perl-incremental-parsing

Incremental parsing infrastructure for efficient re-parsing of Perl source code in response to document edits, designed for LSP integration.

Overview

This crate provides multiple incremental parsing strategies that minimize re-parsing overhead by reusing unaffected AST subtrees when documents change. It converts LSP textDocument/didChange events into efficient partial re-parses using lexer checkpoints, subtree caching with LRU eviction, and content-based node hashing.

Key Types

• IncrementalState -- Rope-backed document state with lexer/parse checkpoints and the apply_edits entry point
• IncrementalDocument -- Arc<Node>-based document with subtree cache, priority-aware eviction, and per-cycle ParseMetrics
• SimpleIncrementalParser -- Lightweight parser tracking reused vs. reparsed node counts
• CheckpointedIncrementalParser -- Lexer-checkpoint-driven parser with token cache reuse
• AdvancedReuseAnalyzer -- Multi-strategy reuse analyzer (structural, position-shifted, content-updated, aggressive matching)
• DocumentParser -- Enum wrapper (Full | Incremental) for transparent LSP integration
• IncrementalEdit / IncrementalEditSet -- Edit representation with byte-shift arithmetic and batch application

Part of the perl-lsp Workspace

This crate is a Tier 3 member of the tree-sitter-perl-rs workspace. It depends on perl-parser-core, perl-edit, and perl-lexer.

License

Licensed under either of MIT or Apache-2.0 at your option.