parse link-shaped clash config to json obects:

clash_extract_subscription.py

import base64
import ssl
import urllib.parse
import urllib.request

def fetch_clash_subscriptions(link: str, timeout: int = 20) -> dict:
    """
    Fetch a Clash/ClashX subscription and try to interpret it.

    Returns a dict:
      {
        "source_url": <resolved HTTP(S) URL>,
        "type": "clash-yaml" | "uri-list" | "text" | "bytes",
        "items": <list of proxies or URIs if detected>,
        "raw": <string or bytes of the fetched content>
      }
    """
    # 1) Normalize clash:// or clashx:// installer schemes to a real URL
    parsed = urllib.parse.urlparse(link)
    if parsed.scheme in ("clash", "clashx"):
        qs = urllib.parse.parse_qs(parsed.query)
        if not qs.get("url"):
            raise ValueError("Scheme URL missing 'url=' parameter.")
        link = urllib.parse.unquote(qs["url"][0])

    # 2) Fetch
    headers = {"User-Agent": "ClashX-Subscription-Fetcher/1.0"}
    req = urllib.request.Request(link, headers=headers)
    ctx = ssl.create_default_context()
    with urllib.request.urlopen(req, timeout=timeout, context=ctx) as resp:
        data = resp.read()
    result = {"source_url": link, "type": "bytes", "items": [], "raw": data}

    # 3) Try to decode as UTF-8 text
    text = None
    try:
        text = data.decode("utf-8")
    except UnicodeDecodeError:
        text = None

    def try_b64_decode(s: str):
        try:
            # tolerate whitespace/newlines; validate=True enforces strict base64 alphabet
            return base64.b64decode(s.strip(), validate=True)
        except Exception:
            return None

    # 4) If we have text, check formats
    if text is not None:
        # 4a) Base64 subscription (V2Ray-style) -> decode, split into URIs
        b = try_b64_decode(text)
        if b:
            decoded = b.decode("utf-8", errors="ignore")
            lines = [ln.strip() for ln in decoded.splitlines() if ln.strip()]
            uri_prefixes = ("vmess://", "vless://", "ss://", "ssr://", "trojan://", "tuic://", "hysteria2://", "hy2://")
            uris = [ln for ln in lines if ln.startswith(uri_prefixes)]
            if uris:
                return {"source_url": link, "type": "uri-list", "items": uris, "raw": decoded}

        # 4b) Clash YAML heuristics
        if any(tag in text for tag in ("proxies:", "proxy-groups:", "proxy-providers:", "Proxy:")):
            proxies = []
            # Try to parse with PyYAML if present (optional)
            try:
                import yaml  # type: ignore
                cfg = yaml.safe_load(text) or {}
                maybe = cfg.get("proxies") or cfg.get("Proxy")
                if isinstance(maybe, list):
                    proxies = maybe
            except Exception:
                pass
            return {"source_url": link, "type": "clash-yaml", "items": proxies, "raw": text}

        # 4c) Fallback: plain text
        return {"source_url": link, "type": "text", "items": [], "raw": text}

    # 5) Fallback: opaque bytes
    return result

# Example:
subs = fetch_clash_subscriptions("clash://install-config?url=https://YOUR_LINK")