This is an index to template dump files from the English Wiktionary. Many of the more common internal linking templates such as {{l}} and {{m}} and other templates that have language code parameters are included.
They contain template instances from main, Reconstruction, and Appendix namespaces. They were generated from the pages-articles.xml.bz2 file in the dump. The folders are named by the date of the dump file: 20191201 was generated from enwiktionary-20191201-pages-articles.xml.bz2 using a file of template redirects compiled from enwiktionary-20191201-page.sql.gz and enwiktionary-20191201-redirect.sql.gz. These files come out sometime after the first and twentieth day of the month, and the template dump files and updated list of dumps on this page are generated by manually running programs on Toolforge. Typically the dump files are available a day or two after the date of the dump, and the template dumps are generated soon after that if Erutuon remembers to run the programs.
The format is a series of CBOR objects, each of which is equivalent to the JSON {"title":"woordenboek","templates":[{"name":"af","parameters":{"1":"nl","2":"woord","3":"-en-","4":"boek","t1":"word","t3":"book"},"text":"{{af|nl|woord|t1=word|-en-|boek|t3=book}}"}]}. It is faster to decode than JSON because strings are encoded without escape sequences and numbers are encoded binarily, but it may be harder to find libraries that parse sequences of CBOR objects as opposed to single CBOR objects.