Europe PMC Complete Date Range Export

Export every article matching a biomedical topic and an inclusive publication-date range into structured bibliographic metadata. This completes a narrow literature curation window without silently substituting a top-results shortlist.

Input

query defaults to TITLE_ABS:asthma and accepts1–500characters of Europe PMC topic/field/Boolean syntax. publishedFrom/publishedTo are inclusive valid ISO dates,default2026-10-01..2026-10-02. FIRST_PDATE and sort_date/sort_cited are reserved and cannot be embedded in query. maxSafetyItems defaults1000 and accepts1–1000. It is a safety cap,not an output limit: if source hitCount exceeds it,the run fails refine-required,saves count/date coverage metadata,and delivers/charges zero article rows. Narrow the date range to retrieve the complete subset. Invalid types,unknown fields,blank queries,malformed/reversed dates fail before retrieval. Source publication-date descending order and tie order are retained. Ordinary full-range example: TITLE_ABS:"SARS-CoV-2",2023-09-01..2023-09-11.

Output

One row per matching source:id including title,PMID/PMCID/DOI,first publication date,ordered author/collective names,journal,publication types,indexed citations,open-access flag,hasTextMinedAccessions,all returned accession type labels,database cross-reference identifiers,full-text link metadata,source attribution and retrieval time. Missing optional fields are null/empty arrays. No abstract,affiliation,PDF or full-text content is copied. This is a complete matching article set within the accepted safety cap,not every field of the API or a guaranteed stable future corpus. Accessions are source-indexed metadata and may be absent; no linked datasets are fetched. Clinical evidence quality/reuse rights are not inferred. COVERAGE/SUMMARY record source hitCount,date range,selected rows,pages and coverageComplete; they are uncharged key-value metadata. Zero matching article ranges fail with no-matching-results and coverage metadata.

Limits

Up to1000articles in ten100-row cursor pages and32HTTP requests including robots. Honor EMBL-EBI robots crawl-delay10seconds,sequential requests. First page establishes total and confirms all matches fit safety/PPE budget before further acquisition. A changed source count,duplicate identity/cursor,invalid date or short page fails the selected run. Acquisition180seconds,delivery90seconds,request30seconds;5MiB per body,64MiB aggregate,24MiB staged output,memory256MiB. Fixed first-party HTTPS API routes,robots gate,public DNS pinning;no credentials,cookies,browser,proxy or paid API. No automatic retry;blocks/rate-limits/timeouts/structure change fail visibly. All matched source data is staged and validated before any row delivery.

Pricing

Local hypothesis:$0.001 per record-delivered unique article row,no start/trial charge. The complete matched range must fit PPE event allowance; if not,the run fails instead of shortening the range. Recheck allowance before each up-to50-row SDK push;short charge acknowledgement fails. Non-PPE saves results and summaries with uncharged SDK pushes. Source failures and SUMMARY are uncharged. SDK delivery has no full-run transaction or rollback guarantee;earlier successful batches can remain after later delivery failure. Repeated runs charge delivered rows again.

Privacy

Ordered public author/collective names are bibliographic attribution only. Affiliations,researcher IDs,contacts,credentials,cookies and raw response headers are excluded. EMBL-EBI attribution is included;publishers retain rights over abstracts/article content,which are not copied.