Switch to a complete UTS #46 implementation

Author: jdesrosiers

.npmignore

@@ -2,6 +2,5 @@
 eslint.config.js
 **/*.test.js
 tsconfig.json
-documents/
 scratch/
 TODO*

documents/idna2008-limitations.md

@@ -27,7 +27,6 @@
   "devDependencies": {
     "@stylistic/eslint-plugin": "*",
     "@types/node": "*",
-    "@types/tr46": "*",
     "eslint-import-resolver-typescript": "*",
     "eslint-plugin-import": "*",
     "json-schema-test-suite": "github:json-schema-org/json-schema-test-suite",
@@ -37,6 +36,6 @@
   },
   "dependencies": {
     "@hyperjump/uri": "^1.3.2",
-    "tr46": "^6.0.0"
+    "idn-hostname": "^15.1.2"
   }
 }

package.json

@@ -1,7 +1,5 @@
 /**
  * @module
- *
- * @document ../documents/idna2008-limitations.md
  */
 
 // JSON Schema Validation - Dates, Times, and Duration
@@ -13,7 +11,7 @@ export { isIdnEmail } from "./rfc6531.js";
 
 // JSON Schema Validation - Hostnames
 export { isHostname } from "./rfc1123.js";
-export { isAsciiIdn, isIdn } from "./idna2008.js";
+export { isAsciiIdn, isIdn } from "./uts46.js";
 
 // JSON Schema Validation - IP Addresses
 export { isIPv4 } from "./rfc2673.js";

src/index.js

@@ -63,23 +63,7 @@ await testSuite("draft2020-12/optional/format/email", (email) => typeof email !=
 await testSuite("draft6/optional/format/hostname", (hostname) => typeof hostname !== "string" || isHostname(hostname));
 await testSuite("draft2020-12/optional/format/hostname", (hostname) => typeof hostname !== "string" || isAsciiIdn(hostname));
 await testSuite("draft2020-12/optional/format/idn-email", (email) => typeof email !== "string" || isIdnEmail(email));
-await testSuite("draft2020-12/optional/format/idn-hostname", (hostname) => typeof hostname !== "string" || isIdn(hostname), {
-  "validation of internationalized host names": new Set([
-    "contains illegal char U+302E Hangul single dot tone mark",
-    "Exceptions that are DISALLOWED, right-to-left chars",
-    "Exceptions that are DISALLOWED, left-to-right chars",
-    "MIDDLE DOT with no preceding 'l'",
-    "MIDDLE DOT with nothing preceding",
-    "MIDDLE DOT with no following 'l'",
-    "MIDDLE DOT with nothing following",
-    "Greek KERAIA not followed by Greek",
-    "Greek KERAIA not followed by anything",
-    "Hebrew GERESH not preceded by anything",
-    "Hebrew GERSHAYIM not preceded by anything",
-    "KATAKANA MIDDLE DOT with no Hiragana, Katakana, or Han",
-    "KATAKANA MIDDLE DOT with no other characters"
-  ])
-});
+await testSuite("draft2020-12/optional/format/idn-hostname", (hostname) => typeof hostname !== "string" || isIdn(hostname));
 await testSuite("draft2020-12/optional/format/ipv4", (ip) => typeof ip !== "string" || isIPv4(ip));
 await testSuite("draft2020-12/optional/format/ipv6", (ip) => typeof ip !== "string" || isIPv6(ip));
 await testSuite("draft2020-12/optional/format/iri-reference", (iri) => typeof iri !== "string" || isIriReference(iri));

src/json-schema-test-suite.test.js

@@ -1,4 +1,4 @@
-import { isIdn } from "./idna2008.js";
+import { isIdn } from "./uts46.js";
 
 const ucschar = `[\\u{A0}-\\u{D7FF}\\u{F900}-\\u{FDCF}\\u{FDF0}-\\u{FFEF}\\u{10000}-\\u{1FFFD}\\u{20000}-\\u{2FFFD}\\u{30000}-\\u{3FFFD}\\u{40000}-\\u{4FFFD}\\u{50000}-\\u{5FFFD}\\u{60000}-\\u{6FFFD}\\u{70000}-\\u{7FFFD}\\u{80000}-\\u{8FFFD}\\u{90000}-\\u{9FFFD}\\u{A0000}-\\u{AFFFD}\\u{B0000}-\\u{BFFFD}\\u{C0000}-\\u{CFFFD}\\u{D0000}-\\u{DFFFD}\\u{E1000}-\\u{EFFFD}]`;
 
@@ -47,10 +47,6 @@ const mailboxPattern = new RegExp(`^${mailbox}$`, "u");
  * The 'idn-email' format. Validates that a string represents an email as
  * defined by the "Mailbox" ABNF rule in [RFC 6531, section 3.3](https://www.rfc-editor.org/rfc/rfc6531.html#section-3.3).
  *
- * **NOTE**: Validation of the domain has some minor limitations. The spec
- * requires the domain be a valid IDNA2008 IDN. See {@link !"IDNA2008
- * Limitations" | IDNA2008 Limitations} for details.
- *
  * @see [JSON Schema Core, section 7.3.2](https://json-schema.org/draft/2020-12/draft-bhutton-json-schema-validation-01#section-7.3.2)
  *
  * @param {string} email

src/rfc6531.js

@@ -1,15 +1,5 @@
-import tr46 from "tr46";
-
-const label = `[A-Za-z0-9-]{1,63}`;
-const domain = `${label}(?:\\.${label})*`;
-
-const domainPattern = new RegExp(`^${domain}$`);
-
-const parserOptions = {
-  checkBidi: true,
-  checkJoiners: true,
-  checkHyphens: true
-};
+import { isIdnHostname } from "idn-hostname";
+import { isHostname } from "./rfc1123.js";
 
 /**
  * The 'hostname' format since draft-07. Validates that a string represents an
@@ -19,38 +9,28 @@ const parserOptions = {
  * **NOTE**: The 'hostname' format changed in draft-07. Use {@link isHostname}
  * for draft-06 and earlier.
  *
- * **WARNING**: This function can't completely validate IDNA2008 hostnames. See
- * {@link !"IDNA2008 Limitations" | IDNA2008 Limitations} for details.
- *
  * @see [JSON Schema Core, section 7.3.3](https://json-schema.org/draft/2020-12/draft-bhutton-json-schema-validation-01#section-7.3.3)
  *
  * @param {string} hostname
  * @returns {boolean}
  */
 export const isAsciiIdn = (hostname) => {
-  return domainPattern.test(hostname)
-    && hostname.length < 256
-    && !tr46.toUnicode(hostname, parserOptions).error;
+  return isHostname(hostname) && isIdn(hostname);
 };
 
 /**
  * The 'idn-hostname' format. Validates that a string represents an IDNA2008
  * internationalized domain name as defined by [RFC 5890, section 2.3.2.1](https://www.rfc-editor.org/rfc/rfc5890.html#section-2.3.2.1).
  *
- * **WARNING**: This function can't completely validate IDNA2008 hostnames. See
- * {@link !"IDNA2008 Limitations" | IDNA2008 Limitations} for details.
- *
  * @see [JSON Schema Core, section 7.3.3](https://json-schema.org/draft/2020-12/draft-bhutton-json-schema-validation-01#section-7.3.3)
  *
  * @param {string} hostname
  * @returns {boolean}
  */
 export const isIdn = (hostname) => {
-  const asciiHostname = tr46.toASCII(hostname);
-
-  if (!asciiHostname) {
+  try {
+    return isIdnHostname(hostname);
+  } catch (_error) {
     return false;
   }
-
-  return isAsciiIdn(asciiHostname);
 };