Fix XQuery 3.1 casting and numeric comparison compliance by joewiz · Pull Request #6165 · eXist-db/exist

joewiz · 2026-03-22T18:49:24Z

Summary

Push XQuery 3.1 XQTS compliance (HEAD test suite) from 85.3% to 88.8% (+652 tests) by fixing error codes, URI resolution, and parser edge cases.

Casting & type fixes

Implement XQuery 3.1 casting table (Type.isCastable): Pre-validates cast operations. Impossible casts now correctly raise XPTY0004 instead of FORG0001. Fixes ~580 tests.
Fix fn:min/fn:max mixed numeric comparisons: Allow cross-numeric-family comparisons by checking numeric union membership. Fixes ~60 tests.
Fix FOCA0002 for NaN/INF → integer/decimal: NaN/INF casts now correctly raise FOCA0002 instead of FORG0001. Fixes ~44 tests.
Fix xs:boolean → integer subtypes: Route through IntegerValue for proper range validation. Fixes ~5 tests.
Fix generic ERROR codes to FORG0001: StringValue.checkType() and Base64BinaryValueType now use ErrorCodes.FORG0001. Fixes ~58 tests.

fn:json-doc URI resolution

Fix relative URI resolution: Resolve relative URIs against the static base URI before falling back to the database URI prefix. Fixes ~295 tests (misc-JsonTestSuite: 0% → 93%).
Fix FOJS0001 error code: JSON parse errors (JsonParseException) now correctly raise FOJS0001 instead of FOUT1170.

fn:parse-ietf-date improvements

Case-insensitive day/month names: Accept "SAT", "aug", etc.
Single-digit hours: Grammar allows 1-2 digits.
Optional whitespace before timezone: Allow "19:36:01GMT".
24:00:00 normalization: Normalize to 00:00:00 of the next day.
Fixes ~8 tests.

XQTS HEAD Results

Metric	Baseline	After
Pass rate	85.3% (26563/31158)	88.8% (27215/30654)
Failures	3528	2459
Errors	160	157
Net improvement	—	+652 tests

Note: test totals vary between runs due to BrokerPool shutdown timeouts.

Test plan

XQTS HEAD full suite run shows improvement from 85.3% to 88.8%
exist-core unit tests: BUILD SUCCESS (0 failures)
Codacy/PMD local analysis: no new warnings introduced
CI verification

🤖 Generated with Claude Code

Per XPath F&O 3.1 Section 19, casting between types that have no valid conversion path (e.g., xs:time to xs:date, xs:anyURI to xs:hexBinary) should raise XPTY0004, not FORG0001. FORG0001 is reserved for when the cast IS allowed but the specific value is invalid. Add Type.isCastable() implementing the XQuery 3.1 casting table to pre-validate cast operations before attempting them. CastExpression now checks castability first and raises XPTY0004 for impossible casts. CastableExpression returns false immediately for impossible casts. Fixes ~580 XQTS 3.1 test failures in prod-CastExpr and related sets. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

fn:min and fn:max threw FORG0006 when given mixed numeric types (e.g., xs:integer and xs:double) because getCommonSuperType() returned ANY_ATOMIC_TYPE for cross-numeric-family types. Per XQuery 3.1, mixed numeric types should be promoted to a common type before comparison. Add a check: if both types are members of the xs:numeric union, allow the comparison to proceed to the existing numeric promotion code. Fixes ~60 XQTS 3.1 test failures in fn-min and fn-max. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

When casting xs:double NaN, INF, or -INF to xs:integer or xs:decimal, eXist incorrectly raised FORG0001. Per XPath F&O 3.1 Section 4.1.16, FOCA0002 should be raised when the cast value is outside the target type's value space (NaN and infinities have no integer/decimal representation). Fixes ~44 XQTS 3.1 test failures in prod-CastExpr. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

xs:boolean cast to integer subtypes like xs:nonPositiveInteger or xs:negativeInteger hit the default case and threw an incorrect error. Now routes through IntegerValue conversion which properly validates the value against the subtype's range (e.g., true=1 is invalid for xs:nonPositiveInteger, producing the correct FORG0001). Also fixes the default error code from XPTY0004 to FORG0001 for BooleanValue, since any cast that reaches convertTo() has already passed the casting table validation. Fixes ~5 XQTS 3.1 test failures in prod-CastExpr. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

reinhapa · 2026-03-22T19:49:47Z


    public DecimalValue toDecimalValue() throws XPathException {
        if (isNaN() || isInfinite()) {
-            throw conversionError(Type.DECIMAL);


Why no longer using the existing conversionError()method?

reinhapa · 2026-03-22T19:49:59Z


    public IntegerValue toIntegerSubType(final int subType) throws XPathException {
        if (isNaN() || isInfinite()) {
-            throw conversionError(subType);


Why no longer using the existing conversionError()method?

reinhapa · 2026-03-22T19:50:10Z


    public IntegerValue toIntegerValue() throws XPathException {
        if (isNaN() || isInfinite()) {
-            throw conversionError(Type.INTEGER);


Why no longer using the existing conversionError()method?

reinhapa · 2026-03-22T19:53:24Z


    public DecimalValue toDecimalValue() throws XPathException {
        if (isNaN() || isInfinite()) {
-            throw conversionError(Type.DECIMAL);


why was the method call conversionError() removed and code duplication introduced? we could pass in an additional error code for the other usages instead in order to reduce code duplication.

reinhapa · 2026-03-22T19:53:39Z


    public IntegerValue toIntegerValue() throws XPathException {
        if (isNaN() || isInfinite()) {
-            throw conversionError(Type.INTEGER);


same as in toDecimalValue()

reinhapa · 2026-03-22T19:54:55Z


    public IntegerValue toIntegerSubType(final int subType) throws XPathException {
        if (isNaN() || isInfinite()) {
-            throw conversionError(subType);


same as in toDecimalValue()

reinhapa · 2026-03-22T19:55:33Z

+     *
+     * @return true if the cast may succeed (for some values), false if the cast can never succeed
+     */
+    public static boolean isCastable(final int sourceType, final int targetType) {


Please address the the NPath complexity of 8960, current threshold is 200.

DoubleValue: reuse conversionError() with an error code parameter instead of inlining XPathException construction in toDecimalValue(), toIntegerValue(), and toIntegerSubType(). Type.isCastable(): extract primitive casting table lookup into isPrimitiveCastable() and factor repeated type-group checks into isNumericTarget(), isDateTimeTarget(), isGregorianTarget(), and isStringOrUntypedAtomic() helpers. Reduces NPath complexity from 8960 to below the 200 threshold. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

joewiz · 2026-03-22T21:16:10Z

[This response was co-authored with Claude Code. -Joe]

Thanks for the review, @reinhapa! Addressed both points in f92f3e6:

DoubleValue: Added an overloaded conversionError(ErrorCode, int) so the NaN/INF paths reuse the existing helper instead of inlining. The original single-arg conversionError(int) now delegates to it.
Type.isCastable(): Extracted the primitive casting table into isPrimitiveCastable() and factored the repeated type-group checks into isNumericTarget(), isDateTimeTarget(), isGregorianTarget(), and isStringOrUntypedAtomic() helpers. NPath complexity drops from 8960 to below the 200 threshold (confirmed via local Codacy/PMD run).

StringValue.checkType() threw XPathException without an ErrorCodes constant for xs:language, xs:Name, xs:NCName, xs:ID, xs:IDREF, xs:ENTITY, and xs:NMTOKEN validation failures. The runner reported these as exerr:ERROR instead of FORG0001. Base64BinaryValueType.verifyString() had the error code as a string in the message ("FORG0001: Invalid base64 data") instead of using ErrorCodes.FORG0001 as the structured error code parameter. Fixes ~58 XQTS HEAD test failures in prod-CastExpr and misc-CombinedErrorCodes. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Two issues prevented fn:json-doc from working with relative URIs: 1. Relative URIs (no scheme) were unconditionally prefixed with the eXist database URI prefix, bypassing filesystem resolution. Now resolves relative URIs against the static base URI first, matching the behavior expected by the XQTS test suite. 2. JSON parse errors (JsonParseException) were caught by the generic IOException handler and reported as FOUT1170 (source not found) instead of FOJS0001 (JSON syntax error). Now catches JsonParseException separately. Fixes ~295 XQTS HEAD test failures in misc-JsonTestSuite (0% to 93%) and improves fn-json-doc results. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Several IETF date parsing issues: - Day and month names were case-sensitive (rejected "SAT", "aug"). Now uses case-insensitive matching. - Hours required exactly 2 digits but the IETF grammar allows 1-2 (digit digit?). Changed parseInt minimum from 2 to 1. - Whitespace between time and timezone was mandatory but should be optional per the grammar. - Seconds detection relied on whitespace check instead of colon check, failing when timezone immediately followed time. - 24:00:00 was rejected; now normalized to 00:00:00 of the next day. - Timezone is now optional (grammar: (S? timezone)?). Fixes ~8 XQTS HEAD test failures in fn-parse-ietf-date. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

joewiz · 2026-03-24T03:34:52Z

[This response was co-authored with Claude Code. -Joe]

The CDataIntergationTest.cdataWebDavApi failure is unrelated to this PR's changes. The test passes locally on both develop and this branch. This appears to be a flaky CI test (likely timing/port-binding in the embedded WebDAV server).

joewiz · 2026-04-06T23:37:21Z

[This comment was co-authored with Claude Code. -Joe]

Closing — superseded by #6207 (v2/xq31-compliance-fixes).

This work has been consolidated into a clean v2/ branch as part of the eXist-db 7.0 PR reorganization. The new PR includes all commits from this PR plus additional related work, with reviewer feedback incorporated where applicable. See the reviewer guide for the full context.

joewiz and others added 4 commits March 22, 2026 14:47

joewiz requested a review from a team as a code owner March 22, 2026 18:49

reinhapa requested changes Mar 22, 2026

View reviewed changes

joewiz and others added 3 commits March 22, 2026 21:01

reinhapa approved these changes Mar 23, 2026

View reviewed changes

duncdrum approved these changes Mar 23, 2026

View reviewed changes

joewiz mentioned this pull request Apr 6, 2026

Fix XQuery 3.1 compliance: casting, error codes, fn:not, path dedup, format-date, and more #6207

Open

3 tasks

joewiz closed this Apr 6, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Fix XQuery 3.1 casting and numeric comparison compliance#6165

Fix XQuery 3.1 casting and numeric comparison compliance#6165
joewiz wants to merge 8 commits intoeXist-db:developfrom
joewiz:feature/xq31-compliance-push

joewiz commented Mar 22, 2026 •

edited

Loading

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

reinhapa Mar 22, 2026

Uh oh!

joewiz commented Mar 22, 2026

Uh oh!

joewiz commented Mar 24, 2026

Uh oh!

joewiz commented Apr 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

joewiz commented Mar 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Casting & type fixes

fn:json-doc URI resolution

fn:parse-ietf-date improvements

XQTS HEAD Results

Test plan

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

reinhapa Mar 22, 2026

Choose a reason for hiding this comment

Uh oh!

joewiz commented Mar 22, 2026

Uh oh!

joewiz commented Mar 24, 2026

Uh oh!

joewiz commented Apr 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

joewiz commented Mar 22, 2026 •

edited

Loading