When multiple languages are required but only some of them are found, the scraper still sets all languages in ZIM Language metadata (or at least it will once https://github.com/openzim/ted/pull/170 will be merged).
The scraper should in fact:
filter out languages for which no videos have been found
order languages by importance (in term of number of videos) inside the ZIM
This is going to be a bit tricky because TED lang codes are different than ISO-639-3 codes, we will have to be careful about that.
Note that once #171 is implemented, the exact list of languages will be dynamic.
When multiple languages are required but only some of them are found, the scraper still sets all languages in ZIM Language metadata (or at least it will once https://github.com/openzim/ted/pull/170 will be merged).
The scraper should in fact:
This is going to be a bit tricky because TED lang codes are different than ISO-639-3 codes, we will have to be careful about that.
Note that once #171 is implemented, the exact list of languages will be dynamic.