ajaxmulticommentary / ajmc-elixir Goto Github PK

To be then shared with Luca and Mahmoud.

URL template: https://ajmc.unil.ch/tei/{commentary_id}.xml

Find a better default zoom level (a bit more zoomed out than it is now).
This is related to the OSD IIIF viewer.

With data ingestion step.

e.g., in the page http://0.0.0.0:4000/bibliography/DeRomilly1976
https://wikidata.org/Q118976611 should rather be https://wikidata.org/wiki/Q118976611

In order for this to happen, it's necessary to ingest all available lemmata annotations into the canonical JSONs. So, all other commentaries (but Jebb) will need to be re-canonified so that these annotations will be picked up by the backend API and finally be displayed in the application. On the inception-side of things, all available annotations were exported to XMI (see this commit https://github.com/AjaxMultiCommentary/lemma-linkage-corpus/commit/4e949614345443d397c3df6386cc095da214219f).

We need to define API endpoints for browsing and searching lemmas for commentaries.

We can start with Jebb, as the data are mostly ready to go.

• /commentaries API endpoint
• /commentaries/:commentary_urn API endpoint
• /commentaries/:commentary_urn/comments API endpoint
• /commentaries/:commentary_urn/lemmas API endpoint
• Expose search API (/commentaries/:commentary_urn/comments?search=my%20query ?)

@pletcher Just an idea, let's discuss
As a user, it'd be useful to know (at a glance), at all times, how many glosses are available for the text section in focus and how many am I currently viewing (e.g., as a result of filtering).

Possible template: Currently displaying X glosses from X commentaries for this section (lines x-y). In total, there are Y glosses (of which Z are lemma-less).

Once #20 is fixed, quite a bit of space will be freed up in the top horizontal part (i.e. above the horizontal grey lines)., so this could be easily added.

JSON data for the synopsis is here.

Behaviour: the entire text will be loaded, but when clicking on a synopsis section the reading focus (= text window in focus) will ‘jump to’ the desired section.

steps to reproduce:

• refresh the application
• select and expand a glossa in the glossa viewer
• click on show page image
• now select a different commentary in the comm. filter
• the glossa should disappear but stays there

Some thoughts:

• this applies only to public domain commentaries
• we should keep the link between text and IIIF images (TEI Publisher will then exploit this for display)
• we don't have paragraphs information, but we do have divisions into glosses (for the commentary part)
• we do have a ToC for each commentary that we can use
• canonical references and tagged entities should be translated into corresponding TEI elements

Needs further discussion:

• should we produce one TEI file per commentary, or split into critical text/commentary/translation (if applicable)?

As a user I'd like to:

• read the introduction in De Romilly’s commentary
• read the bibliography of Finglass’ commentary

Features: OCR + image display

This can perhaps be offloaded to existing tools, e.g. TEI Publisher

⚠️This feature has a dependency with #39

For debugging of data it would be super useful. It could be hidden behind a small button or something. For a given glossa, I'd like to know what is the corresponding page image number.

Even better: make the page number a link to the IIIF page image for a quick check (I know there is the OSD viewer, but it makes the page number transparent, which is instead useful for debugging)

Just happened to notice that Campbell's commentary (cu31924087948174) does not show up in the interface. However, it's listed in https://github.com/AjaxMultiCommentary/ajmc-multicommentary/blob/main/priv/commentaries.toml

(to be sketched out)

• ajmc.unil.ch should point directly to the multi-comm., at the very beginning of the text (no landing page)
• a new page About should be added (with localised versions)
• the menu will then have: Ajax multi-commentary, Meta-commentary, Available commentaries, About

NB: the Meta-Commentary menu item will be a placeholder for functionalities to come in next releases

• There should be three lemmas, but there are only two.
• The first lemma's glossa goes on for way too long (to the end of the commentary region ("column")).

As a user, I'd like to have direct links to individual commentary lemmata. This requirement could be satisfied if we had a CTS URN for each lemma, and to have the reading environment respond when it gets a reference like that in the URL.

TODO: sketch out what the desired behaviour will be

some design changes to the filter (as discussed yesterday):

• put the filter behind a button (when not in use it stays collapsed and takes up less space)
• move the filter/button at the very top of the glossa viewer panel (column)
• make the list of available commentaries scrollable
• add a search field to make the list quickly searchable

PS: can we have in brackets, beside the commentator's name, the total number of glosses existing for the text section in focus?

This refers to the content of the html > head > title element in the various pages. It's a small detail, but it becomes "visible" whenever pages get bookmarked.

• default: "Ajax Multi-Commentary"
• bibliography page: "Ajax Multi-Commentary – Bibliography"
• about page: "Ajax Multi-Commentary – About"

For links to the ajmc platform that point to a specific glossa, I guess the title could/should reflect the target glossa, but we can leave this, in case, for the minimal comp. version.

Bibliography will be one item in the main menu.

In this basic version, I'd like to have displayed bibliographic metadata for the commentaries in our corpus, which should be fetched from this Zotero library. Short labels in the multi-comm. reading environment (e.g. Finglass 2011) should link back to the corresponding record in this section.

The easiest, I guess, is to hide those details behind a button. But I let you decide how to implement this.

There is a dependency with this ingestion pipeline issue: AjaxMultiCommentary/ajmc-pipeline#16

It'd be great to be able to edit some pages via Markdown. For now this applies to the to-come About page but there may be more use cases in the near future.

As an example, Stanford’s glossa to lines 160-1, is rendered as if it was referring to lines 1-160. It seems to be happening also for other glosses.

Add some info buttons to provide the users quick explanations of how key interface components work.
Info buttons for:

• commentary filter
• glossae viewer
• critical text selector
• dynamic apparatus

Explanations attached to info-buttons should be localised

• a good example of this is Wecklein 1894 ad loc. vv. 1–133.
• in order to keep the line heatmap informative, only the starting line (here line 1) should be a highlight
• all other lines, 2-133 will not receive any "heat"

Not sure what would be the best to place this new button. Perhaps in the middle column, on top of the critical text?

At all times, the user should be able to ‘switch on/off’ glosses from a given commentary, filtering of glosses based on comm.