Open chavlji opened 4 years ago
As can be seen here, it seems random. I haven't touched those translations lately. After running script a few times, some new translations gets corrupted, while some of the old errors are gone:
If anybody else runs into this issue:
There's a bug with turning buffer chunks into UTF8 string too early. Buffers should be concat together as raw byte buffers and converted to string at the end:
I use this code, to download translations:
On random I get invalid characters in my jsons:
What could be the cause? It happens in a lot of languages: de, es, it, sl, zh, ru,...