# unpack_logs.py
# Author: David Niedbala Giraudin, ORCID 0009-0009-1526-1178
# Technical assistance: Claude (Anthropic)
#
# The solver logs and search trees of both programs are shipped as one file, logs_bundle.log
# (the archive may contain at most 99 files).  run() recreates the folder logs/ from it, byte for
# byte, and checks the size and the SHA-256 of every file against the bundle header.  It is called
# by verify_logs.py and drat_spot_check.py; it can also be run alone (no arguments, < 1 s).
# Bundle format: for each file, a line "##### FILE <name> <size in bytes> <sha256>", then exactly
# <size> bytes of content, then one newline.
import os, hashlib
HERE = os.path.dirname(os.path.abspath(__file__))

def run(verbose=False):
    data = open(os.path.join(HERE, 'logs_bundle.log'), 'rb').read()
    out = os.path.join(HERE, 'logs'); os.makedirs(out, exist_ok=True)
    pos, n = 0, 0
    while pos < len(data):
        end = data.index(b'\n', pos)
        head = data[pos:end].decode().split()
        assert len(head) == 5 and head[0] == '#####' and head[1] == 'FILE', 'bad bundle header'
        name, size, digest = head[2], int(head[3]), head[4]
        assert '/' not in name and '\\' not in name and not name.startswith('.'), 'bad file name'
        body = data[end + 1:end + 1 + size]
        assert len(body) == size and hashlib.sha256(body).hexdigest() == digest, 'checksum mismatch: ' + name
        assert data[end + 1 + size:end + 2 + size] == b'\n', 'bad bundle separator'
        with open(os.path.join(out, name), 'wb') as f: f.write(body)
        pos = end + 2 + size; n += 1
    if verbose: print('logs/: %d files recreated, all SHA-256 checked' % n, flush=True)
    return n

if __name__ == '__main__':
    run(True)