Files
IPSExtractor/services/services-count-facturas-medicas-serving/scripts/compare_parity.py
T

60 lines
1.9 KiB
Python

"""Compara baseline Python vs respuesta del microservicio Go."""
from __future__ import annotations
import json
import os
import subprocess
import sys
from pathlib import Path
ROOT = Path(__file__).resolve().parents[1]
SC_ROOT = Path(os.environ.get("SERVICES_COUNT_ROOT", ROOT.parent / "services-count"))
BASELINE = ROOT / "data" / "parity_baseline.json"
API_URL = sys.argv[1] if len(sys.argv) > 1 else "http://localhost:8080/v1/count"
def call_api(pdf_path: Path) -> dict:
result = subprocess.run(
["curl", "-s", "-F", f"files=@{pdf_path}", API_URL],
capture_output=True,
text=True,
check=True,
)
payload = json.loads(result.stdout)
for item in payload.get("results", []):
if item.get("filename") == pdf_path.name:
return item
raise RuntimeError(f"sin resultado para {pdf_path.name}")
def main() -> None:
if not BASELINE.exists():
raise SystemExit("Ejecuta primero: python scripts/parity_baseline.py")
import yaml
config_path = SC_ROOT / "config" / "default.yaml"
pdfs_dir = SC_ROOT / yaml.safe_load(config_path.read_text())["paths"]["pdfs_dir"]
baseline = json.loads(BASELINE.read_text(encoding="utf-8"))
mismatches = 0
for expected in baseline:
pdf = pdfs_dir / expected["filename"]
if not pdf.exists():
print(f"SKIP {expected['filename']} (no existe)")
continue
got = call_api(pdf)
if got.get("service_count") != expected.get("service_count"):
mismatches += 1
print(f"MISMATCH {expected['filename']}: py={expected['service_count']} go={got.get('service_count')}")
else:
print(f"OK {expected['filename']}: count={got.get('service_count')}")
if mismatches:
raise SystemExit(f"{mismatches} diferencias encontradas")
print("Paridad OK")
if __name__ == "__main__":
main()