Alignment and Apparatus View

[hlapin]

Updated "index-Wit-Compos.xsl"
@raffazizzi
This runs faster than the original and combines the two steps in app:index-compos into one, gets rid of the need for index.xml, and groupFromIndex.xsl. Also does not need parameters passed.
Resulting XML has an element for every ab whether populated or not.
Assuming that exist caches the results of xslt transformations (docs to transform:transform, if I understood correctly), or if result document is storable, it should be immediately reusable for all the compare page elements: left TOC and wit list.
at heading should also provide text for mouseovers on the witlist if we want them.
Did not want to touch major changes until I understand the templating functions better.

[hlapin]

Have implemented some changes based on new xsl.
Does improve speed. One question: perhaps since ref.xml is effectively an index we can set up a procedure that runs this transform whenever the [header of] ref.xml is updated with new xrefs or new docs are loaded in the database. The saved XML (or json?) can then be queried rather than generated at run time.

[hlapin]

Rather than "variant/invariant" we should do grouping on surface. Same algorithm should work for both alignment and apparatus view. Apparatus view is NOT grouping properly (seems to not group the not-equal-to-group-1 items correctly).
I can work on this, but need @raffazizzi 's help to know how to return data so that it can by styled properly.
Styling is also an issue now that tokenization separates h1/h2; these are in IDs; need styles to distinguish.

[hlapin]

Whole chapter view:
Needs to check each mishnah and see if individual mishnah is attested in that witness.

[hlapin]

Whole chapter view on align/apparatus and synoptic view improved with latest sync.

[hlapin]

http://localhost:8080/exist/apps/digitalmishnah/compare/4.1.2.1/S07326,S00483,S01520/apparatus
Apparatus view fails on reordering. Why?

[raffazizzi]

http://localhost:8080/exist/apps/digitalmishnah/compare/4.1.2.1/S07326,S00483,S01520/apparatus
Apparatus view fails on reordering. Why?

It's because of this: https://github.com/umd-mith/mishnah/blob/exist/xsl/apparatus.xsl#L15
And it only becomes problematic in that comparison when S07326 is the base text.

@hlapin can you have a look at either the XSLT or the XML?

[hlapin]

Rather than "variant/invariant" we should do grouping on surface. Same algorithm should work for both alignment and apparatus view. Apparatus view is NOT grouping properly (seems to not group the not-equal-to-group-1 items correctly).
I can work on this, but need @raffazizzi 's help to know how to return data so that it can by styled properly.
Styling is also an issue now that tokenization separates h1/h2; these are in IDs; need styles to distinguish.

[replied by email, but copying here]
Rather than variant/invariant, we can use the same algorithm that we use in apparatus (also needs work, but I need to experiment a bit and see what is going on) to distinguish groups of distinct values and use colors or styling to do this.
This becomes a teensy bit more complicated than I realized with curated collations and we have some decisions to make and work around those.
For editing alignments we converted adjacent tokens with additions and deletions into distinct runs of text.. E.g.

    [The]     (E)U[SA] 
    tkn1      tkn2  

becomes:

h1:               EU 
                  tkn2-h1 

h2:     The       USA 
        tkn1-h1   tkn2-h2     

The ids with the suffixes -h1 and -h2 correspond to the same tokens.

For DISPLAY, I suspect that the typographical distinctions are both sufficient and in some ways better, because they better correspond to what visually appears in the text.
So this may get messy:

• when things are flagged -h1, -h2 we need the css to pick up and distinguish
• possibility of double alignments (h1/h2 in A align with h1/h2 in witness B)

What is the simplest way to handle this?

[raffazizzi]

when things are flagged -h1, -h2 we need the css to pick up and distinguish

Is it always going to be h1 and h2, or is it incremental? (h3, h4, ...)

[hlapin]

Some day, I would love to have it be incremental, but right now we are by default just distinguishing -h1 -h2. (I did this for the editor).

…

------------------------------ Robert H. Smith Professor of Jewish Studies and Professor of History Department of History University of Maryland 2115 Francis Scott Key Hall College Park, MD 20742 301 405 4296 | hlapin@umd.edu www.digitalmishnah.org | www.eRabbinica.org Director Joseph and Rebecca Meyerhoff Program and Center forJewish Studies University of Maryland 4141 Susquehanna Hall College Park, MD 20742 301 405 4975 | jwst-contact@umd.edu www.jewishstudies.umd.edu ------------------------------

On Tue, Apr 10, 2018 at 2:35 PM, Raffaele Viglianti < ***@***.***> wrote: when things are flagged -h1, -h2 we need the css to pick up and distinguish Is it always going to be h1 and h2, or is it incremental? (h3, h4, ...) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABBRzX2I1Ca1ib-2h0wMl7hWawaqQi_eks5tnPuDgaJpZM4Sy3x6> .

[raffazizzi]

Something like this? (We can pick better colors)

I got the h1/h2 from "resp" in the JSON table rather than the end of the id string. Is that a reliable place? It could be incremental as long as there's data in the "resp" field, but we'll need to add more colors to the CSS.

We'll discuss how to handle this in the TEI stand-off collations when we talk.

[hlapin]

Excellent. For the editor, I just did bold and underline. Since we may use colors for grouping, I though I might want to avoid background color as a designator.

…

------------------------------ Robert H. Smith Professor of Jewish Studies and Professor of History Department of History University of Maryland 2115 Francis Scott Key Hall College Park, MD 20742 301 405 4296 | hlapin@umd.edu www.digitalmishnah.org | www.eRabbinica.org Director Joseph and Rebecca Meyerhoff Program and Center forJewish Studies University of Maryland 4141 Susquehanna Hall College Park, MD 20742 301 405 4975 | jwst-contact@umd.edu www.jewishstudies.umd.edu ------------------------------

On Tue, Apr 10, 2018 at 2:48 PM, Raffaele Viglianti < ***@***.***> wrote: Something like this? (We can pick better colors) [image: screenshot from 2018-04-10 14-45-53] <https://user-images.githubusercontent.com/144770/38577101-f54d8b4c-3ccd-11e8-912e-5aca426c352f.png> I got the h1/h2 from "resp" in the JSON table rather than the end of the id string. Is that a reliable place? It could be incremental as long as there's data in the "resp" field, but we'll need to add more colors to the CSS. We'll discuss how to handle this in the TEI stand-off collations when we talk. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABBRzbGhaqXqJv5PZA7f33zw1QUNPUs4ks5tnP6FgaJpZM4Sy3x6> .

[hlapin]

in terms of resp vs the id, yes. Initially I assumed that resp was all we needed. At one point I found (it might still be true on the updated controller) that for the edit app, having the same id value caused problems, so I added the suffix to the ID.

…

------------------------------ Robert H. Smith Professor of Jewish Studies and Professor of History Department of History University of Maryland 2115 Francis Scott Key Hall College Park, MD 20742 301 405 4296 | hlapin@umd.edu www.digitalmishnah.org | www.eRabbinica.org Director Joseph and Rebecca Meyerhoff Program and Center forJewish Studies University of Maryland 4141 Susquehanna Hall College Park, MD 20742 301 405 4975 | jwst-contact@umd.edu www.jewishstudies.umd.edu ------------------------------

On Tue, Apr 10, 2018 at 2:48 PM, Raffaele Viglianti < ***@***.***> wrote: Something like this? (We can pick better colors) [image: screenshot from 2018-04-10 14-45-53] <https://user-images.githubusercontent.com/144770/38577101-f54d8b4c-3ccd-11e8-912e-5aca426c352f.png> I got the h1/h2 from "resp" in the JSON table rather than the end of the id string. Is that a reliable place? It could be incremental as long as there's data in the "resp" field, but we'll need to add more colors to the CSS. We'll discuss how to handle this in the TEI stand-off collations when we talk. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABBRzbGhaqXqJv5PZA7f33zw1QUNPUs4ks5tnP6FgaJpZM4Sy3x6> .

[raffazizzi]

Ok, I've committed code for this. I'll leave this issue open because we still have the TEI part to sort out.

[hlapin]

Not sure if this is significant. Opted not to update the data and the build failed (below); tried again updating data and succeeded. BUILD FAILED C:\Users\hlapin\Documents\MTExistApp\build.xml:150: XMLDB exception caught while executing query: Failed to invoke method queryPT in class org.exist.xmlrpc.RpcC onnection: null

…

------------------------------ Robert H. Smith Professor of Jewish Studies and Professor of History Department of History University of Maryland 2115 Francis Scott Key Hall College Park, MD 20742 301 405 4296 | hlapin@umd.edu www.digitalmishnah.org | www.eRabbinica.org Director Joseph and Rebecca Meyerhoff Program and Center forJewish Studies University of Maryland 4141 Susquehanna Hall College Park, MD 20742 301 405 4975 | jwst-contact@umd.edu www.jewishstudies.umd.edu ------------------------------

On Tue, Apr 10, 2018 at 3:07 PM, Raffaele Viglianti < ***@***.***> wrote: Ok, I've committed code for this. I'll leave this issue open because we still have the TEI part to sort out. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABBRze-3nKawABo9xuer3hHGTdi3M9bpks5tnQMDgaJpZM4Sy3x6> .

[hlapin]

Added a new collation 4.2.5.1.xml to collations based on my changes to editor and @raffazizzi's new toTEI.js implementation.