feat: Add symbol parsing, fuzzy search, and syntax-based search

[aaronstevenwhite]

Description

This PR introduces symbol parsing capabilities, fuzzy string matching, and syntax-based search functionality to the Glazing package. These features enable users to parse linguistic identifiers, perform approximate string matching, and search across datasets using syntactic patterns.

Type of Change

• New feature (non-breaking change which adds functionality)
• Bug fix
• Breaking change
• Documentation update

Changes Made

Core Features

• Symbol parsers for all four datasets (FrameNet, PropBank, VerbNet, WordNet)
• Fuzzy string matching using Levenshtein distance algorithm
• Syntax-based search interface for pattern matching
• Transitive cross-reference resolution with MappingIndex

Implementation Details

• Symbol Parsing: Parse identifiers like give.01 (PropBank) or 02756558-n (WordNet)
• Fuzzy Matching: Configurable similarity thresholds with Levenshtein distance
• Syntax Search: Pattern-based search (e.g., "NP V NP PP[to]")
• Cross-References: New MappingIndex class with file-based caching

New Files Added (12 new modules)

• src/glazing/syntax/ - Syntactic pattern models and parser
• src/glazing/utils/fuzzy_match.py - Fuzzy matching utilities
• src/glazing/utils/ranking.py - Result ranking functions
• src/glazing/references/index.py - Mapping index implementation
• src/glazing/symbols.py - Symbol type definitions
• src/glazing/{dataset}/symbol_parser.py - Dataset parsers (4 files)

Documentation Added (18 docs)

• docs/user-guide/fuzzy-search.md - Fuzzy search guide
• docs/user-guide/syntax-search.md - Syntax search guide
• docs/api/{dataset}/symbol-parser.md - API docs (4 files)
• docs/api/utils/fuzzy-match.md - Fuzzy match API
• docs/api/symbols.md - Symbols API
• Updated data-formats.md with JSON schemas

Testing

• 1,333 total tests pass
• 28 test files modified/added
• Type checking: mypy --strict src/ passes
• Linting: ruff check passes
• Formatting: ruff format validated

Dependencies

• python-Levenshtein>=0.20.0 added to dependencies

Docker Support

• Dockerfile added with multi-stage build
• Python 3.13 base image
• Automated data download during build

Code Examples

Symbol Parsing

>>> from glazing.propbank.symbol_parser import parse_roleset_id
>>> result = parse_roleset_id("give.01")
>>> result.lemma
'give'
>>> result.sense_number
1

Fuzzy Search

>>> from glazing.search import UnifiedSearch
>>> searcher = UnifiedSearch()
>>> results = searcher.search_with_fuzzy("comunication", fuzzy_threshold=0.8)
>>> # Results include frames with similar names like "Communication"

Syntax Search

>>> from glazing.search import UnifiedSearch
>>> searcher = UnifiedSearch()
>>> # Search for ditransitive patterns (e.g., "NP gives NP to NP")
>>> results = searcher.search_by_syntax(
...     pattern="NP V NP PP[to]",
...     allow_wildcards=True
... )