1 /*
2 * Copyright (c) 2012, 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. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25
26 package build.tools.cldrconverter;
27
28 import java.io.File;
29 import java.io.IOException;
30 import java.util.ArrayList;
31 import java.util.HashMap;
32 import java.util.List;
33 import java.util.Locale;
34 import java.util.Map;
35 import org.xml.sax.Attributes;
36 import org.xml.sax.InputSource;
37 import org.xml.sax.SAXException;
38
39 /**
40 * Handles parsing of files in Locale Data Markup Language and produces a map
41 * that uses the keys and values of JRE locale data.
42 */
43 class LDMLParseHandler extends AbstractLDMLHandler<Object> {
44 private String defaultNumberingSystem;
45 private String currentNumberingSystem = "";
46 private CalendarType currentCalendarType;
47 private String zoneNameStyle; // "long" or "short" for time zone names
48 private String zonePrefix;
49 private final String id;
50
51 LDMLParseHandler(String id) {
52 this.id = id;
53 }
54
55 @Override
56 public InputSource resolveEntity(String publicID, String systemID) throws IOException, SAXException {
57 // avoid HTTP traffic to unicode.org
58 if (systemID.startsWith(CLDRConverter.LDML_DTD_SYSTEM_ID)) {
59 return new InputSource((new File(CLDRConverter.LOCAL_LDML_DTD)).toURI().toString());
60 }
61 return null;
62 }
63
64 @Override
65 public void startElement(String uri, String localName, String qName, Attributes attributes) throws SAXException {
66 switch (qName) {
67 //
68 // Generic information
69 //
70 case "identity":
71 // ignore this element - it has language and territory elements that aren't locale data
72 pushIgnoredContainer(qName);
73 break;
74 case "language":
75 // for LocaleNames
76 // copy string
77 pushStringEntry(qName, attributes, CLDRConverter.LOCALE_NAME_PREFIX + attributes.getValue("type"));
78 break;
79 case "script":
80 // for LocaleNames
81 // copy string
82 pushStringEntry(qName, attributes, CLDRConverter.LOCALE_NAME_PREFIX + attributes.getValue("type"));
83 break;
84 case "territory":
85 // for LocaleNames
86 // copy string
87 pushStringEntry(qName, attributes, CLDRConverter.LOCALE_NAME_PREFIX + attributes.getValue("type"));
88 break;
89
90 //
91 // Currency information
92 //
93 case "currency":
94 // for CurrencyNames
95 // stash away "type" value for nested <symbol>
96 pushKeyContainer(qName, attributes, attributes.getValue("type"));
97 break;
98 case "symbol":
99 // for CurrencyNames
100 // need to get the key from the containing <currency> element
101 pushStringEntry(qName, attributes, CLDRConverter.CURRENCY_SYMBOL_PREFIX + getContainerKey());
102 break;
103 case "displayName":
104 // for CurrencyNames
105 // need to get the key from the containing <currency> element
106 // ignore if is has "count" attribute
107 String containerKey = getContainerKey();
108 if (containerKey != null && attributes.getValue("count") == null) {
109 pushStringEntry(qName, attributes,
110 CLDRConverter.CURRENCY_NAME_PREFIX + containerKey.toLowerCase(Locale.ROOT),
111 attributes.getValue("type"));
112 } else {
113 pushIgnoredContainer(qName);
114 }
115 break;
116
117 //
118 // Calendar information
119 //
120 case "calendar":
121 {
122 // mostly for FormatData (CalendarData items firstDay and minDays are also nested)
123 // use only if it's supported by java.util.Calendar.
124 String calendarName = attributes.getValue("type");
125 currentCalendarType = CalendarType.forName(calendarName);
126 if (currentCalendarType != null) {
127 pushContainer(qName, attributes);
128 } else {
129 pushIgnoredContainer(qName);
130 }
131 }
132 break;
133 case "monthContext":
134 {
135 // for FormatData
136 // need to keep stand-alone and format, to allow for inheritance in CLDR
137 String type = attributes.getValue("type");
138 if ("stand-alone".equals(type) || "format".equals(type)) {
139 pushKeyContainer(qName, attributes, type);
140 } else {
141 pushIgnoredContainer(qName);
142 }
143 }
144 break;
145 case "monthWidth":
146 {
147 // for FormatData
148 // create string array for the two types that the JRE knows
149 // keep info about the context type so we can sort out inheritance later
150 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
151 switch (attributes.getValue("type")) {
152 case "wide":
153 pushStringArrayEntry(qName, attributes, prefix + "MonthNames/" + getContainerKey(), 13);
154 break;
155 case "abbreviated":
156 pushStringArrayEntry(qName, attributes, prefix + "MonthAbbreviations/" + getContainerKey(), 13);
157 break;
158 default:
159 pushIgnoredContainer(qName);
160 break;
161 }
162 }
163 break;
164 case "month":
165 // for FormatData
166 // add to string array entry of monthWidth element
167 pushStringArrayElement(qName, attributes, Integer.parseInt(attributes.getValue("type")) - 1);
168 break;
169 case "dayContext":
170 {
171 // for FormatData
172 // need to keep stand-alone and format, to allow for multiple inheritance in CLDR
173 String type = attributes.getValue("type");
174 if ("stand-alone".equals(type) || "format".equals(type)) {
175 pushKeyContainer(qName, attributes, type);
176 } else {
177 pushIgnoredContainer(qName);
178 }
179 }
180 break;
181 case "dayWidth":
182 {
183 // for FormatData
184 // create string array for the two types that the JRE knows
185 // keep info about the context type so we can sort out inheritance later
186 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
187 switch (attributes.getValue("type")) {
188 case "wide":
189 pushStringArrayEntry(qName, attributes, prefix + "DayNames/" + getContainerKey(), 7);
190 break;
191 case "abbreviated":
192 pushStringArrayEntry(qName, attributes, prefix + "DayAbbreviations/" + getContainerKey(), 7);
193 break;
194 default:
195 pushIgnoredContainer(qName);
196 break;
197 }
198 }
199 break;
200 case "day":
201 // for FormatData
202 // add to string array entry of monthWidth element
203 pushStringArrayElement(qName, attributes, Integer.parseInt(DAY_OF_WEEK_MAP.get(attributes.getValue("type"))) - 1);
204 break;
205 case "dayPeriodContext":
206 // for FormatData
207 // need to keep stand-alone and format, to allow for multiple inheritance in CLDR
208 // for FormatData
209 // need to keep stand-alone and format, to allow for multiple inheritance in CLDR
210 {
211 String type = attributes.getValue("type");
212 if ("stand-alone".equals(type) || "format".equals(type)) {
213 pushKeyContainer(qName, attributes, type);
214 } else {
215 pushIgnoredContainer(qName);
216 }
217 }
218 break;
219 case "dayPeriodWidth":
220 // for FormatData
221 // create string array entry for am/pm. only keeping wide
222 if ("wide".equals(attributes.getValue("type"))) {
223 pushStringArrayEntry(qName, attributes, "AmPmMarkers/" + getContainerKey(), 2);
224 } else {
225 pushIgnoredContainer(qName);
226 }
227 break;
228 case "dayPeriod":
229 // for FormatData
230 // add to string array entry of AmPmMarkers element
231 switch (attributes.getValue("type")) {
232 case "am":
233 pushStringArrayElement(qName, attributes, 0);
234 break;
235 case "pm":
236 pushStringArrayElement(qName, attributes, 1);
237 break;
238 default:
239 pushIgnoredContainer(qName);
240 break;
241 }
242 break;
243 case "eraNames":
244 // CLDR era names are inconsistent in terms of their lengths. For example,
245 // the full names of Japanese imperial eras are eraAbbr, while the full names
246 // of the Julian eras are eraNames.
247 if (currentCalendarType == null) {
248 assert currentContainer instanceof IgnoredContainer;
249 pushIgnoredContainer(qName);
250 } else {
251 String key = currentCalendarType.keyElementName() + "long.Eras"; // for now
252 pushStringArrayEntry(qName, attributes, key, currentCalendarType.getEraLength(qName));
253 }
254 break;
255 case "eraAbbr":
256 // for FormatData
257 // create string array entry
258 if (currentCalendarType == null) {
259 assert currentContainer instanceof IgnoredContainer;
260 pushIgnoredContainer(qName);
261 } else {
262 String key = currentCalendarType.keyElementName() + "Eras";
263 pushStringArrayEntry(qName, attributes, key, currentCalendarType.getEraLength(qName));
264 }
265 break;
266 case "eraNarrow":
267 // mainly used for the Japanese imperial calendar
268 if (currentCalendarType == null) {
269 assert currentContainer instanceof IgnoredContainer;
270 pushIgnoredContainer(qName);
271 } else {
272 String key = currentCalendarType.keyElementName() + "short.Eras";
273 pushStringArrayEntry(qName, attributes, key, currentCalendarType.getEraLength(qName));
274 }
275 break;
276 case "era":
277 // for FormatData
278 // add to string array entry of eraAbbr element
279 if (currentCalendarType == null) {
280 assert currentContainer instanceof IgnoredContainer;
281 pushIgnoredContainer(qName);
282 } else {
283 int index = Integer.parseInt(attributes.getValue("type"));
284 index = currentCalendarType.normalizeEraIndex(index);
285 if (index >= 0) {
286 pushStringArrayElement(qName, attributes, index);
287 } else {
288 pushIgnoredContainer(qName);
289 }
290 if (currentContainer.getParent() == null) {
291 throw new InternalError("currentContainer: null parent");
292 }
293 }
294 break;
295
296 //
297 // Time zone names
298 //
299 case "timeZoneNames":
300 pushContainer(qName, attributes);
301 break;
302 case "zone":
303 {
304 String zone = attributes.getValue("type");
305 zonePrefix = CLDRConverter.TIMEZONE_ID_PREFIX;
306 put(zonePrefix + zone, new HashMap<String, String>());
307 pushKeyContainer(qName, attributes, zone);
308 }
309 break;
310 case "metazone":
311 {
312 String zone = attributes.getValue("type");
313 zonePrefix = CLDRConverter.METAZONE_ID_PREFIX;
314 put(zonePrefix + zone, new HashMap<String, String>());
315 pushKeyContainer(qName, attributes, zone);
316 }
317 break;
318 case "long":
319 zoneNameStyle = "long";
320 pushContainer(qName, attributes);
321 break;
322 case "short":
323 zoneNameStyle = "short";
324 pushContainer(qName, attributes);
325 break;
326 case "generic": // not used in JDK
327 pushIgnoredContainer(qName);
328 break;
329 case "standard": // standard time
330 pushStringEntry(qName, attributes, CLDRConverter.TIMEZONE_NAME_PREFIX + "standard." + zoneNameStyle);
331 break;
332 case "daylight":
333 pushStringEntry(qName, attributes, CLDRConverter.TIMEZONE_NAME_PREFIX + "daylight." + zoneNameStyle);
334 break;
335 case "exemplarCity":
336 pushIgnoredContainer(qName);
337 break;
338
339 //
340 // Number format information
341 //
342 case "decimalFormatLength":
343 if (attributes.getValue("type") == null) {
344 // skipping type="short" data
345 // for FormatData
346 // copy string for later assembly into NumberPatterns
347 pushStringEntry(qName, attributes, "NumberPatterns/decimal");
348 } else {
349 pushIgnoredContainer(qName);
350 }
351 break;
352 case "currencyFormat":
353 // for FormatData
354 // copy string for later assembly into NumberPatterns
355 pushStringEntry(qName, attributes, "NumberPatterns/currency");
356 break;
357 case "percentFormat":
358 // for FormatData
359 // copy string for later assembly into NumberPatterns
360 pushStringEntry(qName, attributes, "NumberPatterns/percent");
361 break;
362 case "defaultNumberingSystem":
363 // default numbering system if multiple numbering systems are used.
364 pushStringEntry(qName, attributes, "DefaultNumberingSystem");
365 break;
366 case "symbols":
367 // for FormatData
368 // look up numberingSystems
369 symbols: {
370 String script = attributes.getValue("numberSystem");
371 if (script == null) {
372 // Has no script. Just ignore.
373 pushIgnoredContainer(qName);
374 break;
375 }
376
377 // Use keys as <script>."NumberElements/<symbol>"
378 currentNumberingSystem = script + ".";
379 String digits = CLDRConverter.handlerNumbering.get(script);
380 if (digits == null) {
381 throw new InternalError("null digits for " + script);
382 }
383 if (Character.isSurrogate(digits.charAt(0))) {
384 // DecimalFormatSymbols doesn't support supplementary characters as digit zero.
385 pushIgnoredContainer(qName);
386 break;
387 }
388 // in case digits are in the reversed order, reverse back the order.
389 if (digits.charAt(0) > digits.charAt(digits.length() - 1)) {
390 StringBuilder sb = new StringBuilder(digits);
391 digits = sb.reverse().toString();
392 }
393 // Check if the order is sequential.
394 char c0 = digits.charAt(0);
395 for (int i = 1; i < digits.length(); i++) {
396 if (digits.charAt(i) != c0 + i) {
397 pushIgnoredContainer(qName);
398 break symbols;
399 }
400 }
401 @SuppressWarnings("unchecked")
402 List<String> numberingScripts = (List<String>) get("numberingScripts");
403 if (numberingScripts == null) {
404 numberingScripts = new ArrayList<>();
405 put("numberingScripts", numberingScripts);
406 }
407 numberingScripts.add(script);
408 put(currentNumberingSystem + "NumberElements/zero", digits.substring(0, 1));
409 pushContainer(qName, attributes);
410 }
411 break;
412 case "decimal":
413 // for FormatData
414 // copy string for later assembly into NumberElements
415 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/decimal");
416 break;
417 case "group":
418 // for FormatData
419 // copy string for later assembly into NumberElements
420 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/group");
421 break;
422 case "list":
423 // for FormatData
424 // copy string for later assembly into NumberElements
425 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/list");
426 break;
427 case "percentSign":
428 // for FormatData
429 // copy string for later assembly into NumberElements
430 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/percent");
431 break;
432 case "nativeZeroDigit":
433 // for FormatData
434 // copy string for later assembly into NumberElements
435 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/zero");
436 break;
437 case "patternDigit":
438 // for FormatData
439 // copy string for later assembly into NumberElements
440 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/pattern");
441 break;
442 case "plusSign":
443 // TODO: DecimalFormatSymbols doesn't support plusSign
444 pushIgnoredContainer(qName);
445 break;
446 case "minusSign":
447 // for FormatData
448 // copy string for later assembly into NumberElements
449 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/minus");
450 break;
451 case "exponential":
452 // for FormatData
453 // copy string for later assembly into NumberElements
454 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/exponential");
455 break;
456 case "perMille":
457 // for FormatData
458 // copy string for later assembly into NumberElements
459 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/permille");
460 break;
461 case "infinity":
462 // for FormatData
463 // copy string for later assembly into NumberElements
464 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/infinity");
465 break;
466 case "nan":
467 // for FormatData
468 // copy string for later assembly into NumberElements
469 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/nan");
470 break;
471 case "timeFormatLength":
472 {
473 // for FormatData
474 // copy string for later assembly into DateTimePatterns
475 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
476 pushStringEntry(qName, attributes, prefix + "DateTimePatterns/" + attributes.getValue("type") + "-time");
477 }
478 break;
479 case "dateFormatLength":
480 {
481 // for FormatData
482 // copy string for later assembly into DateTimePatterns
483 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
484 pushStringEntry(qName, attributes, prefix + "DateTimePatterns/" + attributes.getValue("type") + "-date");
485 }
486 break;
487 case "dateTimeFormat":
488 {
489 // for FormatData
490 // copy string for later assembly into DateTimePatterns
491 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
492 pushStringEntry(qName, attributes, prefix + "DateTimePatterns/date-time");
493 }
494 break;
495 case "localizedPatternChars":
496 {
497 // for FormatData
498 // copy string for later adaptation to JRE use
499 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
500 pushStringEntry(qName, attributes, prefix + "DateTimePatternChars");
501 }
502 break;
503
504 default:
505 // treat anything else as a container
506 pushContainer(qName, attributes);
507 break;
508 }
509 }
510
511 @Override
512 public void endElement(String uri, String localName, String qName) throws SAXException {
513 assert qName.equals(currentContainer.getqName()) : "current=" + currentContainer.getqName() + ", param=" + qName;
514 switch (qName) {
515 case "calendar":
516 assert !(currentContainer instanceof Entry);
517 currentCalendarType = null;
518 break;
519
520 case "defaultNumberingSystem":
521 if (currentContainer instanceof StringEntry) {
522 defaultNumberingSystem = ((StringEntry) currentContainer).getValue();
523 assert defaultNumberingSystem != null;
524 put(((StringEntry) currentContainer).getKey(), defaultNumberingSystem);
525 } else {
526 defaultNumberingSystem = null;
527 }
528 break;
529
530 case "timeZoneNames":
531 zonePrefix = null;
532 break;
533 case "standard":
534 case "daylight":
535 if (zonePrefix != null && (currentContainer instanceof Entry)) {
536 @SuppressWarnings("unchecked")
537 Map<String, String> valmap = (Map<String, String>) get(zonePrefix + getContainerKey());
538 Entry<?> entry = (Entry<?>) currentContainer;
539 valmap.put(entry.getKey(), (String) entry.getValue());
540 }
541 break;
542 default:
543 if (currentContainer instanceof Entry) {
544 Entry<?> entry = (Entry<?>) currentContainer;
545 Object value = entry.getValue();
546 if (value != null) {
547 put(entry.getKey(), value);
548 }
549 }
550 }
551 currentContainer = currentContainer.getParent();
552 }
553 }