Datasets:
File size: 2,904 Bytes
487088f | 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 | """List revivals that have no review yet, with risk flags, so a human reads the risky ones closely.
usage: review_queue.py FINAL_DIR"""
import json, re, sys
from pathlib import Path
F = Path(sys.argv[1] if len(sys.argv) > 1 else "final")
reviews = json.load(open(Path(__file__).parent / "reviews.json"))
REQ = re.compile(r"^[+-]\s*[A-Za-z0-9_.\-\[\]]+\s*(@\s*\S+|(==|>=|<=|~=|<|>)\S*)?\s*$")
SUSP_ADD = re.compile(r"\bexcept\b|\bpass\b|return None|mock|dummy|placeholder|stub|fake|os\.environ\.get\(|file_types|allow_any|skip", re.I)
BAD_OUT = re.compile(r"error|exception|traceback|please (provide|enter|upload|set)|api[ _-]?key|not found|invalid|rate limit|429|unauthori|forbidden|quota", re.I)
rows = []
for d in sorted((F / "repro").iterdir()):
f = d / "case.json"
if not f.exists():
continue
c = json.load(open(f))
if c.get("tier") != "functional" or c["space"] in reviews:
continue
a = next(a for a in c["attempts"] if a["stage"] == "functional")
diff = a.get("diff", "")
files = re.findall(r"^\+\+\+ b/(.+)$", diff, re.M)
code = [l for l in diff.splitlines() if re.match(r"^[+-](?![+-])", l)]
cur = None
code_lines = []
for l in diff.splitlines():
if l.startswith("+++ b/"):
cur = l[6:]
elif re.match(r"^[+-](?![+-])", l) and cur and not cur.endswith(("requirements.txt", "README.md", "packages.txt")):
code_lines.append(l)
probe = next((p for p in a.get("probe") or [] if p.get("ok") and not p.get("helper")), {})
out = str(probe.get("output_summary") or "")
flags = []
if code_lines:
flags.append(f"code:{len(code_lines)}")
if any(SUSP_ADD.search(l) for l in code_lines if l.startswith("+")):
flags.append("SUSPICIOUS-ADD")
if BAD_OUT.search(out):
flags.append("OUTPUT-LOOKS-LIKE-ERROR")
if probe.get("inputs") and any(len(i) > 8 and i in out for i in probe["inputs"]):
flags.append("echo?")
if not probe.get("files") and len(out.strip(" '\"[]()")) < 4:
flags.append("TRIVIAL-OUTPUT")
if (probe.get("endpoint") or "").startswith("fn_index") or probe.get("via") == "legacy":
flags.append("legacy")
rows.append((c["likes"], c["space"], d.name, a.get("lock"), files, flags, code_lines, probe, a.get("diagnosis", "")))
for likes, sid, dn, lock, files, flags, code_lines, probe, diag in sorted(rows, key=lambda r: -r[0]):
print(f"===== {sid} ({likes}) lock={lock['date'] if lock else '-'} files={files} FLAGS={flags or 'none'}")
print(" diag:", diag[:260])
if code_lines:
print(" code:", "\n ".join(code_lines[:14]))
print(f" probe {probe.get('endpoint')} via={probe.get('via')} in={[str(i)[-40:] for i in probe.get('inputs', [])][:3]}")
print(f" out: {str(probe.get('output_summary'))[:240]} files={probe.get('files')}")
print(f"\n{len(rows)} unreviewed revivals")
|