Upload 2 files

Dockerfile

@@ -0,0 +1,16 @@
+FROM python:3.12-slim
+
+WORKDIR /app
+
+ENV PYTHONUNBUFFERED=1 \
+    PIP_NO_CACHE_DIR=1 \
+    ICMR_DATA_DIR=/data \
+    PORT=7860
+
+RUN mkdir -p /data
+
+COPY app.py /app/app.py
+
+EXPOSE 7860
+
+CMD ["python", "/app/app.py"]

app.py

@@ -0,0 +1,131 @@
+#!/usr/bin/env python3
+"""Hugging Face Docker Space bootstrapper."""
+
+import os
+import sys
+import subprocess
+from pathlib import Path
+from urllib.request import Request, urlopen
+from urllib.error import HTTPError, URLError
+
+DATA_REPO = os.getenv("DATA_REPO", "Kzr0xx/icrm-hitek-full-db-mixed")
+INDEX_REPO = os.getenv("INDEX_REPO", "CutehackX/icrm-hitek-full-db-mixed-bucket")
+CODE_REPO = os.getenv("CODE_REPO", "CutehackX/icrm-hitek-full-db-mixed-bucket")
+DATA_DIR = Path(os.getenv("ICMR_DATA_DIR", "/data"))
+PORT = int(os.getenv("PORT", "7860"))
+
+FILES = ["part1.parquet", "part2a.parquet", "part2b_new.parquet"]
+
+IDX = [
+    "idx_phone.0.parquet", "idx_phone.1.parquet", "idx_phone.2.parquet",
+    "idx_phone.3.parquet", "idx_phone.4.parquet", "idx_phone.5.parquet",
+    "idx_phone.6.parquet", "idx_aadhar.0.parquet", "idx_aadhar.1.parquet",
+    "idx_aadhar.2.parquet", "idx_aadhar.3.parquet", "idx_aadhar.4.parquet",
+    "idx_aadhar.5.parquet", "idx_aadhar.6.parquet",
+]
+
+CODE = ["main.py", "build_index.py", "requirements.txt"]
+
+
+def download(repo: str, name: str, destination: Path):
+    """Download directly from Hugging Face resolve/main URL, with resume support."""
+    destination.parent.mkdir(parents=True, exist_ok=True)
+
+    url = f"https://huggingface.co/datasets/{repo}/resolve/main/{name}"
+    token = os.getenv("HF_TOKEN", "").strip()
+    existing = destination.stat().st_size if destination.exists() else 0
+
+    print(f"[DOWNLOAD] {url}", flush=True)
+
+    headers = {"User-Agent": "hf-docker-space-downloader/1.0"}
+    if token:
+        headers["Authorization"] = f"Bearer {token}"
+    if existing:
+        headers["Range"] = f"bytes={existing}-"
+
+    request = Request(url, headers=headers)
+
+    try:
+        with urlopen(request, timeout=120) as response:
+            status = getattr(response, "status", 200)
+
+            # If the server ignores Range, restart the file instead of appending.
+            if existing and status != 206:
+                print("[INFO] Server did not return partial content; restarting download.", flush=True)
+                existing = 0
+
+            mode = "ab" if existing and status == 206 else "wb"
+            total = response.headers.get("Content-Length")
+            total_bytes = int(total) + existing if total and status == 206 else int(total) if total else None
+            downloaded = existing
+
+            with open(destination, mode) as f:
+                while True:
+                    chunk = response.read(16 * 1024 * 1024)
+                    if not chunk:
+                        break
+                    f.write(chunk)
+                    downloaded += len(chunk)
+                    if total_bytes:
+                        pct = downloaded * 100 / total_bytes
+                        print(f"\r  {downloaded:,}/{total_bytes:,} bytes ({pct:.1f}%)", end="", flush=True)
+                    else:
+                        print(f"\r  {downloaded:,} bytes", end="", flush=True)
+
+            print(f"\n[OK] {destination} ({downloaded:,} bytes)", flush=True)
+
+    except HTTPError as e:
+        raise RuntimeError(f"Download failed: HTTP {e.code} for {url}") from e
+    except URLError as e:
+        raise RuntimeError(f"Download failed: {url}: {e.reason}") from e
+
+
+def install_requirements(path: Path):
+    print("[SETUP] Installing API requirements...", flush=True)
+    subprocess.check_call([
+        sys.executable, "-m", "pip", "install", "--no-cache-dir", "-r", str(path)
+    ])
+
+
+def start_api():
+    env = os.environ.copy()
+    env["ICMR_DATA_DIR"] = str(DATA_DIR)
+    print(f"[START] Uvicorn on 0.0.0.0:{PORT}", flush=True)
+    os.execvpe(sys.executable, [
+        sys.executable, "-m", "uvicorn", "main:app",
+        "--host", "0.0.0.0", "--port", str(PORT), "--workers", "1"
+    ], env)
+
+
+def main():
+    DATA_DIR.mkdir(parents=True, exist_ok=True)
+    skip = os.getenv("SKIP_DOWNLOAD", "").lower() in {"1", "true", "yes", "on"}
+
+    if not skip:
+        print("[1/3] Downloading parquet data from Kzr0xx...", flush=True)
+        for name in FILES:
+            download(DATA_REPO, name, DATA_DIR / name)
+
+        print("[2/3] Downloading 14 indexes from CutehackX...", flush=True)
+        for name in IDX:
+            download(INDEX_REPO, name, DATA_DIR / name)
+
+        print("[3/3] Downloading API code from CutehackX...", flush=True)
+        for name in CODE:
+            download(CODE_REPO, name, Path("/app") / name)
+    else:
+        print("[SKIP_DOWNLOAD] Using existing files.", flush=True)
+
+    requirements = Path("/app/requirements.txt")
+    if not requirements.exists():
+        raise FileNotFoundError("/app/requirements.txt was not downloaded.")
+
+    if not Path("/app/main.py").exists():
+        raise FileNotFoundError("/app/main.py was not downloaded.")
+
+    install_requirements(requirements)
+    start_api()
+
+
+if __name__ == "__main__":
+    main()