Fixes & tests for surrogates support in strings encoder/decoder

README.md

@@ -30,13 +30,13 @@ node v0.10.26, 1mb sample:
    deflate-gildas x 4.58 ops/sec ±2.33% (15 runs sampled)
    deflate-imaya x 3.22 ops/sec ±3.95% (12 runs sampled)
  ! deflate-pako x 6.99 ops/sec ±0.51% (21 runs sampled)
-   deflate-pako-string x 6.22 ops/sec ±0.76% (19 runs sampled)
+   deflate-pako-string x 5.89 ops/sec ±0.77% (18 runs sampled)
    deflate-pako-untyped x 4.39 ops/sec ±1.58% (14 runs sampled)
  * deflate-zlib x 14.71 ops/sec ±4.23% (59 runs sampled)
    inflate-dankogai x 32.16 ops/sec ±0.13% (56 runs sampled)
    inflate-imaya x 30.35 ops/sec ±0.92% (53 runs sampled)
  ! inflate-pako x 69.89 ops/sec ±1.46% (71 runs sampled)
-   inflate-pako-string x 64.99 ops/sec ±1.64% (67 runs sampled)
+   inflate-pako-string x 12.46 ops/sec ±5.65% (36 runs sampled)
    inflate-pako-untyped x 17.19 ops/sec ±0.85% (32 runs sampled)
  * inflate-zlib x 70.03 ops/sec ±1.64% (81 runs sampled)
 
@@ -46,13 +46,13 @@ node v0.11.12, 1mb sample:
    deflate-gildas x 5.06 ops/sec ±6.00% (16 runs sampled)
    deflate-imaya x 3.52 ops/sec ±3.71% (13 runs sampled)
  ! deflate-pako x 11.52 ops/sec ±0.22% (32 runs sampled)
-   deflate-pako-string x 9.61 ops/sec ±0.97% (28 runs sampled)
+   deflate-pako-string x 9.53 ops/sec ±1.12% (27 runs sampled)
    deflate-pako-untyped x 5.44 ops/sec ±0.72% (17 runs sampled)
  * deflate-zlib x 14.05 ops/sec ±3.34% (63 runs sampled)
    inflate-dankogai x 42.19 ops/sec ±0.09% (56 runs sampled)
    inflate-imaya x 79.68 ops/sec ±1.07% (68 runs sampled)
  ! inflate-pako x 97.52 ops/sec ±0.83% (80 runs sampled)
-   inflate-pako-string x 89.05 ops/sec ±0.48% (76 runs sampled)
+   inflate-pako-string x 19.07 ops/sec ±12.67% (32 runs sampled)
    inflate-pako-untyped x 24.35 ops/sec ±2.59% (40 runs sampled)
  * inflate-zlib x 60.32 ops/sec ±1.36% (69 runs sampled)
 ```

lib/utils/strings.js

@@ -10,64 +10,66 @@ var STR_APPLY_OK = true;
 try { String.fromCharCode.apply(null, [0]); } catch(__) { STR_APPLY_OK = false; }
 
 
-// Table with utf8 lengths
+// Table with utf8 lengths (calculated by first byte of sequence)
-var utf8len = new utils.Buf8(256);
+// Note, that 5 & 6-byte values and some 4-byte values can not be represented in JS,
+// because max possible codepoiny is 0x10ffff
+var _utf8len = new utils.Buf8(256);
 for (var i=0; i<256; i++) {
-  utf8len[i] = (i >= 252 ? 6 : i >= 248 ? 5 : i >= 240 ? 4 : i >= 224 ? 3 : i >= 192 ? 2 : 1);
+  _utf8len[i] = (i >= 252 ? 6 : i >= 248 ? 5 : i >= 240 ? 4 : i >= 224 ? 3 : i >= 192 ? 2 : 1);
 }
+_utf8len[254]=_utf8len[254]=1; // Invalid sequence start
 
 
 // convert string to array (typed, when possible)
 // src: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Base64_encoding_and_decoding
 exports.string2buf = function (str) {
-  var buf, c, str_len = str.length, buf_len = 0;
+  var buf, c, c2, m_pos, i, str_len = str.length, buf_len = 0;
 
   /* mapping... */
 
-  for (var m_pos = 0; m_pos < str_len; m_pos++) {
+  for (m_pos = 0; m_pos < str_len; m_pos++) {
     c = str.charCodeAt(m_pos);
-    buf_len += c < 0x80 ? 1 : c < 0x800 ? 2 : c < 0x10000 ? 3 : c < 0x200000 ? 4 : c < 0x4000000 ? 5 : 6;
+    if ((c & 0xfc00) === 0xd800 && (m_pos+1 < str_len)) {
+      c2 = str.charCodeAt(m_pos+1);
+      if ((c2 & 0xfc00) === 0xdc00) {
+        c = 0x10000 + ((c - 0xd800) << 10) + (c2 - 0xdc00);
+        m_pos++;
+      }
+    }
+    buf_len += c < 0x80 ? 1 : c < 0x800 ? 2 : c < 0x10000 ? 3 : 4;
   }
 
   buf = new utils.Buf8(buf_len);
 
   /* transcription... */
 
-  for (var i = 0, c_pos = 0; i < buf_len; c_pos++) {
+  for (i=0, m_pos = 0; i < buf_len; m_pos++) {
-    c = str.charCodeAt(c_pos);
+    c = str.charCodeAt(m_pos);
-    if (c < 128) {
+    if ((c & 0xfc00) === 0xd800 && (m_pos+1 < str_len)) {
+      c2 = str.charCodeAt(m_pos+1);
+      if ((c2 & 0xfc00) === 0xdc00) {
+        c = 0x10000 + ((c - 0xd800) << 10) + (c2 - 0xdc00);
+        m_pos++;
+      }
+    }
+    if (c < 0x80) {
       /* one byte */
       buf[i++] = c;
     } else if (c < 0x800) {
       /* two bytes */
-      buf[i++] = 192 + (c >>> 6);
+      buf[i++] = 0xC0 | (c >>> 6);
-      buf[i++] = 128 + (c & 63);
+      buf[i++] = 0x80 | (c & 0x3f);
     } else if (c < 0x10000) {
       /* three bytes */
-      buf[i++] = 224 + (c >>> 12);
+      buf[i++] = 0xE0 | (c >>> 12);
-      buf[i++] = 128 + (c >>> 6 & 63);
+      buf[i++] = 0x80 | (c >>> 6 & 0x3f);
-      buf[i++] = 128 + (c & 63);
+      buf[i++] = 0x80 | (c & 0x3f);
-    } else if (c < 0x200000) {
+    } else {
       /* four bytes */
-      buf[i++] = 240 + (c >>> 18);
+      buf[i++] = 0xf0 | (c >>> 18);
-      buf[i++] = 128 + (c >>> 12 & 63);
+      buf[i++] = 0x80 | (c >>> 12 & 0x3f);
-      buf[i++] = 128 + (c >>> 6 & 63);
+      buf[i++] = 0x80 | (c >>> 6 & 0x3f);
-      buf[i++] = 128 + (c & 63);
+      buf[i++] = 0x80 | (c & 0x3f);
-    } else if (c < 0x4000000) {
-      /* five bytes */
-      buf[i++] = 248 + (c >>> 24);
-      buf[i++] = 128 + (c >>> 18 & 63);
-      buf[i++] = 128 + (c >>> 12 & 63);
-      buf[i++] = 128 + (c >>> 6 & 63);
-      buf[i++] = 128 + (c & 63);
-    } else /* if (c <= 0x7fffffff) */ {
-      /* six bytes */
-      buf[i++] = 252 + /* (c >>> 32) is not possible in ECMAScript! So...: */ (c / 1073741824);
-      buf[i++] = 128 + (c >>> 24 & 63);
-      buf[i++] = 128 + (c >>> 18 & 63);
-      buf[i++] = 128 + (c >>> 12 & 63);
-      buf[i++] = 128 + (c >>> 6 & 63);
-      buf[i++] = 128 + (c & 63);
     }
   }
 
@@ -104,7 +106,7 @@ exports.binstring2buf = function(str) {
 // src: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Base64_encoding_and_decoding
 exports.buf2string = function (buf, max) {
   /*jshint nonstandard:true*/
-  // That's not as fast as via String.fromCharCode.appy
+  // That's not as fast as via String.fromCharCode.apply
   /*return decodeURIComponent(escape(exports.buf2binstring(
     (buf.length === max) ?
       buf
@@ -115,24 +117,24 @@ exports.buf2string = function (buf, max) {
   var str, i, out, part, char_len;
   var len = max || buf.length;
 
-  // Reserve max possibli length
+  // Reserve max possible length (2 words per char)
   var utf16buf = new utils.Buf16(len*2);
 
   for (out=0, i=0; i<len; i++) {
     part = buf[i];
-    char_len = utf8len[part];
+    char_len = _utf8len[part];
 
     // edge case - broken sequence
     if (i + char_len > len) {
       utf16buf[out++] = 0xfffd;
-      break;
+      break; // end of string reached, stop
     }
     switch (char_len) {
       case 1:
         utf16buf[out++] = part;
         break;
       case 2:
-        utf16buf[out++] = ((part & 0x1f) << 6) | (buf[++i] & 0x7f);
+        utf16buf[out++] = ((part & 0x1f) << 6) | (buf[++i] & 0x3f);
         break;
       case 3:
         utf16buf[out++] = ((part & 0x0f) << 12) | ((buf[++i] & 0x3f) << 6) | (buf[++i] & 0x3f);
@@ -144,17 +146,15 @@ exports.buf2string = function (buf, max) {
         utf16buf[out++] = 0xd800 | ((part >> 10) & 0x3ff);
         utf16buf[out++] = 0xdc00 | (part & 0x3ff);
         break;
-      // 5 & 6 bytes uticodes not supported in UTF16 (JS), so fill with dummy symbol
+      // 5 & 6 bytes uticodes not supported in UTF16 (JS),
+      // so fill with dummy symbol & update scan position.
       case 5:
         i += 4;
         utf16buf[out++] = 0xfffd;
-        //utf16buf[out++] = (part - 248 << 24) + (buf[++i] - 128 << 18) + (buf[++i] - 128 << 12) + (buf[++i] - 128 << 6) + buf[++i] - 128;
         break;
       case 6:
         i += 5;
         utf16buf[out++] = 0xfffd;
-                          // (part - 252 << 32) is not possible in ECMAScript! So...:
-        //utf16buf[out++] = (part - 252) * 1073741824 + (buf[++i] - 128 << 24) + (buf[++i] - 128 << 18) + (buf[++i] - 128 << 12) + (buf[++i] - 128 << 6) + buf[++i] - 128;
         break;
     }
   }
@@ -174,5 +174,5 @@ exports.buf2string = function (buf, max) {
 
 // calculate tail size of utf8 char by current byte value
 exports.utf8tail = function(code) {
-  return utf8len[code];
+  return _utf8len[code];
 };

test/strings.js

@@ -9,17 +9,62 @@ var path    = require('path');
 var assert  = require('assert');
 
 var pako    = require('../index');
+var cmp     = require('./helpers').cmpBuf;
+var strings = require('../lib/utils/strings');
 
-var helpers = require('./helpers');
+// fromCharCode, but understands right > 0xffff values
-var cmp     = helpers.cmpBuf;
+function fixedFromCharCode(code) {
+  /*jshint bitwise: false*/
+  if (code > 0xffff) {
+    code -= 0x10000;
+
+    var surrogate1 = 0xd800 + (code >> 10)
+      , surrogate2 = 0xdc00 + (code & 0x3ff);
+
+    return String.fromCharCode(surrogate1, surrogate2);
+  } else {
+    return String.fromCharCode(code);
+  }
+}
+
+// Converts array of codes / chars / strings to utf16 string
+function a2utf16(arr) {
+  var result = '';
+  arr.forEach(function (item) {
+    if (typeof item === 'string') { result += item; return; }
+    result += fixedFromCharCode(item);
+  });
+  return result;
+}
 
 
-var file = path.join(__dirname, 'fixtures/samples/lorem_utf_100k.txt');
+describe('Encode/Decode', function () {
-var sampleString = fs.readFileSync(file, 'utf8');
+
-var sampleArray  = new Uint8Array(fs.readFileSync(file));
+  var utf16sample = a2utf16([0x1f3b5, 'abcd', 0x266a, 0x35, 0xe800, 0x10ffff, 0x0fffff]);
+  var utf8sample = new Uint8Array(new Buffer(utf16sample));
+
+  console.log(utf16sample, utf16sample.length);
+  console.log(new Buffer(utf16sample));
+
+  it('Encode string to utf8 buf', function () {
+    assert.ok(cmp(
+      strings.string2buf(utf16sample),
+      utf8sample
+    ));
+  });
+
+  it('Decode utf8 buf to string', function () {
+    assert.ok(strings.buf2string(utf8sample), utf16sample);
+  });
+
+});
 
 
-describe('Deflate strings', function () {
+describe('Deflate/Inflate strings', function () {
+
+  var file = path.join(__dirname, 'fixtures/samples/lorem_utf_100k.txt');
+  var sampleString = fs.readFileSync(file, 'utf8');
+  var sampleArray  = new Uint8Array(fs.readFileSync(file));
 
   it('Deflate javascript string (utf16) on input', function () {
     assert.ok(cmp(
@@ -35,22 +80,18 @@ describe('Deflate strings', function () {
     assert.ok(cmp(new Buffer(data, 'binary'), pako.deflate(sampleArray)));
   });
 
-});
-
-
-describe('Inflate strings', function () {
-  var deflatedString = pako.deflate(sampleArray, { to: 'string' });
-  var deflatedArray  = pako.deflate(sampleArray);
-
   it('Inflate binary string input', function () {
+    var deflatedString = pako.deflate(sampleArray, { to: 'string' });
+    var deflatedArray  = pako.deflate(sampleArray);
     assert.ok(cmp(pako.inflate(deflatedString), pako.inflate(deflatedArray)));
   });
 
   it('Inflate with javascript string (utf16) output', function () {
+    var deflatedArray  = pako.deflate(sampleArray);
     var data = pako.inflate(deflatedArray, { to: 'string', chunkSize: 99 });
 
     assert.equal(typeof data, 'string');
-    assert(data === sampleString);
+    assert.equal(data, sampleString);
   });
 
 });