run_all_samples.py

# run_all_samples.py
import json
import os
from pathlib import Path
from fastapi.testclient import TestClient
from app import app

# -----------------------------
# HuggingFace dataset URLs
# -----------------------------
SAMPLE_URLS = [
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_1.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_2.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_3.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_4.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_5.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_6.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_7.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_8.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_9.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_10.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_11.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_12.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_13.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_14.pdf",
    "https://huggingface.co/datasets/Sathvik-kota/samples/resolve/main/train_sample_15.pdf",
]

BASE_DIR = Path(__file__).parent
RESULTS_DIR = BASE_DIR / "results"
RESULTS_DIR.mkdir(exist_ok=True)

client = TestClient(app)


def main():
    results = {}

    print("\n🚀 Running extractor on HuggingFace dataset...\n")

    for url in SAMPLE_URLS:
        fname = url.split("/")[-1]

        print(f"🔍 Processing: {fname}")

        resp = client.post("/extract-bill-data", json={"document": url})

        if resp.status_code != 200:
            print(f"❌ Error: {resp.status_code}")
            results[fname] = "API_ERROR"
            continue

        data = resp.json()

        out_file = RESULTS_DIR / f"{fname}.json"
        with open(out_file, "w", encoding="utf-8") as f:
            json.dump(data, f, indent=2)

        item_count = data["data"]["total_item_count"]

        results[fname] = item_count
        print(f"   → Saved: {out_file}")
        print(f"   → Items Extracted: {item_count}")

    print("\n📊 FINAL SUMMARY:")
    print(json.dumps(results, indent=2))


if __name__ == "__main__":
    main()