app.py

import os, sqlite3, threading
from fastapi import FastAPI, Query
from fastapi.responses import JSONResponse
from fastapi.middleware.cors import CORSMiddleware
from huggingface_hub import hf_hub_download
import uvicorn

app = FastAPI(title="InstaNum API v3")
app.add_middleware(CORSMiddleware, allow_origins=["*"], allow_methods=["GET"])

DB1_PATH = "instadb.db"
DB2_PATH = "Instagram.db"

db1_ready = threading.Event()
db1_error = None

def download_db():
    global db1_error
    try:
        print("Downloading instadb.db from HF Dataset...")
        hf_hub_download(
            repo_id="nonamearyan/instanum",
            filename="instadb.db",
            repo_type="dataset",
            local_dir="."
        )
        print("DB1 ready!")
        db1_ready.set()
    except Exception as e:
        db1_error = str(e)
        print(f"DB1 download failed: {e}")
        db1_ready.set()

threading.Thread(target=download_db, daemon=True).start()

def get_columns(db_path):
    try:
        conn = sqlite3.connect(db_path)
        cur = conn.cursor()
        cur.execute("PRAGMA table_info(users)")
        cols = [r[1] for r in cur.fetchall()]
        conn.close()
        return cols
    except:
        return []

def query_db(db_path, field, val, limit):
    if not os.path.exists(db_path):
        return []
    try:
        cols = get_columns(db_path)
        col = field
        if field == "instagram_id":
            if "instagram_id" in cols:
                col = "instagram_id"
            elif "username" in cols:
                col = "username"
                val = val.lstrip("@")
            else:
                return []
        if col not in cols:
            return []

        conn = sqlite3.connect(db_path)
        conn.row_factory = sqlite3.Row
        cur = conn.cursor()

        if col == "phone":
            cur.execute(
                f"SELECT * FROM users WHERE {col} = ? OR {col} = ? LIMIT ?",
                (val, f"'{val}'", limit)
            )
        elif col == "username":
            clean = val.lstrip("@")
            cur.execute(
                f"SELECT * FROM users WHERE {col} = ? OR {col} = ? LIMIT ?",
                (clean, "@" + clean, limit)
            )
        else:
            cur.execute(f"SELECT * FROM users WHERE {col} = ? LIMIT ?", (val, limit))

        rows = [dict(r) for r in cur.fetchall()]
        conn.close()
        return rows
    except Exception as e:
        print(f"Query error on {db_path}: {e}")
        return []

def normalize_row(row):
    """Normalize row to match insta.js expectations"""
    # Get username from instagram_id or username column
    username = row.get("instagram_id") or row.get("username") or ""
    # Build full name
    fname = row.get("fname") or row.get("name") or ""
    lname = row.get("lname") or ""
    name = f"{fname} {lname}".strip() or None
    # Strip quotes from phone
    phone = row.get("phone") or ""
    phone = phone.replace("'", "").strip() or None

    return {
        "id":       row.get("id"),
        "username": username,
        "name":     name,
        "email":    row.get("email") or None,
        "phone":    phone,
        "address":  row.get("address") or None,
    }

def merge_results(r1, r2, limit):
    seen = set()
    merged = []
    for row in r1 + r2:
        key = (row.get("username", ""), row.get("phone", ""), row.get("email", ""))
        if key not in seen:
            seen.add(key)
            merged.append(row)
        if len(merged) >= limit:
            break
    return merged

@app.get("/")
def root():
    db1_ok = db1_ready.is_set() and not db1_error
    db2_ok = os.path.exists(DB2_PATH)
    return {
        "status": "ready",
        "db1_17M": "ready" if db1_ok else ("error: " + str(db1_error) if db1_error else "loading"),
        "db2_4.9M": "ready" if db2_ok else "not found",
        "endpoints": {
            "/api?username=": "search by instagram username",
            "/api?phone=":    "search by phone",
            "/api?email=":    "search by email",
            "/api?name=":     "search by name",
        }
    }

@app.get("/status")
def status():
    return {
        "db1_ready": db1_ready.is_set(),
        "db1_error": db1_error,
        "db1_exists": os.path.exists(DB1_PATH),
        "db1_columns": get_columns(DB1_PATH),
        "db2_exists": os.path.exists(DB2_PATH),
        "db2_columns": get_columns(DB2_PATH),
    }

@app.get("/api")
def search(
    username: str = Query(None),
    phone:    str = Query(None),
    email:    str = Query(None),
    name:     str = Query(None),
    limit:    int = Query(10, ge=1, le=50)
):
    fields = {
        "instagram_id": "@" + username.lstrip("@") if username else None,
        "phone": phone,
        "email": email,
        "fname": name,
    }
    col, val = next(((k, v) for k, v in fields.items() if v), (None, None))

    if not col:
        return {"success": False, "error": "no_query", "hint": "Use ?username= ?phone= ?email= ?name="}

    val = val.strip()
    r1 = query_db(DB1_PATH, col, val, limit) if db1_ready.is_set() and not db1_error else []
    r2 = query_db(DB2_PATH, col, val, limit)
    raw = merge_results(r1, r2, limit)

    # Normalize for insta.js compatibility
    results = [normalize_row(r) for r in raw]

    return {
        "success": True,
        "count": len(results),
        "sources": {"db1_17M": len(r1), "db2_4.9M": len(r2)},
        "results": results
    }

if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=7860)


# 1774199147