For reasons unknown, the logs of a web hook were paginated in memory.
This would result in the "Edit" page of a web hook timing out once it
has more than a few thousand log entries.

This commit makes the following changes:

1. We use LIMIT/OFFSET to paginate the data, instead of doing this in
   memory.

2. We limit the logs to the last two days, just like the documentation
   says (instead of retrieving everything).

3. We change the indexes on "web_hook_logs" so the query to get the data
   can perform a backwards index scan, without the need for a Filter.

These changes combined ensure that Projects::HooksController#edit no
longer times out.

* Group filtering now includes also issues/MRs from
subgroups/subprojects
* fixed due_date
* Also DRYed todo controller specs

repository.can_be_merged? can raise error if diff_head_sha or
target_branch are absent, therefore check those explicitly.

The original API that queries by label (`/rest/api/latest/result?label=#{sha1}`)
only works for results from main plans and not branch plans. The
`/rest/api/latest/result/byChangeset/#{sha1}` endpoint gives results from
branch plans but not for the first push to the branch. Subsequent pushes
work.

For more details, see https://jira.atlassian.com/browse/BAM-16121.

Closes #1355

Signed-off-by: NRémy Coutable <remy@rymai.me>

This specific one isn't used on most machines, therefor low risk.

Closes https://gitlab.com/gitlab-org/gitaly/issues/944

There is still the edge case when 'no commits' changes to 'conflict'
would not trigger notification, which we ignore for now.

Calling can_be_merged? can cause exception (e.g. non-UTF8)
Ignore those by rescueing.

Remove unmergeable_reason as now only conflict is notified

Update spec

Closes https://gitlab.com/gitlab-org/gitaly/issues/889

This significantly improves performance when a user pushes many references.

project.path_locks.any? doesn't cache the output and runs `SELECT 1 AS one
FROM "path_locks" WHERE project_id = N` each time. When there are thousands
of refs being pushed, this can time out the unicorn worker.

CE port for https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/6159.

Use data_source_exists? where possible instead of table_exists? in order to be Rails5 compatible

Closes #42342

Before the push git would make a call to
`/:namespace/:project/git-receive-pack`. This would perform an access
check without a ref. So the `Project#branch_allows_maintainer_push?`
would return false.

This adjusts `Project#branch_allows_maintainer_push?` to return true
when passing no branch name if there are merge requests open that
would allow the user to push.

The actual check then happens when a call to
`/api/v4/internal/allowed` is made from a git hook.

The cache state for Wikis that were imported via GitHub or Bitbucket does
not appear to have been flushed after a successful import.

Closes #47546