doi-utils-pmb

Validate and normalize a Digital Object Identifier (DOI).

Similar projects

Other projects I found that claim to be able to validate DOIs, normalize them, and format them as URI:

• identifiers-doi: Readme says it's only for ancient node versions. Also its only interface (as of v0.2.1) is .extract(), no obvious way to validate and build a URI.
• doi-utils: Seems rather battle-proven, has lots more tests, and can detect a much wider range of DOI-containing URIs.
  • Unfortunately, in versions ≥ 2.x, the author's ESM-only crusade got to levels where my eslint is no longer able to verify that the dependency can be resolved. So I tried using the latest CJS version, but its export is wrapped in a default property and eslint says I cannot use destructuring in import statements. Thus, effectively, I would need almost as much code just for importing his package (either as ESM or CJS) as I had needed for the crappy draft implementation I made before I went searching for a pre-existing module.

API

This module exports an object that holds these properties and methods:

.doiNsBaseUri

A string with the author's latest recommended prefix for constructing a URI from a bare DOI. Currently, in violation of the DOI handbook, it is: https://doi.org/

.whyNotBareDoi(input)

If input is a string with a bare DOI (10.xxx/xxx), return false. Otherwise, return an Error object describing a failed expectation.

.expectBareDoi(input)

If .whyNotBareDoi(input) has a complaint, throw it. Otherwise, return input.

.toBare(input)

Remove known namespace identifier prefixes from a DOI URI, and validate (.expectBareDoi) the remaining text.

.toUri(doi)

Convert any understood representation of a DOI to the recommended URI for that DOI.

Usage

see test/usage.mjs.

Known issues

• Needs more/better tests and docs.

License

MIT