redesign

adlist/validation.py

@@ -0,0 +1,90 @@
+"""Pure input validation and supported host-list line formats."""
+import ipaddress
+import re
+from urllib.parse import urlparse
+
+def should_ignore_domain(domain):
+    return domain.startswith(".") or any(
+        ch in domain for ch in ["~", "=", "$", "'", "^", "_", ">", "<", ":"]
+    )
+
+def should_ignore_line(line):
+    return any(sym in line for sym in ["<", ">", "##", "###", "div", "span"])
+
+def is_valid_domain(domain):
+    return bool(re.compile(r"^(?:[a-zA-Z0-9-]+\.)+[a-zA-Z]{2,}$").match(domain))
+
+def convert_host_line(line: str, target_ip: str):
+    # szybkie odrzucenia
+    if not line:
+        return None
+    line = line.strip()
+
+    if not line or line.startswith(("!", "#", "/", ";")):
+        return None
+
+    for sep in (" #", " ;"):
+        idx = line.find(sep)
+        if idx != -1:
+            line = line[:idx].rstrip()
+
+    if not line:
+        return None
+
+    m = re.match(r"^\|\|([a-z0-9.-]+)\^", line, re.IGNORECASE)
+    if m:
+        domain = m.group(1).strip(".")
+        if not should_ignore_domain(domain) and is_valid_domain(domain):
+            return f"{target_ip} {domain}"
+        return None
+
+    parts = line.split()
+
+    if len(parts) >= 2 and (
+        re.match(r"^\d{1,3}(?:\.\d{1,3}){3}$", parts[0]) or ":" in parts[0]
+    ):
+        domain = parts[1].strip().split("#", 1)[0].strip().strip(".")
+        if not should_ignore_domain(domain) and is_valid_domain(domain):
+            return f"{target_ip} {domain}"
+        return None
+
+    m = re.match(r"^(?:address|server)=/([a-z0-9.-]+)/", line, re.IGNORECASE)
+    if m:
+        domain = m.group(1).strip(".")
+        if not should_ignore_domain(domain) and is_valid_domain(domain):
+            return f"{target_ip} {domain}"
+        return None
+
+    token = parts[0].split("#", 1)[0].strip().strip(".")
+    if token and not should_ignore_domain(token) and is_valid_domain(token):
+        return f"{target_ip} {token}"
+
+    return None
+
+def validate_and_normalize_url(url):
+    url = (url or "").strip()
+
+    # prosta sanity-check: usuń CR/LF
+    if any(c in url for c in ("\r", "\n")):
+        raise ValueError("Invalid characters in URL")
+
+    parsed = urlparse(url)
+    if not parsed.scheme:
+        url = f"https://{url}"
+        parsed = urlparse(url)
+
+    # akceptuj tylko http/https
+    if parsed.scheme not in {"http", "https"}:
+        raise ValueError(f"Unsupported scheme: {parsed.scheme}")
+
+    if not parsed.netloc:
+        raise ValueError("Missing host in URL")
+
+    return parsed.geturl()
+
+def validate_ip(value: str) -> str:
+    v = (value or "").strip()
+    try:
+        return str(ipaddress.ip_address(v))
+    except Exception:
+        raise ValueError("Invalid IP address")