report.py

import argparse
import subprocess
from pathlib import Path
import yaml
import pandas as pd

from config import Config, Workdir


def make_report(workdir: Workdir):
    print(f"Generating coverage report for {workdir.path}")
    
    corpus = workdir.output() / 'queue'
    covdir = workdir.coverage()

    corpus_files = []

    if corpus.exists():
        corpus_files += list(corpus.iterdir())
    else:
        for subdir in workdir.output().iterdir():
            subcorpus = subdir / 'queue'
            if subcorpus.exists():
                corpus_files += list(subcorpus.iterdir())

    # Iterate over corpus in order of edit time
    corpus_files = sorted(
        [x for x in corpus_files if x.is_file() and not x.name.startswith('.')],
        key=lambda x: x.stat().st_mtime
    )

    start_time = corpus_files[0].stat().st_mtime - 1

    # Metrics
    branch_cov = [0]
    corpus_size = [0]
    time = [0]

    cov = set()
    corp_size = 0

    for file in corpus_files:
        cov_file = covdir / f'{file.name}.cov'
        if not cov_file.exists():
            print('Warning: coverage file not found for', file)
            continue

        with open(cov_file) as f:
            branches = set(f.read().strip().split('\n'))
            cov |= branches
            corp_size += 1

        branch_cov.append(len(cov))
        corpus_size.append(corp_size)
        time.append(file.stat().st_mtime - start_time)

    df = pd.DataFrame({
        'time': time,
        'branch_cov': branch_cov,
        'corpus_size': corpus_size
    })

    df.to_csv(workdir.path / 'report.csv', index=False)

    print('Total branches covered:', len(cov))
    print('Total corpus size:', corp_size)


if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='Generate coverage report for a workdir')
    parser.add_argument('workdir', type=str, help='The workdir to use')
    args = parser.parse_args()

    workdir = Workdir(Path(args.workdir), ensure_exists=True)
    
    make_report(workdir)