I made the header sections above all our settings
panel lists more consistent.
Before this change:
* some lists had titles, others didn't
* the placement of the filter box was random
* alerts strangely went between the filter box
and the list
* filter boxes were too large
* CSS was haphazard
* forms were squished against tables
Now all the settings with list have consistent
HTML, CSS, and look-and-feel in the area directly above
their list of items.
With the exception of Custom Profile Fields, all the
lists with headers above them happen to be based on
ListWidget, but the header styling is not coupled
to ListWidget, because we want consistent headers
even if Custom Profile Fields has a non-ListWidget
list (due to its drag&drop features).
Introduce a new class "table-sticky-headers" in the settings
and organisation settings HTML table tags and it is used
to make the table headers fix at the top. This commit also
add the background-color and hover properties to the
settings and organisation settings table to make them look
similar to the recent_topics_table.
We fixed the main issue of this form in CVE-2020-9444, but the audit
done at that time only included links found in rendered_markdown; this
change completes our audit for links with target=_blank anywhere in
the codebase.
Prior to this change, there were reports of 500s in
production due to `export.extra_data` being a
Nonetype. This was reproducible using the s3
backend in development when a row was created in
the `RealmAuditLog` table, but the export failed in
the `DeferredWorker`. This left an entry lying
about that was never updated with an `extra_data`
field.
To fix this, we catch any exceptions in the
`DeferredWorker`, and then update `extra_data` to
encode the failure. We also fix the fact that we
never updated the export UI table with pending exports.
These changes also negated the use for the somewhat
hacky `clear_success_banner` logic.
We already know which list widget a `<th>`
tag is associated with when we set up the
event handler, so it's silly to read data
from the DOM to find that widget again
when the handler runs.
This commit eliminates a whole class of possible
errors and busy work.
02cfb47 removed a couple HTML tags that were
being used to sort the table. We fix this,
but disable filtering exports by marking the
input type as `hidden`. We use this approach as
it seems `list_render` doesn't like an
undefined `opts.filter.element`, which is
what happens if we simply remove the `filter`
key.