[Parser] Support valid Unicode escapes in IRIs and literals, rejecting malformed sequences, surrogates, and out-of-range codepoints

[remiceres]

Summary

Align ANTLR lexer and parser with SPARQL 1.1 §19.2 grammar for Unicode escapes (\uXXXX and \UXXXXXXXX) across IRIs and string literals, accepting valid hex encodings while strictly rejecting malformed sequences, isolated surrogates, and code points exceeding 0x10FFFF.

Target Branch

• feature/corese-next

Specification & Semantics

According to W3C SPARQL 1.1 Query (§19.2):

• \u must be followed by exactly 4 hexadecimal characters representing a valid Unicode scalar value.
• \U must be followed by exactly 8 hexadecimal characters representing a valid Unicode scalar value.
• Code points greater than U+10FFFF must be rejected with a QuerySyntaxException.
• Isolated surrogate code points in the range [U+D800, U+DFFF] must be rejected with a QuerySyntaxException.
• Hexadecimal escapes inside IRI references (e.g. <\u0078>) and literals must be decoded properly to their Unicode code points during lexical analysis / parsing.
• Note: Scope rules for blank nodes across GRAPH, OPTIONAL, and UNION blocks are explicitly tracked in [Query] Enforce SPARQL 1.1 blank node scope validation across graph patterns #571 and are excluded from this task.

Targeted W3C SPARQL 1.0 / 1.1 Test Cases

• syntax-esc-04.rq (Positive syntax test: <\u0078> :p "xx\u0078")
• syntax-esc-05.rq (Positive syntax test: <\u0078> :p "xx\u0078")

Technical Context & Points to Investigate

1. In Sparql.g4 (or ANTLR grammar / lexer tokens for IRI_REF, STRING_LITERAL*), check how escape sequences are recognized and whether \u / \U escapes in IRI brackets <...> are currently rejected or untreated.
2. In lexical decoding utilities or AST builders, verify validation of scalar ranges (> 0x10FFFF) and surrogate rejection (0xD800 - 0xDFFF).

Acceptance Criteria

[ ] Unit tests covering valid and malformed \u / \U escapes in strings, multiline strings, and IRIs.
[ ] Valid Unicode escapes in IRIs and literals compile without syntax errors.
[ ] Out-of-range values (> 0x10FFFF), isolated surrogates, and malformed hex lengths throw QuerySyntaxException.
[ ] Targeted W3C syntax-esc-04.rq and syntax-esc-05.rq tests pass.
[ ] 0 SonarLint issues and cognitive complexity < 15.

[remiceres]

Résolu et validé sur feature/corese-next via la PR #612 (SparqlIriEscapes, SparqlTermResolver, tests W3C syntax-esc-04 et syntax-esc-05).