development_logs/legacy_src/analyze_samples.py

import sqlite3
import requests
from bs4 import BeautifulSoup
import random
import os

DB_PATH = "data/satellites.db"
OUTPUT_DIR = "data/debug_pages"

def analyze_random_satellites(count=15):
    os.makedirs(OUTPUT_DIR, exist_ok=True)
    
    conn = sqlite3.connect(DB_PATH)
    cursor = conn.cursor()
    
    # Get all China satellites
    satellites = cursor.execute("SELECT satellite_name, url FROM satellites WHERE country_name='China'").fetchall()
    conn.close()
    
    if not satellites:
        print("No satellites found in DB.")
        return

    selected = random.sample(satellites, min(count, len(satellites)))
    
    print(f"Analyzing {len(selected)} random satellites...")
    
    for name, url in selected:
        print(f"\nFetching: {name} | {url}")
        try:
            resp = requests.get(url, headers={"User-Agent": "Mozilla/5.0"})
            resp.raise_for_status()
            soup = BeautifulSoup(resp.text, "lxml")
            
            # 1. Check for specification table (id="satdata")
            satdata = soup.find("table", id="satdata")
            has_satdata = "✅ Found" if satdata else "❌ Missing"
            
            # 2. Check for description
            desc = soup.find("div", id="satdescription")
            has_desc = "✅ Found" if desc else "❌ Missing"
            
            # 3. Check for launch list (id="satlist")
            satlist = soup.find("table", id="satlist")
            has_satlist = "✅ Found" if satlist else "❌ Missing"
            
            print(f"  - SatData: {has_satdata}")
            print(f"  - Desc:    {has_desc}")
            print(f"  - Launch:  {has_satlist}")
            
            # Save HTML for deep inspection if needed
            safe_name = "".join(c for c in name if c.isalnum() or c in (' ', '-', '_')).strip()
            with open(f"{OUTPUT_DIR}/{safe_name}.html", "w", encoding="utf-8") as f:
                f.write(resp.text)
                
        except Exception as e:
            print(f"  ❌ Error: {e}")

if __name__ == "__main__":
    analyze_random_satellites()