let schema propose values for some attributes

[arlogriffiths]

@michaelnmmeyer

At the moment, the EGD-based schema does not yet propose permissible values for a number of attributes.

I have noted:

1. <unclear>: no values are proposed for @Reason
2. <seg>: no values are proposed for @type and @subtype

Is it possible for the schema to systematically propose the admitted values for given attributes?

[michaelnmmeyer]

With the toolchain we are using, autocomplete only works for attributes that have a finite set of values. In our schema, we have about thirty attributes that do not, among which the ones you mentioned.

Now, some of them are probably supposed to have restricted values, e.g. unclear/@reason. I will review the list with the EGD to try to spot them. For the others, e.g. milestone/@unit, I will have to find a workaround, which will take some time.

[arlogriffiths]

Thanks Michaël. The specific values I have been finding annoying to have to type out myself are

• "eccentric_ductus" for @Reason on <unclear> (I think there are no other admitted values for @Reason on this element)
• "component", "body" and "aksara" for @type on <seg> (I think there not many other admitted values for @type on @type this element)

[arlogriffiths]

@michaelnmmeyer : I see that the values of @type on <unclear> are now proposed by the schema. Have all other parts of this issue also been dealt with? If so, you could close it.

[danbalogh]

I notice that the schema now suggests "eccentric_ductus" and "illegible" for values of unclear/@reason and when there's a different value entered, the error message says says that these are the permitted values. I think that "illegible" was a permitted value only in Siddham as a hack for the concept of "supplied reason illegible", which EpiDoc does not permit (see EGD §5.1), but we have chosen not to use this in DHARMA and stick to proper EpiDoc, so "illegible" should be removed from the list of permitted values, leaving only "eccentric_ductus".
Or is there some provision in the EGD for <unclear reason="illegible"> that I have forgotten about and cannot find now?

[michaelnmmeyer]

I do not find any mention of unclear/@reason="illegible" in the EGD (nor in the EGC, for that matter). There are still a few texts that use it, maybe because it is incorrectly treated as valid:

>>> texts/DHARMA_DiplEdCandrakiranaPerpusnasL241.xml
>>> texts/DHARMA_DiplEdSiksaGuruPerpusnasL633.xml
>>> texts/DHARMA_DiplEdSiksaKandangKaresianPerpusnasL624.xml
>>> texts/DHARMA_DiplEdSiksaKandangKaresianPerpusnasL630.xml
>>> texts/DHARMA_INSBengalCharters00001.xml
>>> texts/DHARMA_INSBengalCharters00031.xml
>>> texts/DHARMA_INSBengalCharters00032.xml
>>> texts/DHARMA_INSBengalCharters00033.xml
>>> texts/DHARMA_INSBengalCharters00034.xml
>>> texts/DHARMA_INSBengalCharters00052.xml
>>> texts/DHARMA_INSBengalCharters00084.xml
>>> texts/DHARMA_INSTiruvavatuturai00003.xml

Should I drop the attribute from these files?

[danbalogh]

I thought it occurred only in the siddham backup files. We may have endorsed this usage in some very early incarnation of the EGD.
At any rate, please remove it from the schema.

As for dropping the attribute from the files, you can certainly go ahead and do that for the Bengal Charters files (I've checked those, and all are based on earlier Siddham files, so this must be a legacy that was not updated to Dharma standards). It's probably OK to remove it from the Tiruvavatuturai file too, but I'm not sure about the Indonesian manuscript editions.

@arlogriffiths , can you tell us whether it's OK to replace the erroneous <unclear reason="illegible"> to <unclear> in these four editions? Or if uncertain, can you urge Zaki (and/or whoever is in charge of those files) to check the occurrences and correct to <unclear> or <supplied reason="lost"> on a case-by-case basis?

>>> texts/DHARMA_DiplEdCandrakiranaPerpusnasL241.xml
>>> texts/DHARMA_DiplEdSiksaGuruPerpusnasL633.xml
>>> texts/DHARMA_DiplEdSiksaKandangKaresianPerpusnasL624.xml
>>> texts/DHARMA_DiplEdSiksaKandangKaresianPerpusnasL630.xml

By the way, I see many instances of <unclear reason="blackened"> in DiplEdCandrakiranaPerpusnasL241. If this is a usage endorsed by the EGC, then that should probably be added to the schema-permitted values of reason on unclear.

[michaelnmmeyer]

Looking at my own notes, I kept the value illegible to serve as default value for unclear/@reason. It has no effect in practice.

[danbalogh]

Still, I would prefer if none of our files had it explicitly. And if a default value is essential for good practice, then I think it had better be "damage". "Unclear because it is illegible" is, after all, a sort of tautology.

[arlogriffiths]

Zaki has jusr corrected his file texts/DHARMA_DiplEdCandrakiranaPerpusnasL241.xml.

@aditiagunawan : could you please correct your files by changing <unclear reason="illegible"> to <unclear>, if necessary with addition of <!-- A COMMENT --> or <note>A NOTE</note>?

[arlogriffiths]

@danbalogh : if the SIDDHAM hack is going to be made unavailable, will you make all the necessary manual adjustments in the SIDDHAM files?

[danbalogh]

The Siddham files have all been adjusted a long time ago without manual checking. They now have
<supplied reason="lost"><!-- CHECK if <unclear> applies better.--> to call attention to these spots in manual revision. The files where unclear reason illegible is present are only those in the backup folder, the original siddham XML files.

[danbalogh]

I believe this issue can be closed.