Resolve a DOI List to Publication Metadata

A DOI column is useful only when downstream systems can resolve it into consistent title, author, venue, date, and identifier fields. The Crossref DOI Scraper accepts exact DOI strings or full doi.org URLs and returns normalized Crossref records.

Clean and test the input

Remove surrounding whitespace and obvious duplicates before submitting a large list. Begin with known examples.

{
  "queries": [],
  "dois": [
    "10.1038/s41586-021-03819-2",
    "https://doi.org/10.1126/science.abq1158"
  ],
  "maxResultsPerQuery": 20
}

The Actor strips the standard DOI URL prefix and resolves each identifier. The returned DOI should be stored in a normalized lowercase comparison column while retaining the source value separately when auditability matters.

Reconcile rather than overwrite

Join results back to the input DOI and compare existing title, authors, publisher, and date. Do not silently replace curated local metadata: Crossref fields reflect member deposits and can be incomplete or formatted differently.

Create status fields for resolved, missing, conflicting, and review required. Exact DOI equality establishes record identity, but it does not guarantee every descriptive field is correct for your catalog. Open the canonical URL for conflicts.

Preserve authors as structured objects so ORCIDs and affiliations are not lost. Store ISSNs, ISBNs, license links, record creation time, indexing time, and the enrichment run ID. If a batch fails, retry the affected identifiers rather than interpreting absent output as invalid DOI evidence.

For recurring cleanup, save a Task with a manageable batch or trigger runs from a pipeline. Keep source timestamps so later Crossref updates can be distinguished from changes in your transformation code.