Files
rss-proxy/proxy/rss_proxy.py
bacon 30883c2836 update rss_proxy.py
* added check for `CharRef: invalid decimal value`
2025-03-05 10:43:51 +03:00

105 lines
3.7 KiB
Python
Executable File

import urllib.parse
import os
import requests
import redis
import xml.etree.ElementTree as ET
import re
import unicodedata
from flask import request, Response
PROXY_URL = os.getenv("PROXY_URL")
REDIS_URL = os.getenv("REDIS_URL", "redis://localhost:6379/0")
CACHE_TTL = int(os.getenv("CACHE_TTL", 3600))
rdb = redis.from_url(REDIS_URL)
def normalize_text(text):
"""Приводит текст к нормальному виду, устраняя странные символы."""
if text:
return unicodedata.normalize("NFKC", text)
return text
def extract_viewtopic_link(description):
"""Ищет любую ссылку в description"""
match = re.search(r'href="(https?://[^"]+)"', description)
return match.group(1) if match else None
def normalize_xml_texts(elem):
"""Применяет normalize_text ко всем текстовым узлам XML."""
if elem.text:
elem.text = normalize_text(elem.text)
if elem.tail:
elem.tail = normalize_text(elem.tail)
for child in elem:
normalize_xml_texts(child)
def init_proxy(app):
@app.route("/proxy")
def proxy():
"""Proxy RSS feed with per-item caching and GUID replacement."""
raw_query = request.query_string.decode()
if raw_query.startswith("url="):
url = urllib.parse.unquote(raw_query[4:])
else:
return "Missing URL", 400
try:
proxies = {"http": PROXY_URL, "https": PROXY_URL} if PROXY_URL else None
r = requests.get(url, timeout=10, proxies=proxies)
r.encoding = r.apparent_encoding
xml_data = r.text
xml_data = xml_data.replace("&", "&")
if '<?xml version="1.0" encoding="' in xml_data:
xml_data = xml_data.replace(
'<?xml version="1.0" encoding="' + xml_data.split('encoding="')[1].split('"')[0] + '"?>',
'<?xml version="1.0" encoding="UTF-8"?>'
)
if '<?xml version="1.0" encoding="UTF-8"?>' not in xml_data:
xml_data = f'<?xml version="1.0" encoding="UTF-8"?>{xml_data}'
root = ET.fromstring(xml_data)
items = root.findall(".//item")
cached_items = []
new_items = []
for item in items:
guid = item.find("guid")
if guid is None or not guid.text:
continue
cache_key = f"rss:item:{guid.text}"
cached_item = rdb.get(cache_key)
if cached_item:
cached_items.append(cached_item.decode())
else:
description = item.find("description")
if description is not None:
new_guid = extract_viewtopic_link(description.text)
print(new_guid)
if new_guid:
guid.attrib.clear()
guid.text = new_guid
# Нормализуем весь <item>
normalize_xml_texts(item)
item_str = ET.tostring(item, encoding="unicode")
rdb.setex(cache_key, CACHE_TTL, item_str)
new_items.append(item_str)
final_items = cached_items + new_items
response_xml = f'<?xml version="1.0" encoding="UTF-8"?><rss><channel>{"".join(final_items)}</channel></rss>'
return Response(response_xml, content_type="application/xml; charset=utf-8")
except Exception as e:
return f"Error: {e}", 500