author | jiangli |
Wed, 27 Sep 2017 17:55:20 -0400 | |
changeset 47548 | 664b9d44db74 |
parent 47216 | 71c04702a3d5 |
permissions | -rw-r--r-- |
2294 | 1 |
/* |
23010
6dadb192ad81
8029235: Update copyright year to match last edit in jdk8 jdk repository for 2013
lana
parents:
21596
diff
changeset
|
2 |
* Copyright (c) 2000, 2013, 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 |
|
33663 | 25 |
* @bug 6636323 6636319 7040220 7096080 7183053 8080248 8054307 |
30820 | 26 |
* @summary Test if StringCoding and NIO result have the same de/encoding result |
27 |
* @modules java.base/sun.nio.cs |
|
5970
d4e98bbfb0be
6963027: TEST_BUG: channels and buffer tests need to run in samevm mode
alanb
parents:
5506
diff
changeset
|
28 |
* @run main/othervm/timeout=2000 TestStringCoding |
30046 | 29 |
* @key randomness |
2294 | 30 |
*/ |
31 |
||
32 |
import java.util.*; |
|
33 |
import java.nio.*; |
|
34 |
import java.nio.charset.*; |
|
35 |
||
36 |
public class TestStringCoding { |
|
37 |
public static void main(String[] args) throws Throwable { |
|
38 |
||
33663 | 39 |
// full bmp first |
40 |
char[] bmp = new char[0x10000]; |
|
41 |
for (int i = 0; i < 0x10000; i++) { |
|
42 |
bmp[i] = (char)i; |
|
43 |
} |
|
44 |
char[] latin = Arrays.copyOf(bmp, 0x100); |
|
45 |
char[] ascii = Arrays.copyOf(bmp, 0x80); |
|
46 |
||
47 |
byte[] latinBA = new byte[0x100]; |
|
48 |
for (int i = 0; i < 0x100; i++) { |
|
49 |
latinBA[i] = (byte)i; |
|
50 |
} |
|
51 |
byte[] asciiBA = Arrays.copyOf(latinBA, 0x80); |
|
52 |
||
2294 | 53 |
for (Boolean hasSM: new boolean[] { false, true }) { |
33663 | 54 |
if (hasSM) { |
2294 | 55 |
System.setSecurityManager(new PermissiveSecurityManger()); |
33663 | 56 |
} |
2294 | 57 |
for (Charset cs: Charset.availableCharsets().values()) { |
58 |
if ("ISO-2022-CN".equals(cs.name()) || |
|
59 |
"x-COMPOUND_TEXT".equals(cs.name()) || |
|
60 |
"x-JISAutoDetect".equals(cs.name())) |
|
61 |
continue; |
|
62 |
System.out.printf("Testing(sm=%b) " + cs.name() + "....", hasSM); |
|
33663 | 63 |
|
64 |
testNewString(cs, testGetBytes(cs, new String(bmp))); |
|
65 |
testNewString(cs, testGetBytes(cs, new String(latin))); |
|
66 |
testNewString(cs, testGetBytes(cs, new String(ascii))); |
|
67 |
testGetBytes(cs, testNewString(cs, latinBA)); |
|
68 |
testGetBytes(cs, testNewString(cs, asciiBA)); |
|
69 |
||
2294 | 70 |
// "randomed" sizes |
71 |
Random rnd = new Random(); |
|
72 |
for (int i = 0; i < 10; i++) { |
|
73 |
//System.out.printf(" blen=%d, clen=%d%n", blen, clen); |
|
33663 | 74 |
char[] bmp0 = Arrays.copyOf(bmp, rnd.nextInt(0x10000)); |
75 |
testNewString(cs, testGetBytes(cs, new String(bmp0))); |
|
2294 | 76 |
//add a pair of surrogates |
33663 | 77 |
int pos = bmp0.length / 2; |
78 |
if ((pos + 1) < bmp0.length) { |
|
79 |
bmp0[pos] = '\uD800'; |
|
80 |
bmp0[pos+1] = '\uDC00'; |
|
2294 | 81 |
} |
33663 | 82 |
testNewString(cs, testGetBytes(cs, new String(bmp0))); |
83 |
||
84 |
char[] latin0 = Arrays.copyOf(latin, rnd.nextInt(0x100)); |
|
85 |
char[] ascii0 = Arrays.copyOf(ascii, rnd.nextInt(0x80)); |
|
86 |
byte[] latinBA0 = Arrays.copyOf(latinBA, rnd.nextInt(0x100)); |
|
87 |
byte[] asciiBA0 = Arrays.copyOf(asciiBA, rnd.nextInt(0x80)); |
|
88 |
testNewString(cs, testGetBytes(cs, new String(latin0))); |
|
89 |
testNewString(cs, testGetBytes(cs, new String(ascii0))); |
|
90 |
testGetBytes(cs, testNewString(cs, latinBA0)); |
|
91 |
testGetBytes(cs, testNewString(cs, asciiBA0)); |
|
2294 | 92 |
} |
33663 | 93 |
testSurrogates(cs); |
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
94 |
testMixed(cs); |
2294 | 95 |
System.out.println("done!"); |
96 |
} |
|
97 |
} |
|
98 |
} |
|
99 |
||
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
100 |
static void testMixed(Charset cs) throws Throwable { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
101 |
CharsetDecoder dec = cs.newDecoder() |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
102 |
.onMalformedInput(CodingErrorAction.REPLACE) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
103 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
104 |
CharsetEncoder enc = cs.newEncoder() |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
105 |
.onMalformedInput(CodingErrorAction.REPLACE) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
106 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
107 |
List<Integer> cps = new ArrayList<>(0x10000); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
108 |
int off = 0; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
109 |
int cp = 0; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
110 |
while (cp < 0x10000) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
111 |
if (enc.canEncode((char)cp)) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
112 |
cps.add(cp); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
113 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
114 |
cp++; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
115 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
116 |
Collections.shuffle(cps); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
117 |
char[] bmpCA = new char[cps.size()]; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
118 |
for (int i = 0; i < cps.size(); i++) |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
119 |
bmpCA[i] = (char)(int)cps.get(i); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
120 |
String bmpStr = new String(bmpCA); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
121 |
//getBytes(csn); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
122 |
byte[] bmpBA = bmpStr.getBytes(cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
123 |
ByteBuffer bf = enc.reset().encode(CharBuffer.wrap(bmpCA)); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
124 |
byte[] baNIO = new byte[bf.limit()]; |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
125 |
bf.get(baNIO, 0, baNIO.length); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
126 |
if (!Arrays.equals(bmpBA, baNIO)) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
127 |
throw new RuntimeException("getBytes(csn) failed -> " + cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
128 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
129 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
130 |
//getBytes(cs); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
131 |
bmpBA = bmpStr.getBytes(cs); |
33663 | 132 |
if (!Arrays.equals(bmpBA, baNIO)) { |
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
133 |
throw new RuntimeException("getBytes(cs) failed -> " + cs.name()); |
33663 | 134 |
} |
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
135 |
|
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
136 |
//new String(csn); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
137 |
String strSC = new String(bmpBA, cs.name()); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
138 |
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
|
139 |
if(!strNIO.equals(strSC)) { |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
140 |
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
|
141 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
142 |
//new String(cs); |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
143 |
strSC = new String(bmpBA, cs); |
33663 | 144 |
if (!strNIO.equals(strSC)) { |
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
145 |
throw new RuntimeException("new String(cs) failed -> " + cs.name()); |
33663 | 146 |
} |
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
147 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
148 |
|
33663 | 149 |
static byte[] getBytes(CharsetEncoder enc, String str) throws Throwable { |
150 |
ByteBuffer bf = enc.reset().encode(CharBuffer.wrap(str.toCharArray())); |
|
151 |
byte[] ba = new byte[bf.limit()]; |
|
152 |
bf.get(ba, 0, ba.length); |
|
153 |
return ba; |
|
154 |
} |
|
155 |
||
156 |
static byte[] testGetBytes(Charset cs, String str) throws Throwable { |
|
2294 | 157 |
CharsetEncoder enc = cs.newEncoder() |
158 |
.onMalformedInput(CodingErrorAction.REPLACE) |
|
159 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
|
160 |
//getBytes(csn); |
|
33663 | 161 |
byte[] baSC = str.getBytes(cs.name()); |
162 |
byte[] baNIO = getBytes(enc, str); |
|
163 |
if (!Arrays.equals(baSC, baNIO)) { |
|
2294 | 164 |
throw new RuntimeException("getBytes(csn) failed -> " + cs.name()); |
33663 | 165 |
} |
2294 | 166 |
//getBytes(cs); |
33663 | 167 |
baSC = str.getBytes(cs); |
168 |
if (!Arrays.equals(baSC, baNIO)) { |
|
2294 | 169 |
throw new RuntimeException("getBytes(cs) failed -> " + cs.name()); |
33663 | 170 |
} |
171 |
return baSC; |
|
172 |
} |
|
2294 | 173 |
|
33663 | 174 |
static String testNewString(Charset cs, byte[] ba) throws Throwable { |
175 |
CharsetDecoder dec = cs.newDecoder() |
|
176 |
.onMalformedInput(CodingErrorAction.REPLACE) |
|
177 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
|
2294 | 178 |
//new String(csn); |
33663 | 179 |
String strSC = new String(ba, cs.name()); |
180 |
String strNIO = dec.reset().decode(ByteBuffer.wrap(ba)).toString(); |
|
181 |
if(!strNIO.equals(strSC)) { |
|
2294 | 182 |
throw new RuntimeException("new String(csn) failed -> " + cs.name()); |
33663 | 183 |
} |
184 |
//new String(cs); |
|
185 |
strSC = new String(ba, cs); |
|
186 |
if (!strNIO.equals(strSC)) { |
|
187 |
throw new RuntimeException("new String(cs)/bmp failed -> " + cs.name()); |
|
188 |
} |
|
189 |
return strSC; |
|
190 |
} |
|
2294 | 191 |
|
33663 | 192 |
static void testSurrogates(Charset cs) throws Throwable { |
2294 | 193 |
//encode unmappable surrogates |
33663 | 194 |
CharsetEncoder enc = cs.newEncoder() |
195 |
.onMalformedInput(CodingErrorAction.REPLACE) |
|
196 |
.onUnmappableCharacter(CodingErrorAction.REPLACE); |
|
2294 | 197 |
if (enc instanceof sun.nio.cs.ArrayEncoder && |
198 |
cs.contains(Charset.forName("ASCII"))) { |
|
10898 | 199 |
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
|
200 |
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
|
201 |
return; |
2294 | 202 |
enc.replaceWith(new byte[] { (byte)'A'}); |
203 |
sun.nio.cs.ArrayEncoder cae = (sun.nio.cs.ArrayEncoder)enc; |
|
204 |
||
205 |
String str = "ab\uD800\uDC00\uD800\uDC00cd"; |
|
206 |
byte[] ba = new byte[str.length() - 2]; |
|
207 |
int n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
208 |
if (n != 6 || !"abAAcd".equals(new String(ba, cs.name()))) |
|
209 |
throw new RuntimeException("encode1(surrogates) failed -> " |
|
210 |
+ cs.name()); |
|
211 |
||
212 |
ba = new byte[str.length()]; |
|
213 |
n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
214 |
if (n != 6 || !"abAAcd".equals(new String(ba, 0, n, |
|
215 |
cs.name()))) |
|
216 |
throw new RuntimeException("encode2(surrogates) failed -> " |
|
217 |
+ cs.name()); |
|
218 |
str = "ab\uD800B\uDC00Bcd"; |
|
219 |
ba = new byte[str.length()]; |
|
220 |
n = cae.encode(str.toCharArray(), 0, str.length(), ba); |
|
221 |
if (n != 8 || !"abABABcd".equals(new String(ba, 0, n, |
|
222 |
cs.name()))) |
|
223 |
throw new RuntimeException("encode3(surrogates) failed -> " |
|
224 |
+ cs.name()); |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
225 |
/* 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
|
226 |
invoker (StringCoder) allocates enough output buf, utf8 |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
227 |
and double-byte coder does not check the output buffer limit. |
2294 | 228 |
ba = new byte[str.length() - 1]; |
229 |
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
|
230 |
if (n != 7 || !"abABABc".equals(new String(ba, 0, n, cs.name()))) { |
2294 | 231 |
throw new RuntimeException("encode4(surrogates) failed -> " |
232 |
+ cs.name()); |
|
13257
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
233 |
} |
5cf4ff2cbb37
7183053: Optimize DoubleByte charset for String.getBytes()/new String(byte[])
sherman
parents:
10898
diff
changeset
|
234 |
*/ |
2294 | 235 |
} |
236 |
||
30691 | 237 |
//encode mappable surrogates for hkscs |
238 |
if (cs.name().equals("Big5-HKSCS") || cs.name().equals("x-MS950-HKSCS")) { |
|
239 |
String str = "ab\uD840\uDD0Ccd"; |
|
240 |
byte[] expected = new byte[] {(byte)'a', (byte)'b', |
|
241 |
(byte)0x88, (byte)0x45, (byte)'c', (byte)'d' }; |
|
242 |
if (!Arrays.equals(str.getBytes(cs.name()), expected) || |
|
243 |
!Arrays.equals(str.getBytes(cs), expected)) { |
|
244 |
throw new RuntimeException("encode(surrogates) failed -> " |
|
245 |
+ cs.name()); |
|
246 |
} |
|
247 |
} |
|
2294 | 248 |
} |
249 |
||
250 |
static class PermissiveSecurityManger extends SecurityManager { |
|
251 |
@Override public void checkPermission(java.security.Permission p) {} |
|
252 |
} |
|
253 |
} |