Because the post-processing of the rendered README is dependent on the
context (i.e. the current user), do the post-processing when the
README is being displayed.

It does more than solely GitLab Markdown, so rename to MarkupHelper.

Fix https://gitlab.com/gitlab-org/gitlab-ce/issues/29425

 - gl_emoji for falling back to image/css-sprite when the browser
   doesn't support an emoji
 - Markdown rendering (Banzai filter)
 - Autocomplete
 - Award emoji menu
    - Perceived perf
    - Immediate response because we now build client-side
 - Update `digests.json` generation in gemojione rake task to be more
   useful and  include `unicodeVersion`

MR: !9437

See issues

 - #26371
 - #27250
 - #22474

This reverts commit cb10b725c8929b8b4460f89c9d96c773af39ba6b.

This reverts commit 96bef54154e669f9a3e92c3a4bc76c0be3a52e48.

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

Also added relevant specs and refactored to_references in a bunch of places to be more consistent.

This would fix long standing failures running tests on
my development machine, which set `Gitlab.config.gitlab.host`
to another host because it's not my local computer. Now I
finally cannot withstand it and decided to fix them once and
for all.

- Need to pass in a `page_slug` to the filter, so it can rewrite based
  on the current page (all links are rewritten to the level of the app root).
- The earlier `markdown_preview` endpoint was at the level of the wiki.
  We need to know the current page (for rewriting, as above), so this
  commit moves the endpoint to the level of a wiki page.
- Fix all tests

Only usage got removed in !3442

HTML would be stripped in `truncate_if_block` when a comment had multiple lines
and the first wasn't long enough to be truncated. The use of `node.content` would
strip all HTML tags. Using `node.inner_html` retains these tags and puts the
"..." in between paragraph tags.

Closes #2586

Closes #2586

This module is now the sole source of knowledge for *how* we render
Markdown (and GFM).