Hi Jesús, 

The source text has both languages together, some segments are in Spanish and others in English (example attached).
It is a software translation and the colleague preparing the files has them mixed, for some reason. 
I have tried the first option you included, but when applying the filter, nothing happens. 

I thought the tool would have a simple way of recognising different languages in the ST, and it wouldn't be difficult to hide one. 
The doc is longish. I don't think I want to manually lock all Spanish segments. 

Thanks, 

Marta
 

You can try this regex replacement in Notepad++ (make a copy before you proceed with replacing):

Find what: ("[<>\.\,\;\:\?\!\(\)\{\}\[\]0-9\w+_\s-]+"):\s((?>(?!\1))"([^"]+)")
Replace with: \1: "\3 @@@"

Then import the new file into Trados. 
The translated segments in Spanish would be marked with " @@@" . Then you can filter for @@@ in source, copy source to target and lock all with Ctrl+L.
Then translate the unlocked segments and save target as *.json. Open the target file with Notepad++ again and delete all instances of " @@@" by replacing it with nothing (blank Replace with field).

If there are some characters in your strings that are not present in the find what regex, the replacement won't work for that string and you may need to modify the regex by adding the missing characters.

You can try this regex replacement in Notepad++ (make a copy before you proceed with replacing):

Find what: ("[<>\.\,\;\:\?\!\(\)\{\}\[\]0-9\w+_\s-]+"):\s((?>(?!\1))"([^"]+)")
Replace with: \1: "\3 @@@"

Then import the new file into Trados. 
The translated segments in Spanish would be marked with " @@@" . Then you can filter for @@@ in source, copy source to target and lock all with Ctrl+L.
Then translate the unlocked segments and save target as *.json. Open the target file with Notepad++ again and delete all instances of " @@@" by replacing it with nothing (blank Replace with field).

If there are some characters in your strings that are not present in the find what regex, the replacement won't work for that string and you may need to modify the regex by adding the missing characters.