criteria: add build_quantity_criteria_parser for physical-quantity criteria

[m-messer]

Moves the physical-quantity criteria grammar out of compareExpressions into lambdafeedback.criteria. First of a stack: #6 and #7 build on it.

Changes

• build_quantity_criteria_parser(reserved_expressions, token_list=, productions=) in the new quantity_grammar.py, with defaults QUANTITY_TOKENS / QUANTITY_PRODUCTIONS. It parses:
  • response matches answer, and =, <=, >=, <, > between quantities;
  • dimension(response) matches dimension(answer).
• It was generate_criteria_parser in compareExpressions' physical-quantity context, and its parse trees are unchanged (operator labels keep their surrounding spaces, e.g. " matches ").
• It takes the same arguments as build_criteria_parser, so either can fill the same slot. reserved_expressions isn't used yet: only response and answer scan as QUANTITY; other text is INPUT.
• Exported from lambdafeedback.criteria; CHANGELOG entry under Added.

Downstream
Additive. compareExpressions can switch its physical-quantity context to this parser.

Verification

• ruff, format and mypy are clean, and all four test suites pass.
• New tests cover each relation, matching dimensions, and free text read as INPUT.

🤖 Generated with Claude Code