Normally strict should be faster, because whenever you try to match something unmatchable the strict version will fail faster and therefore abort earlier.

Whereas slightly faster parsing of the simpler regex is a one-time gain