CafeTran

 Hi!

I use the pipe character quite often in my glossaries because it is quite useful for German, which is an inflecting language.

Now I'm quite happy that glossaries can be used in both directions now. But I noticed a problem yesterday, when I was using a glossary that was originally German-English for an English-to-German project. Now there were quite a few pipe characters on the target language (German) side of the glossary. When such a target term was inserted into a target segment or suggested during auto-completion, the pipe character was also shown in the term so that I had to delete it each time. I was lucky this happened only a couple of times, but with a different text, this could be different.

So my question is: Would it be possible to make Cafetran not show the pipe characters in terms that are being auto-inserted or suggested for auto-completion? As actually, pipe characters are used in glossaries only to mark the end of a stem, those characters needn't be shown during auto-insertion/auto-completion in any case.

Cheers,
Martin

Would it hurt much to copy the glossary, reverse it, and delete all pipe characters at the target side? And wouldn't that be a hell of a lot simpler that "hiding" them? If possible at all?

H.

Come to think of it, depending upon how you set up your glossary, this may not be possible. And I thought I covered ALL disadvantages/stupidities of tab del glossaries...

H.

Actually, reversing a glossary created in CT is possible (using any spreadsheet programme). I've tried, and it was quite easy, and much easier than I'd thought.
But actually, with CT now offering the possibility to use glossaries in both directions, such reversals should no longer be necessary. So it would still be much more convenient to make CT ignore the pipes.

BTW, I reported a similar problem with pipes in the old CafeTranslators group almost three years ago (here's the link). It was about CT not finding search terms in glossaries when the glossary term contained a pipe. Igor was then able to solve that problem very quickly, so I hope the current problem can be solved easily too.

 

Cheers,

Martin

Thanks a lot, Igor!