Improve pure-Ruby parser caching and literal alternatives

[suleman-uzair]

This PR:

• improves pure-Ruby parser-class cache defaults so recursive grammars can memoize immediately without changing direct atom-level defaults
• scopes parser cache entries by consume_all mode and preserves the Parslet-compatible prefix-success reuse for ordered choice
• never memoizes results that evaluate dynamic {} blocks or write captures, so cache thresholds stay performance-only
• adds internal literal-prefix indexing for large Alternative choices while preserving ordered-choice semantics and complete failure reporting
• keeps unsafe/dynamic branches on the normal scan path and freezes alternatives so the lazily built index cannot go stale
• fixes repetition interval/tree memoization replay, consume-all rechecks, and stale-entry eviction
• improves interval-tree same-start lookup used by interval-cache prefix-success checks
• fixes correctness issues in Slice#hash, Buffer#clear!, and Source::LineCache
• repairs benchmark/benchmark_suite.rb and removes the legacy ffi-hash/ffi-json benchmark plumbing superseded by the unified parsanol-rs parse() API
• updates the benchmark docs and HISTORY.txt to match the current backends and the behavior changes above
• adds cache-threshold, literal-choice, interval-cache, cache-safety, and parser regression coverage

This affects the pure-Ruby parser path only. Native parsing continues through the Rust-backed parser path.

[Copilot]

Pull request overview

This PR refines memoization behavior by introducing a separate adaptive cache threshold fallback for parser-class parsing contexts (defaulting to immediate caching), while keeping the existing atom-level default threshold. It also adds an optimization to Alternative to reduce work for large ordered choices via literal-prefix indexing, plus accompanying specs and a benchmark case to compare cache-threshold behaviors.

Changes:

• Add a :parser_default adaptive cache threshold for unnamed/unnamed-override parser classes (set to 0), while keeping the atom-level default at 1000.
• Adjust cache keying to better separate strict (consume-all) vs prefix behavior while ensuring successful prefix parses are cached consistently.
• Add literal-prefix indexing for large Alternative choices, plus focused specs and a benchmark input/parser to compare cache-threshold approaches.

Reviewed changes

Copilot reviewed 14 out of 14 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
spec/parsanol/source_spec.rb	Adds coverage for Source#remaining.
spec/parsanol/atoms/context_spec.rb	Adds specs for atom vs parser-class cache thresholds and consume-all boundary behavior.
spec/parsanol/atoms/alternative_spec.rb	Adds specs validating literal-prefix indexing behavior and error-detail preservation.
spec/parsanol/atom_results_spec.rb	Adds coverage for cached prefix successes across strict named subexpressions.
lib/parsanol/source.rb	Adds Source#remaining for non-advancing access to the unconsumed input.
lib/parsanol/atoms/context.rb	Introduces :parser_default threshold and refines memoization keying/lookup behavior.
lib/parsanol/atoms/alternative.rb	Adds large-choice literal indexing and lazy choice-error generation.
benchmark/run_all.rb	Adds cache-threshold comparison parsers and input-type selection plumbing.
benchmark/README.md	Documents the new cache-threshold benchmark input type and CLI flag.
benchmark/parsers/cache_threshold_parsanol.rb	Adds a synthetic recursive grammar to compare cache-threshold defaults.
benchmark/inputs/tiny/cache_threshold.txt	Adds tiny cache-threshold benchmark input.
benchmark/inputs/small/cache_threshold.txt	Adds small cache-threshold benchmark input.
benchmark/inputs/medium/cache_threshold.txt	Adds medium cache-threshold benchmark input.
benchmark/inputs/large/cache_threshold.txt	Adds large cache-threshold benchmark input.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 22 out of 22 changed files in this pull request and generated 4 comments.

[Copilot]

Pull request overview

Copilot reviewed 22 out of 22 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 22 out of 22 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 23 out of 23 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 23 out of 23 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 25 out of 25 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 25 out of 25 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 32 out of 32 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 32 out of 32 changed files in this pull request and generated 4 comments.