Hi Helena Grulichova , good question -> short answer: this is not accessible. I have asked the question to the team, to confirm if it makes sense exposing this.  +Andrei Caia might be able to provide more info.

You were able to locate this property in the Trados code?

Hi Helena Grulichova , good question -> short answer: this is not accessible. I have asked the question to the team, to confirm if it makes sense exposing this.  +Andrei Caia might be able to provide more info.

You were able to locate this property in the Trados code?

I found it during debugging — it causes issues with duplicating search results. I was able to work around it by using SearchAndMarkup instead of Search (the plugin doesn't implement ITerminologyProvider, but rather ITerminologyMarkupProvider).

Hello!

I will add here some context on why this value is always true:

RecognizeTerms searches each termbase in turn and stops after finding a match without searching the remaining termbases. For example, suppose there are three termbases in a project TermbaseA, TermbaseB, and TermbaseC. TermbaseB has the term "lead" - "reglette" (metal) whilst TermbaseC has the term "lead" - "verbleien" (to lead). The source document has the segment "The manager lead the team to victory!". RecognizeTerms would only return the term from TermbaseB "lead" - "reglette" (metal). RecognizeTerms with includeHomonyms=true would return the terms from TermbaseB and TermbaseC; "lead"-"reglette" and "lead"-"verbleien". RecognizeTerms is faster when includeHomonyms=false because it does not need an extra lookup but does not return matches from all termbases and does not return any homonyms.

Feel free to add your feedback and any additional information about your scenario so we can better understand the issue and work to improve the feature.

Andre

Helena Grulichova - To avoid duplicated terms, can you try this option, please? Will this fix the issue? If not, can you please describe in more detail what the issue is/what you are trying to achieve?

Hello,

We use two termbases in our project - the first one is our terminology provider plugin and the second one is a real file termbase which is disabled. In our plugin, we read the real termbase, lemmatize it, and override the Search method. We provide results from the real termbase, but it can lemmatize the source/target segment and terms in the real termbase and find the hits. F8 for verification - usage of terms must work, too. F8 fires the same Search method on all segments in both languages and compares results on both sides based on their entry IDs, right? It seems verification works only with the first termbase (our plugin in this case), right? The Search setting does not help - I suppose it works only for search, not for verification. We do not have trouble with standard term recognition but only with verification of term usage (Verification Settings -> Check for possible non-usage of target terms: true).

When one term is found in the source segment, Trados Studio expects one corresponding term in the target segment with the same entry ID - this works fine. However, if it finds two terms with the same ID in the source segment, it expects FOUR terms with the same ID in the translation. If it finds three occurrences in the source, it demands NINE occurrences in the target. I eventually found that if I use SearchAndMarkup instead of Search and provide exact positions of search results, I can avoid this issue.

Best regards,
Helena