1 /*
2 * Copyright (c) 2007, 2017, Oracle and/or its affiliates. All rights reserved.
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
7 * published by the Free Software Foundation.
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 *
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.
22 */
23 /*
24 * @test
25 * @bug 4052473 4052679 4055602 4066550 4067619 4068012 4068073 4070174 4070452
26 * 4070178 4070450 4070695 4070725 4070795 4071003 4071183 4071782 4072013
27 * 4072388 4072773 4075404 4084356 4087238 4092361 4094033 4094371 4098518
28 * 4099810 4103218 4103220 4103861 4112136 4113638 4113654 4117054 4122468
29 * 4122840 4139860 4156708 4175306 4215747 4209960 4290801 4900884 4942982
30 * 4518811 4945388 4936845 4794068 4461740 4965260 4984277 4826794 5032580
31 * 5102005 5074431 6182685 6208712 6277020 6245766 6351682 6386647 6379382
32 * 6414459 6455680 6498742 6558863 6488119 6547501 6497154 6558856 6481177
33 * 6379214 6485516 6486607 4225362 4494727 6533691 6531591 6531593 6570259
34 * 6509039 6609737 6610748 6645271 6507067 6873931 6450945 6645268 6646611
35 * 6645405 6650730 6910489 6573250 6870908 6585666 6716626 6914413 6916787
36 * 6919624 6998391 7019267 7020960 7025837 7020583 7036905 7066203 7101495
37 * 7003124 7085757 7028073 7171028 7189611 8000983 7195759 8004489 8006509
38 * 7114053 7074882 7040556 8008577 8013836 8021121 6192407 6931564 8027695
39 * 8017142 8037343 8055222 8042126 8074791 8075173 8080774 8129361 8134916
40 * 8145136 8145952 8164784 8037111 8081643 7037368 8178872
41 * @summary Verify locale data
42 * @modules java.base/sun.util.resources
43 * @modules jdk.localedata
44 * @run main LocaleDataTest
45 * @run main LocaleDataTest -cldr
46 *
47 */
48
49 /*
50 *
51 * (C) Copyright Taligent, Inc. 1996, 1997 - All Rights Reserved
52 * (C) Copyright IBM Corp. 1996 - 1998 - All Rights Reserved
53 *
54 * Portions copyright (c) 2007 Sun Microsystems, Inc.
55 * All Rights Reserved.
56 *
57 * The original version of this source code and documentation
58 * is copyrighted and owned by Taligent, Inc., a wholly-owned
59 * subsidiary of IBM. These materials are provided under terms
60 * of a License Agreement between Taligent and Sun. This technology
61 * is protected by multiple US and International patents.
62 *
63 * This notice and attribution to Taligent may not be removed.
64 * Taligent is a registered trademark of Taligent, Inc.
65 *
66 * Permission to use, copy, modify, and distribute this software
67 * and its documentation for NON-COMMERCIAL purposes and without
68 * fee is hereby granted provided that this copyright notice
69 * appears in all copies. Please refer to the file "copyright.html"
70 * for further important copyright and licensing information.
71 *
72 * SUN MAKES NO REPRESENTATIONS OR WARRANTIES ABOUT THE SUITABILITY OF
73 * THE SOFTWARE, EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED
74 * TO THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
75 * PARTICULAR PURPOSE, OR NON-INFRINGEMENT. SUN SHALL NOT BE LIABLE FOR
76 * ANY DAMAGES SUFFERED BY LICENSEE AS A RESULT OF USING, MODIFYING OR
77 * DISTRIBUTING THIS SOFTWARE OR ITS DERIVATIVES.
78 *
79 */
80
81 /* This test is a generalized test for verifying changes to the locale data.
82 * It is driven by an external file that specifies the particular pieces of locale
83 * data to check. That file is in .properties file format: a series of key/value
84 * pairs delimited by newline characters, with the keys separated from the values
85 * by = signs. The keys are similar in syntax to a Unix pathname, with keys at
86 * successive levels of containment in the resource-data hierarchy separated by
87 * slashes. The file is in ISO 8859-1 encoding, with control characters and
88 * non-ASCII characters denoted with backslash-u escape sequences. The program also allows
89 * blank lines and comment lines to be interspersed with the data. Comment lines
90 * begin with '#'.
91 *
92 * A data file for this test would look something like this:<pre>
93 * FormatData//MonthNames/0=January
94 * FormatData//MonthNames/1=February
95 * LocaleNames//US=United States
96 * LocaleNames//FR=France
97 * FormatData/fr_FR/MonthNames/0=janvier
98 * FormatData/fr_FR/MonthNames/1=f\u00e9vrier
99 * LocaleNames/fr_FR/US=\u00c9tats-Unis
100 * LocaleNames/fr_FR/FR=France</pre>
101 *
102 * Second field which designates locale is in the form of:
103 * 1) Legacy locale notation using '_' as a locale component(language/country/variant) separator.
104 * language is a mandatory component. country and variant are optional, however,
105 * variant cannot exist without country. So for example, while "ja"/"ja_JP"/"ja_JP_JP" are valid,
106 * "_JP"/"ja__JP" are invalid.
107 *
108 * 2) BCP47 language tag notation in which we can specify language tag with '-' as a subtag
109 * separator. Language tag can be specified with '-' in locale field like this:
110 * <pre>LocaleNames/sr-Latn/SR=Surinam
111 * FormatData/sr-Latn-BA/DayNames/2=utorak</pre>
112 *
113 * The command-line syntax of this test is
114 * <tt>java LocaleDataTest [-w] [{ -s | <filename> }] [-cldr]</tt>
115 *
116 * This program always sends its results to standard output. If -w is not specified,
117 * this program prints out only the differences between the data file and the actual
118 * resource data. If -w is specified, the program prints out every entry, comment,
119 * and blank line from the data file. Where there is a difference between the data
120 * file and the resource data, the data is the data from the resources. This feature
121 * can be used to quickly generate a new data file.
122 *
123 * The user can specify an optional filename or -s. If the user specifies a filename,
124 * the program uses that file as the data file. If the user specifies -s, the program
125 * reads its input from standard input rather than from a file. If the user specifies
126 * neither, the program reads its input from a file called LocaleData in the same
127 * directory the program itself resides in.
128 *
129 * The -nothrow option prevents the program from throwing an exception when it
130 * gets an error. -w implies -nothrow.
131 *
132 * -cldr option specifies to test CLDR locale data. The default data file name for this
133 * option is "LocaleData.cldr".
134 *
135 * Other command-line options can be specified, but are ignored.
136 *
137 * It's important to note what this test will NOT test. Certain changes to the locale
138 * data are meant to have certain effects on the internationalization frameworks. For
139 * instance, we could ensure round-trip formatting/parsing integrity for the full
140 * date/time format of SimpleDateFormat by making sure that the full date and time
141 * patterns include sufficient data. The test of this is not whether changes were
142 * made to the locale data; it's whether using this data gives round-trip integrity.
143 * Likewise, changing the currency patterns to use \u00a4 instead of local currency
144 * symbols isn't something that can be tested by this test; instead, you want to
145 * actually format currency values and make sure the proper currency symbol was used.
146 *
147 * This test by itself doesn't do an exhaustive comparison of locale data. It is
148 * possible to do this manually, however: Use the GenerateKeyList tool to produce
149 * a complete list of keys for the two versions of the locales you want to compare,
150 * and then diff them. This will flag additions and deletions. Generate a data file
151 * for the base version of the data using the -w option and the output from
152 * GenerateKeyList, and then use the resultant file as the data file when you run
153 * this test against the new version of the data.
154 */
155
156 import java.io.BufferedReader;
157 import java.io.File;
158 import java.io.FileInputStream;
159 import java.io.FilterReader;
160 import java.io.FilterWriter;
161 import java.io.IOException;
162 import java.io.InputStreamReader;
163 import java.io.OutputStreamWriter;
164 import java.io.PrintWriter;
165 import java.io.Reader;
166 import java.io.Writer;
167 import java.util.Locale;
168 import java.util.MissingResourceException;
169 import java.util.ResourceBundle;
170 import sun.util.resources.LocaleData;
171
172 public class LocaleDataTest
173 {
174 static final String TEXT_RESOURCES_PACKAGE ="sun.text.resources";
175 static final String UTIL_RESOURCES_PACKAGE ="sun.util.resources";
176 static final String DEFAULT_DATAFILE ="LocaleData";
177 static String cldrSuffix = "";
178
179 public static void main(String[] args) throws Exception {
180
181 // set up our flags and our input and output streams based on the
182 // command-line arguments (exceptions generated here will propagate out
183 // to the environment)
184 BufferedReader in = null;
185 PrintWriter out = null;
186 boolean writeNewFile = false;
187 boolean doThrow = true;
188
189 for (int i = 0; i < args.length; i++) {
190 if (args[i].equals("-w")) {
191 writeNewFile = true;
192 doThrow = false;
193 }
194
195 else if (args[i].equals("-nothrow"))
196 doThrow = false;
197
198 else if (args[i].equals("-cldr")) {
199 cldrSuffix = ".cldr";
200 }
201
202 else if (args[i].equals("-s") && in == null)
203 in = new BufferedReader(new EscapeReader(new InputStreamReader(System.in,
204 "ISO8859_1")));
205 else if (!args[i].startsWith("-") && in == null)
206 in = new BufferedReader(new EscapeReader(new InputStreamReader(new
207 FileInputStream(args[i]), "ISO8859_1")));
208 }
209 if (in == null) {
210 File localeData = new File(System.getProperty("test.src", "."), DEFAULT_DATAFILE + cldrSuffix);
211 in = new BufferedReader(new EscapeReader(new InputStreamReader(new
212 FileInputStream(localeData), "ISO8859_1")));
213 }
214 out = new PrintWriter(new EscapeWriter(new OutputStreamWriter(System.out,
215 "ISO8859_1")), true);
216
217 // perform the actual test
218 int errorCount = doTest(in, out, writeNewFile);
219
220 // write out the error count, and throw an exception out into the environment
221 // if there were any errors
222 if (errorCount != 0) {
223 if (!writeNewFile)
224 out.println("Test failed. " + errorCount + " errors.");
225 if (doThrow)
226 throw new Exception("Test failed. " + errorCount + " errors.");
227 }
228 else if (!writeNewFile)
229 out.println("Test passed.");
230
231 in.close();
232 out.close();
233 }
234
235 static int doTest(BufferedReader in, PrintWriter out, boolean writeNewFile)
236 throws Exception {
237 int errorCount = 0;
238
239 String key = null;
240 String expectedValue = null;
241 String line = in.readLine();
242 while (line != null) {
243 if (line.startsWith("#") || line.length() == 0) {
244 if (writeNewFile)
245 out.println(line);
246 }
247
248 else {
249 int index = line.indexOf("=");
250 if (index == -1) {
251 key = line;
252 expectedValue = "";
253 }
254 else {
255 key = line.substring(0, index);
256 if (index + 1 == line.length())
257 expectedValue = "";
258 else
259 expectedValue = line.substring(index + 1);
260 }
261 if (!processLine(key, expectedValue, out, writeNewFile))
262 ++errorCount;
263 }
264 line = in.readLine();
265 }
266 return errorCount;
267 }
268
269 static boolean processLine(String key, String expectedValue, PrintWriter out,
270 boolean writeNewFile) throws Exception {
271 String rbName, localeName, resTag, qualifier;
272 String language = "", country = "", variant = "";
273 int index, oldIndex;
274
275 index = key.indexOf("/");
276 if (index == -1 || index + 1 == key.length())
277 throw new Exception("Malformed input file: no slashes in \"" + key + "\"");
278 rbName = key.substring(0, index);
279
280 oldIndex = index + 1;
281 index = key.indexOf("/", oldIndex);
282 if (index == -1 || index + 1 == key.length())
283 throw new Exception("Malformed input file: \"" + key + "\" is missing locale name");
284 localeName = key.substring(oldIndex, index);
285 boolean use_tag = localeName.indexOf("-") != -1;
286 if (use_tag == false && localeName.length() > 0) {
287 String[] locDetails = localeName.split("_");
288 switch (locDetails.length) {
289 case 1:
290 language = locDetails[0];
291 break;
292 case 2:
293 language = locDetails[0];
294 country = locDetails[1];
295 break;
296 case 3:
297 language = locDetails[0];
298 country = locDetails[1];
299 variant = locDetails[2];
300 break;
301 default:
302 throw new Exception("locale not specified properly " + locDetails);
303 }
304 }
305 oldIndex = index + 1;
306 index = key.indexOf("/", oldIndex);
307 if (index == -1)
308 index = key.length();
309 resTag = key.substring(oldIndex, index);
310
311 // TimeZone name may have "/" in it, for example "Asia/Taipei", so use "Asia\/Taipei in LocaleData.
312 if(resTag.endsWith("\\")) {
313 resTag = resTag.substring(0, resTag.length() - 1);
314 oldIndex = index;
315 index = key.indexOf("/", oldIndex + 1);
316 if (index == -1) index = key.length();
317 resTag += key.substring(oldIndex, index);
318 }
319
320 if (index < key.length() - 1)
321 qualifier = key.substring(index + 1);
322 else
323 qualifier = "";
324
325 String retrievedValue = null;
326 Object resource = null;
327 try {
328 String fullName = null;
329 if (rbName.equals("CalendarData")
330 || rbName.equals("CurrencyNames")
331 || rbName.equals("LocaleNames")
332 || rbName.equals("TimeZoneNames")) {
333 fullName = UTIL_RESOURCES_PACKAGE + cldrSuffix + "." + rbName;
334 } else {
335 fullName = TEXT_RESOURCES_PACKAGE + cldrSuffix + "." + rbName;
336 }
337 Locale locale;
338 if (use_tag) {
339 locale = Locale.forLanguageTag(localeName);
340 } else {
341 locale = new Locale(language, country, variant);
342 }
343 ResourceBundle bundle = LocaleData.getBundle(fullName, locale);
344 resource = bundle.getObject(resTag);
345 }
346 catch (MissingResourceException e) {
347 }
348
349 if (resource != null) {
350 if (resource instanceof String) {
351 retrievedValue = (String)resource;
352 }
353 else if (resource instanceof String[]) {
354 int element = Integer.valueOf(qualifier).intValue();
355 String[] stringList = (String[])resource;
356 if (element >= 0 && element < stringList.length)
357 retrievedValue = stringList[element];
358 }
359 else if (resource instanceof String[][]) {
360 String[][] stringArray = (String[][])resource;
361 int slash = qualifier.indexOf("/");
362 if (slash == -1) {
363 for (int i = 0; i < stringArray.length; i++) {
364 if (stringArray[i][0].equals(qualifier))
365 retrievedValue = stringArray[i][1];
366 }
367 }
368 else {
369 int row = Integer.valueOf(qualifier.substring(0, slash)).intValue();
370 int column = Integer.valueOf(qualifier.substring(slash + 1)).intValue();
371 if (row >= 0 && row < stringArray.length && column >= 0 && column <
372 stringArray[row].length)
373 retrievedValue = stringArray[row][column];
374 }
375 }
376 }
377
378 if (retrievedValue == null || !retrievedValue.equals(expectedValue)) {
379 if (retrievedValue == null)
380 retrievedValue = "<MISSING!>";
381
382 if (writeNewFile)
383 out.println(key + "=" + retrievedValue);
384 else {
385 out.println("Mismatch in " + key + ":");
386 out.println(" file = \"" + expectedValue + "\"");
387 out.println(" jvm = \"" + retrievedValue + "\"");
388 }
389 return false;
390 }
391 else {
392 if (writeNewFile)
393 out.println(key + "=" + expectedValue);
394 }
395 return true;
396 }
397 }
398
399 class EscapeReader extends FilterReader {
400 public EscapeReader(Reader in) {
401 super(in);
402 }
403
404 public int read() throws IOException {
405 if (buffer != null) {
406 String b = buffer.toString();
407 int result = b.charAt(0);
408 if (b.length() > 1)
409 buffer = new StringBuffer(b.substring(1));
410 else
411 buffer = null;
412 return result;
413 }
414 else {
415 int result = super.read();
416 if (result != '\\')
417 return result;
418 else {
419 buffer = new StringBuffer();
420 result = super.read();
421 buffer.append((char)result);
422 if (result == 'u') {
423 for (int i = 0; i < 4; i++) {
424 result = super.read();
425 if (result == -1)
426 break;
427 buffer.append((char)result);
428 }
429 String number = buffer.toString().substring(1);
430 result = Integer.parseInt(number, 16);
431 buffer = null;
432 return result;
433 }
434 return '\\';
435 }
436 }
437 }
438
439 public int read(char[] cbuf, int start, int len) throws IOException {
440 int p = start;
441 int end = start + len;
442 int c = 0;
443 while (c != -1 && p < end) {
444 c = read();
445 if (c != -1)
446 cbuf[p++] = (char)c;
447 }
448 if (c == -1 && p == start)
449 return -1;
450 else
451 return p - start;
452 }
453
454 private StringBuffer buffer = null;
455 }
456
457 class EscapeWriter extends FilterWriter {
458 public EscapeWriter(Writer out) {
459 super(out);
460 }
461
462 public void write(int c) throws IOException {
463 if ((c >= ' ' && c <= '\u007e') || c == '\r' || c == '\n')
464 super.write(c);
465 else {
466 super.write('\\');
467 super.write('u');
468 String number = Integer.toHexString(c);
469 if (number.length() < 4)
470 number = zeros.substring(0, 4 - number.length()) + number;
471 super.write(number.charAt(0));
472 super.write(number.charAt(1));
473 super.write(number.charAt(2));
474 super.write(number.charAt(3));
475 }
476 }
477
478 public void write(char[] cbuf, int off, int len) throws IOException {
479 int end = off + len;
480 while (off < end)
481 write(cbuf[off++]);
482 }
483
484 public void write(String str, int off, int len) throws IOException {
485 int end = off + len;
486 while (off < end)
487 write(str.charAt(off++));
488 }
489
490 private static String zeros = "0000";
491 }