File size: 8,395 Bytes
b72f98e | 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 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 | #!/usr/bin/env python3
from __future__ import annotations
import argparse
import csv
import json
import re
from datetime import datetime, timezone
from pathlib import Path
from typing import Any, Dict, Iterable, List, Optional
def _safe_str(value: Any) -> str:
if value is None:
return ""
if isinstance(value, (dict, list)):
return json.dumps(value, ensure_ascii=False)
return str(value)
def _extract_why(answer: str) -> str:
text = str(answer or "")
if not text:
return ""
patterns = [
r"\*\*Why\*\*\s*(.*?)(?:\n\s*\*\*Next action\*\*|\Z)",
r"###\s*Why this recommendation\s*(.*?)(?:\n\s*###\s*Next step|\Z)",
r"###\s*Why\s*(.*?)(?:\n\s*###\s*Next|\Z)",
]
for p in patterns:
m = re.search(p, text, flags=re.IGNORECASE | re.DOTALL)
if m:
return re.sub(r"\s+", " ", m.group(1)).strip()
return ""
def _sources_summary(sources: Any) -> str:
if not isinstance(sources, list):
return ""
names: List[str] = []
for src in sources:
if not isinstance(src, dict):
continue
doc = src.get("doc") or src.get("document") or src.get("relative_path") or src.get("href") or src.get("id")
if doc:
names.append(str(doc))
# de-dupe preserve order
seen = set()
out = []
for n in names:
if n in seen:
continue
seen.add(n)
out.append(n)
return " | ".join(out)
def _iter_result_rows(data: Dict[str, Any], source_file: Path) -> Iterable[Dict[str, Any]]:
results = data.get("results")
if not isinstance(results, list):
return
run_generated_at = _safe_str(data.get("generated_at"))
suite_total = _safe_str(data.get("total"))
suite_pass_rate = _safe_str(data.get("pass_rate"))
for r in results:
if not isinstance(r, dict):
continue
assistant = _safe_str(r.get("assistant"))
assistant_preview = _safe_str(r.get("assistant_preview"))
query = _safe_str(r.get("query") or r.get("question") or r.get("message"))
# only include rows where an answer was produced
if not assistant and not assistant_preview:
continue
semantic = r.get("semantic") if isinstance(r.get("semantic"), dict) else {}
timing = r.get("timing_ms") if isinstance(r.get("timing_ms"), dict) else {}
sources = r.get("sources") if isinstance(r.get("sources"), list) else []
issues = r.get("issues") if isinstance(r.get("issues"), list) else []
notes = r.get("notes") if isinstance(r.get("notes"), list) else []
yield {
"source_file": str(source_file),
"run_generated_at": run_generated_at,
"suite_total": suite_total,
"suite_pass_rate": suite_pass_rate,
"case_id": _safe_str(r.get("id")),
"query": query,
"assistant": assistant,
"assistant_preview": assistant_preview,
"why_extracted": _extract_why(assistant),
"domain": _safe_str(r.get("domain")),
"mode": _safe_str(r.get("mode")),
"difficulty": _safe_str(r.get("difficulty")),
"web_assisted": _safe_str(r.get("web_assisted")),
"pass": _safe_str(r.get("pass")),
"grade": _safe_str(r.get("grade")),
"rule_grade": _safe_str(r.get("rule_grade")),
"rule_score": _safe_str(r.get("rule_score")),
"semantic_score": _safe_str(semantic.get("score")),
"semantic_grade": _safe_str(semantic.get("grade")),
"semantic_reason": _safe_str(semantic.get("reason") or semantic.get("notes") or semantic.get("explanation")),
"final_score": _safe_str(r.get("final_score")),
"latency_ms": _safe_str(r.get("latency_ms")),
"timing_total_ms": _safe_str(timing.get("total")),
"timing_ms_json": _safe_str(timing),
"issues": " | ".join(_safe_str(x) for x in issues),
"notes": " | ".join(_safe_str(x) for x in notes),
"source_count": str(len(sources)),
"sources_summary": _sources_summary(sources),
"sources_json": _safe_str(sources),
}
def _iter_cases_rows(data: Dict[str, Any], source_file: Path) -> Iterable[Dict[str, Any]]:
cases = data.get("cases")
if not isinstance(cases, list):
return
for r in cases:
if not isinstance(r, dict):
continue
assistant = _safe_str(r.get("assistant"))
assistant_preview = _safe_str(r.get("assistant_preview"))
query = _safe_str(r.get("query") or r.get("question") or r.get("q") or r.get("message"))
if not assistant and not assistant_preview:
continue
yield {
"source_file": str(source_file),
"run_generated_at": _safe_str(data.get("generated_at")),
"suite_total": "",
"suite_pass_rate": "",
"case_id": _safe_str(r.get("id")),
"query": query,
"assistant": assistant,
"assistant_preview": assistant_preview,
"why_extracted": _extract_why(assistant),
"domain": _safe_str(r.get("domain")),
"mode": _safe_str(r.get("mode")),
"difficulty": _safe_str(r.get("difficulty")),
"web_assisted": _safe_str(r.get("web_assisted")),
"pass": _safe_str(r.get("pass")),
"grade": "",
"rule_grade": "",
"rule_score": "",
"semantic_score": "",
"semantic_grade": "",
"semantic_reason": _safe_str(r.get("weak_reason")),
"final_score": "",
"latency_ms": _safe_str(r.get("latency_ms")),
"timing_total_ms": "",
"timing_ms_json": "",
"issues": "",
"notes": _safe_str(r.get("notes")),
"source_count": _safe_str(r.get("source_count")),
"sources_summary": "",
"sources_json": "",
}
def export_eval_csv(evals_root: Path, out_csv: Path) -> int:
rows: List[Dict[str, Any]] = []
for json_path in sorted(evals_root.rglob("*.json")):
try:
with json_path.open("r", encoding="utf-8") as f:
data = json.load(f)
except Exception:
continue
if not isinstance(data, dict):
continue
rows.extend(_iter_result_rows(data, json_path))
rows.extend(_iter_cases_rows(data, json_path))
# stable ordering for review
def _sort_key(r: Dict[str, Any]):
return (r.get("source_file", ""), str(r.get("case_id", "")))
rows.sort(key=_sort_key)
out_csv.parent.mkdir(parents=True, exist_ok=True)
headers = [
"exported_at",
"source_file",
"run_generated_at",
"suite_total",
"suite_pass_rate",
"case_id",
"query",
"assistant",
"assistant_preview",
"why_extracted",
"domain",
"mode",
"difficulty",
"web_assisted",
"pass",
"grade",
"rule_grade",
"rule_score",
"semantic_score",
"semantic_grade",
"semantic_reason",
"final_score",
"latency_ms",
"timing_total_ms",
"timing_ms_json",
"issues",
"notes",
"source_count",
"sources_summary",
"sources_json",
]
exported_at = datetime.now(timezone.utc).isoformat()
with out_csv.open("w", encoding="utf-8", newline="") as f:
writer = csv.DictWriter(f, fieldnames=headers)
writer.writeheader()
for row in rows:
out = {k: row.get(k, "") for k in headers}
out["exported_at"] = exported_at
writer.writerow(out)
return len(rows)
def main() -> int:
parser = argparse.ArgumentParser(description="Export eval Q/A rows to CSV")
parser.add_argument("--evals-root", default="docs/evals", help="Folder to scan for eval JSON files")
parser.add_argument("--out", default="docs/evals/all_eval_questions_answers.csv", help="Output CSV path")
args = parser.parse_args()
root = Path(args.evals_root)
out = Path(args.out)
count = export_eval_csv(root, out)
print(f"Wrote {count} rows to {out}")
return 0
if __name__ == "__main__":
raise SystemExit(main())
|