modifications to DHARMA_CritEdSarvavajrodaya.xml and CritEd schema

arlogriffiths commented 6 months ago

@ajaniak — thanks for importing this file and its Translation from LaTeX source code. At present, the files are not yet valid and hence cannot be displayed. I will list stuff to be changed in this checklist, to which I will gradually be making additions.

[x] 1. all cases of <rdg type="emn" wit="#XYZ">READING</rdg><witDetail wit="#XYZ" type="silently"/> should become <rdg wit="#XYZ">READING</rdg><witDetail wit="#XYZ" type="silemn"/> — this use of "silemn" is new
[x] 2. <lem type="emn" wit="#([A-Z]+)">(.+)</lem><witDetail wit="#([A-Z]+)" type="silently"/> should become <lem wit="#$1">$2</lem><witDetail wit="#$3" type="silemn"/>.
[x] 3. I think cases like this <app><lem type="emn" wit="#MSK">vajradhātvīśvaryā</lem><witDetail wit="#MSK" corresp="#KSP">($\leftarrow$ KSP, Tib.)</witDetail><rdg wit="#cod">vajradhātvīśvaryādi</rdg></app> can be represented as <app><lem type="emn" wit="#MSK" evidence="external">vajradhātvīśvaryā</lem><witDetail wit="#MSK" corresp="#KSP #Tib"/><rdg wit="#cod">vajradhātvīśvaryādi</rdg></app> — this use of @evidence on <lem> and the use of @corresp in <witDetail> are new. Alternative idea: <app><lem type="emn" wit="#MSK" evidence="external" corresp="#KSP #Tib">vajradhātvīśvaryā</lem><rdg wit="#cod">vajradhātvīśvaryādi</rdg></app>.
[x] 4. I think all cases of <hi rend="italic">$1</hi> can be replaced by <foreign xml:lang="san-Latn">$1</foreign>.
[x] 5. I have found a few cases where a little chunk of information from the LatEX source seems to have been lost in conversion. See XML line 3334 <note>incorrectly emends <hi rend="italic">raṅgaṁ rekhāyeta | paraṁ</hi></note> for last LaTeX-generated pdf "EdT incorrectly emends raṅgaṁrekhāyeta | paraṁ." Can you track down what has gone wrong here?
[x] 6. All cases like <app><lem>svahṛdgatavajre</lem><rdg>svahṛdgata[21v1]vajre</rdg></app> are actually only a way to mark the folio transitions in the codex unicus. If it is possible to isolate all these cases (<app>s without attributes on <lem> and <rdg> and with a sequence [##r/v#] inside <rdg>), they can be replaced by regular use of <pb edRef="#cod" n="$1"> inside the edited text (or a <lem>) and where relevant also in a <rdg> with @wit="#cod". Since this point does not lead to file invalidation, it is not a priority.
[x] 7. When we have a case like this: `STTS 72. The MS and the editions read the first half as follows: aho hy anupaman tejaḥ satvadhātvavabhāsanam aho hy anupamaṁ tejaḥ sattvadhātvavabhāsanam
, do you see any convienent way to represent what is meant other thanSTTS 72. The MS and the editions read the first half as follows:
: aho hy anupaman tejaḥ satvadhātvavabhāsanam : aho hy anupamaṁ tejaḥ sattvadhātvavabhāsanam `?
[x] 8. Cases like <rdg type="emn" wit="#S1">$2</rdg> should be replaced by <rdg wit="#$1">$2</rdg><witDetail wit="$1" type="emn"/> — this use of @type on <witDetail> is new, display should be "$2 $1 (em.)"
[x] 9. I now suggest that cases like <lem wit="$1">$2</lem><witDetail wit="$1" corresp="#$3"/> can b replaced by <lem wit="$1">$2</lem><note corresp="$3"/> — this use of <note> immediately after <lem> is new, display should be "$2] em. $1 (∥ $3)"
[x] 10. all cases of @wit="" inside <lem> can be removed if @wit="#MSK" and @wit="#cod" both appear in <rdg>

Of course I would appreciate your advice on the new encoding strategies that I propose. I am not sure @evidence="external" in (2) is necessary as I don't suppose there will be any computational use.

As a next step, I will need your or @michaelnmmeyer's help to make the necessary adjustments to the schema and XLST transformation in order for the usages not yet foreseen in EGC to be validated by the schema and properely displayed.

ajaniak commented 6 months ago

@arlogriffiths,

what about

[x] <lem type="emn" wit="#([A-Z]+)">(.+)</lem><witDetail wit="#([A-Z]+)" type="silently"/> would you like to modify it? Should it become <lem wit="#$1">$2</lem><witDetail wit="#$3" type="silemn"/> or <lem type="emn" wit="#$1">$2</lem><witDetail wit="#$3" type="silemn"/>?

arlogriffiths commented 6 months ago

Thanks. My answer is now inserted into the list above.

ajaniak commented 6 months ago

For 3. I have handled all the \leftarrow cases except 3 because there were not matching the case you have identify.

For 5. I think the pattern implying a MSK was confused in some cases. There are several ways to express it in the file and one of the pattern must have created a conflict deleting the ms sigla, probably also linked with the <note>. I have worked to reintroduce most of them, but some might still be missing.

Regarding @evidence, the main issue is that this attribute isn't allowed on <lem>. I think @type or <witDetail/>, even adding a <note> seem enough but I might not see what you are trying to achieve adding it.

arlogriffiths commented 6 months ago

Thanks for all of that. So let's forget about use of @evidence. I have made a new suggestion in no. 9 above. If that is implemented together with no. 8, I think the file will start to be close to validation.

ajaniak commented 6 months ago

For 6. I have changed all the folios, but some are still inside <note>.

For 9. Do you mean to say, I should change all <witDetail/> for <note> (also for <rdg> and for <witDetail type="silemn"/>?

arlogriffiths commented 6 months ago

For 6. thanks. We'll handle the rest manually.

For 9. I can think of: </rdg><witDetail wit="$1" type="silemn"/>. This does not need to be changed, I think. Please explain which other scenarios you have in mind.

arlogriffiths commented 6 months ago

I have added no. 10.

ajaniak commented 6 months ago

For 9. I was wondering if it was the @corresp requiring the change for <witDetail/> to <note>. And if yes, should I change those when linked with the <rdg>. Also what about the mix : <witDetail type="silemn" wit="#MSK" corresp="#KSP #Ied"/> should it become <witDetail type="silemn" wit="#MSK"/><note corresp="#KSP #Ied"/>. All in all, I have just checked in the file with an XPath, it concerns only 39 cases counting both cases.

arlogriffiths commented 6 months ago

Now I am home without my computer so a bit more difficult to answer. But yes, I think that use of @corresp was not allowed on witDetail and it also seemed semantically inappropriate to encode corresp on witDetail, so I thought of note. It seems elegant by way of analogy with how we use note in listApp for parallels.

I can only imagine that we have cases of rdg where corresp becomes relevant if MSK made an emn based on a parallel while we overrule the force of that parallel with other arguments. In such cases, yes witDetail with type=emn as well as note with corresp will be required for the rdg.

Yes, that mixed case should be changed as you suggest.

From: ajaniak @.> Sent: Monday, April 29, 2024 6:54 PM To: erc-dharma/tfd-sanskrit-philology @.> Cc: Arlo Griffiths @.>; Mention @.> Subject: Re: [erc-dharma/tfd-sanskrit-philology] modifications to DHARMA_CritEdSarvavajrodaya.xml and CritEd schema (Issue #2)

For 9. I was wondering if it was the @corresp requiring the change for to . And if yes, should I change those when linked with the . Also what about the mix : should it become . All in all, I have just checked in the file with an XPath, it concerns only 39 cases counting both cases.

— Reply to this email directly, view it on GitHubhttps://github.com/erc-dharma/tfd-sanskrit-philology/issues/2#issuecomment-2082527825, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AAGMAE2RLOPPJRXEEFQ2ECLY7YYG5AVCNFSM6AAAAABGX5CRGSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOBSGUZDOOBSGU. You are receiving this because you were mentioned.Message ID: @.***>

arlogriffiths commented 6 months ago

Thanks Axelle!

erc-dharma / tfd-sanskrit-philology

modifications to DHARMA_CritEdSarvavajrodaya.xml and CritEd schema #2