Changeset 74225 in webkit

Timestamp:

Dec 16, 2010 6:28:48 PM (13 years ago)

Author:

commit-queue@webkit.org

Message:

2010-12-16 Yong Li <​yoli@rim.com>

Reviewed by Alexey Proskuryakov.

​https://bugs.webkit.org/show_bug.cgi?id=51199
Add decoding tests for UTF-16 LE/BE and their variants.
Also, check the full decoded text but not only the first one.

• fast/encoding/char-decoding.html:
• fast/encoding/char-decoding-expected.txt:
• fast/encoding/resources/char-decoding-utils.js:

Location:

trunk/LayoutTests

Files:

• ChangeLog (modified) (1 diff)
• fast/encoding/char-decoding-expected.txt (modified) (1 diff)
• fast/encoding/char-decoding.html (modified) (1 diff)
• fast/encoding/resources/char-decoding-utils.js (modified) (2 diffs)

trunk/LayoutTests/ChangeLog

@@ +1 @@
+2010-12-16  Yong Li  <yoli@rim.com>
+
+        Reviewed by Alexey Proskuryakov.
+
+        https://bugs.webkit.org/show_bug.cgi?id=51199
+        Add decoding tests for UTF-16 LE/BE and their variants.
+        Also, check the full decoded text but not only the first one.
+
+        * fast/encoding/char-decoding.html:
+        * fast/encoding/char-decoding-expected.txt:
+        * fast/encoding/resources/char-decoding-utils.js:
+
 2010-12-16  Jian Li  <jianli@chromium.org>
 

trunk/LayoutTests/fast/encoding/char-decoding-expected.txt

@@ -156 +156 @@
 PASS decode('dos-874', '%A1') is 'U+0E01'
 PASS decode('dos-874', '%DB') is 'U+F8C1'
-PASS decode('UTF-7', '+AD4') is 'U+002B'
-PASS decode('utf-7', '+AD4') is 'U+002B'
+PASS decode('UTF-7', '+AD4') is 'U+002B/U+0041/U+0044/U+0034'
+PASS decode('utf-7', '+AD4') is 'U+002B/U+0041/U+0044/U+0034'
+PASS decode('UTF-16LE', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('unicodeFEFF', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('UTF-16', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('ISO-10646-UCS-2', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('UCS-2', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('Unicode', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('csUnicode', '%69%D8%D6%DE') is 'U+D869/U+DED6'
+PASS decode('UTF-16BE', '%D8%69%DE%D6') is 'U+D869/U+DED6'
+PASS decode('unicodeFFFE', '%D8%69%DE%D6') is 'U+D869/U+DED6'
 PASS successfullyParsed is true
 

trunk/LayoutTests/fast/encoding/char-decoding.html

@@ -90 +90 @@
 // UTF-7 is expressly forbidden, so decoding it should not work correctly.
 // This attempts to decode '<' as UTF-7 (+AD4) but it ends up being decoded
-// as a '+'.
-testDecode('UTF-7', '+AD4', 'U+002B');
-testDecode('utf-7', '+AD4', 'U+002B');
+// as a '+AD4'.
+testDecode('UTF-7', '+AD4', 'U+002B/U+0041/U+0044/U+0034');
+testDecode('utf-7', '+AD4', 'U+002B/U+0041/U+0044/U+0034');
+
+// UTF-16LE and variants.
+testDecode('UTF-16LE', '%69%D8%D6%DE', 'U+D869/U+DED6');
+testDecode('unicodeFEFF', '%69%D8%D6%DE', 'U+D869/U+DED6');
+// According to HTML5 and for IE compatibility, UTF-16 is treated as little endian. The following tests fail as of Firefox 3.6.13.
+testDecode('UTF-16', '%69%D8%D6%DE', 'U+D869/U+DED6');
+testDecode('ISO-10646-UCS-2', '%69%D8%D6%DE', 'U+D869/U+DED6');
+testDecode('UCS-2', '%69%D8%D6%DE', 'U+D869/U+DED6');
+testDecode('Unicode', '%69%D8%D6%DE', 'U+D869/U+DED6');
+testDecode('csUnicode', '%69%D8%D6%DE', 'U+D869/U+DED6');
+
+// UTF-16BE and variants.
+testDecode('UTF-16BE', '%D8%69%DE%D6', 'U+D869/U+DED6');
+testDecode('unicodeFFFE', '%D8%69%DE%D6', 'U+D869/U+DED6');
 
 successfullyParsed = true;

trunk/LayoutTests/fast/encoding/resources/char-decoding-utils.js

@@ -10 +10 @@
 }
 
-function decode(charsetName, characterSequence)
+function decodeText(charsetName, characterSequence)
 {
     var req = new XMLHttpRequest;
@@ -16 +16 @@
     req.overrideMimeType('text/plain; charset="' + charsetName + '"');
     req.send('');
-    var code = hex(req.responseText.charCodeAt(0));
-    return "U+" + ("0000" + code).substr(code.length, 4);
+    return req.responseText;
+}
+
+function decode(charsetName, characterSequence)
+{
+    var decodedText = decodeText(charsetName, characterSequence);
+    var result = "";
+    for (var i = 0; i < decodedText.length; ++i) {
+        var code = hex(decodedText.charCodeAt(i));
+        if (i)
+            result += "/";
+        result += "U+" + ("0000" + code).substr(code.length, 4);
+    }
+    return result;
 }