Update rules + add fast path for experimental bundle as well

remusao · Jun 6, 2019 · e2555ba · e2555ba
1 parent 7a7f1f7
commit e2555ba
Show file tree

Hide file tree

Showing 10 changed files with 152 additions and 118 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,6 +2,12 @@
 
 ### Not Released
 
+### 5.1.0
+
+*2019-06-06*
+
+- [#164](https://github.com/remusao/tldts/pull/164) Update rules + add fast path for experimental bundle as well
+
 ### 5.0.3
 
 *2019-05-29*

diff --git a/bench/benchmark.js b/bench/benchmark.js
@@ -3,7 +3,8 @@ const chalk = require('chalk');
 const { URL } = require('url');
 const fs = require('fs');
 const path = require('path');
-const tldts = require('..');
+const tldtsDefault = require('../dist/tldts.cjs.min.js');
+const tldtsExperimental = require('../dist/tldts-experimental.cjs.min.js');
 
 function main() {
   const urls = Array.from(
@@ -31,52 +32,58 @@ function main() {
       .run({ async: false });
   }
 
-  for (const method of [
-    'parse',
-    'getHostname',
-    'getPublicSuffix',
-    'getDomain',
-    'getSubdomain',
-  ]) {
-    console.log(`= ${chalk.bold(method)}`);
-    const fn = tldts[method];
-
-    for (const options of [
-      undefined, // defaults
-      { validateHostname: false },
-      { validateHostname: false, detectIp: false, mixedInputs: false },
+  function run(bundleName, tldts) {
+    console.log(bundleName);
+    for (const method of [
+      'parse',
+      'getHostname',
+      'getPublicSuffix',
+      'getDomain',
+      'getSubdomain',
     ]) {
-      bench(
-        `#${chalk.bold(method)}(url, ${chalk.underline(
-          JSON.stringify(options),
-        )})`,
-        urls,
-        urls => {
-          for (let i = 0; i < urls.length; i += 1) {
-            fn(urls[i], options);
-          }
-        },
-      );
-    }
+      console.log(`= ${chalk.bold(method)}`);
+      const fn = tldts[method];
 
-    for (const options of [
-      undefined, // defaults
-      { validateHostname: false },
-      { validateHostname: false, detectIp: false, extractHostname: false },
-    ]) {
-      bench(
-        `#${chalk.bold(method)}(hostname, ${chalk.underline(
-          JSON.stringify(options),
-        )})`,
-        hostnames,
-        hostnames => {
-          for (let i = 0; i < hostnames.length; i += 1) {
-            fn(hostnames[i], options);
-          }
-        },
-      );
+      for (const options of [
+        undefined, // defaults
+        { validateHostname: false },
+        { validateHostname: false, detectIp: false, mixedInputs: false },
+      ]) {
+        bench(
+          `#${chalk.bold(method)}(url, ${chalk.underline(
+            JSON.stringify(options),
+          )})`,
+          urls,
+          urls => {
+            for (let i = 0; i < urls.length; i += 1) {
+              fn(urls[i], options);
+            }
+          },
+        );
+      }
+
+      for (const options of [
+        undefined, // defaults
+        { validateHostname: false },
+        { validateHostname: false, detectIp: false, extractHostname: false },
+      ]) {
+        bench(
+          `#${chalk.bold(method)}(hostname, ${chalk.underline(
+            JSON.stringify(options),
+          )})`,
+          hostnames,
+          hostnames => {
+            for (let i = 0; i < hostnames.length; i += 1) {
+              fn(hostnames[i], options);
+            }
+          },
+        );
+      }
     }
   }
+
+  run('tldts-experimental', tldtsExperimental);
+  run('tldts', tldtsDefault);
 }
 
 main();
diff --git a/lib/lookup/data/hashes.ts b/lib/lookup/data/hashes.ts
diff --git a/lib/lookup/data/trie.ts b/lib/lookup/data/trie.ts
diff --git a/lib/lookup/fast-path.ts b/lib/lookup/fast-path.ts
@@ -0,0 +1,80 @@
+import { IPublicSuffix, ISuffixLookupOptions } from './interface';
+
+export default function(
+  hostname: string,
+  options: ISuffixLookupOptions,
+  out: IPublicSuffix,
+): boolean {
+  // Fast path for very popular suffixes; this allows to by-pass lookup
+  // completely as well as any extra allocation or string manipulation.
+  if (options.allowPrivateDomains === false && hostname.length > 3) {
+    const last: number = hostname.length - 1;
+    const c3: number = hostname.charCodeAt(last);
+    const c2: number = hostname.charCodeAt(last - 1);
+    const c1: number = hostname.charCodeAt(last - 2);
+    const c0: number = hostname.charCodeAt(last - 3);
+
+    if (
+      c3 === 109 /* 'm' */ &&
+      c2 === 111 /* 'o' */ &&
+      c1 === 99 /* 'c' */ &&
+      c0 === 46 /* '.' */
+    ) {
+      out.isIcann = true;
+      out.isPrivate = false;
+      out.publicSuffix = 'com';
+      return true;
+    } else if (
+      c3 === 103 /* 'g' */ &&
+      c2 === 114 /* 'r' */ &&
+      c1 === 111 /* 'o' */ &&
+      c0 === 46 /* '.' */
+    ) {
+      out.isIcann = true;
+      out.isPrivate = false;
+      out.publicSuffix = 'org';
+      return true;
+    } else if (
+      c3 === 117 /* 'u' */ &&
+      c2 === 100 /* 'd' */ &&
+      c1 === 101 /* 'e' */ &&
+      c0 === 46 /* '.' */
+    ) {
+      out.isIcann = true;
+      out.isPrivate = false;
+      out.publicSuffix = 'edu';
+      return true;
+    } else if (
+      c3 === 118 /* 'v' */ &&
+      c2 === 111 /* 'o' */ &&
+      c1 === 103 /* 'g' */ &&
+      c0 === 46 /* '.' */
+    ) {
+      out.isIcann = true;
+      out.isPrivate = false;
+      out.publicSuffix = 'gov';
+      return true;
+    } else if (
+      c3 === 116 /* 't' */ &&
+      c2 === 101 /* 'e' */ &&
+      c1 === 110 /* 'n' */ &&
+      c0 === 46 /* '.' */
+    ) {
+      out.isIcann = true;
+      out.isPrivate = false;
+      out.publicSuffix = 'net';
+      return true;
+    } else if (
+      c3 === 101 /* 'e' */ &&
+      c2 === 100 /* 'd' */ &&
+      c1 === 46 /* '.' */
+    ) {
+      out.isIcann = true;
+      out.isPrivate = false;
+      out.publicSuffix = 'de';
+      return true;
+    }
+  }
+
+  return false;
+}
diff --git a/lib/lookup/packed-hashes.ts b/lib/lookup/packed-hashes.ts
@@ -1,4 +1,5 @@
 import packed from './data/hashes';
+import fastPathLookup from './fast-path';
 import { IPublicSuffix, ISuffixLookupOptions } from './interface';
 
 /**
@@ -100,9 +101,15 @@ const enum Result {
  */
 export default function suffixLookup(
   hostname: string,
-  { allowIcannDomains, allowPrivateDomains }: ISuffixLookupOptions,
+  options: ISuffixLookupOptions,
   out: IPublicSuffix,
 ): void {
+  if (fastPathLookup(hostname, options, out) === true) {
+    return;
+  }
+
+  const { allowIcannDomains, allowPrivateDomains } = options;
+
   // Keep track of longest match
   let matchIndex = -1;
   let matchKind = Result.NO_MATCH;

diff --git a/lib/lookup/suffix-trie.ts b/lib/lookup/suffix-trie.ts
@@ -1,4 +1,5 @@
 import { exceptions, ITrie, rules } from './data/trie';
+import fastPathLookup from './fast-path';
 import { IPublicSuffix, ISuffixLookupOptions } from './interface';
 
 // Flags used to know if a rule is ICANN or Private
@@ -57,75 +58,8 @@ export default function suffixLookup(
   options: ISuffixLookupOptions,
   out: IPublicSuffix,
 ): void {
-  // Fast path for very popular suffixes; this allows to by-pass lookup
-  // completely as well as any extra allocation or string manipulation.
-  if (options.allowPrivateDomains === false && hostname.length > 3) {
-    const last: number = hostname.length - 1;
-    const c3: number = hostname.charCodeAt(last);
-    const c2: number = hostname.charCodeAt(last - 1);
-    const c1: number = hostname.charCodeAt(last - 2);
-    const c0: number = hostname.charCodeAt(last - 3);
-
-    if (
-      c3 === 109 /* 'm' */ &&
-      c2 === 111 /* 'o' */ &&
-      c1 === 99 /* 'c' */ &&
-      c0 === 46 /* '.' */
-    ) {
-      out.isIcann = true;
-      out.isPrivate = false;
-      out.publicSuffix = 'com';
-      return;
-    } else if (
-      c3 === 103 /* 'g' */ &&
-      c2 === 114 /* 'r' */ &&
-      c1 === 111 /* 'o' */ &&
-      c0 === 46 /* '.' */
-    ) {
-      out.isIcann = true;
-      out.isPrivate = false;
-      out.publicSuffix = 'org';
-      return;
-    } else if (
-      c3 === 117 /* 'u' */ &&
-      c2 === 100 /* 'd' */ &&
-      c1 === 101 /* 'e' */ &&
-      c0 === 46 /* '.' */
-    ) {
-      out.isIcann = true;
-      out.isPrivate = false;
-      out.publicSuffix = 'edu';
-      return;
-    } else if (
-      c3 === 118 /* 'v' */ &&
-      c2 === 111 /* 'o' */ &&
-      c1 === 103 /* 'g' */ &&
-      c0 === 46 /* '.' */
-    ) {
-      out.isIcann = true;
-      out.isPrivate = false;
-      out.publicSuffix = 'gov';
-      return;
-    } else if (
-      c3 === 116 /* 't' */ &&
-      c2 === 101 /* 'e' */ &&
-      c1 === 110 /* 'n' */ &&
-      c0 === 46 /* '.' */
-    ) {
-      out.isIcann = true;
-      out.isPrivate = false;
-      out.publicSuffix = 'net';
-      return;
-    } else if (
-      c3 === 101 /* 'e' */ &&
-      c2 === 100 /* 'd' */ &&
-      c1 === 46 /* '.' */
-    ) {
-      out.isIcann = true;
-      out.isPrivate = false;
-      out.publicSuffix = 'de';
-      return;
-    }
+  if (fastPathLookup(hostname, options, out) === true) {
+    return;
   }
 
   const hostnameParts = hostname.split('.');

diff --git a/package-lock.json b/package-lock.json
diff --git a/package.json b/package.json
@@ -1,7 +1,7 @@
 {
   "name": "tldts",
   "description": "Library to work against complex domain names, subdomains and URIs.",
-  "version": "5.0.3",
+  "version": "5.1.0",
   "homepage": "https://github.com/remusao/tldts",
   "author": "Rémi Berson",
   "contributors": [

diff --git a/publicsuffix b/publicsuffix