1 /*
2 * Copyright (c) 2012, 2013, 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 "type":
75 if ("calendar".equals(attributes.getValue("key"))) {
76 pushStringEntry(qName, attributes, CLDRConverter.CALENDAR_NAME_PREFIX + attributes.getValue("type"));
77 } else {
78 pushIgnoredContainer(qName);
79 }
80 break;
81 case "language":
82 // for LocaleNames
83 // copy string
84 pushStringEntry(qName, attributes, CLDRConverter.LOCALE_NAME_PREFIX + attributes.getValue("type"));
85 break;
86 case "script":
87 // for LocaleNames
88 // copy string
89 pushStringEntry(qName, attributes, CLDRConverter.LOCALE_NAME_PREFIX + attributes.getValue("type"));
90 break;
91 case "territory":
92 // for LocaleNames
93 // copy string
94 pushStringEntry(qName, attributes, CLDRConverter.LOCALE_NAME_PREFIX + attributes.getValue("type"));
95 break;
96
97 //
98 // Currency information
99 //
100 case "currency":
101 // for CurrencyNames
102 // stash away "type" value for nested <symbol>
103 pushKeyContainer(qName, attributes, attributes.getValue("type"));
104 break;
105 case "symbol":
106 // for CurrencyNames
107 // need to get the key from the containing <currency> element
108 pushStringEntry(qName, attributes, CLDRConverter.CURRENCY_SYMBOL_PREFIX
109 + getContainerKey());
110 break;
111
112 // Calendar or currency
113 case "displayName":
114 {
115 if (currentCalendarType != null) {
116 pushStringEntry(qName, attributes,
117 currentCalendarType.keyElementName() + "field." + getContainerKey());
118 } else {
119 // for CurrencyNames
120 // need to get the key from the containing <currency> element
121 // ignore if is has "count" attribute
122 String containerKey = getContainerKey();
123 if (containerKey != null && attributes.getValue("count") == null) {
124 pushStringEntry(qName, attributes,
125 CLDRConverter.CURRENCY_NAME_PREFIX
126 + containerKey.toLowerCase(Locale.ROOT),
127 attributes.getValue("type"));
128 } else {
129 pushIgnoredContainer(qName);
130 }
131 }
132 }
133 break;
134
135 //
136 // Calendar information
137 //
138 case "calendar":
139 {
140 // mostly for FormatData (CalendarData items firstDay and minDays are also nested)
141 // use only if it's supported by java.util.Calendar.
142 String calendarName = attributes.getValue("type");
143 currentCalendarType = CalendarType.forName(calendarName);
144 if (currentCalendarType != null) {
145 pushContainer(qName, attributes);
146 } else {
147 pushIgnoredContainer(qName);
148 }
149 }
150 break;
151 case "fields":
152 if (currentCalendarType != null) {
153 pushContainer(qName, attributes);
154 } else {
155 pushIgnoredContainer(qName);
156 }
157 break;
158 case "field":
159 {
160 String type = attributes.getValue("type");
161 switch (type) {
162 case "era":
163 case "year":
164 case "month":
165 case "week":
166 case "weekday":
167 case "dayperiod":
168 case "hour":
169 case "minute":
170 case "second":
171 case "zone":
172 pushKeyContainer(qName, attributes, type);
173 break;
174 default:
175 pushIgnoredContainer(qName);
176 break;
177 }
178 }
179 break;
180 case "monthContext":
181 {
182 // for FormatData
183 // need to keep stand-alone and format, to allow for inheritance in CLDR
184 String type = attributes.getValue("type");
185 if ("stand-alone".equals(type) || "format".equals(type)) {
186 pushKeyContainer(qName, attributes, type);
187 } else {
188 pushIgnoredContainer(qName);
189 }
190 }
191 break;
192 case "monthWidth":
193 {
194 // for FormatData
195 // create string array for the two types that the JRE knows
196 // keep info about the context type so we can sort out inheritance later
197 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
198 switch (attributes.getValue("type")) {
199 case "wide":
200 pushStringArrayEntry(qName, attributes, prefix + "MonthNames/" + getContainerKey(), 13);
201 break;
202 case "abbreviated":
203 pushStringArrayEntry(qName, attributes, prefix + "MonthAbbreviations/" + getContainerKey(), 13);
204 break;
205 case "narrow":
206 pushStringArrayEntry(qName, attributes, prefix + "MonthNarrows/" + getContainerKey(), 13);
207 break;
208 default:
209 pushIgnoredContainer(qName);
210 break;
211 }
212 }
213 break;
214 case "month":
215 // for FormatData
216 // add to string array entry of monthWidth element
217 pushStringArrayElement(qName, attributes, Integer.parseInt(attributes.getValue("type")) - 1);
218 break;
219 case "dayContext":
220 {
221 // for FormatData
222 // need to keep stand-alone and format, to allow for multiple inheritance in CLDR
223 String type = attributes.getValue("type");
224 if ("stand-alone".equals(type) || "format".equals(type)) {
225 pushKeyContainer(qName, attributes, type);
226 } else {
227 pushIgnoredContainer(qName);
228 }
229 }
230 break;
231 case "dayWidth":
232 {
233 // for FormatData
234 // create string array for the two types that the JRE knows
235 // keep info about the context type so we can sort out inheritance later
236 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
237 switch (attributes.getValue("type")) {
238 case "wide":
239 pushStringArrayEntry(qName, attributes, prefix + "DayNames/" + getContainerKey(), 7);
240 break;
241 case "abbreviated":
242 pushStringArrayEntry(qName, attributes, prefix + "DayAbbreviations/" + getContainerKey(), 7);
243 break;
244 case "narrow":
245 pushStringArrayEntry(qName, attributes, prefix + "DayNarrows/" + getContainerKey(), 7);
246 break;
247 default:
248 pushIgnoredContainer(qName);
249 break;
250 }
251 }
252 break;
253 case "day":
254 // for FormatData
255 // add to string array entry of monthWidth element
256 pushStringArrayElement(qName, attributes, Integer.parseInt(DAY_OF_WEEK_MAP.get(attributes.getValue("type"))) - 1);
257 break;
258 case "dayPeriodContext":
259 // for FormatData
260 // need to keep stand-alone and format, to allow for multiple inheritance in CLDR
261 // for FormatData
262 // need to keep stand-alone and format, to allow for multiple inheritance in CLDR
263 {
264 String type = attributes.getValue("type");
265 if ("stand-alone".equals(type) || "format".equals(type)) {
266 pushKeyContainer(qName, attributes, type);
267 } else {
268 pushIgnoredContainer(qName);
269 }
270 }
271 break;
272 case "dayPeriodWidth":
273 // for FormatData
274 // create string array entry for am/pm. only keeping wide
275 switch (attributes.getValue("type")) {
276 case "wide":
277 pushStringArrayEntry(qName, attributes, "AmPmMarkers/" + getContainerKey(), 2);
278 break;
279 case "narrow":
280 pushStringArrayEntry(qName, attributes, "narrow.AmPmMarkers/" + getContainerKey(), 2);
281 break;
282 default:
283 pushIgnoredContainer(qName);
284 break;
285 }
286 break;
287 case "dayPeriod":
288 // for FormatData
289 // add to string array entry of AmPmMarkers element
290 if (attributes.getValue("alt") == null) {
291 switch (attributes.getValue("type")) {
292 case "am":
293 pushStringArrayElement(qName, attributes, 0);
294 break;
295 case "pm":
296 pushStringArrayElement(qName, attributes, 1);
297 break;
298 default:
299 pushIgnoredContainer(qName);
300 break;
301 }
302 } else {
303 // discard alt values
304 pushIgnoredContainer(qName);
305 }
306 break;
307 case "eraNames":
308 // CLDR era names are inconsistent in terms of their lengths. For example,
309 // the full names of Japanese imperial eras are eraAbbr, while the full names
310 // of the Julian eras are eraNames.
311 if (currentCalendarType == null) {
312 assert currentContainer instanceof IgnoredContainer;
313 pushIgnoredContainer(qName);
314 } else {
315 String key = currentCalendarType.keyElementName() + "long.Eras"; // for now
316 pushStringArrayEntry(qName, attributes, key, currentCalendarType.getEraLength(qName));
317 }
318 break;
319 case "eraAbbr":
320 // for FormatData
321 // create string array entry
322 if (currentCalendarType == null) {
323 assert currentContainer instanceof IgnoredContainer;
324 pushIgnoredContainer(qName);
325 } else {
326 String key = currentCalendarType.keyElementName() + "Eras";
327 pushStringArrayEntry(qName, attributes, key, currentCalendarType.getEraLength(qName));
328 }
329 break;
330 case "eraNarrow":
331 // mainly used for the Japanese imperial calendar
332 if (currentCalendarType == null) {
333 assert currentContainer instanceof IgnoredContainer;
334 pushIgnoredContainer(qName);
335 } else {
336 String key = currentCalendarType.keyElementName() + "narrow.Eras";
337 pushStringArrayEntry(qName, attributes, key, currentCalendarType.getEraLength(qName));
338 }
339 break;
340 case "era":
341 // for FormatData
342 // add to string array entry of eraAbbr element
343 if (currentCalendarType == null) {
344 assert currentContainer instanceof IgnoredContainer;
345 pushIgnoredContainer(qName);
346 } else {
347 int index = Integer.parseInt(attributes.getValue("type"));
348 index = currentCalendarType.normalizeEraIndex(index);
349 if (index >= 0) {
350 pushStringArrayElement(qName, attributes, index);
351 } else {
352 pushIgnoredContainer(qName);
353 }
354 if (currentContainer.getParent() == null) {
355 throw new InternalError("currentContainer: null parent");
356 }
357 }
358 break;
359
360 //
361 // Time zone names
362 //
363 case "timeZoneNames":
364 pushContainer(qName, attributes);
365 break;
366 case "zone":
367 {
368 String tzid = attributes.getValue("type"); // Olson tz id
369 zonePrefix = CLDRConverter.TIMEZONE_ID_PREFIX;
370 put(zonePrefix + tzid, new HashMap<String, String>());
371 pushKeyContainer(qName, attributes, tzid);
372 }
373 break;
374 case "metazone":
375 {
376 String zone = attributes.getValue("type"); // LDML meta zone id
377 zonePrefix = CLDRConverter.METAZONE_ID_PREFIX;
378 put(zonePrefix + zone, new HashMap<String, String>());
379 pushKeyContainer(qName, attributes, zone);
380 }
381 break;
382 case "long":
383 zoneNameStyle = "long";
384 pushContainer(qName, attributes);
385 break;
386 case "short":
387 zoneNameStyle = "short";
388 pushContainer(qName, attributes);
389 break;
390 case "generic": // generic name
391 case "standard": // standard time name
392 case "daylight": // daylight saving (summer) time name
393 pushStringEntry(qName, attributes, CLDRConverter.ZONE_NAME_PREFIX + qName + "." + zoneNameStyle);
394 break;
395 case "exemplarCity": // not used in JDK
396 pushIgnoredContainer(qName);
397 break;
398
399 //
400 // Number format information
401 //
402 case "decimalFormatLength":
403 if (attributes.getValue("type") == null) {
404 // skipping type="short" data
405 // for FormatData
406 // copy string for later assembly into NumberPatterns
407 pushStringEntry(qName, attributes, "NumberPatterns/decimal");
408 } else {
409 pushIgnoredContainer(qName);
410 }
411 break;
412 case "currencyFormat":
413 // for FormatData
414 // copy string for later assembly into NumberPatterns
415 pushStringEntry(qName, attributes, "NumberPatterns/currency");
416 break;
417 case "percentFormat":
418 // for FormatData
419 // copy string for later assembly into NumberPatterns
420 pushStringEntry(qName, attributes, "NumberPatterns/percent");
421 break;
422 case "defaultNumberingSystem":
423 // default numbering system if multiple numbering systems are used.
424 pushStringEntry(qName, attributes, "DefaultNumberingSystem");
425 break;
426 case "symbols":
427 // for FormatData
428 // look up numberingSystems
429 symbols: {
430 String script = attributes.getValue("numberSystem");
431 if (script == null) {
432 // Has no script. Just ignore.
433 pushIgnoredContainer(qName);
434 break;
435 }
436
437 // Use keys as <script>."NumberElements/<symbol>"
438 currentNumberingSystem = script + ".";
439 String digits = CLDRConverter.handlerNumbering.get(script);
440 if (digits == null) {
441 throw new InternalError("null digits for " + script);
442 }
443 if (Character.isSurrogate(digits.charAt(0))) {
444 // DecimalFormatSymbols doesn't support supplementary characters as digit zero.
445 pushIgnoredContainer(qName);
446 break;
447 }
448 // in case digits are in the reversed order, reverse back the order.
449 if (digits.charAt(0) > digits.charAt(digits.length() - 1)) {
450 StringBuilder sb = new StringBuilder(digits);
451 digits = sb.reverse().toString();
452 }
453 // Check if the order is sequential.
454 char c0 = digits.charAt(0);
455 for (int i = 1; i < digits.length(); i++) {
456 if (digits.charAt(i) != c0 + i) {
457 pushIgnoredContainer(qName);
458 break symbols;
459 }
460 }
461 @SuppressWarnings("unchecked")
462 List<String> numberingScripts = (List<String>) get("numberingScripts");
463 if (numberingScripts == null) {
464 numberingScripts = new ArrayList<>();
465 put("numberingScripts", numberingScripts);
466 }
467 numberingScripts.add(script);
468 put(currentNumberingSystem + "NumberElements/zero", digits.substring(0, 1));
469 pushContainer(qName, attributes);
470 }
471 break;
472 case "decimal":
473 // for FormatData
474 // copy string for later assembly into NumberElements
475 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/decimal");
476 break;
477 case "group":
478 // for FormatData
479 // copy string for later assembly into NumberElements
480 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/group");
481 break;
482 case "list":
483 // for FormatData
484 // copy string for later assembly into NumberElements
485 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/list");
486 break;
487 case "percentSign":
488 // for FormatData
489 // copy string for later assembly into NumberElements
490 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/percent");
491 break;
492 case "nativeZeroDigit":
493 // for FormatData
494 // copy string for later assembly into NumberElements
495 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/zero");
496 break;
497 case "patternDigit":
498 // for FormatData
499 // copy string for later assembly into NumberElements
500 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/pattern");
501 break;
502 case "plusSign":
503 // TODO: DecimalFormatSymbols doesn't support plusSign
504 pushIgnoredContainer(qName);
505 break;
506 case "minusSign":
507 // for FormatData
508 // copy string for later assembly into NumberElements
509 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/minus");
510 break;
511 case "exponential":
512 // for FormatData
513 // copy string for later assembly into NumberElements
514 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/exponential");
515 break;
516 case "perMille":
517 // for FormatData
518 // copy string for later assembly into NumberElements
519 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/permille");
520 break;
521 case "infinity":
522 // for FormatData
523 // copy string for later assembly into NumberElements
524 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/infinity");
525 break;
526 case "nan":
527 // for FormatData
528 // copy string for later assembly into NumberElements
529 pushStringEntry(qName, attributes, currentNumberingSystem + "NumberElements/nan");
530 break;
531 case "timeFormatLength":
532 {
533 // for FormatData
534 // copy string for later assembly into DateTimePatterns
535 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
536 pushStringEntry(qName, attributes, prefix + "DateTimePatterns/" + attributes.getValue("type") + "-time");
537 }
538 break;
539 case "dateFormatLength":
540 {
541 // for FormatData
542 // copy string for later assembly into DateTimePatterns
543 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
544 pushStringEntry(qName, attributes, prefix + "DateTimePatterns/" + attributes.getValue("type") + "-date");
545 }
546 break;
547 case "dateTimeFormat":
548 {
549 // for FormatData
550 // copy string for later assembly into DateTimePatterns
551 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
552 pushStringEntry(qName, attributes, prefix + "DateTimePatterns/date-time");
553 }
554 break;
555 case "localizedPatternChars":
556 {
557 // for FormatData
558 // copy string for later adaptation to JRE use
559 String prefix = (currentCalendarType == null) ? "" : currentCalendarType.keyElementName();
560 pushStringEntry(qName, attributes, prefix + "DateTimePatternChars");
561 }
562 break;
563
564 default:
565 // treat anything else as a container
566 pushContainer(qName, attributes);
567 break;
568 }
569 }
570
571 @Override
572 public void endElement(String uri, String localName, String qName) throws SAXException {
573 assert qName.equals(currentContainer.getqName()) : "current=" + currentContainer.getqName() + ", param=" + qName;
574 switch (qName) {
575 case "calendar":
576 assert !(currentContainer instanceof Entry);
577 currentCalendarType = null;
578 break;
579
580 case "defaultNumberingSystem":
581 if (currentContainer instanceof StringEntry) {
582 defaultNumberingSystem = ((StringEntry) currentContainer).getValue();
583 assert defaultNumberingSystem != null;
584 put(((StringEntry) currentContainer).getKey(), defaultNumberingSystem);
585 } else {
586 defaultNumberingSystem = null;
587 }
588 break;
589
590 case "timeZoneNames":
591 zonePrefix = null;
592 break;
593 case "generic":
594 case "standard":
595 case "daylight":
596 if (zonePrefix != null && (currentContainer instanceof Entry)) {
597 @SuppressWarnings("unchecked")
598 Map<String, String> valmap = (Map<String, String>) get(zonePrefix + getContainerKey());
599 Entry<?> entry = (Entry<?>) currentContainer;
600 valmap.put(entry.getKey(), (String) entry.getValue());
601 }
602 break;
603 default:
604 if (currentContainer instanceof Entry) {
605 Entry<?> entry = (Entry<?>) currentContainer;
606 Object value = entry.getValue();
607 if (value != null) {
608 put(entry.getKey(), value);
609 }
610 }
611 }
612 currentContainer = currentContainer.getParent();
613 }
614 }