perl-corpus

Test corpus management, property-based generators, and edge case fixtures for Perl parsers.

Part of the perl-lsp workspace.

Overview

perl-corpus provides curated Perl code samples, randomized code generators (via proptest), and a CLI tool for corpus linting, indexing, and statistics. It is used across the workspace for parser testing, LSP feature validation, and DAP variable inspection.

Key Features

• Corpus file parsing: Load and query section-based .txt corpus files with metadata (tags, IDs, flags)
• Edge case fixtures: 100 static edge cases and 32 complex data structure samples, filterable by tag
• Property-based generators: 21 generator categories (regex, heredoc, glob, tie, I/O, expressions, OOP, and more)
• Randomized codegen: Deterministic Perl code generation with seed control and coverage options
• Specialized fixtures: Continue/redo, format statements, glob expressions, tie interface test cases
• CLI tool: perl-corpus lint, perl-corpus index, perl-corpus stats, perl-corpus gen

Usage

use perl_corpus::{parse_dir, find_by_tag, EdgeCaseGenerator, generate_perl_code_with_seed};

// Load and query corpus sections
let sections = parse_dir(std::path::Path::new("test_corpus"))?;
let regex_tests = find_by_tag(&sections, "regex");

// Generate deterministic Perl code
let code = generate_perl_code_with_seed(10, 42);

// Query edge case fixtures
let heredocs = EdgeCaseGenerator::by_tag("heredoc");

License

MIT OR Apache-2.0