Unsure about where translation files come from

I’m trying to understand the fbt demo app a little better. I think I understand what most scripts are doing, but I don’t understand where any of the ./translations/*.json files come from. They seem to be auto generated due to there being base64 strings as keys.

Are these json files auto generated? Or compiled by hand?

(to be clear - I’m not asking about the actual translated text, I’m asking about how these files get generated for me to fill in with my own translations for phrases 😊)

[jrwats]

HI @mikedloss,

This is my fault for not yet adding an overview section in our docs explaining the high-level architecture of FBT. I’ll message our engineers internally about adding some overview documentation.

These particular files were manually translated by our vendors and I did a query to fetch them from our internal DB. I then generated the JSON from the Object in the expected format which is described here:

https://facebookincubator.github.io/fbt/docs/translating

To get to the root of your question, those translation JSON files are not generated for “free”.

The translation “cloud” and “Storage” in the above diagram are NOT provided by the FBT framework. And that magical dotted arrow going from the “storage” to the “translation dictionaries” are basically what calling the “translate” script on those JSON translations input represents.

CC @kayhadrin @w1106c

[koistya]

I’m building a crowdsourced translation tool / manager that would allow you to have a similar workflow demonstrated above by @jrwats. You can check it out here -> https://helptranslate.com While it’s in early preview, I’m willing to help you integrate it with FBT and set all the things up.