Complex insertions not working

[aewingate]

It appears that the functionality described here in the encoding guide (both inserting entire <item> elements and inserting <add>) is not working.

In the diplomatic version, text correctly appears in the place where it was originally written, but in the normalized version, the text is not being inserted into where Newton meant it to go and removed from where it was originally written.

You can see this with "Affalia" here from 3v of ALCH00200

Diplomatic

Normalized

There's no change. The text shouldn't appear on 3v in the normalized and should instead by inserted into 4r.

Affalia is an example of moving an entire <item>. See 34v/35r of ALCH00200 for an example of moving an <add>

[aewingate]

Also, it feels like this is a new bug? I think I would have noticed this problem when I was re-writing the encoding guidelines!

[tubesoft]

@aewingate
Sorry for reacting to this thread really late!

First of all, the moving of the element is indicated by
line 628: <item rend="ti-2" corresp="#folio4r" xml:id="folio3v"> and
line 728: <item xml:id="folio4r" copyOf="#folio3v" rend="ti-2"/>.

There is no XSLT code dealing with those codes above so far. Should I add this feature to the XSLT code?

[aewingate]

Hi @tubesoft! No worries, and sorry for not getting back to you either. Yes, we do need the movement because it's in the live site. This is what the live site has:
Diplomatic:

Normalized

[tubesoft]

@aewingate Okay. I can work on that. I have one question before starting to code.

This is line 628-642 of ALCH00200:

<item rend="ti-2" corresp="#folio4r" xml:id="folio3v">
	<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The text on
		this page is inserted in folio 4r.</note>
	<add>
		<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The
			inserted item "Affalia" is physically located on folio
			3v.</note>
		<note type="editorial" resp="#cpastori" rend="diplomatic" anchored="true">The item
			"Affalia" is a subsequent insertion by Isaac Newton to folio
			4r.</note>
		<label>Affalia</label>, spiritus cum anima conjunctus. <bibl corresp="CHYM000001.xml#Zetzner1660">Senior p. 80</bibl><!--p.226, MA-->
		<lb/></add>
	<pb n="7"/>
	<milestone unit="folio" n="4r" xml:id="f4r" facs="#surface-4r"/>
</item>

Now, @corresp and @xml:id are at the item element, but if we want to link the add element to line 728: <item xml:id="folio4r" copyOf="#folio3v" rend="ti-2"/>, we would want to move those @corresp and @xml:id to the add element. Does this still conform to the coding guidelines?

Although I tried multiple ways, I could not render it as you showed above. However, if the TEI code is the following, I can manage to do it.

<item rend="ti-2">
	<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The text on
		this page is inserted in folio 4r.</note>
	<add  corresp="#folio4r" xml:id="folio3v">
		<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The
			inserted item "Affalia" is physically located on folio
			3v.</note>
		<note type="editorial" resp="#cpastori" rend="diplomatic" anchored="true">The item
			"Affalia" is a subsequent insertion by Isaac Newton to folio
			4r.</note>
		<label>Affalia</label>, spiritus cum anima conjunctus. <bibl corresp="CHYM000001.xml#Zetzner1660">Senior p. 80</bibl><!--p.226, MA-->
		<lb/></add>
	<pb n="7"/>
	<milestone unit="folio" n="4r" xml:id="f4r" facs="#surface-4r"/>
</item>

and as for the moving destination,

<item rend="ti-2">
	<add xml:id="folio4r" copyOf="#folio3v"></add>
</item>

Then, the rendition goes like these:
normalized:

diplomatic:

[aewingate]

Hi @tubesoft, sorry again for the delay.

So the way you have it conforms to how we do insertions into items, but not the movement of entire items.

One quick question before I bring up this issue to the big group (I don't feel like I can unilaterally change how we encode something like this without bringing it up!!): if we move this first note outside the <item>, does that make the whole item easier to move? So line 628-642 would look like this:

<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The text on
		this page is inserted in folio 4r.</note>
<item rend="ti-2" corresp="#folio4r" xml:id="folio3v">

	<add>
		<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The
			inserted item "Affalia" is physically located on folio
			3v.</note>
		<note type="editorial" resp="#cpastori" rend="diplomatic" anchored="true">The item
			"Affalia" is a subsequent insertion by Isaac Newton to folio
			4r.</note>
		<label>Affalia</label>, spiritus cum anima conjunctus. <bibl corresp="CHYM000001.xml#Zetzner1660">Senior p. 80</bibl><!--p.226, MA-->
		<lb/></add>
	<pb n="7"/>
	<milestone unit="folio" n="4r" xml:id="f4r" facs="#surface-4r"/>
</item>

Since that note should stay on 3v, it seems like it shouldn't be part of the <item> that is being moved.

Also, what happens if we did the following? It seems semantically closer to the original in that we're moving the whole <item> but maybe more consistent for the purposes of XSLT?

	<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The text on
		this page is inserted in folio 4r.</note>
	<add  corresp="#folio4r" xml:id="folio3v">
               <item rend="ti-2">

		   <note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The
			inserted item "Affalia" is physically located on folio
			3v.</note>
		   <note type="editorial" resp="#cpastori" rend="diplomatic" anchored="true">The item
			"Affalia" is a subsequent insertion by Isaac Newton to folio
			4r.</note>
		   <label>Affalia</label>, spiritus cum anima conjunctus. <bibl corresp="CHYM000001.xml#Zetzner1660">Senior p. 80</bibl><!--p.226, MA-->
		   <lb/>
              </item>
         </add>
<!---alewinga, there seems to be the persistent issue of items ending with the <pb> and <milestone> inside the <item> which
 seems like it would make your life harder. If this way works, I may work on that for the whole manuscript-->
	<pb n="7"/>
	<milestone unit="folio" n="4r" xml:id="f4r" facs="#surface-4r"/>

and then moving destination:

<add xml:id="folio4r" copyOf="#folio3v"></add> 

and on the moving destination, would an empty element work too?

<add xml:id="folio4r" copyOf="#folio3v"/> 

Thanks! -Alex

[tubesoft]

Hello @aewingate
I tried both of your suggestions. It seems that only the latter one is working (at least with my solution), so could you bring that one to the meeting?

If I may suggest more, I would say:

	<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The text on
		this page is inserted in folio 4r.</note>
        <note type="editorial" resp="#cpastori" rend="diplomatic" anchored="true">The item
			"Affalia" is a subsequent insertion by Isaac Newton to folio
			4r.</note>
	<add  corresp="#folio4r" xml:id="folio3v">
               <item rend="ti-2">
                   <note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The
			inserted item "Affalia" is physically located on folio
			3v.</note>
		   <label>Affalia</label>, spiritus cum anima conjunctus. <bibl corresp="CHYM000001.xml#Zetzner1660">Senior p. 80</bibl><!--p.226, MA-->
		   <lb/>
              </item>
        </add>
	<pb n="7"/>
	<milestone unit="folio" n="4r" xml:id="f4r" facs="#surface-4r"/>

I think the note ["Affalia" is a subsequent insertion by Isaac Newton to folio 4r.] is always rendered on 3v in the diplomatic, so we can treat the same as [The text on this page is inserted in folio 4r.] in the normalized.

Also, as you suggested, I feel it makes sense to put <pb> and <milestone> outside of because this is not a part of insertion.

As for the destination part, <add xml:id="folio4r" copyOf="#folio3v"/> works as well.

Thank you!

[aewingate]

@tubesoft, everything that you say makes sense to me! I will bring all of this to the big meeting. But just a heads up that it won't be until March 31 because next week is Spring Break and I'll be gone the week after that.

Thanks!

[aewingate]

@tubesoft , I talked in the meeting today, and we are good with the latter solution. I will work on re-encoding the parts across the corpus to make it work with how we have proposed if you make whatever changes are necessary on the back end. I think it's better all around because now there's just one consistent way of moving pieces between pages.

[aewingate]

@tubesoft
I ran into an unanticipated problem because I guess I wasn't testing my encodings within a TEI schema. <item> can't be a direct child of <add> according to the TEI guidelines. I'm going to email the TEI listserv to see if this could get changed because it's nuts that you can have a <list> inside of <add> but not an individual <item>.

But this also explains why we had a separate way of encoding the movement of whole items!

[aewingate]

@tubesoft Alright, I talked to some TEI people. I don't think we have any chance of getting them to change it so <item> can go inside <add>, but their suggestion was to use <addSpan>. https://tei-c.org/release/doc/tei-p5-doc/en/html/ref-addSpan.html. If we were to use , I think it would look something like this:

	<note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The text on
		this page is inserted in folio 4r.</note>
        <note type="editorial" resp="#cpastori" rend="diplomatic" anchored="true">The item
			"Affalia" is a subsequent insertion by Isaac Newton to folio
			4r.</note>
	<addSpan  corresp="#folio4r" xml:id="folio3v" spanTo="#end-folio3v"/><!-- addSpan marking beginning of insertion-->
               <item rend="ti-2">
                   <note type="editorial" resp="#cpastori" rend="normalized" anchored="true">The
			inserted item "Affalia" is physically located on folio
			3v.</note>
		   <label>Affalia</label>, spiritus cum anima conjunctus. <bibl corresp="CHYM000001.xml#Zetzner1660">Senior p. 80</bibl><!--p.226, MA-->
		   <lb/>
              </item>
        <anchor xml:id="end-folio3v"/><!--empty anchor marking end of insertion-->
	<pb n="7"/>
	<milestone unit="folio" n="4r" xml:id="f4r" facs="#surface-4r"/>
...
<add xml:id="folio4r" copyOf="#folio3v"/><!--destination of insertion in diplomatic-->

Could we get it to move like this?

Another solution that was brought up in talking to TEI people was to use <orig>/<reg> to make things appear and disappear between the diplomatic and normalized. I wonder if that would be simpler and maybe even semantically better. I will talk to John Walsh when he gets back from France about it in mid-May to see what he thinks and then go to the larger team.

Thanks for sticking with me on this issue!

[tubesoft]

@aewingate
Yeah, I am with the <orig> and <reg> solution as well. It also seems better for XSLT implementation.

[aewingate]

@tubesoft I talked with John Walsh today. He seems to think that getting the @copyOf attribute to work is the key, and he pointed me to this point in his Swinburne project XSLT that might be useful in getting it to work:
See line 338ff for an example of code using @copyOf to process the copied element: https://github.com/jawalsh/swinburne/blob/main/common/xslt/p5-to-html.xsl

I did figure out an orig/reg solution too, but Jim brought up some valid concerns with encoding it that way, so I'd like to see first if John's code can help us make @copyOf work.