Supported reference formats
Run your searches in the databases you already use, export the results, and bring them into Littview. Each upload becomes a library in the project, so you always know which search produced which records — and full-text PDFs can arrive whenever you have them, references first.
Littview imports references in four formats: BibTeX, RIS, PubMed MEDLINE, and EndNote XML. Every major database and reference manager exports at least one of them, so nothing needs retyping.
The four formats
Section titled “The four formats”| Format | Extension | Typical sources |
|---|---|---|
| BibTeX | .bib |
Zotero, JabRef, Google Scholar, Mendeley |
| RIS | .ris |
Scopus, Web of Science, EndNote, ProQuest, most publisher sites |
| PubMed MEDLINE | .txt |
PubMed’s Save → Format: PubMed export |
| EndNote XML | .xml |
EndNote’s File → Export with the XML output style |
If your database offers several export options, any of the four works — pick whichever is listed. For PubMed specifically, use the PubMed format (the MEDLINE-style text file), not the CSV or citation-manager options.
How detection works
Section titled “How detection works”Littview identifies the format by reading the file’s content, not its extension. The first non-empty line decides it:
- Starts with
@(such as@article{) → BibTeX - Starts with
TY -→ RIS - Starts with
PMID-→ PubMed MEDLINE - Starts with
<?xmland the file contains a<records>element → EndNote XML
This means a correctly exported file always imports, even if it was renamed — and a file with the right extension but the wrong content (for example a CSV renamed to .ris) is rejected as unsupported. If a file fails detection, re-export it rather than editing the extension; How to fix import errors covers the messages you might see.
What gets captured
Section titled “What gets captured”All four formats are normalized to the same internal record, so it makes no difference which one you use. For each entry Littview captures the title, authors, year, journal, volume, issue, pages, DOI, URL, abstract, keywords, and publisher — whatever of those the export actually contains. Records with sparse metadata still import; a missing abstract or DOI just stays empty on the reference.
Files upload in batches of up to 20 at a time, 10 MB each — more than enough headroom, since even a several-thousand-record export is typically a few megabytes. The next article walks through the upload itself: How to import references from a file.