Bulk import

Bulk import turns a list of citations into library items — one row at a time, with a clear report of what could not be placed.

This is separate from Importing references via the search bar (single DOI, ISBN or title) and from the CLI batch import.

From the home page

On the signed-in home, in Find mode, open the search bar menu and choose Import from document. Paste markdown or plain text — a bibliography, reading list or notes — or choose a .md / .txt file.

Pangur saves the source as a Research note (the text is kept), then scans it for references. Each candidate appears with a checkbox; import only the ones you want. The source document stays in Research either way.

You can also start from an existing Research document in the same dialog.

From a reference file

A reference file exported from another reference manager is read entry by entry, so nothing has to be looked up. Pangur reads BibTeX and BibLaTeX (.bib), RIS (.ris), CSL-JSON (.json), PubMed (.nbib), EndNote’s tagged format (.enw) and Zotero RDF (.rdf): what Zotero, Mendeley, EndNote, JabRef, Papers and Paperpile export, and the zotero.rdf in a Pangur project export.

In the same dialog, choose the file. Each entry appears as a row; entries already in the library are marked and left alone. Tick the ones you want, or import them all, and choose where they go: the library root, a collection, or a new collection named after the file.

• Each reference keeps its authors, title, date, identifiers, tags and abstract.
• Notes in the file (BibTeX annote, RIS N1) are saved as the reference’s notes.
• A BibTeX key is kept as a Citation Key: line in the item’s Extra field, so an export gives the same key back.
• A Zotero RDF file’s collections are rebuilt as they were, nested, and each reference goes into its own. They are made inside the collection you import into, or at the top of the library. A collection already there with the same name and place is used again, so importing the same file twice does not double the tree.
• The PDF search is off for a file of more than 100 entries, so a large import finishes in minutes. Turn it on if you want it, or look afterwards: when the import finishes, Find PDFs searches for the references it saved, on the import’s own row and under Bulk imports in Settings.

A file holds up to 2,000 references and 5 MB; split anything larger. The files an entry names (a BibTeX file field, a RIS L1 line) are on the computer the list came from, so a browser cannot bring them in. The command line can: zcp import refs library.bib --attach-files uploads each one from beside the reference file.

From an item’s references

When a PDF’s Study → References lists citations Pangur found in the text, check the ones to import and save them in batch — the same selectable list as the document workflow.

From a web page

If you have a URL to a bibliography page, the document import can save the page as a Research note and scan it the same way.

Progress and history

Long scans run in the background; the list updates as matches arrive. Settings → Recent imports shows bulk import jobs and their counters (resolved, skipped, failed).

Failed entries stay visible so you can fix a typo, add a DOI by hand, or search for the work in the main bar.

Assistants and CLI

Assistants can drive the same pipeline over MCP (import_search, import_save, and related tools). Text from a reference file passed to scan_references is read the same way as a file chosen here. The CLI documents batch patterns in Command line (zcp).