Bibliography Entry Counter

Some journals ask for a bibliography entry count before submission, and counting them by hand in a long thesis or review article is the kind of task that begs for automation.

Paste your bibliography (Chicago, APA, MLA, IEEE, Vancouver, Harvard — any format) and the tool counts entries by detecting consistent separator patterns. It splits on blank lines, numbered lists ([1], [2]…), or author-year prefixes depending on what it sees in the text.

False positives happen with broken formatting — a citation split mid-line by a manual line break can be miscounted as two entries. The tool flags any entries that look unusually short or long compared to the median, which is usually where the formatting trouble lives. For most clean bibliographies pasted from Word or Zotero, the count is accurate first try.

The detection heuristic is pragmatic rather than strict. Numbered citations ([1] First Author, [2] Second Author...) are easy — split on bracketed numbers. Author-year systems (APA, Harvard) usually have entries separated by blank lines or paragraph breaks; the parser splits there. Footnote-style citations use superscripts in body text and full citations in footnotes, which is a different beast — those should be counted from the bibliography list at the end, not the inline marks. The parser tries all common patterns and picks the one with the most consistent results.

The pain this exists to solve: you've spent six months on a literature review and the journal you're submitting to wants a count of references in the bibliography. You could open Zotero or EndNote and look at the count there, but maybe you handed in your draft as a Word doc with bibliography compiled by hand. Counting 200+ entries manually with the inevitable miscounts and recounts is the kind of task that makes people lose an hour to something a script can do in a second.

Worked example: paste 87 APA-style references separated by blank lines. The counter reports 87 entries, and flags one entry that's noticeably shorter than the median (the parser thinks it might be a partial entry that got separated from the rest). You investigate and realize that entry is correctly short — it's a software citation. The flag was a false alarm, but the median-length check did exactly what it was supposed to: surface anomalies for human review.

Where it's wrong: hand-formatted bibliographies with inconsistent separators. If half your entries are separated by blank lines and the other half are continuous text with hanging indents that didn't survive copy-paste, the counter will undercount. The fix is upstream — clean up your bibliography in Zotero or paste from a properly formatted source. The tool can't recover information that was lost in formatting, only count what's actually distinguishable as separate entries.