National Diet Library (Japan) DOIs and BibTex/JSON data all importing as preprints

nathan.hopson
Regardless of item type, National Diet Library (Japan) DOIs and downloaded BibTex/JSON data all import as preprints. This is true via all of the NDL's search functions:

https://ndlsearch.ndl.go.jp/
https://mina.ndl.go.jp/
https://dl.ndl.go.jp/

Fwiw, all of the items are properly catalogued by the NDL, i.e., books are listed as 図書 (book), journals are 雑誌 (journal/magazine), etc.

This wouldn't be an issue except that the Zotero extension can't import correctly from the NDL directly (all "Webpage"), which makes adding via DOI the quickest and most convenient solution.

Is there a workaround for this? Or is there something missing, corrupted, etc., in the DOI and/or BibTex/JSON data? (In other words, is this just an NDL problem?)
  • adamsmith
    We should be supporting NDL better (I thought we did, I'll take a look) but can you point to some examples? If I do random searches on https://ndlsearch.ndl.go.jp/ and get the BibTeX data from the result page, it's all labeled book and imports as such (even things that probably are papers, but nothing we can do about that). Whare are some DOIs?
  • nathan.hopson
    Thanks for the fast reply @adamsmith!

    All of the sources below import as Preprints:

    Two recent articles:

    10.11501/1148989
    10.11501/3373505

    (I did note that the DOIs appear to point to the journals, not the articles, so I think these are an NDL problem)

    Books imported by DOI:
    10.11501/1258566
    10.11501/1143785
    10.11501/1094626
    10.11501/1916436
    10.11501/1148989
    10.11501/9644625
    10.11501/12114094
    10.11501/1209913

    Books imported by BibTex (no DOI available):
    https://ndlsearch.ndl.go.jp/books/R100000002-I000001912963

    The exported BibTex file text, which I used to import this reference, is:
    @book {R100000002-I000001912963,
    author = "甘楽町史編さん委員会 編",
    title = "甘楽町史",
    publisher = "甘楽町",
    year = "1979.9",
    URL = "https://ndlsearch.ndl.go.jp/books/R100000002-I000001912963"
    }

    The JSON, which I did not use, is:
    {"link":"000001912963","identifier":{"JPNO":["88031540"],"NDLBibID":["000001912963"]},"title":[{"value":"甘楽町史","transcription":"カンラ チョウシ"}],"creator":[{"name":"甘楽町 (群馬県)","transcription":"カンラマチ (グンマケン)"}],"dc_creator":[{"name":"甘楽町史編さん委員会 編"}],"publisher":[{"name":"甘楽町","location":"甘楽町 (群馬県)"}],"date":"1979.9","issued":"1979","description":["付 (図2枚 袋入) : 甘楽町文化財分布図・甘楽町中世石造物分布図","甘楽町年表・参考文献: p1461～1482"],"subject":{"NDLSH":["甘楽町 (群馬県)"],"NDLC":["GC54"],"NDC8":["213.3"]},"price":"非売品","extent":["1493p ; 22cm"],"materialType":["図書","政府刊行物","地方公共団体刊行物"]}

    I see that the "materialType" attribute starts with "book" (図書).
  • adamsmith
    edited February 2, 2024
    The JSON wouldn't import at all. That's not a JSON format Zotero knows.
    The BibTeX should import as a book and does for me
    The DOIs do all import as preprint -- we can see if JALC (the registration agency) has something better (I haven't found that so far), but the CSL JSON they offer has 'article' as type throughout, and that's what CSL uses for preprints, so Zotero's behavior is correct and the issue is on their end.

    Edit: Note that Zotero should also import MARC
  • nathan.hopson
    @adamsmith
    My apologies, the BibTex did import correctly. I thought I had double checked that.

    And thanks for the explanation about the CSL JSON.

    Btw/fwiw, I imported 10.11501/1258566 from BibTex and MARC (with the output language as English (US) and Japanese), and the results are different:

    BibTex:
    (English)
    編群馬県. 昭和9. 栄養改善ト婦人労働トノ関係ニ就テノ実験成績報告. 群馬県. https://ndlsearch.ndl.go.jp/books/R100000002-I000000739945.

    (Japanese)
    編群馬県. 昭和9. 栄養改善ト婦人労働トノ関係ニ就テノ実験成績報告. 群馬県. https://ndlsearch.ndl.go.jp/books/R100000002-I000000739945.


    MARC
    (English)
    群馬県, ed. 9. 栄養改善ト婦人労働トノ関係ニ就テノ実験成績報告. 前橋: 群馬県.

    (Japanese)
    群馬県, 編. 9. 栄養改善ト婦人労働トノ関係ニ就テノ実験成績報告. 前橋: 群馬県.


    BibTex:
    * "editor" (編) as the author surname
    * full Japanese date (昭和9)

    MARC
    * no surname (corporate name) and correctly interpreted as "editor"
    * date is just the numeral 9, without the reign name (昭和)
    * includes place of publication (前橋)

    I don't know whether that will help in diagnosing the issue, but just in case.

    Thanks again!
Sign In or Register to comment.