|
| 1 | +import cloudscraper |
| 2 | +import time |
| 3 | + |
| 4 | +from urllib.parse import unquote |
| 5 | +from flask import Flask, request, Response |
| 6 | + |
| 7 | +scraper = cloudscraper.create_scraper( |
| 8 | + browser={ |
| 9 | + 'browser': 'chrome', |
| 10 | + 'platform': 'windows', |
| 11 | + 'desktop': True |
| 12 | + }, |
| 13 | + delay=1, |
| 14 | + allow_brotli=True |
| 15 | +) |
| 16 | + |
| 17 | + |
| 18 | +def set_user_agent(headers): |
| 19 | + # this needs to match Sec-Ch-Ua (cloudflare will flag as a bot seeing two different user agents) |
| 20 | + # make sure to change it accordingly if you change this |
| 21 | + headers['User-Agent'] = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) ' \ |
| 22 | + 'Chrome/120.0.0.0 Safari/537.36' |
| 23 | + return headers |
| 24 | + |
| 25 | + |
| 26 | +def set_security_headers(headers): |
| 27 | + headers['Sec-Ch-Ua'] = '"Not_A Brand";v="8", "Chromium";v="120", "Google Chrome";v="120"' |
| 28 | + headers['Sec-Ch-Ua-Mobile'] = '?0' |
| 29 | + headers['Sec-Ch-Ua-Platform'] = '"Windows"' |
| 30 | + headers['Sec-Fetch-Dest'] = 'empty' |
| 31 | + headers['Sec-Fetch-Mode'] = 'cors' |
| 32 | + headers['Sec-Fetch-Site'] = 'same-origin' |
| 33 | + return headers |
| 34 | + |
| 35 | + |
| 36 | +def set_origin_and_ref(headers, origin, ref): |
| 37 | + headers['Origin'] = origin |
| 38 | + headers['Referer'] = ref |
| 39 | + return headers |
| 40 | + |
| 41 | + |
| 42 | +def generate_origin_and_ref(url, headers): |
| 43 | + data = url.split('/') |
| 44 | + first = data[0] |
| 45 | + base = data[2] |
| 46 | + c_url = f"{first}//{base}/" |
| 47 | + headers = set_origin_and_ref(headers, c_url, c_url) |
| 48 | + return headers |
| 49 | + |
| 50 | + |
| 51 | +app = Flask(__name__) |
| 52 | + |
| 53 | +HOP_BY_HOP_HEADERS = { |
| 54 | + 'connection', |
| 55 | + 'keep-alive', |
| 56 | + 'proxy-authenticate', |
| 57 | + 'proxy-authorization', |
| 58 | + 'te', |
| 59 | + 'trailers', |
| 60 | + 'transfer-encoding', |
| 61 | + 'upgrade', |
| 62 | +} |
| 63 | + |
| 64 | + |
| 65 | +def clean_headers(response): |
| 66 | + headers = {} |
| 67 | + for name, value in response.headers.items(): |
| 68 | + if name.lower() not in HOP_BY_HOP_HEADERS: |
| 69 | + headers[name] = value |
| 70 | + headers.pop('content-encoding', None) |
| 71 | + headers.pop('content-length', None) |
| 72 | + return headers |
| 73 | + |
| 74 | + |
| 75 | +def generate_proxy_response(response) -> Response: |
| 76 | + content_type = response.headers.get('content-type', '') |
| 77 | + |
| 78 | + if 'text' in content_type or 'html' in content_type: |
| 79 | + content = response.text |
| 80 | + else: |
| 81 | + content = response.content |
| 82 | + |
| 83 | + headers = clean_headers(response) |
| 84 | + |
| 85 | + # For JSON content |
| 86 | + if 'application/json' in content_type: |
| 87 | + return Response(content, status=response.status_code, content_type='application/json') |
| 88 | + |
| 89 | + # For HTML content |
| 90 | + if 'text/html' in content_type: |
| 91 | + return Response(content, status=response.status_code, content_type='text/html; charset=utf-8') |
| 92 | + |
| 93 | + # For all other content types |
| 94 | + return Response( |
| 95 | + content, |
| 96 | + status=response.status_code, |
| 97 | + headers=headers |
| 98 | + ) |
| 99 | + |
| 100 | + |
| 101 | +def get_headers(): |
| 102 | + headers = { |
| 103 | + 'Accept': 'application/json, text/plain, */*', |
| 104 | + 'Accept-Language': 'en-US,en;q=0.9', |
| 105 | + 'Accept-Encoding': 'gzip, deflate, br', |
| 106 | + 'Connection': 'keep-alive' |
| 107 | + } |
| 108 | + headers = set_user_agent(headers) |
| 109 | + headers = set_security_headers(headers) |
| 110 | + return headers |
| 111 | + |
| 112 | + |
| 113 | +def get_proxy_request_url(req, url): |
| 114 | + full_url = unquote(url.strip('"')) |
| 115 | + if req.query_string: |
| 116 | + full_url = f"{full_url}?{req.query_string.decode('utf-8')}" |
| 117 | + return full_url |
| 118 | + |
| 119 | + |
| 120 | +def get_proxy_request_headers(req, url): |
| 121 | + headers = get_headers() |
| 122 | + headers['Accept-Encoding'] = 'gzip, deflate, br' |
| 123 | + |
| 124 | + for header in req.headers: |
| 125 | + if header[0].lower() not in ['host', 'connection', 'content-length']: |
| 126 | + headers[header[0]] = header[1] |
| 127 | + headers = generate_origin_and_ref(url, headers) |
| 128 | + return headers |
| 129 | + |
| 130 | + |
| 131 | +# Cloudflare bypassed request |
| 132 | +@app.route("/api/proxy/<path:url>", methods=["GET"]) |
| 133 | +def handle_proxy(url): |
| 134 | + if request.method == 'GET': |
| 135 | + full_url = get_proxy_request_url(request, url) # parse request url |
| 136 | + headers = get_proxy_request_headers(request, url) # generate headers for the request |
| 137 | + |
| 138 | + try: |
| 139 | + start = time.time() |
| 140 | + response = scraper.get(full_url, headers=headers) |
| 141 | + end = time.time() |
| 142 | + elapsed = end - start |
| 143 | + print(f"Proxied request for {full_url.split('?')[0]} in {elapsed:.6f} seconds") |
| 144 | + response.raise_for_status() |
| 145 | + |
| 146 | + return generate_proxy_response(response) |
| 147 | + |
| 148 | + except Exception as e: |
| 149 | + print(f"Proxy Request Error: {str(e)}") |
| 150 | + return {'error': str(e)}, 500 |
| 151 | + |
| 152 | + |
| 153 | +if __name__ == "__main__": |
| 154 | + print('Starting cloudflare bypass proxy server') |
| 155 | + from waitress import serve |
| 156 | + serve(app, host="0.0.0.0", port=5000) |
0 commit comments