Key terms: how to disable automatic linking for unrelated renderings?

Question

Let's assume the translators have chosen 2 context-sensitive renderings for προσκαλέω: which we might gloss as "call.to.come" and "send.someone.to.bring".
Root-based matching is working well, so entering both renderings into the key terms renderings tool gets lots of nice ticks, and the translators are happy.
As call.to.come  is a slightly rare word which not everyone understands,  after testing it's decided to add a glossary entry for it.

At present, (at least, in 9.3.103.16), if I instruct paratext to automatically link keyterm entries, then we end up with weird entries like:

and they \w sent.someone.to.bring|call.to.come\w* the apostles back in. 

That link shouldn't have been inserted. At that point I have 2 options:
1. Hit undo and do all the linking manually.
2. I need to search-and-replace for things with a pipe/bar (|) and/or manually delete the wrong entries.
I'd love to be proven wrong!

Also, if there's an actual call.to.come later on in the section, then it gets skipped because of the false positive, resulting in readers wondering what that word means, and/or why there's no 'go hunt in the glossary' marking.

Because we're using stem-based matching, Paratext has the information it ought to need to get this right - e.g. it could restrict the matches it links to only ones that have the same stem as the glossary entry - it just doesn't seem to be able to do that at the moment.... a
Please tell me I'm wrong, and there's a control lurking somewhere?
 

Answer 1

When Paratext links terms it is looking at the Biblical Terms tool and considering what words have been linked to terms in the Glossary. From your example there is a glossary entry: call.to.come. In the text you have sent.someone.to.bring

Based on the output you need to review the linkages in the Biblical Terms tool. If you in fact have two different renderings linked to the same glossary entry there is no way for Paratext to distinguish that one should not be used.

As to whether is is better to fix this one way or the other, for me it depends on how many issues need to be fixed. If you only have to look for a few wrong issues then that is probably the better way to go. Please note that if you later go back and re-link then you may need to fix these again.

Comments

I'm not sure what you mean by "two distinct renderings linked to the same glossary entry". I certainly have 2 distinct stems in 'key terms renderings' for προσκαλέω, and some key terms where it's more like 4 or five, e.g. where the greek uses a nominal form and we need to use a full verb or expression.

Is there some way I can tell paratext which renderings link to the glossary and which don't, or some way (without rewriting xml files) that I can split a key term into multiple senses?

Except for where the noun/verb has an irregular form, as this is a semi-aggluntinative language and as we're using "Match based on stems", paratext could distinguish between which rendering matches the same stem as the glossary entry, and which do not match.
I realise that for irregular stems or for people who are not using stem-based matching, that does not help, so it would be nice to have a way of telling paratext to use some entries in the renderings tool for checking occurrences, but not for linking.

I also (of course) have glossary entries that are not tied to a key term, and it would be nice to have a way of linking those automatically or semi-automatically too.

---

DG - no if you have multiple renderings for the same biblical term there is no way to separate them out. What you can do is to create separate entries in the project biblical terms list so that each rendering is separated. There are several videos on using the project biblical terms list on Paratext.org: https://paratext.org/videos/en/paratext-features-training-videos/biblical-terms-videos/