Skip to content

Latest commit

 

History

History
168 lines (128 loc) · 6.72 KB

CHANGELOG.md

File metadata and controls

168 lines (128 loc) · 6.72 KB

Change Log

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog and this project adheres to Semantic Versioning.

Be aware that this project is still v0.y.z which means that anything can change anytime:

"4. Major version zero (0.y.z) is for initial development. Anything MAY change at any time. The public API SHOULD NOT be considered stable."

(Semantic Versioning Specification)

Indicating incompatible changes on major version zero

We defined for this project that while being on major version zero we mark incompatible changes with new minor version numbers. Please note that this is no version handling covered by Semver.

0.8.0 - Not release yet

  • Refactor internal structs

    • Refactored CompiledNfa to CompiledDfa for clarity
    • Refactored ScannerNfaImpl to ScannerImpl for clarity

  • Minimization of the CompiledDfa for enhanced scanning performance

  • Introduce the feature regex_automata. Using the default feature set, which is actually empty, usually results in a slower scanner, but it is faster at compiling the regexes. The regex_automata feature is faster at scanning the input, but it is possibly slower at compiling the regexes. This depends on the size of your scanner modes, i.e. the number of regexes you use.

    Both features are mutually exclusive.

    Using the default feature set is straight forward:

    scnr = "0.8.0"

    For the feature regex_automata to be enabled use this variant:

    scnr = { version = "0.8.0", default-features = false, features = [ "regex_automata" ] }

0.7.1 - 2025-01-21

  • Handle iterator exhaustion in FindMatchesImpl as partial fix for jsinger67/parol#558. This fix updates the line offset vector when a newline character immediately precedes the end of input.

0.7.0 - 2025-01-17

  • Improved benchmarking with throughput measurements
  • Performance: several optimizations
    • CompiledNfa is now able to handle multiple terminals in one automaton per scanner mode
    • To make creation feasible a new struct MultiPatternNfa was introduced as intermediate creation step
    • Each Nfa contains a Pattern now instead of a plain String to have terminal ids and optional lookahead data available
    • Introduced a ScannerCache that saves time when the same scanner is built multiple times during the lifetime of a parser process
  • Two public methods on the Scanner struct have changed their arguments, thus this release is potentially breaking

0.6.0 - 2024-12-14

  • performance: Refactor NFA end states representation to use a boolean vector for accepting states
  • doc: Update UML diagram
  • Simplify internal structure and use the fact that only one implementation of ScannerImplTrait exists. The trait has lost its purpose.
  • test: Add hundreds of match tests in integration test match_test.

0.5.0 - 2024-11-24

  • Cleanup of the library
  • Removed DFA scanner variant and making NFA implementation the default, hence use_nfa() on ScannerBuilder is no more needed and was removed
  • Update of documentation
  • Performance optimization of scanner phase of compiled NFA

0.4.0 - 2024-11-12

  • Support for lookahead, negative and positive. Please see README.md for details.
  • Support for Scanners based on NFAs. These scanners can handle overlapping character classes. Call use_nfa() on the scanner builder before calling build().
let scanner = ScannerBuilder::new()
    .add_scanner_modes(&*MODES)
    .use_nfa()
    .build()
    .unwrap();
let find_iter = scanner.find_iter(INPUT).with_positions();
let matches: Vec<MatchExt> = find_iter.collect();

0.3.3 - 2024-10-11

  • Provide an iterator adapter WithPositions to convert the iterator over type Match to an iterator over types MatchExt which contains line and column information for the start position as well as the end position of each match.
let scanner = ScannerBuilder::new().add_scanner_modes(&*MODES).build().unwrap();
let find_iter = scanner.find_iter(INPUT).with_positions();
let matches: Vec<MatchExt> = find_iter.collect();

  • Fixed handling of current scanner mode. There was a bug that scanner mode switching from the outside had no effect on cloned ScannerImpl instances. This was fixed by removing the mode from the Scanner and leaving it only on the ScannerImpl.

  • We also allow now to set the scanner mode on a FindMatches and even on a WithPositions by implementing the new trait ScannerModeSwitcher for both of them.

  • Add some documentation like PlantUML overview diagram to the doc folder. Also moved matching_state.dot into this folder to have anything in one place. For viewing the PlantUML diagram in Visual Studio Code I recommend the excellent PlantUML extension. Let me add that this overview diagram is in no way complete. It should just give a rough overview.

0.3.2 - 2024-09-09

  • Performance: Scanner no more holds ScannerImpl in a Rc<RefCell<>> to save time during creation of a new find_iter. Instead ScannerImpl is now Clone by wrapping the match functions array in an Arc. This makes the Scanner usable as static global again and has the same effect regarding performance.
  • Scanner::mode_name returns a Option<&str> again, instead of Option<String> which saves an additional heap allocation.

0.3.1 - 2024-09-07

  • Add support for lots of unicode named classes like XID_Start and XID_Continue by the help of the seshat-unicode crate
  • Performance: Scanner holds ScannerImpl in a Rc<RefCell<>> to save time during creation of a new find_iter
  • Add support for generating compiled DFAs as DOT files to scanner implementation

0.3.0 - 2024-08-29

Breaking changes

  • Renamed Scanner::trace_compiled_dfa_as_dot to Scanner::log_compiled_dfas_as_dot

Non-breaking changes

  • Fixed some help comments
  • Fixed the Display implementation of DFA
  • Added a new test to module internal::match_function
  • Added new function FindMatches::with_offset to support resetting the input test
  • Added new function FindMatches::offset to retrieve the total offset of the char indices iterator in bytes.

0.2.0 - 2024-08-19

Breaking changes

  • Scanner::find_iter now returns a FindMatches directly instead of Result<FindMatches> because the construction is basically infallible.

Non-breaking changes

  • Add a new API add_patterns to ScannerBuilder to support simple use cases with only one scanner state.
  • Add derive Debug trait to Scanner
  • Add CHANGELOG

0.1.1 - 2024-08-17

  • Changed description in Cargo.toml

0.1.0 - 2024-08-17

  • First release