Datasets:
File size: 3,145 Bytes
84a242a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 | #!/usr/bin/env python3
"""Materialize original CSVs for the recovered analysis notebooks; no experiments.
Default: estimate only. --write requires a new or empty destination. Seed 0
is the paper's downstream archive; all three acquisition seeds are available.
"""
import argparse
import collections
import gzip
import hashlib
import json
from pathlib import Path
import tarfile
from read_acquisition_artifact import METHODS, git_blob_id
def main():
p = argparse.ArgumentParser(description=__doc__)
p.add_argument("--root", type=Path, default=Path(__file__).resolve().parents[1])
p.add_argument("--destination", type=Path, required=True)
p.add_argument("--methods", nargs="+", choices=METHODS, default=list(METHODS))
p.add_argument("--folds", nargs="+", type=int, choices=range(5), default=list(range(5)))
p.add_argument("--seeds", nargs="+", type=int, choices=range(3), default=[0])
p.add_argument("--write", action="store_true")
args = p.parse_args()
groups = collections.defaultdict(list)
with gzip.open(args.root / "provenance/pinned_acquisition_files.jsonl.gz", "rt") as f:
for line in f:
rec = json.loads(line)
method, fold, seed = rec["path"].split("/")[:3]
if method in args.methods and int(fold[4:]) in args.folds and int(seed[4:]) in args.seeds:
groups[rec["archive"]].append(rec)
rows = [r for group in groups.values() for r in group]
print(json.dumps({"csv_files": len(rows), "expanded_bytes": sum(r["bytes"] for r in rows),
"write_requested": args.write, "acquisition_seeds": args.seeds}, indent=2))
if not args.write:
return
dest = args.destination.resolve()
if dest.exists() and any(dest.iterdir()):
raise FileExistsError("Use a new or empty destination; existing files will not be overwritten.")
dest.mkdir(parents=True, exist_ok=True)
for archive, group in sorted(groups.items()):
needed = {r["member"] for r in group}
contents = {}
with tarfile.open(args.root / archive, "r:xz") as tf:
for item in tf:
if item.name in needed:
f = tf.extractfile(item)
if f is None:
raise ValueError("Expected a regular CSV member")
contents[item.name] = f.read()
for rec in group:
raw = contents[rec["member"]]
if rec.get("encoding") == "final_ledger_prefix":
raw = b"".join(raw.splitlines(keepends=True)[:rec["data_rows"] + 1])
if git_blob_id(raw) != rec["git_blob"] or hashlib.sha256(raw).hexdigest() != rec["sha256"]:
raise ValueError("Archive integrity mismatch: " + rec["path"])
target = dest / rec["path"]
if not target.resolve().is_relative_to(dest):
raise ValueError("Unsafe artifact path")
target.parent.mkdir(parents=True, exist_ok=True)
with target.open("xb") as f:
f.write(raw)
print("Materialized " + archive, flush=True)
if __name__ == "__main__":
main()
|