Closes gitaly#929

The default 404 handler would return the Content-Type format based on the
given format extension. This would cause the Rails CSRF protection to flag an
error, since the .js extension gets mapped to text/javascript format.

Closes #40771

Search query is especially slow if a user searches a generic string
which matches many records, in such case search can take tens of
seconds or time out. To speed up the search query, we search only for
first 1000 records, if there is >1000 matching records we just display
"1000+" instead of precise total count supposing that with such amount
the exact count is not so important for the user.

Because for issues even limited search was not fast enough, 2-phase
approach is used for issues: first we use simpler/faster query to get
all public issues, if this exceeds the limit, we just return the limit.
If the amount of matching results is lower than limit, we re-run more
complex search query (which includes also confidential issues).
Re-running the complex query should be fast enough in such case because the
amount of matching issues is lower than limit.

Because exact total_count is now limited, this patch also switches to
to "prev/next" pagination.

Related #40540

Otherwise we'd only preload the ancestors that would fit the
page. That way, when a search result was on the first page, but the
ancestor didn't fit the page anymore. We would not have the preloaded
ancestor when rendering the hierarchy.

When a project is shared with a group, it should not be included as a
descendant on the group dashboard.

When searching we would limit the scope of ancestors to load because
the filter would be applied to the ancestors. Instead, we need to load
_all_ ancestors for matching projects.

Html code now works in commits atom feed `title` and `summary`.

Given no search term, the `search` and `search_with_secondary_emails` methods will yield an empty result set

Without this, it's very hard to locate it.