author | serb |
Tue, 12 Nov 2013 20:24:25 +0400 | |
changeset 21596 | 0e3a39f29dbc |
parent 14342 | 8435a30053c1 |
child 23010 | 6dadb192ad81 |
permissions | -rw-r--r-- |
2294 | 1 |
/* |
14342
8435a30053c1
7197491: update copyright year to match last edit in jdk8 jdk repository
alanb
parents:
13257
diff
changeset
|
2 |
* Copyright (c) 2000, 2012, Oracle and/or its affiliates. All rights reserved. |
2294 | 3 |
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. |
4 |
* |
|
5 |
* This code is free software; you can redistribute it and/or modify it |
|
6 |
* under the terms of the GNU General Public License version 2 only, as |
|
21596 | 7 |
* published by the Free Software Foundation. |
2294 | 8 |
* |
9 |
* This code is distributed in the hope that it will be useful, but WITHOUT |
|
10 |
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
|
11 |
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
|
12 |
* version 2 for more details (a copy is included in the LICENSE file that |
|
13 |
* accompanied this code). |
|
14 |
* |
|
15 |
* You should have received a copy of the GNU General Public License version |
|
16 |
* 2 along with this work; if not, write to the Free Software Foundation, |
|
17 |
* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. |
|
18 |
* |
|
5506 | 19 |
* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA |
20 |
* or visit www.oracle.com if you need additional information or have any |
|
21 |
* questions. |
|
2294 | 22 |
*/ |
23 |
||
24 |
/* @test |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
25 |
@bug 6636323 6636319 7040220 7096080 7183053 |
2294 | 26 |
@summary Test if StringCoding and NIO result have the same de/encoding result |
5970
d4e98bbfb0be
6963027: TEST_BUG: channels and buffer tests need to run in samevm mode
alanb
parents:
5506
diff
changeset
|
27 |
* @run main/othervm/timeout=2000 TestStringCoding |
2294 | 28 |
*/ |
29 |
||
30 |
import java.util.*; |
|
31 |
import java.nio.*; |
|
32 |
import java.nio.charset.*; |
|
33 |
||
34 |
public class TestStringCoding { |
|
35 |
public static void main(String[] args) throws Throwable { |
|
36 |
||
37 |
for (Boolean hasSM: new boolean[] { false, true }) { |
|
38 |
if (hasSM) |
|
39 |
System.setSecurityManager(new PermissiveSecurityManger()); |
|
40 |
for (Charset cs: Charset.availableCharsets().values()) { |
|
41 |
if ("ISO-2022-CN".equals(cs.name()) || |
|
42 |
"x-COMPOUND_TEXT".equals(cs.name()) || |
|
43 |
"x-JISAutoDetect".equals(cs.name())) |
|
44 |
continue; |
|
45 |
System.out.printf("Testing(sm=%b) " + cs.name() + "....", hasSM); |
|
46 |
// full bmp first |
|
47 |
char[] bmpCA = new char[0x10000]; |
|
48 |
for (int i = 0; i < 0x10000; i++) { |
|
49 |
bmpCA[i] = (char)i; |
|
50 |
} |
|
51 |
byte[] sbBA = new byte[0x100]; |
|
52 |
for (int i = 0; i < 0x100; i++) { |
|
53 |
sbBA[i] = (byte)i; |
|
54 |
} |
|
55 |
test(cs, bmpCA, sbBA); |
|
56 |
// "randomed" sizes |
|
57 |
Random rnd = new Random(); |
|
58 |
for (int i = 0; i < 10; i++) { |
|
59 |
int clen = rnd.nextInt(0x10000); |
|
60 |
int blen = rnd.nextInt(0x100); |
|
61 |
//System.out.printf(" blen=%d, clen=%d%n", blen, clen); |
|
62 |
test(cs, Arrays.copyOf(bmpCA, clen), Arrays.copyOf(sbBA, blen)); |
|
63 |
//add a pair of surrogates |
|
64 |
int pos = clen / 2; |
|
65 |
if ((pos + 1) < blen) { |
|
66 |
bmpCA[pos] = '\uD800'; |
|
67 |
bmpCA[pos+1] = '\uDC00'; |
|
68 |
} |
|
69 |
test(cs, Arrays.copyOf(bmpCA, clen), Arrays.copyOf(sbBA, blen)); |
|
70 |
} |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
71 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
72 |
testMixed(cs); |
2294 | 73 |
System.out.println("done!"); |
74 |
} |
|
75 |
} |
|
76 |
} |
|
77 |
||
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
78 |
static void testMixed(Charset cs) throws Throwable { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
79 |
CharsetDecoder dec = cs.newDecoder() |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
80 |
.onMalformedInput(CodingErrorAction.REPLACE) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
81 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
82 |
CharsetEncoder enc = cs.newEncoder() |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
83 |
.onMalformedInput(CodingErrorAction.REPLACE) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
84 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
85 |
List<Integer> cps = new ArrayList<>(0x10000); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
86 |
int off = 0; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
87 |
int cp = 0; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
88 |
while (cp < 0x10000) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
89 |
if (enc.canEncode((char)cp)) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
90 |
cps.add(cp); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
91 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
92 |
cp++; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
93 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
94 |
Collections.shuffle(cps); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
95 |
char[] bmpCA = new char[cps.size()]; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
96 |
for (int i = 0; i < cps.size(); i++) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
97 |
bmpCA[i] = (char)(int)cps.get(i); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
98 |
String bmpStr = new String(bmpCA); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
99 |
//getBytes(csn); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
100 |
byte[] bmpBA = bmpStr.getBytes(cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
101 |
ByteBuffer bf = enc.reset().encode(CharBuffer.wrap(bmpCA)); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
102 |
byte[] baNIO = new byte[bf.limit()]; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
103 |
bf.get(baNIO, 0, baNIO.length); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
104 |
if (!Arrays.equals(bmpBA, baNIO)) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
105 |
throw new RuntimeException("getBytes(csn) failed -> " + cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
106 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
107 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
108 |
//getBytes(cs); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
109 |
bmpBA = bmpStr.getBytes(cs); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
110 |
if (!Arrays.equals(bmpBA, baNIO)) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
111 |
throw new RuntimeException("getBytes(cs) failed -> " + cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
112 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
113 |
//new String(csn); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
114 |
String strSC = new String(bmpBA, cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
115 |
String strNIO = dec.reset().decode(ByteBuffer.wrap(bmpBA)).toString(); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
116 |
if(!strNIO.equals(strSC)) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
117 |
throw new RuntimeException("new String(csn) failed -> " + cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
118 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
119 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
120 |
//new String(cs); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
121 |
strSC = new String(bmpBA, cs); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
122 |
if (!strNIO.equals(strSC)) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
123 |
throw new RuntimeException("new String(cs) failed -> " + cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
124 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
125 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
126 |
|
2294 | 127 |
static void test(Charset cs, char[] bmpCA, byte[] sbBA) throws Throwable { |
128 |
String bmpStr = new String(bmpCA); |
|
129 |
CharsetDecoder dec = cs.newDecoder() |
|
130 |
.onMalformedInput(CodingErrorAction.REPLACE) |
|
131 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
|
132 |
CharsetEncoder enc = cs.newEncoder() |
|
133 |
.onMalformedInput(CodingErrorAction.REPLACE) |
|
134 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
|
135 |
||
136 |
//getBytes(csn); |
|
137 |
byte[] baSC = bmpStr.getBytes(cs.name()); |
|
138 |
ByteBuffer bf = enc.reset().encode(CharBuffer.wrap(bmpCA)); |
|
139 |
byte[] baNIO = new byte[bf.limit()]; |
|
140 |
bf.get(baNIO, 0, baNIO.length); |
|
141 |
if (!Arrays.equals(baSC, baNIO)) |
|
142 |
throw new RuntimeException("getBytes(csn) failed -> " + cs.name()); |
|
143 |
||
144 |
//getBytes(cs); |
|
145 |
baSC = bmpStr.getBytes(cs); |
|
146 |
if (!Arrays.equals(baSC, baNIO)) |
|
147 |
throw new RuntimeException("getBytes(cs) failed -> " + cs.name()); |
|
148 |
||
149 |
//new String(csn); |
|
150 |
String strSC = new String(sbBA, cs.name()); |
|
151 |
String strNIO = dec.reset().decode(ByteBuffer.wrap(sbBA)).toString(); |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
152 |
|
2294 | 153 |
if(!strNIO.equals(strSC)) |
154 |
throw new RuntimeException("new String(csn) failed -> " + cs.name()); |
|
155 |
||
156 |
//new String(cs); |
|
157 |
strSC = new String(sbBA, cs); |
|
158 |
if (!strNIO.equals(strSC)) |
|
159 |
throw new RuntimeException("new String(cs) failed -> " + cs.name()); |
|
160 |
||
161 |
//encode unmappable surrogates |
|
162 |
if (enc instanceof sun.nio.cs.ArrayEncoder && |
|
163 |
cs.contains(Charset.forName("ASCII"))) { |
|
10898 | 164 |
if (cs.name().equals("UTF-8") || // utf8 handles surrogates |
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
165 |
cs.name().equals("CESU-8")) // utf8 handles surrogates |
9547
454881baaca0
7040220: java/char_encodin Optimize UTF-8 charset for String.getBytes()/new String(byte[])
sherman
parents:
7668
diff
changeset
|
166 |
return; |
2294 | 167 |
enc.replaceWith(new byte[] { (byte)'A'}); |
168 |
sun.nio.cs.ArrayEncoder cae = (sun.nio.cs.ArrayEncoder)enc; |
|
169 |
||
170 |
String str = "ab\uD800\uDC00\uD800\uDC00cd"; |
|
171 |
byte[] ba = new byte[str.length() - 2]; |
|
172 |
int n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
173 |
if (n != 6 || !"abAAcd".equals(new String(ba, cs.name()))) |
|
174 |
throw new RuntimeException("encode1(surrogates) failed -> " |
|
175 |
+ cs.name()); |
|
176 |
||
177 |
ba = new byte[str.length()]; |
|
178 |
n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
179 |
if (n != 6 || !"abAAcd".equals(new String(ba, 0, n, |
|
180 |
cs.name()))) |
|
181 |
throw new RuntimeException("encode2(surrogates) failed -> " |
|
182 |
+ cs.name()); |
|
183 |
str = "ab\uD800B\uDC00Bcd"; |
|
184 |
ba = new byte[str.length()]; |
|
185 |
n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
186 |
if (n != 8 || !"abABABcd".equals(new String(ba, 0, n, |
|
187 |
cs.name()))) |
|
188 |
throw new RuntimeException("encode3(surrogates) failed -> " |
|
189 |
+ cs.name()); |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
190 |
/* sun.nio.cs.ArrayDeEncoder works on the assumption that the |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
191 |
invoker (StringCoder) allocates enough output buf, utf8 |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
192 |
and double-byte coder does not check the output buffer limit. |
2294 | 193 |
ba = new byte[str.length() - 1]; |
194 |
n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
195 |
if (n != 7 || !"abABABc".equals(new String(ba, 0, n, cs.name()))) { |
2294 | 196 |
throw new RuntimeException("encode4(surrogates) failed -> " |
197 |
+ cs.name()); |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
198 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
199 |
*/ |
2294 | 200 |
} |
201 |
||
202 |
} |
|
203 |
||
204 |
static class PermissiveSecurityManger extends SecurityManager { |
|
205 |
@Override public void checkPermission(java.security.Permission p) {} |
|
206 |
} |
|
207 |
} |