Hello @gchevillot!
Apologies for the delay in reply. We got a reply from our devs and they suggest you make the next changes:
in the find field - \"(no)\"(\s)*:(\s)*\{
and in replace field - "nb" : {
Please give it a try and let us know how it goes!
Well sadly, it didn’t go well. It doesn’t work and I’m under the impression the regex content processor doesn’t work at all for the .xcstrings for some reason. I’ve tried a very simple regex like no to be replaced by nb and even that doesn’t change a thing 
(see capture).
Honestly what bothers me the most is that we shouldn’t have to use regex at all as language mappings should work for .xcstrings. It’s been the official format supported by Apple for a year now !
Hi @gchevillot ,
It’s unfortunate to hear that the expression didn’t work. We’ll double-check with the development team!
Hi @gchevillot just wanted to let you know that we have created a task for our engineers to check everything (int.51558)
Thank you for the update @Ira . To sum up I see 2 issues so far in generated .xcstrings :
"en-AU": {
"stringUnit": {
"state": "needs_review",
"value": ""
}
},
Thanks for your feedback! Added as a comment to int.51558
Hello all ! No news on this topic ?
Hi @gchevillot , as we can see, devs are investigating everything
Hello team !
In addition to the issues already mentioned, I spotted another thing that should be fixed on xcstrings generation. Currently we can set the "extraction state"in the settings of the exporter. It is defaulted to “manual”.
However for XCode to extract automatically strings from the code, the extraction state should be omitted in the JSON of the xcstrings. I tried not setting anything in the field but it resets to “manual”
We could maybe have a checkbox to omit extraction state entirely ?
Thank you in advance !
Hi @gchevillot,
Sorry for the delay in reply!
Discussed it with the tech team and we will add this improvement,
Thank you 
