Updates the worklist to the current settled-expectation baseline
(72/2461/308/165) and records the twelve fixes landed. The remaining instance
deviations are root-caused individually: the particle clusters need
multi-document schema loading and xsi:schemaLocation honoring under a strict
wildcard (a new capability, the largest single lever); reI is pattern-attribute
whitespace normalization; test/elemT are heterogeneous grab-bags needing
per-case triage; the rest are identity-XPath and wildcard/simple-type edges.
The clean single-root-cause fixes are exhausted; the tail is deliberate work.
Refs #129, #122.
The runner read <expected validity> but ignored <current status>, so it counted
disagreements with W3C-disputed expectations as deviations. The 2006-11-06
archive marks 748 instanceTests (plus schemaTests) status="queried" with the
expected validity contested in W3C Bugzilla (for example reJ11.i: \p{Lu}* over
two uppercase letters, prior expectation valid, current invalid+queried, where
PureXML's accept is defensible). Those are now skipped: neither agreement nor
deviation, since the expectation is not authoritative.
Recalibrated baseline (settled expectations only): valid schemas rejected 74 to
72, invalid schemas accepted 2467 to 2461, valid instances rejected 369 to 338,
invalid instances accepted 210 to 179. About 70 apparent deviations were
disagreements with contested expectations, never bugs.
Refs #129, #122.
XSD 1.0 Datatypes 4.3.1 defines the unit of length only for string, anyURI,
binary, and list; for QName it is implementation-defined. We had counted
characters, rejecting valid QName values. Like Xerces and the XSTS NIST set we
now treat length/minLength/maxLength on QName as always satisfied. Lexical
QName validity is still enforced.
The suite's contributors disagree here (NIST: non-constraining; Microsoft: a
character count), which the spec does not resolve, so a single behavior cannot
satisfy both. Net XSTS: valid instances rejected 547 to 502; 4 Microsoft cases
move to invalid instances accepted (257 to 261), recorded as a deliberate
interpretation in docs/xsts-deviations.md.
Refs #147, #129.
Turns the raw failure list into a worklist driven by understanding each rule,
not by chasing the count. Three categories: (A) genuine correctness bugs in
instance validation worth fixing (union restriction facets ~200, QName length
~36, regex Unicode blocks ~155, duration ordering ~50, particles/int facets);
(B) schema-for-schemas structural validity, the bulk of the 2467 invalid
schemas accepted, unimplemented by design and documented as a deliberate
exclusion; (C) arguable cases needing investigation. Records the evidence
sampled per cluster.
Refs #145, #146, #147, #148.