I am using the API to upload source texts to Crowdin. Many of these sourcetexts already have been translated before, either a perfect match or a 100% match. Currently these are still checked by a translator who just pick the 100% match option. This is tedious work.
Is it possible to automatically check if a 100% match is available in the translation memory, add this as translation and approve the translation? All of this with an API call?
It would save us a lot of work, but I haven’ been able to find it in the documentation. I am using Crowdin API v2.
@Dima Amazing reply. We are working with CrowdIn for 2 years now and I did not know this was a possibility. This is going to save everyone so much time, thanks!
(To others, the ‘MT Pre-translate’ in Dima’s example isn’t even necessary, Translation Memory alone works too)
About MT Pre-translate - it depends on your preference, you can have it or you can not.
Machine translation picks up all incoming strings, while Translation memory only those that have similar records in TM itself, so it’s a different approach.
Some users prefer proofreading suggestions pre-translate by MT, and some like to do translation from scratch, without MT assistance. Others have trained AI models and review project from time to time, not on a regular basis.
