In the job I'm doing right now it sometimes happens that new entries (completely new for that glossary) get listed twice. They are exactly the same in everything. The mistery is that this doesn't happen always.
Michael Beijer: The absolute best is Ron's Editor
Isn't this an excellent piece of software! I am just trying it out right now to fine tune my glossaries, and I think I'll buy the license.
I'd like to thank you for suggesting Ron's Editor. I'm happy to report it installs and works well on GNU/Linux with Wine, after installing the .Net Framework 4.5.2. I'm testing it right now, it seems excellent. I've added it to the new Terminology management section of the TranslateOnLinux wiki.
I'm only making a moderate use of glossaries myself, so far I've used LibreOffice Calc for editing tab dels after renaming them to .csv (just tried with .tsv, also works fine). It is quite good actually, and also features a step to remove exact duplicates.
Wow, cool that you can get it to work under Linux. Ron's CSV editor is one of those amazing Windows tools that makes the prospect of switching to an OS other than Windows very unlikely for me. It makes working with tab-delimited files so much easier.
You should also check out it’s amazing Filter tool. You can e.g. create a quick display filter to show only all lines with certain field, such as Source, Client, Subject. You can then edit only this selection, and then just delete the filter. Very useful.
Or Layout templates: tweak the way the columns are displayed until it’s just right, and then save it as a Display template. You can create a template specifically for your CafeTran glossary so it opens the way you like it each time.
Michael: As you will see it is much easier to work in a CSV editor
I've just installed Ron's editor and yes, it's a much better way to work with tabbed TXT files. However, at first sight I noticed no corruption signs, all columns are neatly displayed.
By the way, I don't know if this might be related with my issue somehow, but Ron's Editor reports that the encoding is Utf-8 without BOM: shouldn't be simply Utf-8 instead?
Igor: Can you add the screenshot of the glossary pane right after the addition of a new entry?
Please find attached a screenshot showing the two instances of "spese di elettricità, acqua e gas". Contrary to other times, this time there was only one instance in the glossary file although there were two in the pane.
I also tried to make a screen recording, but the online software I used for some reason didn't allow me to store the files. But it was interesting, because I could also record the double entry on the bottom of the glossary file.
So, CTE behaviour is not always exactly the same, but if it's only me I think I should try to live with this issue.
What language code did you specify for Japanese? I'm asking this, because in contrast to the two other entries, the country 'Japan' isn't listed:
TDB: Still sticking to Déjà Vu terminology? ;)
There is only one flavour of Japanese. In any case this is the setting that CTE allows.
TDB: Still sticking to Déjà Vu terminology? ;)
Déjà Vu was my first and only love...
I can't remember what I saw when I first created an Italian ice - Japanese sushi - British lager glossary, but does 'allow' mean that CafeTran Espresso 2018 uses 'ja' itself?
Ah, Proz.com, part of my external memory:
Heat didn't attack my memory that much.
Could you try wha happens when you add the country code 'ja-'?
alwayslockyourbike: ...but does 'allow' mean that CafeTran Espresso 2018 uses 'ja' itself?
I didn't change anything, just selected the Japanese language for the glossary. But, in the glossary header I see "ja-JP", so it should be OK.
Has your question anything to do with the duplicate entry issue that I am experiencing with CTE? I ask because it occurs only occasionally.
Thank you for your help!
>Has your question anything to do with the duplicate entry issue that I am experiencing with CTE? I ask because it occurs only occasionally.
So I now pay special care to these headers. That's why I'm 'warning' you.
I also sometimes encounter double entries in glossaries.
Double, triple or even more similar entries (with long segments) appear on the glossary tab only, while on the Matchboard all is ok.
I've been having the glossary double-entry issue since starting out with CTE more than one year ago. It happens "randomly" right when creating new entries in my tri-lingual glossary, and when I open it to delete the redundant entry I often find that it's not identical to the first one.
I think Nicolas's identification of the duplicate entry mechanism in
CTE is basically right, although CTE is somewhat inconsistent. For instance,
during the job I was doing this morning I entered a bunch of new terms
in my tri-lingual glossary (providing the translation of the inactive
language too) and they were created twice in the glossary, except one
term. It's really difficult to see a pattern here.
So, it seems that CTE's behavior is quite erratic
in this respect. It's not a big deal, for it's easy to open the
glossary in an editor and delete the duplicate entries. Yet, this
duplication issue is irritating and should not be happening.
I also have 3 languages on the Definitions page in Preferences, but I use only 2 in each glossary. Might be somehow connected?
I'm not sure what I need to do for the issue to show itself. It happens even in a glossary I never update.
If your glossary contains duplicate entries, CafeTran will display them twice in the glossary pane. The Prefix matching option might cause that as well - two matches with the same prefix might fetch the same entry. The Matchboard can sort the duplicate entries out. A future update might deal with the display of duplicates in the glossary pane too.