url: update IDNA error conditions

This commit contains three separate changes:

- Always return a string from ToUnicode no matter if an error occurred.
- Disable CheckHyphens boolean flag. This flag will soon be enabled in
  the URL Standard, but is implemented manually by selectively ignoring
  certain errors.
- Enable CheckBidi boolean flag per URL Standard update.

This allows domain names with hyphens at 3 and 4th position, as well as
those with leading and trailing hyphens. They are technically invalid,
but seen in the wild.

Tests are updated and simplified accordingly.

PR-URL: #12966
Fixes: #12965
Refs: whatwg/url#309
Reviewed-By: Daijiro Wachi <[email protected]>
Reviewed-By: Timothy Gu <[email protected]>

Author: zimbabao

src/node_i18n.cc

@@ -436,11 +436,9 @@ bool InitializeICUDirectory(const std::string& path) {
 
 int32_t ToUnicode(MaybeStackBuffer<char>* buf,
                   const char* input,
-                  size_t length,
-                  bool lenient) {
+                  size_t length) {
   UErrorCode status = U_ZERO_ERROR;
-  uint32_t options = UIDNA_DEFAULT;
-  options |= UIDNA_NONTRANSITIONAL_TO_UNICODE;
+  uint32_t options = UIDNA_NONTRANSITIONAL_TO_UNICODE;
   UIDNA* uidna = uidna_openUTS46(options, &status);
   if (U_FAILURE(status))
     return -1;
@@ -462,14 +460,10 @@ int32_t ToUnicode(MaybeStackBuffer<char>* buf,
                                   &status);
   }
 
-  // UTS #46's ToUnicode operation applies no validation of domain name length
-  // (nor a flag requesting it to do so, like VerifyDnsLength for ToASCII). For
-  // that reason, unlike ToASCII below, ICU4C correctly accepts long domain
-  // names. However, ICU4C still sets the EMPTY_LABEL error in contrary to UTS
-  // #46. Therefore, explicitly filters out that error here.
-  info.errors &= ~UIDNA_ERROR_EMPTY_LABEL;
+  // info.errors is ignored as UTS #46 ToUnicode always produces a Unicode
+  // string, regardless of whether an error occurred.
 
-  if (U_FAILURE(status) || (!lenient && info.errors != 0)) {
+  if (U_FAILURE(status)) {
     len = -1;
     buf->SetLength(0);
   } else {
@@ -485,8 +479,7 @@ int32_t ToASCII(MaybeStackBuffer<char>* buf,
                 size_t length,
                 bool lenient) {
   UErrorCode status = U_ZERO_ERROR;
-  uint32_t options = UIDNA_DEFAULT;
-  options |= UIDNA_NONTRANSITIONAL_TO_ASCII;
+  uint32_t options = UIDNA_NONTRANSITIONAL_TO_ASCII | UIDNA_CHECK_BIDI;
   UIDNA* uidna = uidna_openUTS46(options, &status);
   if (U_FAILURE(status))
     return -1;
@@ -518,6 +511,21 @@ int32_t ToASCII(MaybeStackBuffer<char>* buf,
   info.errors &= ~UIDNA_ERROR_LABEL_TOO_LONG;
   info.errors &= ~UIDNA_ERROR_DOMAIN_NAME_TOO_LONG;
 
+  // These error conditions are mandated unconditionally by UTS #46 version
+  // 9.0.0 (rev. 17), but were found to be incompatible with actual domain
+  // names in the wild. As such, in the current UTS #46 draft (rev. 18) these
+  // checks are made optional depending on the CheckHyphens flag, which will be
+  // disabled in WHATWG URL's "domain to ASCII" algorithm soon.
+  // Refs:
+  // - https://github.com/whatwg/url/issues/53
+  // - https://github.com/whatwg/url/pull/309
+  // - http://www.unicode.org/review/pri317/
+  // - http://www.unicode.org/reports/tr46/tr46-18.html
+  // - https://www.icann.org/news/announcement-2000-01-07-en
+  info.errors &= ~UIDNA_ERROR_HYPHEN_3_4;
+  info.errors &= ~UIDNA_ERROR_LEADING_HYPHEN;
+  info.errors &= ~UIDNA_ERROR_TRAILING_HYPHEN;
+
   if (U_FAILURE(status) || (!lenient && info.errors != 0)) {
     len = -1;
     buf->SetLength(0);
@@ -534,11 +542,9 @@ static void ToUnicode(const FunctionCallbackInfo<Value>& args) {
   CHECK_GE(args.Length(), 1);
   CHECK(args[0]->IsString());
   Utf8Value val(env->isolate(), args[0]);
-  // optional arg
-  bool lenient = args[1]->BooleanValue(env->context()).FromJust();
 
   MaybeStackBuffer<char> buf;
-  int32_t len = ToUnicode(&buf, *val, val.length(), lenient);
+  int32_t len = ToUnicode(&buf, *val, val.length());
 
   if (len < 0) {
     return env->ThrowError("Cannot convert name to Unicode");

src/node_i18n.h

@@ -43,8 +43,7 @@ int32_t ToASCII(MaybeStackBuffer<char>* buf,
                 bool lenient = false);
 int32_t ToUnicode(MaybeStackBuffer<char>* buf,
                   const char* input,
-                  size_t length,
-                  bool lenient = false);
+                  size_t length);
 
 }  // namespace i18n
 }  // namespace node

test/fixtures/url-idna.js

@@ -191,22 +191,33 @@ module.exports = {
     {
       ascii: `${`${'a'.repeat(64)}.`.repeat(4)}com`,
       unicode: `${`${'a'.repeat(64)}.`.repeat(4)}com`
-    }
-  ],
-  invalid: [
-    // invalid character
+    },
+    // URLs with hyphen
+    {
+      ascii: 'r4---sn-a5mlrn7s.gevideo.com',
+      unicode: 'r4---sn-a5mlrn7s.gevideo.com'
+    },
+    {
+      ascii: '-sn-a5mlrn7s.gevideo.com',
+      unicode: '-sn-a5mlrn7s.gevideo.com'
+    },
     {
-      url: '\ufffd.com',
-      mode: 'ascii'
+      ascii: 'sn-a5mlrn7s-.gevideo.com',
+      unicode: 'sn-a5mlrn7s-.gevideo.com'
     },
     {
-      url: '\ufffd.com',
-      mode: 'unicode'
+      ascii: '-sn-a5mlrn7s-.gevideo.com',
+      unicode: '-sn-a5mlrn7s-.gevideo.com'
     },
-    // invalid Punycode
     {
-      url: 'xn---abc.com',
-      mode: 'unicode'
+      ascii: '-sn--a5mlrn7s-.gevideo.com',
+      unicode: '-sn--a5mlrn7s-.gevideo.com'
     }
+  ],
+  invalid: [
+    // invalid character
+    '\ufffd.com',
+    // invalid bi-directional character
+    'تشادرlatin.icom.museum'
   ]
 }

test/parallel/test-icu-punycode.js

@@ -23,19 +23,13 @@ const tests = require('../fixtures/url-idna.js');
 }
 
 {
-  const errorRe = {
-    ascii: /^Error: Cannot convert name to ASCII$/,
-    unicode: /^Error: Cannot convert name to Unicode$/
-  };
-  const convertFunc = {
-    ascii: icu.toASCII,
-    unicode: icu.toUnicode
-  };
-
-  for (const [i, { url, mode }] of tests.invalid.entries()) {
-    assert.throws(() => convertFunc[mode](url), errorRe[mode],
-                  `Invalid case ${i + 1}`);
-    assert.doesNotThrow(() => convertFunc[mode](url, true),
-                        `Invalid case ${i + 1} in lenient mode`);
+  for (const [i, url] of tests.invalid.entries()) {
+    assert.throws(() => icu.toASCII(url),
+                  /^Error: Cannot convert name to ASCII$/,
+                  `ToASCII invalid case ${i + 1}`);
+    assert.doesNotThrow(() => icu.toASCII(url, true),
+                        `ToASCII invalid case ${i + 1} in lenient mode`);
+    assert.doesNotThrow(() => icu.toUnicode(url),
+                        `ToUnicode invalid case ${i + 1}`);
   }
 }

test/parallel/test-url-domain-ascii-unicode.js

@@ -8,7 +8,8 @@ const domainToASCII = url.domainToASCII;
 const domainToUnicode = url.domainToUnicode;
 
 const domainWithASCII = [
-  ['ıídيٴ', 'xn--d-iga7ro0q9f'],
+  ['ıíd', 'xn--d-iga7r'],
+  ['يٴ', 'xn--mhb8f'],
   ['www.ϧƽəʐ.com', 'www.xn--cja62apfr6c.com'],
   ['новини.com', 'xn--b1amarcd.com'],
   ['名がドメイン.com', 'xn--v8jxj3d1dzdz08w.com'],

test/parallel/test-whatwg-url-domainto.js

@@ -35,11 +35,8 @@ const tests = require('../fixtures/url-idna.js');
 }
 
 {
-  const convertFunc = {
-    ascii: domainToASCII,
-    unicode: domainToUnicode
-  };
-
-  for (const [i, { url, mode }] of tests.invalid.entries())
-    assert.strictEqual(convertFunc[mode](url), '', `Invalid case ${i + 1}`);
+  for (const [i, url] of tests.invalid.entries()) {
+    assert.strictEqual(domainToASCII(url), '', `Invalid case ${i + 1}`);
+    assert.strictEqual(domainToUnicode(url), '', `Invalid case ${i + 1}`);
+  }
 }