Skip to content

Journal article import (tool_bibliography_acquisition)

See also: Tools user guide · Developer reference

Paste a journal's OAI-PMH URL (or a normal OJS article/journal URL), review every article the tool found, then import the kept ones into the Publications section as new records, each linked to its Series and Authors, with the article's PDF attached when one can be found.

What it's for

Building a bibliography of a journal's output by hand — title, authors, abstract, pages, the journal/series it belongs to — is slow and repetitive. This tool speaks OAI-PMH, the metadata- harvesting protocol almost every OJS (Open Journal Systems) journal and many institutional repositories expose, so it can pull structured article metadata straight from the source instead of scraping free text.

Concrete scenario: a librarian wants a Spanish academic journal's current issue in the Publications section. They paste the issue's normal landing-page URL (not the raw OAI endpoint — the tool derives that itself from OJS's own URL convention); the tool resolves every article linked from that page, fetches each one's OAI-PMH record, and the review screen shows a checklist with each article's title, authors, date and page range. The librarian keeps the relevant ones and clicks Confirm import. The tool creates one Publication record per kept article, resolves (or creates) the journal's Series record and each author's Person record and links them, writes the abstract in every language the source carries that your installation uses, and — when a downloadable copy can be found — imports the PDF directly onto the record.

When to use it

  • You have a journal's URL (an OAI-PMH endpoint, or a normal OJS homepage/issue/article URL) and want its articles as Publication records.
  • You want the Series and Authors linked automatically, and the PDF attached when one is available, without re-typing bibliographic metadata by hand.

When NOT to use it:

  • To import a Zotero-exported bibliography you already curated outside Dédalo, use Zotero import.
  • For a library's MARC21 catalogue export, use MARC21 import.
  • To harvest an entire journal's back catalogue in one go: this tool deliberately resolves one bounded set of articles per run (a single article, or every article linked from one listing page such as a current-issue page) — run it again per issue for the rest.

Where to find it

The tool is scoped to Publication sections — it does not appear on an arbitrary section even though its registration names the generic section model. Where your profile is authorized for it, its button opens in a modal (not a separate window).

Using it, step by step

  1. Open the tool on the Publications section.
  2. Paste the journal's URL — either its real OAI-PMH base URL (ending /oai), or a normal OJS URL (a journal homepage, a current-issue page, an individual article's .../article/view/<id> page): the tool derives the OAI endpoint from OJS's own journal-routing URL convention automatically. A listing page (not a single article) is scanned for every real article link it carries.
  3. If the source blocks automated fetching (confirmed for some OJS journal landing pages sitting behind a Cloudflare challenge), Preview refuses and tells you so. Save the page yourself in your own browser, then pick the saved OAI-PMH response (or HTML) file in the upload field below the URL field — the URL is still required, since it picks the right parser and is part of the Series dedup key.
  4. Click Preview. A pasted URL runs as a background job — one OAI-PMH fetch per article found, so a full issue can take a little while; an uploaded file is parsed instantly. Either way you get:
  5. the detected Series (the journal's own name), and whether it will link to an existing record or a new one will be created;
  6. a checklist of every article found, checked by default, each with its title, authors, publication date and page range.
  7. if some articles could not be fetched, a notice listing each one and why, above the articles that were.
  8. Curate the list. Use Select all / Deselect all, an include/exclude keyword filter against the title and authors, or an include/exclude year range (parsed from each article's publication date).
  9. Click Confirm import. The tool creates one record per kept article in the background, with live progress, and reports one result line per article: the fields written, whether the Series was linked (created or reused, or why it could not be), whether the authors were linked, and whether a PDF was imported.

Tips and gotchas

Re-importing the same batch does not duplicate it

Each article is matched by its own stable identifier before anything is created; an article already imported is reported as skipped, not duplicated. Re-running Confirm import — including after stopping a batch partway — picks up only what is left.

The abstract is written once per language the source carries and your installation uses

When a source provides the abstract in more than one language, each variant is written into its own language slot on the record rather than merged into a single one — open the record's language switcher to see all of them. A variant in a language your installation does not use, or a second variant for a language already written, is left out, and the result line names it; the publication itself is still imported.

A PDF is imported best-effort, separately from the record itself

OAI-PMH metadata never carries a direct PDF link, so the tool follows the article's landing page to find one. That step — and the PDF download itself — can fail independently of everything else (a blocked landing page, no PDF on it, a host this tool doesn't recognise); the record is still created with all its other fields, and the result line says specifically why the PDF did not come along.

This is one bounded set per run, not a whole-journal harvest

Pasting a journal's homepage or an issue page resolves the articles linked from that page — typically one issue's worth — capped at 200 articles found and 200 kept per batch, not the journal's entire publication history. Run it again for another issue.

Stopping a long import is safe

Cancelling a running Confirm import stops it before starting the next article — it never leaves a half-written record behind. The summary reports how many of the total were actually committed, and re-running the same batch afterwards only imports what is left (see the dedup note above).

  • Zotero import — the other Publications-section importer, for a bibliography already curated and exported from Zotero rather than harvested live from a journal.
  • MARC21 import — a library MARC21 catalogue import into the same section.
  • Auction import — built the same way (paste a public URL, review a checklist, commit in the background) for a different domain: auction lots into Numismatic Objects, rather than journal articles into Publications.
  • Developer reference — the preview_url/preview_html/commit_publications actions, the OAI-PMH adapter, and the dedup/lock rules.