Final frontend updates before launch.

[vitorgomateus]

• Fix missing converted item: one item format was changed and unaccounted for. Changed sorting method for item formats in listing to account for unaccounted formats.
• Change display of (:unav) populated fields to unknown or unk.
• Force message to show on an item where by request and when there is no cnet_id.
• Improve and simplify logic for dismissible message.
• General UI improvements.
• Updated translations using Google Sheets auto translation.
• Fixed missing date linking in item metadata.

[johnjung]

This looks pretty good so far. I tested some timings with the following three items- these are some sample items with lots of descendant formats:

item/b2051c02741f (5 descendants) item/b20h8hc2sf44 (6 desscendants) item/b20q4gb3zp83 (27 descendants)

The last one returned in about 1.25 seconds on my dev machine- this seems fast enough to me.

In mlc_ucla_search.py, I would change the sortDictByFormat() function a bit. RIght now, it starts with:

if( type(item) != str):
    item = item[0]

I would be explicit and check what the item is:

if isinstance(item, tuple):
    item = item[0]

Or even better, I'd use some assertions to make the function catch more errors, like this:

assert type(item) in (str, tuple)
if isinstance(item, tuple):
    assert len(item)
    assert isinstance(item[0], str)
    item = item[0]

Otherwise the changes look pretty good. Please let me know what you think- if you like those changes please add them in a new commit and I'll merge everything right away.

[vitorgomateus]

Your code looks better. I committed it along with a typo fix.