BG: Feedback

[matyaskopp]

Nice work, I am impressed by how well the reading order is done. Here are some suggestions for improvements

missing link to source data in bibl element

• link to source

I believe this file https://github.com/stefanlmarinov99-hub/PressMint-BG/blob/755d48797515933bdbb63e6497edec89bee1b4a6/Samples/PressMint-BG/1908/PressMint-BG_1908-06-14-6SEP.xml
corresponds to this newspaper issue: https://digital.libplovdiv.com/bg/view/644a11aadff46e028dec6456

Can you link the source pdf in bibl element?
You can get the source link by:

• https://digital.libplovdiv.com/api/object/644a1199f62d02cdaa00a018, where you find the pdfname:

curl --silent https://digital.libplovdiv.com/api/object/644a1199f62d02cdaa00a018 | jq -r '._files'
[
  {
    "fn": "4be89ec3-bf24-4d9f-8620-68ce2cb846d4.pdf",
    "path": "media.0.file"
  }
]

and then you can use this https://digital.libplovdiv.com/asset/default/4be89ec3-bf24-4d9f-8620-68ce2cb846d4.pdf

consider preserving local ids

• local ids

We allow using local ids, so you can have better pointer to the source. it seems that you have one obvious:

• Сигнатура: 63 3/18

But be careful, these IDs, can be just technical for the web app

Some info is present here:

 curl --silent https://digital.libplovdiv.com/api/object/644a111004a8a25b8d415db2?type=ref | jq 
{
  "_id": "644a111004a8a25b8d415db2",
  "identification": {
    "signature": {
      "_v": "63 3/18"
    },
    "kind": {
      "_v": "Вестник",
      "_v:en": "Newspaper"
    },
    "name": {
      "_v": "6-Й СЕПТЕМВРИЙ",
      "_v:en": "6-TH OF SEPTEMBER"
    },
    "subtitle": {
      "_v": "Орган на стамболовистите",
      "_v:en": "Organ of the Stambolovists"
    },
    "publishing": {
      "place": {
        "_v": "Пловдив",
        "_v:en": "Plovdiv"
      },
      "year": {
        "_v": [
          "1908"
        ],
        "_d": [
          1908
        ]
      }
    },
    "print": {
      "year": {
        "_d": []
      }
    },
    "subject": {
      "_v": [
        "Политика",
        "Новини"
      ],
      "_v:en": [
        "Politics",
        "News"
      ]
    },
    "geography": {
      "_v": [
        "Пловдив"
      ],
      "_v:en": [
        "Plovdiv"
      ]
    },
    "language": {
      "_v": "Български език",
      "_v:en": "Bulgarian"
    },
    "annotation": {
      "_v": "Пловдивски стамболовистки вестник. Издава се от отцепници народнолиберали в Пловдив, начело с Ив. Андонов. Противопоставя се на д-р Никола Генадиев, лидер на стамболовистите.",
      "_v:en": "Stambolovist newspaper in Plovdiv. It is published by renegade National Liberals in Plovdiv, headed by Iv. Andonov. Opposes Dr. Nikola Genadiev, leader of the Stambolovists."
    },
    "_title": {
      "_v": "6-Й СЕПТЕМВРИЙ",
      "_v:en": "6-TH OF SEPTEMBER"
    },
    "_thumb": null
  },
  "description": {
    "periodicity": {
      "_v": "week"
    }
  },
  "_title": {
    "_v": "6-Й СЕПТЕМВРИЙ",
    "_v:en": "6-TH OF SEPTEMBER"
  },
  "_thumb": null,
  "_t": "ct:Periodical",
  "_meta": {
    "state": null,
    "user": "60bf3f532ecf5724d8150c76",
    "creator": "633c2ab86db8bc633b664a4c",
    "time": 1690352199,
    "action": "update",
    "ctime": 1690352199
  },
  "_version": 1,
  "_refs": []
}

invalid url in graphic

• graphic url

https://github.com/stefanlmarinov99-hub/PressMint-BG/blob/755d48797515933bdbb63e6497edec89bee1b4a6/Samples/PressMint-BG/1908/PressMint-BG_1908-06-14-6SEP.xml#L42-L54

        <surface xml:id="PressMint-BG_1908-06-14-6SEP.page1">
            <graphic url="1.jpg"></graphic>
        </surface>

My suggestion is to use:

        <surface xml:id="PressMint-BG_1908-06-14-6SEP.page1">
            <graphic url="https://digital.libplovdiv.com/asset/default/4be89ec3-bf24-4d9f-8620-68ce2cb846d4.pdf#page=1" mimeType="application/pdf"></graphic>
        </surface>

and it would probably be good to support it with mimeType="application/pdf"
(@TomazErjavec, @maartenpt, do you have a different suggestion for handling non-image references in graphic?)

missplaced page beginning

• pb in the middle of the column

https://github.com/stefanlmarinov99-hub/PressMint-BG/blob/755d48797515933bdbb63e6497edec89bee1b4a6/Samples/PressMint-BG/1908/PressMint-BG_1908-06-14-6SEP.xml#L108

consider using lb

• lb element

You have provided XML with new-line formatting. There is a lb element for beginnings of lines and cb element for columns. Please consider adding them. Your data seems to be in good reading order, so this would be a nice improvement.

empty urls

• empty urls

https://github.com/stefanlmarinov99-hub/PressMint-BG/blob/755d48797515933bdbb63e6497edec89bee1b4a6/Samples/PressMint-BG/PressMint-BG.xml#L70-L71

                <bibl>
                    <title xml:lang="en">Scanning and OCR processing from Bulgarian National Library Plovdiv &quot;Ivan Vazov&quot;</title>
                    <idno type="URI">https://</idno>
                    <idno type="URI">https://</idno>
                    <date from="1878-01-01" to="2025-12-01"></date>
                </bibl>

ParlaMint garbage

• garbage in teiCorpus

be careful about reusing code, not sure if our finalization scripts successfully remove all improper metadata.
example: https://github.com/stefanlmarinov99-hub/PressMint-BG/blob/755d48797515933bdbb63e6497edec89bee1b4a6/Samples/PressMint-BG/PressMint-BG.xml#L99-L110
This is a copy-paste of information - we don't have u in PressMint

taxonomy placement

• separate file for taxonomies

please extract this taxonomy to separate file
https://github.com/stefanlmarinov99-hub/PressMint-BG/blob/755d48797515933bdbb63e6497edec89bee1b4a6/Samples/PressMint-BG/PressMint-BG.xml#L110-L113

file should be named according to these rules: https://clarin-eric.github.io/PressMint/#sec-files
so sth like PressMint-BG-taxonomy-newspapertypes.xml

Are you planning to add more categories? There is only one category xml:id="press.partyNewspaper" which is a bit weird.

editorial description

• update editorialDecl to be more relevant