net: use icu's punycode implementation

ICU has a punycode implementation built in. Use it instead of the javascript implementation because it's much faster. PR-URL: #7355 Reviewed-By: Trevor Norris <trev.norris@gmail.com> Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
nodejs · Jun 30, 2016 · 7de59ef · 7de59ef
1 parent 3d6a01e
commit 7de59ef
Show file tree

Hide file tree

Showing 6 changed files with 291 additions and 12 deletions.
diff --git a/benchmark/net/punycode.js b/benchmark/net/punycode.js
@@ -0,0 +1,75 @@
+'use strict';
+
+const common = require('../common.js');
+const icu = process.binding('icu');
+const punycode = require('punycode');
+
+const bench = common.createBenchmark(main, {
+  method: ['punycode', 'icu'],
+  n: [1024],
+  val: [
+    'افغانستا.icom.museum',
+    'الجزائر.icom.museum',
+    'österreich.icom.museum',
+    'বাংলাদেশ.icom.museum',
+    'беларусь.icom.museum',
+    'belgië.icom.museum',
+    'българия.icom.museum',
+    'تشادر.icom.museum',
+    '中国.icom.museum',
+    'القمر.icom.museum',
+    'κυπρος.icom.museum',
+    'českárepublika.icom.museum',
+    'مصر.icom.museum',
+    'ελλάδα.icom.museum',
+    'magyarország.icom.museum',
+    'ísland.icom.museum',
+    'भारत.icom.museum',
+    'ايران.icom.museum',
+    'éire.icom.museum',
+    'איקו״ם.ישראל.museum',
+    '日本.icom.museum',
+    'الأردن.icom.museum'
+  ]
+});
+
+function usingPunycode(val) {
+  punycode.toUnicode(punycode.toASCII(val));
+}
+
+function usingICU(val) {
+  icu.toUnicode(icu.toASCII(val));
+}
+
+function runPunycode(n, val) {
+  common.v8ForceOptimization(usingPunycode, val);
+  var i = 0;
+  bench.start();
+  for (; i < n; i++)
+    usingPunycode(val);
+  bench.end(n);
+}
+
+function runICU(n, val) {
+  common.v8ForceOptimization(usingICU, val);
+  var i = 0;
+  bench.start();
+  for (; i < n; i++)
+    usingICU(val);
+  bench.end(n);
+}
+
+function main(conf) {
+  const n = +conf.n;
+  const val = conf.val;
+  switch (conf.method) {
+    case 'punycode':
+      runPunycode(n, val);
+      break;
+    case 'icu':
+      runICU(n, val);
+      break;
+    default:
+      throw new Error('Unexpected method');
+  }
+}
diff --git a/lib/url.js b/lib/url.js
@@ -1,6 +1,14 @@
 'use strict';
 
-const punycode = require('punycode');
+function importPunycode() {
+  try {
+    return process.binding('icu');
+  } catch (e) {
+    return require('punycode');
+  }
+}
+
+const { toASCII } = importPunycode();
 
 exports.parse = urlParse;
 exports.resolve = urlResolve;
@@ -309,7 +317,7 @@ Url.prototype.parse = function(url, parseQueryString, slashesDenoteHost) {
       // It only converts parts of the domain name that
       // have non-ASCII characters, i.e. it doesn't matter if
       // you call it with a domain that already is ASCII-only.
-      this.hostname = punycode.toASCII(this.hostname);
+      this.hostname = toASCII(this.hostname);
     }
 
     var p = this.port ? ':' + this.port : '';

diff --git a/src/node_i18n.cc b/src/node_i18n.cc
@@ -23,8 +23,16 @@
 
 #if defined(NODE_HAVE_I18N_SUPPORT)
 
+#include "node.h"
+#include "env.h"
+#include "env-inl.h"
+#include "util.h"
+#include "util-inl.h"
+#include "v8.h"
+
 #include <unicode/putil.h>
 #include <unicode/udata.h>
+#include <unicode/uidna.h>
 
 #ifdef NODE_HAVE_SMALL_ICU
 /* if this is defined, we have a 'secondary' entry point.
@@ -43,6 +51,13 @@ extern "C" const char U_DATA_API SMALL_ICUDATA_ENTRY_POINT[];
 
 namespace node {
 
+using v8::Context;
+using v8::FunctionCallbackInfo;
+using v8::Local;
+using v8::Object;
+using v8::String;
+using v8::Value;
+
 bool flag_icu_data_dir = false;
 
 namespace i18n {
@@ -64,7 +79,124 @@ bool InitializeICUDirectory(const char* icu_data_path) {
   }
 }
 
+static int32_t ToUnicode(MaybeStackBuffer<char>* buf,
+                         const char* input,
+                         size_t length) {
+  UErrorCode status = U_ZERO_ERROR;
+  uint32_t options = UIDNA_DEFAULT;
+  options |= UIDNA_NONTRANSITIONAL_TO_UNICODE;
+  UIDNA* uidna = uidna_openUTS46(options, &status);
+  if (U_FAILURE(status))
+    return -1;
+  UIDNAInfo info = UIDNA_INFO_INITIALIZER;
+
+  int32_t len = uidna_nameToUnicodeUTF8(uidna,
+                                        input, length,
+                                        **buf, buf->length(),
+                                        &info,
+                                        &status);
+
+  if (status == U_BUFFER_OVERFLOW_ERROR) {
+    status = U_ZERO_ERROR;
+    buf->AllocateSufficientStorage(len);
+    len = uidna_nameToUnicodeUTF8(uidna,
+                                  input, length,
+                                  **buf, buf->length(),
+                                  &info,
+                                  &status);
+  }
+
+  if (U_FAILURE(status))
+    len = -1;
+
+  uidna_close(uidna);
+  return len;
+}
+
+static int32_t ToASCII(MaybeStackBuffer<char>* buf,
+                       const char* input,
+                       size_t length) {
+  UErrorCode status = U_ZERO_ERROR;
+  uint32_t options = UIDNA_DEFAULT;
+  options |= UIDNA_NONTRANSITIONAL_TO_ASCII;
+  UIDNA* uidna = uidna_openUTS46(options, &status);
+  if (U_FAILURE(status))
+    return -1;
+  UIDNAInfo info = UIDNA_INFO_INITIALIZER;
+
+  int32_t len = uidna_nameToASCII_UTF8(uidna,
+                                       input, length,
+                                       **buf, buf->length(),
+                                       &info,
+                                       &status);
+
+  if (status == U_BUFFER_OVERFLOW_ERROR) {
+    status = U_ZERO_ERROR;
+    buf->AllocateSufficientStorage(len);
+    len = uidna_nameToASCII_UTF8(uidna,
+                                 input, length,
+                                 **buf, buf->length(),
+                                 &info,
+                                 &status);
+  }
+
+  if (U_FAILURE(status))
+    len = -1;
+
+  uidna_close(uidna);
+  return len;
+}
+
+static void ToUnicode(const FunctionCallbackInfo<Value>& args) {
+  Environment* env = Environment::GetCurrent(args);
+  CHECK_GE(args.Length(), 1);
+  CHECK(args[0]->IsString());
+  Utf8Value val(env->isolate(), args[0]);
+  MaybeStackBuffer<char> buf;
+  int32_t len = ToUnicode(&buf, *val, val.length());
+
+  if (len < 0) {
+    return env->ThrowError("Cannot convert name to Unicode");
+  }
+
+  args.GetReturnValue().Set(
+      String::NewFromUtf8(env->isolate(),
+                          *buf,
+                          v8::NewStringType::kNormal,
+                          len).ToLocalChecked());
+}
+
+static void ToASCII(const FunctionCallbackInfo<Value>& args) {
+  Environment* env = Environment::GetCurrent(args);
+  CHECK_GE(args.Length(), 1);
+  CHECK(args[0]->IsString());
+  Utf8Value val(env->isolate(), args[0]);
+  MaybeStackBuffer<char> buf;
+  int32_t len = ToASCII(&buf, *val, val.length());
+
+  if (len < 0) {
+    return env->ThrowError("Cannot convert name to ASCII");
+  }
+
+  args.GetReturnValue().Set(
+      String::NewFromUtf8(env->isolate(),
+                          *buf,
+                          v8::NewStringType::kNormal,
+                          len).ToLocalChecked());
+}
+
+void Init(Local<Object> target,
+          Local<Value> unused,
+          Local<Context> context,
+          void* priv) {
+  Environment* env = Environment::GetCurrent(context);
+  env->SetMethod(target, "toUnicode", ToUnicode);
+  env->SetMethod(target, "toASCII", ToASCII);
+}
+
 }  // namespace i18n
 }  // namespace node
 
+NODE_MODULE_CONTEXT_AWARE_BUILTIN(icu, node::i18n::Init)
+
 #endif  // NODE_HAVE_I18N_SUPPORT
diff --git a/test/parallel/test-icu-punycode.js b/test/parallel/test-icu-punycode.js
@@ -0,0 +1,72 @@
+'use strict';
+
+const common = require('../common');
+const icu = getPunycode();
+const assert = require('assert');
+
+function getPunycode() {
+  try {
+    return process.binding('icu');
+  } catch (err) {
+    return undefined;
+  }
+}
+
+if (!icu) {
+  common.skip('icu punycode tests because ICU is not present.');
+  return;
+}
+
+// Credit for list: http://www.i18nguy.com/markup/idna-examples.html
+const tests = [
+  'افغانستا.icom.museum',
+  'الجزائر.icom.museum',
+  'österreich.icom.museum',
+  'বাংলাদেশ.icom.museum',
+  'беларусь.icom.museum',
+  'belgië.icom.museum',
+  'българия.icom.museum',
+  'تشادر.icom.museum',
+  '中国.icom.museum',
+  'القمر.icom.museum',
+  'κυπρος.icom.museum',
+  'českárepublika.icom.museum',
+  'مصر.icom.museum',
+  'ελλάδα.icom.museum',
+  'magyarország.icom.museum',
+  'ísland.icom.museum',
+  'भारत.icom.museum',
+  'ايران.icom.museum',
+  'éire.icom.museum',
+  'איקו״ם.ישראל.museum',
+  '日本.icom.museum',
+  'الأردن.icom.museum',
+  'қазақстан.icom.museum',
+  '한국.icom.museum',
+  'кыргызстан.icom.museum',
+  'ລາວ.icom.museum',
+  'لبنان.icom.museum',
+  'македонија.icom.museum',
+  'موريتانيا.icom.museum',
+  'méxico.icom.museum',
+  'монголулс.icom.museum',
+  'المغرب.icom.museum',
+  'नेपाल.icom.museum',
+  'عمان.icom.museum',
+  'قطر.icom.museum',
+  'românia.icom.museum',
+  'россия.иком.museum',
+  'србијаицрнагора.иком.museum',
+  'இலங்கை.icom.museum',
+  'españa.icom.museum',
+  'ไทย.icom.museum',
+  'تونس.icom.museum',
+  'türkiye.icom.museum',
+  'украина.icom.museum',
+  'việtnam.icom.museum'
+];
+
+// Testing the roundtrip
+tests.forEach((i) => {
+  assert.strictEqual(i, icu.toUnicode(icu.toASCII(i)));
+});
diff --git a/tools/icu/icu-generic.gyp b/tools/icu/icu-generic.gyp
@@ -37,8 +37,7 @@
              'defines': [
                 # ICU cannot swap the initial data without this.
                 # http://bugs.icu-project.org/trac/ticket/11046
-                'UCONFIG_NO_LEGACY_CONVERSION=1',
-                'UCONFIG_NO_IDNA=1',
+                'UCONFIG_NO_LEGACY_CONVERSION=1'
              ],
           }],
         ],
@@ -428,9 +427,6 @@
           #'<(icu_path)/source/common/ubidi_props_data.h',
           # and the callers
           '<(icu_path)/source/common/ushape.cpp',
-          '<(icu_path)/source/common/usprep.cpp',
-          '<(icu_path)/source/common/uts46.cpp',
-          '<(icu_path)/source/common/uidna.cpp',
         ]}],
         [ 'icu_ver_major == 57', { 'sources!': [
           # work around http://bugs.icu-project.org/trac/ticket/12451
@@ -447,9 +443,6 @@
           #'<(icu_path)/source/common/ubidi_props_data.h',
           # and the callers
           '<(icu_path)/source/common/ushape.cpp',
-          '<(icu_path)/source/common/usprep.cpp',
-          '<(icu_path)/source/common/uts46.cpp',
-          '<(icu_path)/source/common/uidna.cpp',
         ]}],
         [ 'OS == "solaris"', { 'defines': [
           '_XOPEN_SOURCE_EXTENDED=0',

diff --git a/tools/icu/icu_small.json b/tools/icu/icu_small.json
@@ -24,7 +24,7 @@
     "region": "none",
     "zone": "locales",
     "converters": "none",
-    "stringprep": "none",
+    "stringprep": "locales",
     "translit": "none",
     "brkfiles": "none",
     "brkdict": "none",
@@ -34,7 +34,6 @@
   "remove": [
     "cnvalias.icu",
     "postalCodeData.res",
-    "uts46.nrm",
     "genderList.res",
     "brkitr/root.res",
     "unames.icu"