1 /*
2 * Copyright (c) 1999, 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 /*
27 * (C) Copyright Taligent, Inc. 1996, 1997 - All Rights Reserved
28 * (C) Copyright IBM Corp. 1996-1998 - All Rights Reserved
29 *
30 * The original version of this source code and documentation is copyrighted
31 * and owned by Taligent, Inc., a wholly-owned subsidiary of IBM. These
32 * materials are provided under terms of a License Agreement between Taligent
33 * and Sun. This technology is protected by multiple US and International
34 * patents. This notice and attribution to Taligent may not be removed.
35 * Taligent is a registered trademark of Taligent, Inc.
36 *
37 */
38
39 package java.text;
40
41 import java.util.Vector;
42 import sun.text.UCompactIntArray;
43 import sun.text.IntHashtable;
44 import sun.text.ComposedCharIter;
45 import sun.text.CollatorUtilities;
46 import sun.text.normalizer.NormalizerImpl;
47
48 /**
49 * This class contains all the code to parse a RuleBasedCollator pattern
50 * and build a RBCollationTables object from it. A particular instance
51 * of tis class exists only during the actual build process-- once an
52 * RBCollationTables object has been built, the RBTableBuilder object
53 * goes away. This object carries all of the state which is only needed
54 * during the build process, plus a "shadow" copy of all of the state
55 * that will go into the tables object itself. This object communicates
56 * with RBCollationTables through a separate class, RBCollationTables.BuildAPI,
57 * this is an inner class of RBCollationTables and provides a separate
58 * private API for communication with RBTableBuilder.
59 * This class isn't just an inner class of RBCollationTables itself because
60 * of its large size. For source-code readability, it seemed better for the
61 * builder to have its own source file.
62 */
63 final class RBTableBuilder {
64
65 public RBTableBuilder(RBCollationTables.BuildAPI tables) {
66 this.tables = tables;
67 }
68
69 /**
70 * Create a table-based collation object with the given rules.
71 * This is the main function that actually builds the tables and
72 * stores them back in the RBCollationTables object. It is called
73 * ONLY by the RBCollationTables constructor.
74 * @see RuleBasedCollator#RuleBasedCollator
75 * @exception ParseException If the rules format is incorrect.
76 */
77
78 public void build(String pattern, int decmp) throws ParseException {
79 String expChars;
80 String groupChars;
81 if (pattern.isEmpty())
82 throw new ParseException("Build rules empty.", 0);
83
84 // This array maps Unicode characters to their collation ordering
85 mapping = new UCompactIntArray(RBCollationTables.UNMAPPED);
86 // Normalize the build rules. Find occurances of all decomposed characters
87 // and normalize the rules before feeding into the builder. By "normalize",
88 // we mean that all precomposed Unicode characters must be converted into
89 // a base character and one or more combining characters (such as accents).
90 // When there are multiple combining characters attached to a base character,
91 // the combining characters must be in their canonical order
92 //
93 // sherman/Note:
94 //(1)decmp will be NO_DECOMPOSITION only in ko locale to prevent decompose
95 //hangual syllables to jamos, so we can actually just call decompose with
96 //normalizer's IGNORE_HANGUL option turned on
97 //
98 //(2)just call the "special version" in NormalizerImpl directly
99 //pattern = Normalizer.decompose(pattern, false, Normalizer.IGNORE_HANGUL, true);
100 //
101 //Normalizer.Mode mode = CollatorUtilities.toNormalizerMode(decmp);
102 //pattern = Normalizer.normalize(pattern, mode, 0, true);
103
104 pattern = NormalizerImpl.canonicalDecomposeWithSingleQuotation(pattern);
105
106 // Build the merged collation entries
107 // Since rules can be specified in any order in the string
108 // (e.g. "c , C < d , D < e , E .... C < CH")
109 // this splits all of the rules in the string out into separate
110 // objects and then sorts them. In the above example, it merges the
111 // "C < CH" rule in just before the "C < D" rule.
112 //
113
114 mPattern = new MergeCollation(pattern);
115
116 int order = 0;
117
118 // Now walk though each entry and add it to my own tables
119 for (int i = 0; i < mPattern.getCount(); ++i) {
120 PatternEntry entry = mPattern.getItemAt(i);
121 if (entry != null) {
122 groupChars = entry.getChars();
123 if (groupChars.length() > 1) {
124 switch(groupChars.charAt(groupChars.length()-1)) {
125 case '@':
126 frenchSec = true;
127 groupChars = groupChars.substring(0, groupChars.length()-1);
128 break;
129 case '!':
130 seAsianSwapping = true;
131 groupChars = groupChars.substring(0, groupChars.length()-1);
132 break;
133 }
134 }
135
136 order = increment(entry.getStrength(), order);
137 expChars = entry.getExtension();
138
139 if (!expChars.isEmpty()) {
140 addExpandOrder(groupChars, expChars, order);
141 } else if (groupChars.length() > 1) {
142 char ch = groupChars.charAt(0);
143 if (Character.isHighSurrogate(ch) && groupChars.length() == 2) {
144 addOrder(Character.toCodePoint(ch, groupChars.charAt(1)), order);
145 } else {
146 addContractOrder(groupChars, order);
147 }
148 } else {
149 char ch = groupChars.charAt(0);
150 addOrder(ch, order);
151 }
152 }
153 }
154 addComposedChars();
155
156 commit();
157 mapping.compact();
158 /*
159 System.out.println("mappingSize=" + mapping.getKSize());
160 for (int j = 0; j < 0xffff; j++) {
161 int value = mapping.elementAt(j);
162 if (value != RBCollationTables.UNMAPPED)
163 System.out.println("index=" + Integer.toString(j, 16)
164 + ", value=" + Integer.toString(value, 16));
165 }
166 */
167 tables.fillInTables(frenchSec, seAsianSwapping, mapping, contractTable, expandTable,
168 contractFlags, maxSecOrder, maxTerOrder);
169 }
170
171 /** Add expanding entries for pre-composed unicode characters so that this
172 * collator can be used reasonably well with decomposition turned off.
173 */
174 private void addComposedChars() throws ParseException {
175 // Iterate through all of the pre-composed characters in Unicode
176 ComposedCharIter iter = new ComposedCharIter();
177 int c;
178 while ((c = iter.next()) != ComposedCharIter.DONE) {
179 if (getCharOrder(c) == RBCollationTables.UNMAPPED) {
180 //
181 // We don't already have an ordering for this pre-composed character.
182 //
183 // First, see if the decomposed string is already in our
184 // tables as a single contracting-string ordering.
185 // If so, just map the precomposed character to that order.
186 //
187 // TODO: What we should really be doing here is trying to find the
188 // longest initial substring of the decomposition that is present
189 // in the tables as a contracting character sequence, and find its
190 // ordering. Then do this recursively with the remaining chars
191 // so that we build a list of orderings, and add that list to
192 // the expansion table.
193 // That would be more correct but also significantly slower, so
194 // I'm not totally sure it's worth doing.
195 //
196 String s = iter.decomposition();
197
198 //sherman/Note: if this is 1 character decomposed string, the
199 //only thing need to do is to check if this decomposed character
200 //has an entry in our order table, this order is not necessary
201 //to be a contraction order, if it does have one, add an entry
202 //for the precomposed character by using the same order, the
203 //previous impl unnecessarily adds a single character expansion
204 //entry.
205 if (s.length() == 1) {
206 int order = getCharOrder(s.charAt(0));
207 if (order != RBCollationTables.UNMAPPED) {
208 addOrder(c, order);
209 }
210 continue;
211 } else if (s.length() == 2) {
212 char ch0 = s.charAt(0);
213 if (Character.isHighSurrogate(ch0)) {
214 int order = getCharOrder(s.codePointAt(0));
215 if (order != RBCollationTables.UNMAPPED) {
216 addOrder(c, order);
217 }
218 continue;
219 }
220 }
221 int contractOrder = getContractOrder(s);
222 if (contractOrder != RBCollationTables.UNMAPPED) {
223 addOrder(c, contractOrder);
224 } else {
225 //
226 // We don't have a contracting ordering for the entire string
227 // that results from the decomposition, but if we have orders
228 // for each individual character, we can add an expanding
229 // table entry for the pre-composed character
230 //
231 boolean allThere = true;
232 for (int i = 0; i < s.length(); i++) {
233 if (getCharOrder(s.charAt(i)) == RBCollationTables.UNMAPPED) {
234 allThere = false;
235 break;
236 }
237 }
238 if (allThere) {
239 addExpandOrder(c, s, RBCollationTables.UNMAPPED);
240 }
241 }
242 }
243 }
244 }
245
246 /**
247 * Look up for unmapped values in the expanded character table.
248 *
249 * When the expanding character tables are built by addExpandOrder,
250 * it doesn't know what the final ordering of each character
251 * in the expansion will be. Instead, it just puts the raw character
252 * code into the table, adding CHARINDEX as a flag. Now that we've
253 * finished building the mapping table, we can go back and look up
254 * that character to see what its real collation order is and
255 * stick that into the expansion table. That lets us avoid doing
256 * a two-stage lookup later.
257 */
258 private final void commit()
259 {
260 if (expandTable != null) {
261 for (int i = 0; i < expandTable.size(); i++) {
262 int[] valueList = expandTable.elementAt(i);
263 for (int j = 0; j < valueList.length; j++) {
264 int order = valueList[j];
265 if (order < RBCollationTables.EXPANDCHARINDEX && order > CHARINDEX) {
266 // found a expanding character that isn't filled in yet
267 int ch = order - CHARINDEX;
268
269 // Get the real values for the non-filled entry
270 int realValue = getCharOrder(ch);
271
272 if (realValue == RBCollationTables.UNMAPPED) {
273 // The real value is still unmapped, maybe it's ignorable
274 valueList[j] = IGNORABLEMASK & ch;
275 } else {
276 // just fill in the value
277 valueList[j] = realValue;
278 }
279 }
280 }
281 }
282 }
283 }
284 /**
285 * Increment of the last order based on the comparison level.
286 */
287 private final int increment(int aStrength, int lastValue)
288 {
289 switch(aStrength)
290 {
291 case Collator.PRIMARY:
292 // increment priamry order and mask off secondary and tertiary difference
293 lastValue += PRIMARYORDERINCREMENT;
294 lastValue &= RBCollationTables.PRIMARYORDERMASK;
295 isOverIgnore = true;
296 break;
297 case Collator.SECONDARY:
298 // increment secondary order and mask off tertiary difference
299 lastValue += SECONDARYORDERINCREMENT;
300 lastValue &= RBCollationTables.SECONDARYDIFFERENCEONLY;
301 // record max # of ignorable chars with secondary difference
302 if (!isOverIgnore)
303 maxSecOrder++;
304 break;
305 case Collator.TERTIARY:
306 // increment tertiary order
307 lastValue += TERTIARYORDERINCREMENT;
308 // record max # of ignorable chars with tertiary difference
309 if (!isOverIgnore)
310 maxTerOrder++;
311 break;
312 }
313 return lastValue;
314 }
315
316 /**
317 * Adds a character and its designated order into the collation table.
318 */
319 private final void addOrder(int ch, int anOrder)
320 {
321 // See if the char already has an order in the mapping table
322 int order = mapping.elementAt(ch);
323
324 if (order >= RBCollationTables.CONTRACTCHARINDEX) {
325 // There's already an entry for this character that points to a contracting
326 // character table. Instead of adding the character directly to the mapping
327 // table, we must add it to the contract table instead.
328 int length = 1;
329 if (Character.isSupplementaryCodePoint(ch)) {
330 length = Character.toChars(ch, keyBuf, 0);
331 } else {
332 keyBuf[0] = (char)ch;
333 }
334 addContractOrder(new String(keyBuf, 0, length), anOrder);
335 } else {
336 // add the entry to the mapping table,
337 // the same later entry replaces the previous one
338 mapping.setElementAt(ch, anOrder);
339 }
340 }
341
342 private final void addContractOrder(String groupChars, int anOrder) {
343 addContractOrder(groupChars, anOrder, true);
344 }
345
346 /**
347 * Adds the contracting string into the collation table.
348 */
349 private final void addContractOrder(String groupChars, int anOrder,
350 boolean fwd)
351 {
352 if (contractTable == null) {
353 contractTable = new Vector<>(INITIALTABLESIZE);
354 }
355
356 //initial character
357 int ch = groupChars.codePointAt(0);
358 /*
359 char ch0 = groupChars.charAt(0);
360 int ch = Character.isHighSurrogate(ch0)?
361 Character.toCodePoint(ch0, groupChars.charAt(1)):ch0;
362 */
363 // See if the initial character of the string already has a contract table.
364 int entry = mapping.elementAt(ch);
365 Vector<EntryPair> entryTable = getContractValuesImpl(entry - RBCollationTables.CONTRACTCHARINDEX);
366
367 if (entryTable == null) {
368 // We need to create a new table of contract entries for this base char
369 int tableIndex = RBCollationTables.CONTRACTCHARINDEX + contractTable.size();
370 entryTable = new Vector<>(INITIALTABLESIZE);
371 contractTable.addElement(entryTable);
372
373 // Add the initial character's current ordering first. then
374 // update its mapping to point to this contract table
375 entryTable.addElement(new EntryPair(groupChars.substring(0,Character.charCount(ch)), entry));
376 mapping.setElementAt(ch, tableIndex);
377 }
378
379 // Now add (or replace) this string in the table
380 int index = RBCollationTables.getEntry(entryTable, groupChars, fwd);
381 if (index != RBCollationTables.UNMAPPED) {
382 EntryPair pair = entryTable.elementAt(index);
383 pair.value = anOrder;
384 } else {
385 EntryPair pair = entryTable.lastElement();
386
387 // NOTE: This little bit of logic is here to speed CollationElementIterator
388 // .nextContractChar(). This code ensures that the longest sequence in
389 // this list is always the _last_ one in the list. This keeps
390 // nextContractChar() from having to search the entire list for the longest
391 // sequence.
392 if (groupChars.length() > pair.entryName.length()) {
393 entryTable.addElement(new EntryPair(groupChars, anOrder, fwd));
394 } else {
395 entryTable.insertElementAt(new EntryPair(groupChars, anOrder,
396 fwd), entryTable.size() - 1);
397 }
398 }
399
400 // If this was a forward mapping for a contracting string, also add a
401 // reverse mapping for it, so that CollationElementIterator.previous
402 // can work right
403 if (fwd && groupChars.length() > 1) {
404 addContractFlags(groupChars);
405 addContractOrder(new StringBuffer(groupChars).reverse().toString(),
406 anOrder, false);
407 }
408 }
409
410 /**
411 * If the given string has been specified as a contracting string
412 * in this collation table, return its ordering.
413 * Otherwise return UNMAPPED.
414 */
415 private int getContractOrder(String groupChars)
416 {
417 int result = RBCollationTables.UNMAPPED;
418 if (contractTable != null) {
419 int ch = groupChars.codePointAt(0);
420 /*
421 char ch0 = groupChars.charAt(0);
422 int ch = Character.isHighSurrogate(ch0)?
423 Character.toCodePoint(ch0, groupChars.charAt(1)):ch0;
424 */
425 Vector<EntryPair> entryTable = getContractValues(ch);
426 if (entryTable != null) {
427 int index = RBCollationTables.getEntry(entryTable, groupChars, true);
428 if (index != RBCollationTables.UNMAPPED) {
429 EntryPair pair = entryTable.elementAt(index);
430 result = pair.value;
431 }
432 }
433 }
434 return result;
435 }
436
437 private final int getCharOrder(int ch) {
438 int order = mapping.elementAt(ch);
439
440 if (order >= RBCollationTables.CONTRACTCHARINDEX) {
441 Vector<EntryPair> groupList = getContractValuesImpl(order - RBCollationTables.CONTRACTCHARINDEX);
442 EntryPair pair = groupList.firstElement();
443 order = pair.value;
444 }
445 return order;
446 }
447
448 /**
449 * Get the entry of hash table of the contracting string in the collation
450 * table.
451 * @param ch the starting character of the contracting string
452 */
453 private Vector<EntryPair> getContractValues(int ch)
454 {
455 int index = mapping.elementAt(ch);
456 return getContractValuesImpl(index - RBCollationTables.CONTRACTCHARINDEX);
457 }
458
459 private Vector<EntryPair> getContractValuesImpl(int index)
460 {
461 if (index >= 0)
462 {
463 return contractTable.elementAt(index);
464 }
465 else // not found
466 {
467 return null;
468 }
469 }
470
471 /**
472 * Adds the expanding string into the collation table.
473 */
474 private final void addExpandOrder(String contractChars,
475 String expandChars,
476 int anOrder) throws ParseException
477 {
478 // Create an expansion table entry
479 int tableIndex = addExpansion(anOrder, expandChars);
480
481 // And add its index into the main mapping table
482 if (contractChars.length() > 1) {
483 char ch = contractChars.charAt(0);
484 if (Character.isHighSurrogate(ch) && contractChars.length() == 2) {
485 char ch2 = contractChars.charAt(1);
486 if (Character.isLowSurrogate(ch2)) {
487 //only add into table when it is a legal surrogate
488 addOrder(Character.toCodePoint(ch, ch2), tableIndex);
489 }
490 } else {
491 addContractOrder(contractChars, tableIndex);
492 }
493 } else {
494 addOrder(contractChars.charAt(0), tableIndex);
495 }
496 }
497
498 private final void addExpandOrder(int ch, String expandChars, int anOrder)
499 throws ParseException
500 {
501 int tableIndex = addExpansion(anOrder, expandChars);
502 addOrder(ch, tableIndex);
503 }
504
505 /**
506 * Create a new entry in the expansion table that contains the orderings
507 * for the given characers. If anOrder is valid, it is added to the
508 * beginning of the expanded list of orders.
509 */
510 private int addExpansion(int anOrder, String expandChars) {
511 if (expandTable == null) {
512 expandTable = new Vector<>(INITIALTABLESIZE);
513 }
514
515 // If anOrder is valid, we want to add it at the beginning of the list
516 int offset = (anOrder == RBCollationTables.UNMAPPED) ? 0 : 1;
517
518 int[] valueList = new int[expandChars.length() + offset];
519 if (offset == 1) {
520 valueList[0] = anOrder;
521 }
522
523 int j = offset;
524 for (int i = 0; i < expandChars.length(); i++) {
525 char ch0 = expandChars.charAt(i);
526 char ch1;
527 int ch;
528 if (Character.isHighSurrogate(ch0)) {
529 if (++i == expandChars.length() ||
530 !Character.isLowSurrogate(ch1=expandChars.charAt(i))) {
531 //ether we are missing the low surrogate or the next char
532 //is not a legal low surrogate, so stop loop
533 break;
534 }
535 ch = Character.toCodePoint(ch0, ch1);
536
537 } else {
538 ch = ch0;
539 }
540
541 int mapValue = getCharOrder(ch);
542
543 if (mapValue != RBCollationTables.UNMAPPED) {
544 valueList[j++] = mapValue;
545 } else {
546 // can't find it in the table, will be filled in by commit().
547 valueList[j++] = CHARINDEX + ch;
548 }
549 }
550 if (j < valueList.length) {
551 //we had at least one supplementary character, the size of valueList
552 //is bigger than it really needs...
553 int[] tmpBuf = new int[j];
554 while (--j >= 0) {
555 tmpBuf[j] = valueList[j];
556 }
557 valueList = tmpBuf;
558 }
559 // Add the expanding char list into the expansion table.
560 int tableIndex = RBCollationTables.EXPANDCHARINDEX + expandTable.size();
561 expandTable.addElement(valueList);
562
563 return tableIndex;
564 }
565
566 private void addContractFlags(String chars) {
567 char c0;
568 int c;
569 int len = chars.length();
570 for (int i = 0; i < len; i++) {
571 c0 = chars.charAt(i);
572 c = Character.isHighSurrogate(c0)
573 ?Character.toCodePoint(c0, chars.charAt(++i))
574 :c0;
575 contractFlags.put(c, 1);
576 }
577 }
578
579 // ==============================================================
580 // constants
581 // ==============================================================
582 static final int CHARINDEX = 0x70000000; // need look up in .commit()
583
584 private static final int IGNORABLEMASK = 0x0000ffff;
585 private static final int PRIMARYORDERINCREMENT = 0x00010000;
586 private static final int SECONDARYORDERINCREMENT = 0x00000100;
587 private static final int TERTIARYORDERINCREMENT = 0x00000001;
588 private static final int INITIALTABLESIZE = 20;
589 private static final int MAXKEYSIZE = 5;
590
591 // ==============================================================
592 // instance variables
593 // ==============================================================
594
595 // variables used by the build process
596 private RBCollationTables.BuildAPI tables = null;
597 private MergeCollation mPattern = null;
598 private boolean isOverIgnore = false;
599 private char[] keyBuf = new char[MAXKEYSIZE];
600 private IntHashtable contractFlags = new IntHashtable(100);
601
602 // "shadow" copies of the instance variables in RBCollationTables
603 // (the values in these variables are copied back into RBCollationTables
604 // at the end of the build process)
605 private boolean frenchSec = false;
606 private boolean seAsianSwapping = false;
607
608 private UCompactIntArray mapping = null;
609 private Vector<Vector<EntryPair>> contractTable = null;
610 private Vector<int[]> expandTable = null;
611
612 private short maxSecOrder = 0;
613 private short maxTerOrder = 0;
614 }