File "probe_final_cpp.py"

Full path: /home/algopkco/public_html/scraper/probe_final_cpp.py
File size: 1.22 B
MIME-type: text/x-script.python
Charset: utf-8

Download   Open   Edit   Advanced Editor &nnbsp; Back

import re

import requests

headers = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0 Safari/537.36"}

probes = [
    ("sanfoundry wp-json", "https://www.sanfoundry.com/wp-json/wp/v2/posts?per_page=5"),
    ("csestudy cpp mcq", "https://cselectricalandelectronics.com/c-plus-plus-mcq-questions/"),
    ("csestudy cpp list", "https://cselectricalandelectronics.com/category/cpp/"),
    ("easyengineering cpp", "https://www.easyengineering.net/c-plus-plus-mcq/"),
    ("gateoverflow cpp", "https://gateoverflow.in/tag/c++/mcq"),
    ("engineeringmcqs feed", "https://engineeringmcqs.blogspot.com/feeds/posts/default?max-results=50"),
    ("examrace cpp", "https://www.examrace.com/IAS-Mains/C-Plus-Plus/"),
]

for name, url in probes:
    try:
        r = requests.get(url, headers=headers, timeout=30)
        txt = r.text[:400].lower()
        ok = r.status_code == 200 and "just a moment" not in txt and "captcha" not in txt
        t = re.search(r"<title>(.*?)</title>", r.text, re.S)
        print(f"{name:28s} {r.status_code:4d} len={len(r.text):7d} ok={ok} | {(t.group(1).strip()[:60] if t else '')}")
    except Exception as e:
        print(f"{name:28s} ERR {type(e).__name__}")