admin', 'lastname',
         'ip_addresses', 'extern_type', 'extern_name',
         'email', 'api_keys', 'last_login',
         'full_name', 'active', 'password', 'emails',
         'inherit_default_permissions', 'created_by', 'created_on'))
 
 
 log_delete_user = ExtensionCallback(
     hook_name='DELETE_USER_HOOK',
     kwargs_keys=(
         'username', 'full_name_or_username', 'full_contact', 'user_id',
         'name', 'firstname', 'short_contact', 'admin', 'lastname',
         'ip_addresses',
         'email', 'last_login',
         'full_name', 'active', 'password', 'emails',
         'inherit_default_permissions', 'deleted_by'))
 
 
 log_create_repository = ExtensionCallback(
     hook_name='CREATE_REPO_HOOK',
     kwargs_keys=(
         'repo_name', 'repo_type', 'description', 'private', 'created_on',
         'enable_downloads', 'repo_id', 'user_id', 'enable_statistics',
         'clone_uri', 'fork_id', 'group_id', 'created_by'))
 
 
 log_delete_repository = ExtensionCallback(
     hook_name='DELETE_REPO_HOOK',
     kwargs_keys=(
         'repo_name', 'repo_type', 'description', 'private', 'created_on',
         'enable_downloads', 'repo_id', 'user_id', 'enable_statistics',
         'clone_uri', 'fork_id', 'group_id', 'deleted_by', 'deleted_on'))
 
 
 log_create_repository_group = ExtensionCallback(
     hook_name='CREATE_REPO_GROUP_HOOK',
     kwargs_keys=(
         'group_name', 'group_parent_id', 'group_description',
         'group_id', 'user_id', 'created_by', 'created_on',
         'enable_locking'))

rhodecode/tests/lib/test_hooks_base.py

@@ -1,141 +1,141 b''
 # -*- coding: utf-8 -*-
 
 # Copyright (C) 2010-2017 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/
 
 import mock
 import pytest
 from rhodecode.model.db import Session, UserLog
 from rhodecode.lib import hooks_base, utils2
 
 
 def test_post_push_truncates_commits(user_regular, repo_stub):
     extras = {
         'ip': '127.0.0.1',
         'username': user_regular.username,
         'action': 'push_local',
         'repository': repo_stub.repo_name,
         'scm': 'git',
         'config': '',
         'server_url': 'http://example.com',
         'make_lock': None,
         'user_agent': 'some-client',
         'locked_by': [None],
         'commit_ids': ['abcde12345' * 4] * 30000,
         'is_shadow_repo': False,
     }
     extras = utils2.AttributeDict(extras)
 
     hooks_base.post_push(extras)
 
     # Calculate appropriate action string here
-    commit_ids = extras.commit_ids[:10000]
+    commit_ids = extras.commit_ids[:400]
 
     entry = UserLog.query().order_by('-user_log_id').first()
     assert entry.action == 'user.push'
     assert entry.action_data['commit_ids'] == commit_ids
     Session().delete(entry)
     Session().commit()
 
 
 def assert_called_with_mock(callable_, expected_mock_name):
     mock_obj = callable_.call_args[0][0]
     mock_name = mock_obj._mock_new_parent._mock_new_name
     assert mock_name == expected_mock_name
 
 
 @pytest.fixture
 def hook_extras(user_regular, repo_stub):
     extras = utils2.AttributeDict({
         'ip': '127.0.0.1',
         'username': user_regular.username,
         'action': 'push',
         'repository': repo_stub.repo_name,
         'scm': '',
         'config': '',
         'server_url': 'http://example.com',
         'make_lock': None,
         'user_agent': 'some-client',
         'locked_by': [None],
         'commit_ids': [],
         'is_shadow_repo': False,
     })
     return extras
 
 
 @pytest.mark.parametrize('func, extension, event', [
     (hooks_base.pre_push, 'pre_push_extension', 'RepoPrePushEvent'),
     (hooks_base.post_push, 'post_pull_extension', 'RepoPushEvent'),
     (hooks_base.pre_pull, 'pre_pull_extension', 'RepoPrePullEvent'),
     (hooks_base.post_pull, 'post_push_extension', 'RepoPullEvent'),
 ])
 def test_hooks_propagate(func, extension, event, hook_extras):
     """
     Tests that our hook code propagates to rhodecode extensions and triggers
     the appropriate event.
     """
     extension_mock = mock.Mock()
     events_mock = mock.Mock()
     patches = {
         'Repository': mock.Mock(),
         'events': events_mock,
         extension: extension_mock,
     }
 
     # Clear shadow repo flag.
     hook_extras.is_shadow_repo = False
 
     # Execute hook function.
     with mock.patch.multiple(hooks_base, **patches):
         func(hook_extras)
 
     # Assert that extensions are called and event was fired.
     extension_mock.called_once()
     assert_called_with_mock(events_mock.trigger, event)
 
 
 @pytest.mark.parametrize('func, extension, event', [
     (hooks_base.pre_push, 'pre_push_extension', 'RepoPrePushEvent'),
     (hooks_base.post_push, 'post_pull_extension', 'RepoPushEvent'),
     (hooks_base.pre_pull, 'pre_pull_extension', 'RepoPrePullEvent'),
     (hooks_base.post_pull, 'post_push_extension', 'RepoPullEvent'),
 ])
 def test_hooks_propagates_not_on_shadow(func, extension, event, hook_extras):
     """
     If hooks are called by a request to a shadow repo we only want to run our
     internal hooks code but not external ones like rhodecode extensions or
     trigger an event.
     """
     extension_mock = mock.Mock()
     events_mock = mock.Mock()
     patches = {
         'Repository': mock.Mock(),
         'events': events_mock,
         extension: extension_mock,
     }
 
     # Set shadow repo flag.
     hook_extras.is_shadow_repo = True
 
     # Execute hook function.
     with mock.patch.multiple(hooks_base, **patches):
         func(hook_extras)
 
     # Assert that extensions are *not* called and event was *not* fired.
     assert not extension_mock.called
     assert not events_mock.trigger.called