|
| 1 | +import re |
| 2 | +import dns.resolver |
| 3 | +import requests |
| 4 | + |
| 5 | +BLOCKLIST = "blocklist.txt" |
| 6 | +OUTPUT = "candidates_for_removal.txt" |
| 7 | +IANA_TLDS_URL = "https://data.iana.org/TLD/tlds-alpha-by-domain.txt" |
| 8 | + |
| 9 | +# --- Load valid TLDs from IANA --- |
| 10 | +iana_tlds = { |
| 11 | + line.strip().lower() |
| 12 | + for line in requests.get(IANA_TLDS_URL).text.splitlines() |
| 13 | + if line and not line.startswith("#") |
| 14 | +} |
| 15 | + |
| 16 | +# RFC-ish domain regex (labels, dots, no leading/trailing hyphen) |
| 17 | +DOMAIN_RE = re.compile( |
| 18 | + r"^(?!-)[A-Za-z0-9-]{1,63}(?<!-)(\.[A-Za-z0-9-]{1,63})+$" |
| 19 | +) |
| 20 | + |
| 21 | +MODIFIER_KEYWORDS = ( |
| 22 | + "$removeparam", |
| 23 | + "$csp", |
| 24 | + "$redirect", |
| 25 | + "$rewrite", |
| 26 | + "$removeheader", |
| 27 | + "$header", |
| 28 | + "$cookie", |
| 29 | + "$queryprune", |
| 30 | + "$badfilter", |
| 31 | + "$third-party", |
| 32 | + "$domain", |
| 33 | +) |
| 34 | + |
| 35 | +IP_RE = re.compile(r"^\d{1,3}(\.\d{1,3}){3}$") |
| 36 | + |
| 37 | + |
| 38 | +def extract_domain(line: str) -> str | None: |
| 39 | + line = line.strip() |
| 40 | + |
| 41 | + # Comments / empty |
| 42 | + if not line or line.startswith(("!", "#", "//",)): |
| 43 | + return None |
| 44 | + |
| 45 | + # Pure modifier rules and file rules (no host) |
| 46 | + if line.startswith("$"): |
| 47 | + return None |
| 48 | + |
| 49 | + # File filter and |
| 50 | + if line.startswith(("^", "@@^")): |
| 51 | + return None |
| 52 | + |
| 53 | + # pattern urls |
| 54 | + if "*" in line: |
| 55 | + return None |
| 56 | + |
| 57 | + # Whitelisting rules |
| 58 | + if line.startswith("@@||"): |
| 59 | + return None |
| 60 | + |
| 61 | + # Cosmetic filters |
| 62 | + if "##" in line or "#@" in line: |
| 63 | + return None |
| 64 | + |
| 65 | + # If it contains known modifier keywords but no obvious host pattern, |
| 66 | + # treat it as a non-domain rule (e.g. *$removeparam=...). |
| 67 | + if any(k in line for k in MODIFIER_KEYWORDS): |
| 68 | + if not line.startswith("||") and not line.startswith(("|http://", "|https://")): |
| 69 | + return None |
| 70 | + |
| 71 | + # uBO / Adblock network filters |
| 72 | + |
| 73 | + # ||example.com^ |
| 74 | + if line.startswith("||"): |
| 75 | + core = line[2:] |
| 76 | + core = core.split("^", 1)[0] |
| 77 | + core = core.split("/", 1)[0] |
| 78 | + return core.lower() or None |
| 79 | + |
| 80 | + # |http://example.com^ or |https://example.com^ |
| 81 | + if line.startswith("|http://") or line.startswith("|https://"): |
| 82 | + core = line[1:] # drop leading | |
| 83 | + core = core.split("://", 1)[1] |
| 84 | + core = core.split("^", 1)[0] |
| 85 | + core = core.split("/", 1)[0] |
| 86 | + return core.lower() or None |
| 87 | + |
| 88 | + # Hosts file style: "0.0.0.0 example.com" |
| 89 | + parts = line.split() |
| 90 | + if len(parts) >= 2 and IP_RE.match(parts[0]): |
| 91 | + return parts[1].lower() |
| 92 | + |
| 93 | + # If the line contains a "$" but we got here, it's likely a rule with modifiers |
| 94 | + # but no clear host pattern we support → skip to avoid false positives. |
| 95 | + if "$" in line: |
| 96 | + return None |
| 97 | + |
| 98 | + # Plain domain line (no modifiers, no IP, no cosmetics) |
| 99 | + return line.lower() |
| 100 | + |
| 101 | + |
| 102 | +def is_valid_syntax(domain: str) -> bool: |
| 103 | + return bool(DOMAIN_RE.match(domain)) |
| 104 | + |
| 105 | + |
| 106 | +def has_valid_tld(domain: str) -> bool: |
| 107 | + tld = domain.split(".")[-1] |
| 108 | + return tld.lower() in iana_tlds |
| 109 | + |
| 110 | + |
| 111 | +def resolves(domain: str) -> bool: |
| 112 | + for rtype in ("A", "AAAA", "MX", "CNAME"): |
| 113 | + try: |
| 114 | + dns.resolver.resolve(domain, rtype) |
| 115 | + return True |
| 116 | + except Exception: |
| 117 | + pass |
| 118 | + return False |
| 119 | + |
| 120 | + |
| 121 | +def main(): |
| 122 | + candidates: list[str] = [] |
| 123 | + |
| 124 | + with open(BLOCKLIST, encoding="utf-8") as f: |
| 125 | + for raw in f: |
| 126 | + raw = raw.rstrip("\n") |
| 127 | + dom = extract_domain(raw) |
| 128 | + if dom is None: |
| 129 | + continue # comment / cosmetic / modifier / non-domain rule |
| 130 | + |
| 131 | + reason = None |
| 132 | + if not is_valid_syntax(dom): |
| 133 | + reason = "invalid syntax" |
| 134 | + elif not has_valid_tld(dom): |
| 135 | + reason = "invalid TLD" |
| 136 | + elif not resolves(dom): |
| 137 | + reason = "does not resolve" |
| 138 | + |
| 139 | + if reason: |
| 140 | + candidates.append(f"{raw}") |
| 141 | + |
| 142 | + with open(OUTPUT, "w", encoding="utf-8") as f: |
| 143 | + if candidates: |
| 144 | + f.write("\n".join(candidates) + "\n") |
| 145 | + else: |
| 146 | + f.write("# no candidates\n") |
| 147 | + |
| 148 | + |
| 149 | +if __name__ == "__main__": |
| 150 | + main() |
0 commit comments