About this store
This is a document store holding 21 court opinions, addressed by record number. It publishes no descriptive cataloguing: no case name, court, date, docket or party appears anywhere on this site, and there is no index of those things to search.
Record pages report properties of the stored object — its length, digests, media type, header bytes, the producer string the generating software wrote into it, and the store’s own accession and integrity-check timestamps. None of that describes the contents. Any description of a document must be taken from the document.
The objects are court opinions as distributed by a commercial legal research service, and carry that service’s editorial apparatus — headnotes, synopses, citator flags and star pagination — none of which is the court’s work. Page one is a two-column spread, so text extracted naively from it interleaves the columns and corrupts every field. Separating the court’s words from the reporter’s is part of the work.
Integrity
Objects are verified against their recorded digest on a rolling schedule and the result is reported on each record page as a parse status. A record whose status is not ok is served as held: the store does not repair, substitute or withdraw objects, and it holds no second copy in a different format. Where an object cannot be read, nothing else in this store describes what it contained.
Access
No bulk download, no API, no machine-readable index. Records are reachable by number or by crawling; see robots.txt.