Parse reject with IP support [skip ci]

This commit is contained in:
SukkaW 2025-06-06 23:23:59 +08:00
parent 74654d1886
commit b1592051ac
2 changed files with 27 additions and 7 deletions

View File

@ -10,12 +10,9 @@ export type TldTsParsed = ReturnType<typeof tldts.parse>;
/**
* Skipped the input non-empty check, the `domain` should not be empty.
*/
export function fastNormalizeDomainWithoutWww(domain: string, parsed: TldTsParsed | null = null) {
export function fastNormalizeDomainWithoutWwwNoIP(domain: string, parsed: TldTsParsed | null = null) {
// We don't want tldts to call its own "extractHostname" on ip, bail out ip first.
// Now ip has been bailed out, we can safely set normalizeTldtsOpt.detectIp to false.
if (isProbablyIpv4(domain) || isProbablyIpv6(domain)) {
return null;
}
// This function won't run with IP, we can safely set normalizeTldtsOpt.detectIp to false.
parsed ??= tldts.parse(domain, normalizeTldtsOpt);
// Private invalid domain (things like .tor, .dn42, etc)
@ -33,6 +30,19 @@ export function fastNormalizeDomainWithoutWww(domain: string, parsed: TldTsParse
return parsed.hostname;
}
/**
* Skipped the input non-empty check, the `domain` should not be empty.
*/
export function fastNormalizeDomainWithoutWww(domain: string, parsed: TldTsParsed | null = null) {
// We don't want tldts to call its own "extractHostname" on ip, bail out ip first.
// Now ip has been bailed out, we can safely set normalizeTldtsOpt.detectIp to false.
if (isProbablyIpv4(domain) || isProbablyIpv6(domain)) {
return null;
}
return fastNormalizeDomainWithoutWwwNoIP(domain, parsed);
}
/**
* Skipped the input non-empty check, the `domain` should not be empty.
*/

View File

@ -6,7 +6,8 @@ import { createRetrieKeywordFilter as createKeywordFilter } from 'foxts/retrie';
import { looseTldtsOpt } from '../../constants/loose-tldts-opt';
import tldts from 'tldts-experimental';
import { NetworkFilter } from '@ghostery/adblocker';
import { fastNormalizeDomain, fastNormalizeDomainWithoutWww } from '../normalize-domain';
import { fastNormalizeDomain, fastNormalizeDomainWithoutWww, fastNormalizeDomainWithoutWwwNoIP } from '../normalize-domain';
import { isProbablyIpv4, isProbablyIpv6 } from 'foxts/is-probably-ip';
const enum ParseType {
WhiteIncludeSubdomain = 0,
@ -14,6 +15,7 @@ const enum ParseType {
BlackAbsolute = 1,
BlackIncludeSubdomain = 2,
ErrorMessage = 10,
BlackIP = 20,
Null = 1000,
NotParsed = 2000
}
@ -230,7 +232,15 @@ export function parse($line: string, result: [string, ParseType], includeThirdPa
&& filter.isPlain() // isPlain() === !isRegex()
&& (!filter.isFullRegex())
) {
const hostname = fastNormalizeDomainWithoutWww(filter.hostname);
// We don't want tldts to call its own "extractHostname" on ip, bail out ip first.
// Now ip has been bailed out, we can safely set normalizeTldtsOpt.detectIp to false.
if (isProbablyIpv4(filter.hostname) || isProbablyIpv6(filter.hostname)) {
result[0] = filter.hostname;
result[1] = ParseType.BlackIP;
return result;
}
const hostname = fastNormalizeDomainWithoutWwwNoIP(filter.hostname);
if (!hostname) {
result[1] = ParseType.Null;
return result;