Should we add a migration to replace with en any lang codes that are invalid?

Added a migration, but it still needs some testing. Should also be merged after #2601 due to migration order

@ikreymer Reworked the migration and tested it locally. Now includes crawls as well. Didn't find a library that would give us a good ISO-639-1 list in a convenient format so I hardcoded it - I think should be okay since changes are infrequent and almost never happen for the most commonly used languages. Also includes crawls now :)

For testing, I used the API off of main (without the backend validation added in this branch) to create some invalid data, e.g.:

curl -X POST -H "Authorization: Bearer TOKEN" -H "Content-type: application/json" http://localhost:30870/api/orgs/OID/crawlconfigs/ --data "{\"runNow\": true, \"name\": \"lang test\", \"config\": {\"seeds\": [{\"url\": \"https://webrecorder.net/\"}], \"limit\": 3, \"lang\": \"f\"}}"

curl -X POST -H "Authorization: Bearer TOKEN" -H "Content-type: application/json" http://localhost:30870/api/orgs/OID/defaults/crawling --data "{\"lang\": \"e\"}"

then switched to this branch and ran the migration. I verified in mongo before and after that the invalid language codes were present on the workflow, crawl, and org, and then that they were fixed.