This release implements shortest and longest trace versions for both pre* and post* algorithms and the combination 'dual*'. Using dual* (-e 3) gives overall the best performance in our benchmarks.
Parsing now supports wildcard rules and vector weights.
Various bug-fixes and algorithmic improvements.
What's Changed
- CMake improvement by @MortenSchou in #30
- Wildcard rules by @MortenSchou in #31
- GitHub Actions build and test workflow by @MortenSchou in #33
- Upload artifacts in GitHub Actions by @MortenSchou in #34
- Add option to include dependencies in the binary by @MortenSchou in #35
- Parse vector weights by @MortenSchou in #36
- Structured json output by @MortenSchou in #37
- Fix bug in post* shortest trace by @MortenSchou in #38
- Fix bug in P-automaton JSON parser by @MortenSchou in #39
- Improve data structure use + various simplification and clean up by @MortenSchou in #40
- Fix PAutomaton to_dot with vector_printer by @MortenSchou in #41
- Add longest trace post* engine + interleaved pre*/post* fixed-point by @MortenSchou in #42
- Fix post* shortest trace (Dijkstra version) by @MortenSchou in #43
- Add file format examples to readme by @MortenSchou in #44
- Add pre* and dual* for shortest trace by @petergjoel in #48
- Lazy pop longest prestar by @MortenSchou and @petergjoel in #47
Full Changelog: v1.0.0...v1.1.0
Contributors
petergjoel and MortenSchou