Translation: keywords in Display Name Editor

Is this the appropriate place to talk about translation? Should there be a dedicated area?

Context: starting up Gramps after a couple of months to find an Irish translation (my OS language), but discovering that it’s not so great, getting sucked into trying to fix some of it, then collecting a lot of questions & problems along the way.

But is this the place to ask? I’ve fixed up the suggested command line examples since some of the parameters were also translated, making it useless. Now I’m struggling with the Name Format Editor keywords in Preferences. Should all the keywords be translated? It seems to be what other languages have done. Is there significance to the lack of spaces in e.g. Primary[pre]? I suppose this is just help text and the actual keywords are defined elsewhere, so we should search out each one? Is there any length limitation? Any reason I couldn’t translate it as PrimarySurname[prefix]? I honestly think that would be more intelligible than inventing abbreviations.

Currently Primary is translated as Bunscoil “primary school; lit. base school”, which is not exactly helpful.

I really, really, really do not have time to devote to this, though I would like a good localisation.

2 Likes

Yes this development section using the localization & translation tags is the right place.

My question is not limited to the Irish translation at all. People often do these translations with very limited time and don’t always give a lot of thought to context (which can be hard to assess anyway). The Irish translation is not the only one to translate command line parameters and many teams have assumed an ambiguous use of the word string is part of a command, rather than programmers just using unnecessary jargon.

For a consistent, quality translation there needs to be a very obvious place to ask questions and collect answers that will be useful to all translations.

1 Like

Agreed. New Translation section created.

2 Likes

After years of being at a very low level of translation, someone provided an almost complete Irish translation. It caused a few fatal errors which I fixed mainly by removing the offending translated strings. In the process I noticed that some translations were obviously bad (country names were duplicated where they didn’t make sense) and I removed some of them. Suspecting malicious activity I checked some random longish strings and they looked good.

Possibly an AI was used for some of the translations or perhaps they were rushed with some copy and paste. I decided to make the translation live anyway and wait for Irish speakers to give feedback.

Thanks for your contributions. The better the translation gets the more it will encourage other people to use and improve it.

I’ll have a look at the code and give you some advice on translating the name formats.

I wonder if some of it might be to do with Weblate too. I’ve seen it warn a couple of times about the translation being updated (presumably the source) ‘but your work was saved’ (but it’s not) and then either the filter queue ends abruptly or it ends up on a completely different set of strings.

This might explain translations ending up attached to unrelated strings, but not the duplication, e.g. <b>Ainm</b> “name” appeared in lots of places as does Dearbhú “confirmation” (3 times on the General Preferences tab).

In fact I now see <b>Ainm</b> next to Automate Place format even in English. Looks like it was written to ~/.config/gramps/place_formats.xml. OK, after renaming place_formats.xml it gets rewritten, but using whatever locale Gramps is running it, so unless I delete it, it will stay as Full.

The translation ranges from the reasonable but awkward to the completely unintelligible, e.g. help tooltip for Automate Place format: “It enables/empowers generation of an automatic place title harvesting”.

The following table lists the keywords which represent parts of names used by the name displayer. I suggest that all the keywords are translated.

Code Keyword Description
t title title
f given given (first names)
l surname full surname (lastname)
c call callname
x common nick name, call, otherwise first first name (common name)
i initials initials of the first names
m primary primary surname (main)
0m primary[pre] primary surname prefix
1m primary[sur] primary surname surname
2m primary[con] primary surname connector
y patronymic pa/matronymic surname (father/mother) - assumed unique
0y patronymic[pre] pa/matronymic surname prefix
1y patronymic[sur] pa/matronymic surname surname
2y patronymic[con] pa/matronymic surname connector
o notpatronymic surnames without pa/matronymic and primary
r rest non primary surnames
p prefix list of all prefixes
q rawsurnames surnames without prefixes and connectors
s suffix suffix
n nickname nick name
g familynick family nick name

These are defined in the gramps/gen/display/name.py file. They should be consistent with the translation of the description in the gramps/gui/configure.py file.

I can’t see any length restriction, or reason to avoid spaces. In English, the keywords have been kept reasonably short for the convenience of users.

The keywords also appear in both lower and upper case in the gramps/gen/utils/keyword.py file. I haven’t had time to investigate where this is used yet.