Avoiding untranslatable text to be added to TM data base?

We have some content that is currently pushed to Transifex, but it shouldn’t be translated as it is part of the code.
For the time being, these strings are manually translated by copying the source text to the translation string’s area.
The problem is that doing this we are introducing English content in the Translation Memories affecting other source strings that need to be translated into other languages properly.
I wonder if there is a way in Transifex to avoid introducing source content as translations in the TMs?

You can always clean up your TM database in different ways as explainedhere. \

It might be really time-consuming, so in this case, if you cannot avoid sending those strings to Transifex, I would recommend tagging them as “Locked”, as explained here, so translators cannot translate them. You can find here information on how to tag strings via API. You can also tag them in the editor, in bulk, filtering by Key.bThose strings will remain untranslated.

If you are using the Client to pull translations, you can use the SourceAsTranslation mode as described here:
sourceastranslation: The files will include the translated strings. The untranslated ones will be filled with the corresponding source strings.

That way, you will get the source as translation in your file, but it won’t affect your Translation Memory.