Validation shouldn't fail on missing translations (anymore)

[TomazErjavec]

@matyaskopp, I think we now got all the translations we are likely to, so could you pls. relax the validation in Actions so it doesn't fail for missing translations? Once done, I'd push the V5 samples in the hope that all wil be green.

[matyaskopp]

@TomazErjavec it does not fail on missing taxonomies, it fails on country-specific taxonomies: #934

I guess you haven't triggered the action, so you looked at the old actions' result.

[TomazErjavec]

@matyaskopp, sorry, I don't really understand what you are saying, for sure my fault..
Anyway, I now tried to push to data the new set of samples & got email about failing run - if I see correctly it is because of a too big TEI file (somewhere, doesn't say which).
After that no other validation seems to be run.

[matyaskopp]

I will fix it, it seems validation of specific taxonomies is still buggy

[matyaskopp]

@TomazErjavec IS and TR need to be reprocessed (https://github.com/clarin-eric/ParlaMint/actions/runs/15922243878/job/44911573334) because their ids contain unicode character and my script did not handle it, sorry for that.
Also, taxonomies patching needs to be redone. It is possible to run it over current version (I believe), because it ignores already patched ids:

ParlaMint/Makefile

Line 244 in 78081df

| grep -vP 'xml:id="${CORPUS}-.*"' \

[TomazErjavec]

IS and TR need to be reprocessed (https://github.com/clarin-eric/ParlaMint/actions/runs/15922243878/job/44911573334) because their ids contain unicode character and my script did not handle it

@matyaskopp, what do you mean by "reprocessed"? Made again from scatch? Only the sample remade? Only samples pushed again? I would imagine the last, if the problem is only with your validation script, but in which case, even if I re-make the samples, they will be identical, hm.

taxonomies patching needs to be redone

Yikes, this would be very difficult, as this now happens before the build pipeline, and changing everything would be really time consuming and error prone. I also don't undestand why they should be patched, as they have been already?

[matyaskopp]

Not all IDs have been patched due to my fault:

ParlaMint/Samples/ParlaMint-IS/ParlaMint-IS-taxonomy-parla.constituencies.xml

Lines 41 to 49 in 78081df

	<category xml:id="const.Ísfirdinga">
	<catDesc><term>Ísfirðinga</term></catDesc>
	</category>
	<category xml:id="const.Ísafjardar">
	<catDesc><term>Ísafjarðar</term></catDesc>
	</category>
	<category xml:id="IS-const.konungkjorinn_althingismadur">
	<catDesc><term>konungkjörinn alþingismaður</term></catDesc>
	</category>

But in fact, we can preserve them as they are and fix them in future releases, because they are just IDs.
It happened because of not counting on Unicode characters in IDs 78081df

The corpus(sample) is valid, but it does not follow the recommendation for making IDs (which causes the errors in validation with GitHub actions).

[TomazErjavec]

OK, thanks, I understand now @matyaskopp. If I see correctly, it is only IS and TR that need to be re-done. Will try to do that then, will report when done (IS takes forever).

[TomazErjavec]

OK, IS & TR redone (60 hrs!), and all checks are ok, thanks & closing.

(btw. the TR-specific taxonomy already had "TR" in the filename / IDs, so now we have the somewhat weird "ParlaMint-TR-taxonomy-parla.constituencies-TR.xml" and "TR-constituency-TR.Uşak". Maybe something to consider for 5.1...)