Import a reading list or reference file

Pangur can read a list of references and import the ones you choose. The list might be text you paste, such as a bibliography copied from an article or a reading list from a course page, or it might be a reference file exported from another reference manager. In either case Pangur reads the list, matches each entry to a record, shows you what it found, and imports only the entries you tick.

The two kinds of input are handled a little differently. Pasted text is first saved as an output, which is a document in Research, and Pangur then reads that document for references. The text stays in Research afterwards, so you can come back to it. A reference file, on the other hand, is read entry by entry as records and is never saved as a document.

Pasting a bibliography or reading listCopy Anchor Link

1. On Home, choose From document beside the search bar. The Import references from a document dialog opens.
2. Leave Existing Output at Create one from Markdown or text, or choose an output you already have in Research.
3. Give the new output a Title, or leave it blank and Pangur calls it "Imported notes".
4. Paste the list into Markdown or text.
5. Choose Save Output and find references.

Pangur saves the output and then reads it for references. It recognises the usual layouts of a bibliography and looks each entry up, so that it can tell you whether the work is already in your library and whether it can be imported. When the reading is done, the results appear in the dialog (see Choosing what to import).

An administrator can also switch on a second pass that uses a model to find references in untidy notes. When it is on, it runs in the background after the first reading, and its model use counts against your account's daily model allowance. If it fails, you keep the first reading's results.

Importing a reference fileCopy Anchor Link

1. On Home, choose From document.
2. Choose Choose a file and select the file.

Pangur reads the file and shows its entries as results. It accepts these formats:

Files exported from Zotero, Mendeley, EndNote, JabRef and Papers can all be read, and so can Pangur's own export. A Markdown or text file (.md, .markdown, .txt) chosen here is treated as pasted text. The file can be up to 5 MB.

A reference file carries some details that ordinary importing would lose, and Pangur keeps them. A BibTeX entry's citation key is stored in the item's Extra field as Citation Key: …, so the same key comes back when you export the reference. The collections in a Zotero RDF file are rebuilt in your library, nested as they were, and the dialog tells you how many it will rebuild. Some files also name a PDF on the computer that made them; the browser cannot reach those files, so the web import brings the references without them. The dialog says how many entries name files and gives the command that brings them in: pangur import refs <file> --attach-files.

Choosing what to importCopy Anchor Link

The results list every reference Pangur found, each with a status:

To import:

1. Tick the references you want, or use Select all.
2. Choose Import selected (or Import all when every importable row is ticked).
3. Choose where to file them under File into collection: No collection (library root), an existing collection, or, for a reference file, a New collection named after the file.
4. Tick or untick fetch open-access PDFs. It starts ticked for lists of up to 100 references and unticked above that.
5. Choose the import button again to start.

It is best to import soon after reading the list. Pangur holds the matched records for about an hour, and after that an entry without a DOI or ISBN is skipped as an expired candidate, which means reading the list again.

How a bulk import runsCopy Anchor Link

The import runs in the background on Pangur's server, one reference after another with a short pause between them, so you can leave the page. It appears in Activity while it runs, and when it ends Activity reports how many references were saved, how many PDFs were attached, how many were already in the library and how many failed.

Pangur skips some references and says why: the work is already in the library, the entry is a duplicate within the batch, the entry has no DOI or ISBN, or the project is scoped to one collection and the entry falls outside it. If every reference is skipped, nothing is imported and Pangur lists the reasons.

Each reference is saved as soon as it is done, so an import that stops part-way keeps what it has saved. The Bulk imports table on a project's settings page shows each import's progress. Retry unfinished carries on with the references that failed or were held back by a rate limit, and Find PDFs looks for PDFs for references that a finished import saved without a search (Where PDFs come from).

LimitsCopy Anchor Link

Because only one bulk import runs at a time in a project, Pangur refuses a second one until the first has finished. These figures are defaults that an administrator can change (plans and limits).

The references a PDF citesCopy Anchor Link

To import the works cited by a PDF already in your library, open the item's Study tab and choose References. This is described in References and connections.

Other ways inCopy Anchor Link

• The command line. pangur import refs <file> imports a reference file, with -c or --new-collection to choose a collection, --pdfs or --no-pdfs, --attach-files for linked files and --dry-run to see what would happen. pangur refs scan reads a document or address for references. pangur refs status, pangur refs resume and pangur refs find-pdfs follow and continue an import. See the command-line reference.
• An assistant. The MCP tools scan_references and import_references read a list and import from it. See MCP tools.