How mbuit82/sponge works
Sponge is a modular checker for paper-like propositional proofs. Users write proof files, select a logic, and check the proof; the documented design compiles abstract proof steps into Datalog for axiomatic checking. Logic specifications and derived rules provide the main abstraction layers. The README also documents logic inspection. Source excerpts were unavailable, so the graph limits relationships to those explicitly described in the README; implementation details and unverified calls are omitted.
Main entry points: Core.hs Checker core · Specs.hs Logic specs · ToDatalog Datalog conversion · user_proofs Proof files
Components 12
Proof workflow
- Proof files
user_proofs - Proof parsing
Parse - Checker core
Core.hs
Logic model
- Logic specs
Defs/Specs.hs - Logic components
Defs - Logic display
Commands/ShowLogic.hs
Proof abstraction
- Derived rules
DerivedRules
Datalog checking
- Datalog conversion
ToDatalog - Logic engines
datalog/datalog_engines - Check result
Commands/CheckDatalogOutput.hs
Other components
- Proof author
- Souffle
Connections 11
- Proof author → Proof files: writes proofs
README.md - Proof files → Proof parsing: provides input
- Proof parsing → Checker core: parses proofs
- Checker core → Datalog conversion: compiles proofs
README.md - Logic specs → Logic components: specifies logics
README.md - Checker core → Logic specs: selects logic
- Derived rules → Datalog conversion: abstracts steps
README.md - Datalog conversion → Logic engines: emits Datalog
README.md - Logic engines → Souffle: checked by
README.md - Souffle → Check result: produces output
- Logic specs → Logic display: displayed by