This script accepts an ANTLR4 grammar file and converts the parser rules within that file into Perses Normal Form (PNF). PNF can be used to assist in test case reduction as presented in:
Perses: Syntax-Guided Program Reduction
Chengnian Sun, Yuanbo Li, Qirun Zhang, Tianxiao Gu, Zhendong Su
ICSE 2018
The resulting grammars should generally be valid ANTLR4 grammars as well.
./antlr2pnf.py <path to .g4 file> <name of starting rule>
This will print out the modified grammar (also in ANTLR4 form).
The modified grammar will have the new starting rule pnf_start.
Minor modifications were made to the process, as the printed algorithm doesn't quite produce grammars in PNF form. Errors in this implementation are solely my own. Where possible, variable names match the names in the paper.
Implementation of quantifying optionals can be slow and a bit dependent the random ordering of rules. If it seems to be taking a long time on a grammar, simply killing the process and rerunning it usually does the trick. This has been tested on a handful of the grammars in the official ANTLR4 grammar repository.