Previously, an attempt was made to speed up the most naive case of
EzLink conversion, where there is a unique filename across the entire
site. This would allow a simple dictionary lookup instead of a full
Prefix Trie search (which is more expensive).
Previously, it incorrectly trimmed all files from the file cache
when reducing it to unique items, because the dictionary comprehension
had an inverted comparison operator.
Now, unique file names will be directly converted via the fast file
cache.
Previously, an attempt was made to speed up the most naive case of EzLink conversion, where there is a unique filename across the entire site. This would allow a simple dictionary lookup instead of a full Prefix Trie search (which is more expensive).
Previously, it incorrectly trimmed all files from the file cache when reducing it to unique items, because the dictionary comprehension had an inverted comparison operator.
Now, unique file names will be directly converted via the fast file cache.