← Contents

Importing and Exporting References

The library uses standard CSL-JSON, and files going in and out of the library are handled by Zotero's translators, so the library interoperates with Zotero and other tools.

Supported formats

Import (File > Import…) recognizes these formats (those supported by the bundled Zotero translators): BibTeX (including files written for BibLaTeX), RIS, EndNote XML, Refer/BibIX, RefWorks Tagged, MEDLINE/nbib, PubMed XML, Web of Science Tagged, OVID Tagged, MODS, MARC, MARCXML, MAB2, Citavi 5 XML, CSL JSON, Crossref Unixref XML, Datacite JSON, OpenAlex JSON, RDF (including Zotero RDF), Bibliontology RDF, METS, DSpace Intermediate Metadata, Primo Normalized XML, XML ContextObject and bookmark files. The application judges the format from the content of the file, not from its extension. When a file is not UTF-8, the application reads it with the encoding declared in the file or ISO-8859-1, and says so when the import is done.

When importing, File Handling in the Options window lets you choose whether attachment files referenced by the file are copied into the library or only linked in their original location.

Export (File > Export Library…, or exporting the selection from the context menu of the reference table) can write BibTeX, BibLaTeX, RIS, CSL JSON, Zotero RDF, Bibliontology RDF, MODS, TEI, CSV, EndNote XML, Refer/BibIX, RefWorks Tagged, Unqualified Dublin Core RDF, CFF, COinS, Wikipedia citation templates, Wikidata QuickStatements and others; when only notes are selected there are also two note formats, Markdown and HTML. You can choose the character encoding (UTF-8, UTF-8 without BOM, Western).

Unsupported formats

If you are told that the selected file is not in a supported format, no translator recognized the file and it is not CSL-JSON either. Common reasons:

  • the file is the private database of another program rather than a citation file exported from it (for example, the library file of some program was chosen). Export BibTeX, RIS or CSL JSON from that program and import that;
  • the file is an archive or a PDF. For a PDF use File > Add Attachment, and the application tries to recognize it;
  • the file was damaged in transfer, or only part of it arrived.

CSL-JSON

Each reference in the library is one standard CSL-JSON record. CiteDesk's own data lives only in custom.citedesk, so an exported CSL-JSON file can be imported back unchanged. When importing CSL-JSON:

  • entries without an id, with an invalid id, or with an id repeated earlier in the same batch get a new id, which is reported when the import finishes;
  • entries whose id is already in the library are neither imported nor overwritten;
  • top-level keys that the CSL specification does not define are moved unchanged into custom, so nothing is lost;
  • entries of an unknown type are recorded as a Document, with the original type kept in custom.type.

Importing from the clipboard

File > Import from Clipboard reads the text on the clipboard and judges and imports it by the same rules as Import…; it suits citation text copied from a web page or a literature database.

Exporting the library

After choosing a format in the export window you can choose whether to export notes, whether to Export Files (copying the attachment files out into a new folder), whether to write annotations into the copied PDFs, and the character encoding; which options exist depends on the chosen format. What is exported depends on where you started the export: the selected references, one collection, or the whole library. If the path you choose already contains something, the application moves the old content to the system trash instead of deleting it.