DOI Extractor: Find DOI-Shaped Candidates in Text
Paste reference lists, citations, or document text into the DOI Extractor to find DOI-shaped candidates. It recognizes bare values such as 10.1000/example, doi.org and dx.doi.org URLs, and doi: prefixes. Results include the candidate, a doi.org URL, source line, and surrounding context. The matches are pattern detections, not verified registrations, and the tool does not resolve them or retrieve publication metadata.
Recognized Forms and Result Details
A match must begin with 10., contain 4–9 digits, a slash, and a non-space suffix. The matcher accepts bare values, doi.org or dx.doi.org URLs with an optional protocol and optional www., and doi: prefixes. It removes those prefixes, trims common trailing citation punctuation, and keeps matches in source order. Each result gets an https://doi.org/ URL.
The results table shows DOI, doi.org URL, Line, and Context. The displayed DOI uses its original case unless you enable Normalize (lowercase, strip URL prefix).
Controls and Automatic Extraction
Extraction runs automatically when you edit the input or change either checkbox. There is no Run or Extract button.
- Unique only is enabled by default. It keeps the first occurrence of each lowercase comparison key, so repeated identifiers are collapsed in the displayed list.
- Disable Unique only to keep repeated matches in source order.
- Normalize (lowercase, strip URL prefix) displays each extracted DOI in lowercase. URL and
doi:prefixes are already removed during extraction; the label does not add a separate prefix-removal step.
How to Extract, Review, and Export DOI Candidates
- Paste text into the input area, or select Upload. Uploaded text files can use
.txt,.csv,.json,.log,.bib, or.ris. - You can also drop a file onto the input area. Review the output, count, and results table as the text or options change.
- Use Copy to copy the non-empty output.
- Use Download to save the non-empty output as plain text named
dois.txt.
When no DOI-shaped candidates are found, the result workspace stays visible with a status message, the output remains empty, and the count reports 0 DOIs found. Copy and Download remain disabled while the output is empty. The count reports the number of displayed items.
Limitations and Data-Handling Scope
This extractor performs pattern detection only. It does not contact doi.org or a registry, check registration or resolution, or retrieve titles, authors, or other publication metadata. Selected and dropped files are read as text with FileReader.readAsText(file). The picker lists specific extensions, but dropped files are not extension-filtered; binary, unsupported, or unreadable content may produce no useful matches, and the implementation does not show a user-facing file-read error. It contains no upload request, but that does not establish how other site-wide scripts handle entered text. No explicit input-size, file-size, or character-encoding policy is defined here.
Frequently Asked Questions (FAQs)
What formats of DOI does this tool recognize?
It recognizes DOI-shaped candidates written as bare 10.xxxx/suffix strings, doi.org or dx.doi.org URLs with optional protocol and optional www., and identifiers prefixed with doi:. URL and doi: prefixes are removed during extraction.
How does normalization work?
Extraction already removes URL and doi: prefixes. When Normalize (lowercase, strip URL prefix) is enabled, the displayed DOI is also lowercased.
Does this tool verify or resolve a DOI?
No. It detects DOI-shaped strings only; it does not contact doi.org, resolve links, check registration, or retrieve publication metadata.
What does Unique only do?
With Unique only enabled, repeated identifiers are compared using lowercase keys and only the first occurrence is shown. Disable it to keep repeated matches in source order.