1 // © 2016 and later: Unicode, Inc. and others.
2 // License & terms of use: http://www.unicode.org/copyright.html
3 /*
4 ***************************************************************************
5 * Copyright (C) 1999-2016, International Business Machines Corporation
6 * and others. All Rights Reserved.
7 ***************************************************************************
8 * Date Name Description
9 * 10/20/99 alan Creation.
10 ***************************************************************************
11 */
12
13 #ifndef UNICODESET_H
14 #define UNICODESET_H
15
16 #include "unicode/utypes.h"
17
18 #if U_SHOW_CPLUSPLUS_API
19
20 #include "unicode/ucpmap.h"
21 #include "unicode/unifilt.h"
22 #include "unicode/unistr.h"
23 #include "unicode/uset.h"
24
25 /**
26 * \file
27 * \brief C++ API: Unicode Set
28 */
29
30 U_NAMESPACE_BEGIN
31
32 // Forward Declarations.
33 class BMPSet;
34 class ParsePosition;
35 class RBBIRuleScanner;
36 class SymbolTable;
37 class UnicodeSetStringSpan;
38 class UVector;
39 class RuleCharacterIterator;
40
41 /**
42 * A mutable set of Unicode characters and multicharacter strings. Objects of this class
43 * represent <em>character classes</em> used in regular expressions.
44 * A character specifies a subset of Unicode code points. Legal
45 * code points are U+0000 to U+10FFFF, inclusive.
46 *
47 * <p>The UnicodeSet class is not designed to be subclassed.
48 *
49 * <p><code>UnicodeSet</code> supports two APIs. The first is the
50 * <em>operand</em> API that allows the caller to modify the value of
51 * a <code>UnicodeSet</code> object. It conforms to Java 2's
52 * <code>java.util.Set</code> interface, although
53 * <code>UnicodeSet</code> does not actually implement that
54 * interface. All methods of <code>Set</code> are supported, with the
55 * modification that they take a character range or single character
56 * instead of an <code>Object</code>, and they take a
57 * <code>UnicodeSet</code> instead of a <code>Collection</code>. The
58 * operand API may be thought of in terms of boolean logic: a boolean
59 * OR is implemented by <code>add</code>, a boolean AND is implemented
60 * by <code>retain</code>, a boolean XOR is implemented by
61 * <code>complement</code> taking an argument, and a boolean NOT is
62 * implemented by <code>complement</code> with no argument. In terms
63 * of traditional set theory function names, <code>add</code> is a
64 * union, <code>retain</code> is an intersection, <code>remove</code>
65 * is an asymmetric difference, and <code>complement</code> with no
66 * argument is a set complement with respect to the superset range
67 * <code>MIN_VALUE-MAX_VALUE</code>
68 *
69 * <p>The second API is the
70 * <code>applyPattern()</code>/<code>toPattern()</code> API from the
71 * <code>java.text.Format</code>-derived classes. Unlike the
72 * methods that add characters, add categories, and control the logic
73 * of the set, the method <code>applyPattern()</code> sets all
74 * attributes of a <code>UnicodeSet</code> at once, based on a
75 * string pattern.
76 *
77 * <p><b>Pattern syntax</b></p>
78 *
79 * Patterns are accepted by the constructors and the
80 * <code>applyPattern()</code> methods and returned by the
81 * <code>toPattern()</code> method. These patterns follow a syntax
82 * similar to that employed by version 8 regular expression character
83 * classes. Here are some simple examples:
84 *
85 * \htmlonly<blockquote>\endhtmlonly
86 * <table>
87 * <tr align="top">
88 * <td nowrap valign="top" align="left"><code>[]</code></td>
89 * <td valign="top">No characters</td>
90 * </tr><tr align="top">
91 * <td nowrap valign="top" align="left"><code>[a]</code></td>
92 * <td valign="top">The character 'a'</td>
93 * </tr><tr align="top">
94 * <td nowrap valign="top" align="left"><code>[ae]</code></td>
95 * <td valign="top">The characters 'a' and 'e'</td>
96 * </tr>
97 * <tr>
98 * <td nowrap valign="top" align="left"><code>[a-e]</code></td>
99 * <td valign="top">The characters 'a' through 'e' inclusive, in Unicode code
100 * point order</td>
101 * </tr>
102 * <tr>
103 * <td nowrap valign="top" align="left"><code>[\\u4E01]</code></td>
104 * <td valign="top">The character U+4E01</td>
105 * </tr>
106 * <tr>
107 * <td nowrap valign="top" align="left"><code>[a{ab}{ac}]</code></td>
108 * <td valign="top">The character 'a' and the multicharacter strings "ab" and
109 * "ac"</td>
110 * </tr>
111 * <tr>
112 * <td nowrap valign="top" align="left"><code>[\\p{Lu}]</code></td>
113 * <td valign="top">All characters in the general category Uppercase Letter</td>
114 * </tr>
115 * </table>
116 * \htmlonly</blockquote>\endhtmlonly
117 *
118 * Any character may be preceded by a backslash in order to remove any special
119 * meaning. White space characters, as defined by UCharacter.isWhitespace(), are
120 * ignored, unless they are escaped.
121 *
122 * <p>Property patterns specify a set of characters having a certain
123 * property as defined by the Unicode standard. Both the POSIX-like
124 * "[:Lu:]" and the Perl-like syntax "\\p{Lu}" are recognized. For a
125 * complete list of supported property patterns, see the User's Guide
126 * for UnicodeSet at
127 * <a href="https://unicode-org.github.io/icu/userguide/strings/unicodeset">
128 * https://unicode-org.github.io/icu/userguide/strings/unicodeset</a>.
129 * Actual determination of property data is defined by the underlying
130 * Unicode database as implemented by UCharacter.
131 *
132 * <p>Patterns specify individual characters, ranges of characters, and
133 * Unicode property sets. When elements are concatenated, they
134 * specify their union. To complement a set, place a '^' immediately
135 * after the opening '['. Property patterns are inverted by modifying
136 * their delimiters; "[:^foo]" and "\\P{foo}". In any other location,
137 * '^' has no special meaning.
138 *
139 * <p>Since ICU 70, "[^...]", "[:^foo]", "\\P{foo}", and "[:binaryProperty=No:]"
140 * perform a “code point complement” (all code points minus the original set),
141 * removing all multicharacter strings,
142 * equivalent to <code>.complement().removeAllStrings()</code>.
143 * The complement() API function continues to perform a
144 * symmetric difference with all code points and thus retains all multicharacter strings.
145 *
146 * <p>Ranges are indicated by placing two a '-' between two
147 * characters, as in "a-z". This specifies the range of all
148 * characters from the left to the right, in Unicode order. If the
149 * left character is greater than or equal to the
150 * right character it is a syntax error. If a '-' occurs as the first
151 * character after the opening '[' or '[^', or if it occurs as the
152 * last character before the closing ']', then it is taken as a
153 * literal. Thus "[a\-b]", "[-ab]", and "[ab-]" all indicate the same
154 * set of three characters, 'a', 'b', and '-'.
155 *
156 * <p>Sets may be intersected using the '&' operator or the asymmetric
157 * set difference may be taken using the '-' operator, for example,
158 * "[[:L:]&[\\u0000-\\u0FFF]]" indicates the set of all Unicode letters
159 * with values less than 4096. Operators ('&' and '|') have equal
160 * precedence and bind left-to-right. Thus
161 * "[[:L:]-[a-z]-[\\u0100-\\u01FF]]" is equivalent to
162 * "[[[:L:]-[a-z]]-[\\u0100-\\u01FF]]". This only really matters for
163 * difference; intersection is commutative.
164 *
165 * <table>
166 * <tr valign=top><td nowrap><code>[a]</code><td>The set containing 'a'
167 * <tr valign=top><td nowrap><code>[a-z]</code><td>The set containing 'a'
168 * through 'z' and all letters in between, in Unicode order
169 * <tr valign=top><td nowrap><code>[^a-z]</code><td>The set containing
170 * all characters but 'a' through 'z',
171 * that is, U+0000 through 'a'-1 and 'z'+1 through U+10FFFF
172 * <tr valign=top><td nowrap><code>[[<em>pat1</em>][<em>pat2</em>]]</code>
173 * <td>The union of sets specified by <em>pat1</em> and <em>pat2</em>
174 * <tr valign=top><td nowrap><code>[[<em>pat1</em>]&[<em>pat2</em>]]</code>
175 * <td>The intersection of sets specified by <em>pat1</em> and <em>pat2</em>
176 * <tr valign=top><td nowrap><code>[[<em>pat1</em>]-[<em>pat2</em>]]</code>
177 * <td>The asymmetric difference of sets specified by <em>pat1</em> and
178 * <em>pat2</em>
179 * <tr valign=top><td nowrap><code>[:Lu:] or \\p{Lu}</code>
180 * <td>The set of characters having the specified
181 * Unicode property; in
182 * this case, Unicode uppercase letters
183 * <tr valign=top><td nowrap><code>[:^Lu:] or \\P{Lu}</code>
184 * <td>The set of characters <em>not</em> having the given
185 * Unicode property
186 * </table>
187 *
188 * <p><b>Formal syntax</b></p>
189 *
190 * \htmlonly<blockquote>\endhtmlonly
191 * <table>
192 * <tr align="top">
193 * <td nowrap valign="top" align="right"><code>pattern := </code></td>
194 * <td valign="top"><code>('[' '^'? item* ']') |
195 * property</code></td>
196 * </tr>
197 * <tr align="top">
198 * <td nowrap valign="top" align="right"><code>item := </code></td>
199 * <td valign="top"><code>char | (char '-' char) | pattern-expr<br>
200 * </code></td>
201 * </tr>
202 * <tr align="top">
203 * <td nowrap valign="top" align="right"><code>pattern-expr := </code></td>
204 * <td valign="top"><code>pattern | pattern-expr pattern |
205 * pattern-expr op pattern<br>
206 * </code></td>
207 * </tr>
208 * <tr align="top">
209 * <td nowrap valign="top" align="right"><code>op := </code></td>
210 * <td valign="top"><code>'&' | '-'<br>
211 * </code></td>
212 * </tr>
213 * <tr align="top">
214 * <td nowrap valign="top" align="right"><code>special := </code></td>
215 * <td valign="top"><code>'[' | ']' | '-'<br>
216 * </code></td>
217 * </tr>
218 * <tr align="top">
219 * <td nowrap valign="top" align="right"><code>char := </code></td>
220 * <td valign="top"><em>any character that is not</em><code> special<br>
221 * | ('\' </code><em>any character</em><code>)<br>
222 * | ('\\u' hex hex hex hex)<br>
223 * </code></td>
224 * </tr>
225 * <tr align="top">
226 * <td nowrap valign="top" align="right"><code>hex := </code></td>
227 * <td valign="top"><code>'0' | '1' | '2' | '3' | '4' | '5' | '6' | '7' | '8' | '9' |<br>
228 * 'A' | 'B' | 'C' | 'D' | 'E' | 'F' | 'a' | 'b' | 'c' | 'd' | 'e' | 'f'</code></td>
229 * </tr>
230 * <tr>
231 * <td nowrap valign="top" align="right"><code>property := </code></td>
232 * <td valign="top"><em>a Unicode property set pattern</em></td>
233 * </tr>
234 * </table>
235 * <br>
236 * <table border="1">
237 * <tr>
238 * <td>Legend: <table>
239 * <tr>
240 * <td nowrap valign="top"><code>a := b</code></td>
241 * <td width="20" valign="top"> </td>
242 * <td valign="top"><code>a</code> may be replaced by <code>b</code> </td>
243 * </tr>
244 * <tr>
245 * <td nowrap valign="top"><code>a?</code></td>
246 * <td valign="top"></td>
247 * <td valign="top">zero or one instance of <code>a</code><br>
248 * </td>
249 * </tr>
250 * <tr>
251 * <td nowrap valign="top"><code>a*</code></td>
252 * <td valign="top"></td>
253 * <td valign="top">one or more instances of <code>a</code><br>
254 * </td>
255 * </tr>
256 * <tr>
257 * <td nowrap valign="top"><code>a | b</code></td>
258 * <td valign="top"></td>
259 * <td valign="top">either <code>a</code> or <code>b</code><br>
260 * </td>
261 * </tr>
262 * <tr>
263 * <td nowrap valign="top"><code>'a'</code></td>
264 * <td valign="top"></td>
265 * <td valign="top">the literal string between the quotes </td>
266 * </tr>
267 * </table>
268 * </td>
269 * </tr>
270 * </table>
271 * \htmlonly</blockquote>\endhtmlonly
272 *
273 * <p>Note:
274 * - Most UnicodeSet methods do not take a UErrorCode parameter because
275 * there are usually very few opportunities for failure other than a shortage
276 * of memory, error codes in low-level C++ string methods would be inconvenient,
277 * and the error code as the last parameter (ICU convention) would prevent
278 * the use of default parameter values.
279 * Instead, such methods set the UnicodeSet into a "bogus" state
280 * (see isBogus()) if an error occurs.
281 *
282 * @author Alan Liu
283 * @stable ICU 2.0
284 */
285 class U_COMMON_API UnicodeSet U_FINAL : public UnicodeFilter {
286 private:
287 /**
288 * Enough for sets with few ranges.
289 * For example, White_Space has 10 ranges, list length 21.
290 */
291 static constexpr int32_t INITIAL_CAPACITY = 25;
292 // fFlags constant
293 static constexpr uint8_t kIsBogus = 1; // This set is bogus (i.e. not valid)
294
295 UChar32* list = stackList; // MUST be terminated with HIGH
296 int32_t capacity = INITIAL_CAPACITY; // capacity of list
297 int32_t len = 1; // length of list used; 1 <= len <= capacity
298 uint8_t fFlags = 0; // Bit flag (see constants above)
299
300 BMPSet *bmpSet = nullptr; // The set is frozen iff either bmpSet or stringSpan is not NULL.
301 UChar32* buffer = nullptr; // internal buffer, may be NULL
302 int32_t bufferCapacity = 0; // capacity of buffer
303
304 /**
305 * The pattern representation of this set. This may not be the
306 * most economical pattern. It is the pattern supplied to
307 * applyPattern(), with variables substituted and whitespace
308 * removed. For sets constructed without applyPattern(), or
309 * modified using the non-pattern API, this string will be empty,
310 * indicating that toPattern() must generate a pattern
311 * representation from the inversion list.
312 */
313 char16_t *pat = nullptr;
314 int32_t patLen = 0;
315
316 UVector* strings = nullptr; // maintained in sorted order
317 UnicodeSetStringSpan *stringSpan = nullptr;
318
319 /**
320 * Initial list array.
321 * Avoids some heap allocations, and list is never nullptr.
322 * Increases the object size a bit.
323 */
324 UChar32 stackList[INITIAL_CAPACITY];
325
326 public:
327 /**
328 * Determine if this object contains a valid set.
329 * A bogus set has no value. It is different from an empty set.
330 * It can be used to indicate that no set value is available.
331 *
332 * @return true if the set is bogus/invalid, false otherwise
333 * @see setToBogus()
334 * @stable ICU 4.0
335 */
336 inline UBool isBogus(void) const;
337
338 /**
339 * Make this UnicodeSet object invalid.
340 * The string will test true with isBogus().
341 *
342 * A bogus set has no value. It is different from an empty set.
343 * It can be used to indicate that no set value is available.
344 *
345 * This utility function is used throughout the UnicodeSet
346 * implementation to indicate that a UnicodeSet operation failed,
347 * and may be used in other functions,
348 * especially but not exclusively when such functions do not
349 * take a UErrorCode for simplicity.
350 *
351 * @see isBogus()
352 * @stable ICU 4.0
353 */
354 void setToBogus();
355
356 public:
357
358 enum {
359 /**
360 * Minimum value that can be stored in a UnicodeSet.
361 * @stable ICU 2.4
362 */
363 MIN_VALUE = 0,
364
365 /**
366 * Maximum value that can be stored in a UnicodeSet.
367 * @stable ICU 2.4
368 */
369 MAX_VALUE = 0x10ffff
370 };
371
372 //----------------------------------------------------------------
373 // Constructors &c
374 //----------------------------------------------------------------
375
376 public:
377
378 /**
379 * Constructs an empty set.
380 * @stable ICU 2.0
381 */
382 UnicodeSet();
383
384 /**
385 * Constructs a set containing the given range. If <code>end <
386 * start</code> then an empty set is created.
387 *
388 * @param start first character, inclusive, of range
389 * @param end last character, inclusive, of range
390 * @stable ICU 2.4
391 */
392 UnicodeSet(UChar32 start, UChar32 end);
393
394 #ifndef U_HIDE_INTERNAL_API
395 /**
396 * @internal
397 */
398 enum ESerialization {
399 kSerialized /* result of serialize() */
400 };
401
402 /**
403 * Constructs a set from the output of serialize().
404 *
405 * @param buffer the 16 bit array
406 * @param bufferLen the original length returned from serialize()
407 * @param serialization the value 'kSerialized'
408 * @param status error code
409 *
410 * @internal
411 */
412 UnicodeSet(const uint16_t buffer[], int32_t bufferLen,
413 ESerialization serialization, UErrorCode &status);
414 #endif /* U_HIDE_INTERNAL_API */
415
416 /**
417 * Constructs a set from the given pattern. See the class
418 * description for the syntax of the pattern language.
419 * @param pattern a string specifying what characters are in the set
420 * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
421 * contains a syntax error.
422 * @stable ICU 2.0
423 */
424 UnicodeSet(const UnicodeString& pattern,
425 UErrorCode& status);
426
427 #ifndef U_HIDE_INTERNAL_API
428 /**
429 * Constructs a set from the given pattern. See the class
430 * description for the syntax of the pattern language.
431 * @param pattern a string specifying what characters are in the set
432 * @param options bitmask for options to apply to the pattern.
433 * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
434 * @param symbols a symbol table mapping variable names to values
435 * and stand-in characters to UnicodeSets; may be NULL
436 * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
437 * contains a syntax error.
438 * @internal
439 */
440 UnicodeSet(const UnicodeString& pattern,
441 uint32_t options,
442 const SymbolTable* symbols,
443 UErrorCode& status);
444 #endif /* U_HIDE_INTERNAL_API */
445
446 /**
447 * Constructs a set from the given pattern. See the class description
448 * for the syntax of the pattern language.
449 * @param pattern a string specifying what characters are in the set
450 * @param pos on input, the position in pattern at which to start parsing.
451 * On output, the position after the last character parsed.
452 * @param options bitmask for options to apply to the pattern.
453 * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
454 * @param symbols a symbol table mapping variable names to values
455 * and stand-in characters to UnicodeSets; may be NULL
456 * @param status input-output error code
457 * @stable ICU 2.8
458 */
459 UnicodeSet(const UnicodeString& pattern, ParsePosition& pos,
460 uint32_t options,
461 const SymbolTable* symbols,
462 UErrorCode& status);
463
464 /**
465 * Constructs a set that is identical to the given UnicodeSet.
466 * @stable ICU 2.0
467 */
468 UnicodeSet(const UnicodeSet& o);
469
470 /**
471 * Destructs the set.
472 * @stable ICU 2.0
473 */
474 virtual ~UnicodeSet();
475
476 /**
477 * Assigns this object to be a copy of another.
478 * A frozen set will not be modified.
479 * @stable ICU 2.0
480 */
481 UnicodeSet& operator=(const UnicodeSet& o);
482
483 /**
484 * Compares the specified object with this set for equality. Returns
485 * <tt>true</tt> if the two sets
486 * have the same size, and every member of the specified set is
487 * contained in this set (or equivalently, every member of this set is
488 * contained in the specified set).
489 *
490 * @param o set to be compared for equality with this set.
491 * @return <tt>true</tt> if the specified set is equal to this set.
492 * @stable ICU 2.0
493 */
494 virtual bool operator==(const UnicodeSet& o) const;
495
496 /**
497 * Compares the specified object with this set for equality. Returns
498 * <tt>true</tt> if the specified set is not equal to this set.
499 * @stable ICU 2.0
500 */
501 inline bool operator!=(const UnicodeSet& o) const;
502
503 /**
504 * Returns a copy of this object. All UnicodeFunctor objects have
505 * to support cloning in order to allow classes using
506 * UnicodeFunctors, such as Transliterator, to implement cloning.
507 * If this set is frozen, then the clone will be frozen as well.
508 * Use cloneAsThawed() for a mutable clone of a frozen set.
509 * @see cloneAsThawed
510 * @stable ICU 2.0
511 */
512 virtual UnicodeSet* clone() const override;
513
514 /**
515 * Returns the hash code value for this set.
516 *
517 * @return the hash code value for this set.
518 * @see Object#hashCode()
519 * @stable ICU 2.0
520 */
521 virtual int32_t hashCode(void) const;
522
523 /**
524 * Get a UnicodeSet pointer from a USet
525 *
526 * @param uset a USet (the ICU plain C type for UnicodeSet)
527 * @return the corresponding UnicodeSet pointer.
528 *
529 * @stable ICU 4.2
530 */
531 inline static UnicodeSet *fromUSet(USet *uset);
532
533 /**
534 * Get a UnicodeSet pointer from a const USet
535 *
536 * @param uset a const USet (the ICU plain C type for UnicodeSet)
537 * @return the corresponding UnicodeSet pointer.
538 *
539 * @stable ICU 4.2
540 */
541 inline static const UnicodeSet *fromUSet(const USet *uset);
542
543 /**
544 * Produce a USet * pointer for this UnicodeSet.
545 * USet is the plain C type for UnicodeSet
546 *
547 * @return a USet pointer for this UnicodeSet
548 * @stable ICU 4.2
549 */
550 inline USet *toUSet();
551
552
553 /**
554 * Produce a const USet * pointer for this UnicodeSet.
555 * USet is the plain C type for UnicodeSet
556 *
557 * @return a const USet pointer for this UnicodeSet
558 * @stable ICU 4.2
559 */
560 inline const USet * toUSet() const;
561
562
563 //----------------------------------------------------------------
564 // Freezable API
565 //----------------------------------------------------------------
566
567 /**
568 * Determines whether the set has been frozen (made immutable) or not.
569 * See the ICU4J Freezable interface for details.
570 * @return true/false for whether the set has been frozen
571 * @see freeze
572 * @see cloneAsThawed
573 * @stable ICU 3.8
574 */
575 inline UBool isFrozen() const;
576
577 /**
578 * Freeze the set (make it immutable).
579 * Once frozen, it cannot be unfrozen and is therefore thread-safe
580 * until it is deleted.
581 * See the ICU4J Freezable interface for details.
582 * Freezing the set may also make some operations faster, for example
583 * contains() and span().
584 * A frozen set will not be modified. (It remains frozen.)
585 * @return this set.
586 * @see isFrozen
587 * @see cloneAsThawed
588 * @stable ICU 3.8
589 */
590 UnicodeSet *freeze();
591
592 /**
593 * Clone the set and make the clone mutable.
594 * See the ICU4J Freezable interface for details.
595 * @return the mutable clone
596 * @see freeze
597 * @see isFrozen
598 * @stable ICU 3.8
599 */
600 UnicodeSet *cloneAsThawed() const;
601
602 //----------------------------------------------------------------
603 // Public API
604 //----------------------------------------------------------------
605
606 /**
607 * Make this object represent the range `start - end`.
608 * If `start > end` then this object is set to an empty range.
609 * A frozen set will not be modified.
610 *
611 * @param start first character in the set, inclusive
612 * @param end last character in the set, inclusive
613 * @stable ICU 2.4
614 */
615 UnicodeSet& set(UChar32 start, UChar32 end);
616
617 /**
618 * Return true if the given position, in the given pattern, appears
619 * to be the start of a UnicodeSet pattern.
620 * @stable ICU 2.4
621 */
622 static UBool resemblesPattern(const UnicodeString& pattern,
623 int32_t pos);
624
625 /**
626 * Modifies this set to represent the set specified by the given
627 * pattern, ignoring Unicode Pattern_White_Space characters.
628 * See the class description for the syntax of the pattern language.
629 * A frozen set will not be modified.
630 * @param pattern a string specifying what characters are in the set
631 * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
632 * contains a syntax error.
633 * <em> Empties the set passed before applying the pattern.</em>
634 * @return a reference to this
635 * @stable ICU 2.0
636 */
637 UnicodeSet& applyPattern(const UnicodeString& pattern,
638 UErrorCode& status);
639
640 #ifndef U_HIDE_INTERNAL_API
641 /**
642 * Modifies this set to represent the set specified by the given
643 * pattern, optionally ignoring Unicode Pattern_White_Space characters.
644 * See the class description for the syntax of the pattern language.
645 * A frozen set will not be modified.
646 * @param pattern a string specifying what characters are in the set
647 * @param options bitmask for options to apply to the pattern.
648 * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
649 * @param symbols a symbol table mapping variable names to
650 * values and stand-ins to UnicodeSets; may be NULL
651 * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
652 * contains a syntax error.
653 *<em> Empties the set passed before applying the pattern.</em>
654 * @return a reference to this
655 * @internal
656 */
657 UnicodeSet& applyPattern(const UnicodeString& pattern,
658 uint32_t options,
659 const SymbolTable* symbols,
660 UErrorCode& status);
661 #endif /* U_HIDE_INTERNAL_API */
662
663 /**
664 * Parses the given pattern, starting at the given position. The
665 * character at pattern.charAt(pos.getIndex()) must be '[', or the
666 * parse fails. Parsing continues until the corresponding closing
667 * ']'. If a syntax error is encountered between the opening and
668 * closing brace, the parse fails. Upon return from a successful
669 * parse, the ParsePosition is updated to point to the character
670 * following the closing ']', and a StringBuffer containing a
671 * pairs list for the parsed pattern is returned. This method calls
672 * itself recursively to parse embedded subpatterns.
673 *<em> Empties the set passed before applying the pattern.</em>
674 * A frozen set will not be modified.
675 *
676 * @param pattern the string containing the pattern to be parsed.
677 * The portion of the string from pos.getIndex(), which must be a
678 * '[', to the corresponding closing ']', is parsed.
679 * @param pos upon entry, the position at which to being parsing.
680 * The character at pattern.charAt(pos.getIndex()) must be a '['.
681 * Upon return from a successful parse, pos.getIndex() is either
682 * the character after the closing ']' of the parsed pattern, or
683 * pattern.length() if the closing ']' is the last character of
684 * the pattern string.
685 * @param options bitmask for options to apply to the pattern.
686 * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE.
687 * @param symbols a symbol table mapping variable names to
688 * values and stand-ins to UnicodeSets; may be NULL
689 * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern
690 * contains a syntax error.
691 * @return a reference to this
692 * @stable ICU 2.8
693 */
694 UnicodeSet& applyPattern(const UnicodeString& pattern,
695 ParsePosition& pos,
696 uint32_t options,
697 const SymbolTable* symbols,
698 UErrorCode& status);
699
700 /**
701 * Returns a string representation of this set. If the result of
702 * calling this function is passed to a UnicodeSet constructor, it
703 * will produce another set that is equal to this one.
704 * A frozen set will not be modified.
705 * @param result the string to receive the rules. Previous
706 * contents will be deleted.
707 * @param escapeUnprintable if true then convert unprintable
708 * character to their hex escape representations, \\uxxxx or
709 * \\Uxxxxxxxx. Unprintable characters are those other than
710 * U+000A, U+0020..U+007E.
711 * @stable ICU 2.0
712 */
713 virtual UnicodeString& toPattern(UnicodeString& result,
714 UBool escapeUnprintable = false) const override;
715
716 /**
717 * Modifies this set to contain those code points which have the given value
718 * for the given binary or enumerated property, as returned by
719 * u_getIntPropertyValue. Prior contents of this set are lost.
720 * A frozen set will not be modified.
721 *
722 * @param prop a property in the range UCHAR_BIN_START..UCHAR_BIN_LIMIT-1
723 * or UCHAR_INT_START..UCHAR_INT_LIMIT-1
724 * or UCHAR_MASK_START..UCHAR_MASK_LIMIT-1.
725 *
726 * @param value a value in the range u_getIntPropertyMinValue(prop)..
727 * u_getIntPropertyMaxValue(prop), with one exception. If prop is
728 * UCHAR_GENERAL_CATEGORY_MASK, then value should not be a UCharCategory, but
729 * rather a mask value produced by U_GET_GC_MASK(). This allows grouped
730 * categories such as [:L:] to be represented.
731 *
732 * @param ec error code input/output parameter
733 *
734 * @return a reference to this set
735 *
736 * @stable ICU 2.4
737 */
738 UnicodeSet& applyIntPropertyValue(UProperty prop,
739 int32_t value,
740 UErrorCode& ec);
741
742 /**
743 * Modifies this set to contain those code points which have the
744 * given value for the given property. Prior contents of this
745 * set are lost.
746 * A frozen set will not be modified.
747 *
748 * @param prop a property alias, either short or long. The name is matched
749 * loosely. See PropertyAliases.txt for names and a description of loose
750 * matching. If the value string is empty, then this string is interpreted
751 * as either a General_Category value alias, a Script value alias, a binary
752 * property alias, or a special ID. Special IDs are matched loosely and
753 * correspond to the following sets:
754 *
755 * "ANY" = [\\u0000-\\U0010FFFF],
756 * "ASCII" = [\\u0000-\\u007F],
757 * "Assigned" = [:^Cn:].
758 *
759 * @param value a value alias, either short or long. The name is matched
760 * loosely. See PropertyValueAliases.txt for names and a description of
761 * loose matching. In addition to aliases listed, numeric values and
762 * canonical combining classes may be expressed numerically, e.g., ("nv",
763 * "0.5") or ("ccc", "220"). The value string may also be empty.
764 *
765 * @param ec error code input/output parameter
766 *
767 * @return a reference to this set
768 *
769 * @stable ICU 2.4
770 */
771 UnicodeSet& applyPropertyAlias(const UnicodeString& prop,
772 const UnicodeString& value,
773 UErrorCode& ec);
774
775 /**
776 * Returns the number of elements in this set (its cardinality).
777 * Note than the elements of a set may include both individual
778 * codepoints and strings.
779 *
780 * This is slower than getRangeCount() because
781 * it counts the code points of all ranges.
782 *
783 * @return the number of elements in this set (its cardinality).
784 * @stable ICU 2.0
785 * @see getRangeCount
786 */
787 virtual int32_t size(void) const;
788
789 /**
790 * Returns <tt>true</tt> if this set contains no elements.
791 *
792 * @return <tt>true</tt> if this set contains no elements.
793 * @stable ICU 2.0
794 */
795 virtual UBool isEmpty(void) const;
796
797 /**
798 * @return true if this set contains multi-character strings or the empty string.
799 * @stable ICU 70
800 */
801 UBool hasStrings() const;
802
803 /**
804 * Returns true if this set contains the given character.
805 * This function works faster with a frozen set.
806 * @param c character to be checked for containment
807 * @return true if the test condition is met
808 * @stable ICU 2.0
809 */
810 virtual UBool contains(UChar32 c) const override;
811
812 /**
813 * Returns true if this set contains every character
814 * of the given range.
815 * @param start first character, inclusive, of the range
816 * @param end last character, inclusive, of the range
817 * @return true if the test condition is met
818 * @stable ICU 2.0
819 */
820 virtual UBool contains(UChar32 start, UChar32 end) const;
821
822 /**
823 * Returns <tt>true</tt> if this set contains the given
824 * multicharacter string.
825 * @param s string to be checked for containment
826 * @return <tt>true</tt> if this set contains the specified string
827 * @stable ICU 2.4
828 */
829 UBool contains(const UnicodeString& s) const;
830
831 /**
832 * Returns true if this set contains all the characters and strings
833 * of the given set.
834 * @param c set to be checked for containment
835 * @return true if the test condition is met
836 * @stable ICU 2.4
837 */
838 virtual UBool containsAll(const UnicodeSet& c) const;
839
840 /**
841 * Returns true if this set contains all the characters
842 * of the given string.
843 * @param s string containing characters to be checked for containment
844 * @return true if the test condition is met
845 * @stable ICU 2.4
846 */
847 UBool containsAll(const UnicodeString& s) const;
848
849 /**
850 * Returns true if this set contains none of the characters
851 * of the given range.
852 * @param start first character, inclusive, of the range
853 * @param end last character, inclusive, of the range
854 * @return true if the test condition is met
855 * @stable ICU 2.4
856 */
857 UBool containsNone(UChar32 start, UChar32 end) const;
858
859 /**
860 * Returns true if this set contains none of the characters and strings
861 * of the given set.
862 * @param c set to be checked for containment
863 * @return true if the test condition is met
864 * @stable ICU 2.4
865 */
866 UBool containsNone(const UnicodeSet& c) const;
867
868 /**
869 * Returns true if this set contains none of the characters
870 * of the given string.
871 * @param s string containing characters to be checked for containment
872 * @return true if the test condition is met
873 * @stable ICU 2.4
874 */
875 UBool containsNone(const UnicodeString& s) const;
876
877 /**
878 * Returns true if this set contains one or more of the characters
879 * in the given range.
880 * @param start first character, inclusive, of the range
881 * @param end last character, inclusive, of the range
882 * @return true if the condition is met
883 * @stable ICU 2.4
884 */
885 inline UBool containsSome(UChar32 start, UChar32 end) const;
886
887 /**
888 * Returns true if this set contains one or more of the characters
889 * and strings of the given set.
890 * @param s The set to be checked for containment
891 * @return true if the condition is met
892 * @stable ICU 2.4
893 */
894 inline UBool containsSome(const UnicodeSet& s) const;
895
896 /**
897 * Returns true if this set contains one or more of the characters
898 * of the given string.
899 * @param s string containing characters to be checked for containment
900 * @return true if the condition is met
901 * @stable ICU 2.4
902 */
903 inline UBool containsSome(const UnicodeString& s) const;
904
905 /**
906 * Returns the length of the initial substring of the input string which
907 * consists only of characters and strings that are contained in this set
908 * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
909 * or only of characters and strings that are not contained
910 * in this set (USET_SPAN_NOT_CONTAINED).
911 * See USetSpanCondition for details.
912 * Similar to the strspn() C library function.
913 * Unpaired surrogates are treated according to contains() of their surrogate code points.
914 * This function works faster with a frozen set and with a non-negative string length argument.
915 * @param s start of the string
916 * @param length of the string; can be -1 for NUL-terminated
917 * @param spanCondition specifies the containment condition
918 * @return the length of the initial substring according to the spanCondition;
919 * 0 if the start of the string does not fit the spanCondition
920 * @stable ICU 3.8
921 * @see USetSpanCondition
922 */
923 int32_t span(const char16_t *s, int32_t length, USetSpanCondition spanCondition) const;
924
925 /**
926 * Returns the end of the substring of the input string according to the USetSpanCondition.
927 * Same as <code>start+span(s.getBuffer()+start, s.length()-start, spanCondition)</code>
928 * after pinning start to 0<=start<=s.length().
929 * @param s the string
930 * @param start the start index in the string for the span operation
931 * @param spanCondition specifies the containment condition
932 * @return the exclusive end of the substring according to the spanCondition;
933 * the substring s.tempSubStringBetween(start, end) fulfills the spanCondition
934 * @stable ICU 4.4
935 * @see USetSpanCondition
936 */
937 inline int32_t span(const UnicodeString &s, int32_t start, USetSpanCondition spanCondition) const;
938
939 /**
940 * Returns the start of the trailing substring of the input string which
941 * consists only of characters and strings that are contained in this set
942 * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
943 * or only of characters and strings that are not contained
944 * in this set (USET_SPAN_NOT_CONTAINED).
945 * See USetSpanCondition for details.
946 * Unpaired surrogates are treated according to contains() of their surrogate code points.
947 * This function works faster with a frozen set and with a non-negative string length argument.
948 * @param s start of the string
949 * @param length of the string; can be -1 for NUL-terminated
950 * @param spanCondition specifies the containment condition
951 * @return the start of the trailing substring according to the spanCondition;
952 * the string length if the end of the string does not fit the spanCondition
953 * @stable ICU 3.8
954 * @see USetSpanCondition
955 */
956 int32_t spanBack(const char16_t *s, int32_t length, USetSpanCondition spanCondition) const;
957
958 /**
959 * Returns the start of the substring of the input string according to the USetSpanCondition.
960 * Same as <code>spanBack(s.getBuffer(), limit, spanCondition)</code>
961 * after pinning limit to 0<=end<=s.length().
962 * @param s the string
963 * @param limit the exclusive-end index in the string for the span operation
964 * (use s.length() or INT32_MAX for spanning back from the end of the string)
965 * @param spanCondition specifies the containment condition
966 * @return the start of the substring according to the spanCondition;
967 * the substring s.tempSubStringBetween(start, limit) fulfills the spanCondition
968 * @stable ICU 4.4
969 * @see USetSpanCondition
970 */
971 inline int32_t spanBack(const UnicodeString &s, int32_t limit, USetSpanCondition spanCondition) const;
972
973 /**
974 * Returns the length of the initial substring of the input string which
975 * consists only of characters and strings that are contained in this set
976 * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
977 * or only of characters and strings that are not contained
978 * in this set (USET_SPAN_NOT_CONTAINED).
979 * See USetSpanCondition for details.
980 * Similar to the strspn() C library function.
981 * Malformed byte sequences are treated according to contains(0xfffd).
982 * This function works faster with a frozen set and with a non-negative string length argument.
983 * @param s start of the string (UTF-8)
984 * @param length of the string; can be -1 for NUL-terminated
985 * @param spanCondition specifies the containment condition
986 * @return the length of the initial substring according to the spanCondition;
987 * 0 if the start of the string does not fit the spanCondition
988 * @stable ICU 3.8
989 * @see USetSpanCondition
990 */
991 int32_t spanUTF8(const char *s, int32_t length, USetSpanCondition spanCondition) const;
992
993 /**
994 * Returns the start of the trailing substring of the input string which
995 * consists only of characters and strings that are contained in this set
996 * (USET_SPAN_CONTAINED, USET_SPAN_SIMPLE),
997 * or only of characters and strings that are not contained
998 * in this set (USET_SPAN_NOT_CONTAINED).
999 * See USetSpanCondition for details.
1000 * Malformed byte sequences are treated according to contains(0xfffd).
1001 * This function works faster with a frozen set and with a non-negative string length argument.
1002 * @param s start of the string (UTF-8)
1003 * @param length of the string; can be -1 for NUL-terminated
1004 * @param spanCondition specifies the containment condition
1005 * @return the start of the trailing substring according to the spanCondition;
1006 * the string length if the end of the string does not fit the spanCondition
1007 * @stable ICU 3.8
1008 * @see USetSpanCondition
1009 */
1010 int32_t spanBackUTF8(const char *s, int32_t length, USetSpanCondition spanCondition) const;
1011
1012 /**
1013 * Implement UnicodeMatcher::matches()
1014 * @stable ICU 2.4
1015 */
1016 virtual UMatchDegree matches(const Replaceable& text,
1017 int32_t& offset,
1018 int32_t limit,
1019 UBool incremental) override;
1020
1021 private:
1022 /**
1023 * Returns the longest match for s in text at the given position.
1024 * If limit > start then match forward from start+1 to limit
1025 * matching all characters except s.charAt(0). If limit < start,
1026 * go backward starting from start-1 matching all characters
1027 * except s.charAt(s.length()-1). This method assumes that the
1028 * first character, text.charAt(start), matches s, so it does not
1029 * check it.
1030 * @param text the text to match
1031 * @param start the first character to match. In the forward
1032 * direction, text.charAt(start) is matched against s.charAt(0).
1033 * In the reverse direction, it is matched against
1034 * s.charAt(s.length()-1).
1035 * @param limit the limit offset for matching, either last+1 in
1036 * the forward direction, or last-1 in the reverse direction,
1037 * where last is the index of the last character to match.
1038 * @param s
1039 * @return If part of s matches up to the limit, return |limit -
1040 * start|. If all of s matches before reaching the limit, return
1041 * s.length(). If there is a mismatch between s and text, return
1042 * 0
1043 */
1044 static int32_t matchRest(const Replaceable& text,
1045 int32_t start, int32_t limit,
1046 const UnicodeString& s);
1047
1048 /**
1049 * Returns the smallest value i such that c < list[i]. Caller
1050 * must ensure that c is a legal value or this method will enter
1051 * an infinite loop. This method performs a binary search.
1052 * @param c a character in the range MIN_VALUE..MAX_VALUE
1053 * inclusive
1054 * @return the smallest integer i in the range 0..len-1,
1055 * inclusive, such that c < list[i]
1056 */
1057 int32_t findCodePoint(UChar32 c) const;
1058
1059 public:
1060
1061 /**
1062 * Implementation of UnicodeMatcher API. Union the set of all
1063 * characters that may be matched by this object into the given
1064 * set.
1065 * @param toUnionTo the set into which to union the source characters
1066 * @stable ICU 2.4
1067 */
1068 virtual void addMatchSetTo(UnicodeSet& toUnionTo) const override;
1069
1070 /**
1071 * Returns the index of the given character within this set, where
1072 * the set is ordered by ascending code point. If the character
1073 * is not in this set, return -1. The inverse of this method is
1074 * <code>charAt()</code>.
1075 * @return an index from 0..size()-1, or -1
1076 * @stable ICU 2.4
1077 */
1078 int32_t indexOf(UChar32 c) const;
1079
1080 /**
1081 * Returns the character at the given index within this set, where
1082 * the set is ordered by ascending code point. If the index is
1083 * out of range for characters, returns (UChar32)-1.
1084 * The inverse of this method is <code>indexOf()</code>.
1085 *
1086 * For iteration, this is slower than UnicodeSetIterator or
1087 * getRangeCount()/getRangeStart()/getRangeEnd(),
1088 * because for each call it skips linearly over <code>index</code>
1089 * characters in the ranges.
1090 *
1091 * @param index an index from 0..size()-1
1092 * @return the character at the given index, or (UChar32)-1.
1093 * @stable ICU 2.4
1094 */
1095 UChar32 charAt(int32_t index) const;
1096
1097 /**
1098 * Adds the specified range to this set if it is not already
1099 * present. If this set already contains the specified range,
1100 * the call leaves this set unchanged. If <code>start > end</code>
1101 * then an empty range is added, leaving the set unchanged.
1102 * This is equivalent to a boolean logic OR, or a set UNION.
1103 * A frozen set will not be modified.
1104 *
1105 * @param start first character, inclusive, of range to be added
1106 * to this set.
1107 * @param end last character, inclusive, of range to be added
1108 * to this set.
1109 * @stable ICU 2.0
1110 */
1111 virtual UnicodeSet& add(UChar32 start, UChar32 end);
1112
1113 /**
1114 * Adds the specified character to this set if it is not already
1115 * present. If this set already contains the specified character,
1116 * the call leaves this set unchanged.
1117 * A frozen set will not be modified.
1118 *
1119 * @param c the character (code point)
1120 * @return this object, for chaining
1121 * @stable ICU 2.0
1122 */
1123 UnicodeSet& add(UChar32 c);
1124
1125 /**
1126 * Adds the specified multicharacter to this set if it is not already
1127 * present. If this set already contains the multicharacter,
1128 * the call leaves this set unchanged.
1129 * Thus "ch" => {"ch"}
1130 * A frozen set will not be modified.
1131 *
1132 * @param s the source string
1133 * @return this object, for chaining
1134 * @stable ICU 2.4
1135 */
1136 UnicodeSet& add(const UnicodeString& s);
1137
1138 private:
1139 /**
1140 * @return a code point IF the string consists of a single one.
1141 * otherwise returns -1.
1142 * @param s string to test
1143 */
1144 static int32_t getSingleCP(const UnicodeString& s);
1145
1146 void _add(const UnicodeString& s);
1147
1148 public:
1149 /**
1150 * Adds each of the characters in this string to the set. Note: "ch" => {"c", "h"}
1151 * If this set already contains any particular character, it has no effect on that character.
1152 * A frozen set will not be modified.
1153 * @param s the source string
1154 * @return this object, for chaining
1155 * @stable ICU 2.4
1156 */
1157 UnicodeSet& addAll(const UnicodeString& s);
1158
1159 /**
1160 * Retains EACH of the characters in this string. Note: "ch" == {"c", "h"}
1161 * A frozen set will not be modified.
1162 * @param s the source string
1163 * @return this object, for chaining
1164 * @stable ICU 2.4
1165 */
1166 UnicodeSet& retainAll(const UnicodeString& s);
1167
1168 /**
1169 * Complement EACH of the characters in this string. Note: "ch" == {"c", "h"}
1170 * A frozen set will not be modified.
1171 * @param s the source string
1172 * @return this object, for chaining
1173 * @stable ICU 2.4
1174 */
1175 UnicodeSet& complementAll(const UnicodeString& s);
1176
1177 /**
1178 * Remove EACH of the characters in this string. Note: "ch" == {"c", "h"}
1179 * A frozen set will not be modified.
1180 * @param s the source string
1181 * @return this object, for chaining
1182 * @stable ICU 2.4
1183 */
1184 UnicodeSet& removeAll(const UnicodeString& s);
1185
1186 /**
1187 * Makes a set from a multicharacter string. Thus "ch" => {"ch"}
1188 *
1189 * @param s the source string
1190 * @return a newly created set containing the given string.
1191 * The caller owns the return object and is responsible for deleting it.
1192 * @stable ICU 2.4
1193 */
1194 static UnicodeSet* U_EXPORT2 createFrom(const UnicodeString& s);
1195
1196
1197 /**
1198 * Makes a set from each of the characters in the string. Thus "ch" => {"c", "h"}
1199 * @param s the source string
1200 * @return a newly created set containing the given characters
1201 * The caller owns the return object and is responsible for deleting it.
1202 * @stable ICU 2.4
1203 */
1204 static UnicodeSet* U_EXPORT2 createFromAll(const UnicodeString& s);
1205
1206 /**
1207 * Retain only the elements in this set that are contained in the
1208 * specified range. If <code>start > end</code> then an empty range is
1209 * retained, leaving the set empty. This is equivalent to
1210 * a boolean logic AND, or a set INTERSECTION.
1211 * A frozen set will not be modified.
1212 *
1213 * @param start first character, inclusive, of range
1214 * @param end last character, inclusive, of range
1215 * @stable ICU 2.0
1216 */
1217 virtual UnicodeSet& retain(UChar32 start, UChar32 end);
1218
1219
1220 /**
1221 * Retain the specified character from this set if it is present.
1222 * A frozen set will not be modified.
1223 *
1224 * @param c the character (code point)
1225 * @return this object, for chaining
1226 * @stable ICU 2.0
1227 */
1228 UnicodeSet& retain(UChar32 c);
1229
1230 /**
1231 * Retains only the specified string from this set if it is present.
1232 * Upon return this set will be empty if it did not contain s, or
1233 * will only contain s if it did contain s.
1234 * A frozen set will not be modified.
1235 *
1236 * @param s the source string
1237 * @return this object, for chaining
1238 * @stable ICU 69
1239 */
1240 UnicodeSet& retain(const UnicodeString &s);
1241
1242 /**
1243 * Removes the specified range from this set if it is present.
1244 * The set will not contain the specified range once the call
1245 * returns. If <code>start > end</code> then an empty range is
1246 * removed, leaving the set unchanged.
1247 * A frozen set will not be modified.
1248 *
1249 * @param start first character, inclusive, of range to be removed
1250 * from this set.
1251 * @param end last character, inclusive, of range to be removed
1252 * from this set.
1253 * @stable ICU 2.0
1254 */
1255 virtual UnicodeSet& remove(UChar32 start, UChar32 end);
1256
1257 /**
1258 * Removes the specified character from this set if it is present.
1259 * The set will not contain the specified range once the call
1260 * returns.
1261 * A frozen set will not be modified.
1262 *
1263 * @param c the character (code point)
1264 * @return this object, for chaining
1265 * @stable ICU 2.0
1266 */
1267 UnicodeSet& remove(UChar32 c);
1268
1269 /**
1270 * Removes the specified string from this set if it is present.
1271 * The set will not contain the specified character once the call
1272 * returns.
1273 * A frozen set will not be modified.
1274 * @param s the source string
1275 * @return this object, for chaining
1276 * @stable ICU 2.4
1277 */
1278 UnicodeSet& remove(const UnicodeString& s);
1279
1280 /**
1281 * This is equivalent to
1282 * <code>complement(MIN_VALUE, MAX_VALUE)</code>.
1283 *
1284 * <strong>Note:</strong> This performs a symmetric difference with all code points
1285 * <em>and thus retains all multicharacter strings</em>.
1286 * In order to achieve a “code point complement” (all code points minus this set),
1287 * the easiest is to <code>.complement().removeAllStrings()</code>.
1288 *
1289 * A frozen set will not be modified.
1290 * @stable ICU 2.0
1291 */
1292 virtual UnicodeSet& complement();
1293
1294 /**
1295 * Complements the specified range in this set. Any character in
1296 * the range will be removed if it is in this set, or will be
1297 * added if it is not in this set. If <code>start > end</code>
1298 * then an empty range is complemented, leaving the set unchanged.
1299 * This is equivalent to a boolean logic XOR.
1300 * A frozen set will not be modified.
1301 *
1302 * @param start first character, inclusive, of range
1303 * @param end last character, inclusive, of range
1304 * @stable ICU 2.0
1305 */
1306 virtual UnicodeSet& complement(UChar32 start, UChar32 end);
1307
1308 /**
1309 * Complements the specified character in this set. The character
1310 * will be removed if it is in this set, or will be added if it is
1311 * not in this set.
1312 * A frozen set will not be modified.
1313 *
1314 * @param c the character (code point)
1315 * @return this object, for chaining
1316 * @stable ICU 2.0
1317 */
1318 UnicodeSet& complement(UChar32 c);
1319
1320 /**
1321 * Complement the specified string in this set.
1322 * The string will be removed if it is in this set, or will be added if it is not in this set.
1323 * A frozen set will not be modified.
1324 *
1325 * @param s the string to complement
1326 * @return this object, for chaining
1327 * @stable ICU 2.4
1328 */
1329 UnicodeSet& complement(const UnicodeString& s);
1330
1331 /**
1332 * Adds all of the elements in the specified set to this set if
1333 * they're not already present. This operation effectively
1334 * modifies this set so that its value is the <i>union</i> of the two
1335 * sets. The behavior of this operation is unspecified if the specified
1336 * collection is modified while the operation is in progress.
1337 * A frozen set will not be modified.
1338 *
1339 * @param c set whose elements are to be added to this set.
1340 * @see #add(UChar32, UChar32)
1341 * @stable ICU 2.0
1342 */
1343 virtual UnicodeSet& addAll(const UnicodeSet& c);
1344
1345 /**
1346 * Retains only the elements in this set that are contained in the
1347 * specified set. In other words, removes from this set all of
1348 * its elements that are not contained in the specified set. This
1349 * operation effectively modifies this set so that its value is
1350 * the <i>intersection</i> of the two sets.
1351 * A frozen set will not be modified.
1352 *
1353 * @param c set that defines which elements this set will retain.
1354 * @stable ICU 2.0
1355 */
1356 virtual UnicodeSet& retainAll(const UnicodeSet& c);
1357
1358 /**
1359 * Removes from this set all of its elements that are contained in the
1360 * specified set. This operation effectively modifies this
1361 * set so that its value is the <i>asymmetric set difference</i> of
1362 * the two sets.
1363 * A frozen set will not be modified.
1364 *
1365 * @param c set that defines which elements will be removed from
1366 * this set.
1367 * @stable ICU 2.0
1368 */
1369 virtual UnicodeSet& removeAll(const UnicodeSet& c);
1370
1371 /**
1372 * Complements in this set all elements contained in the specified
1373 * set. Any character in the other set will be removed if it is
1374 * in this set, or will be added if it is not in this set.
1375 * A frozen set will not be modified.
1376 *
1377 * @param c set that defines which elements will be xor'ed from
1378 * this set.
1379 * @stable ICU 2.4
1380 */
1381 virtual UnicodeSet& complementAll(const UnicodeSet& c);
1382
1383 /**
1384 * Removes all of the elements from this set. This set will be
1385 * empty after this call returns.
1386 * A frozen set will not be modified.
1387 * @stable ICU 2.0
1388 */
1389 virtual UnicodeSet& clear(void);
1390
1391 /**
1392 * Close this set over the given attribute. For the attribute
1393 * USET_CASE, the result is to modify this set so that:
1394 *
1395 * 1. For each character or string 'a' in this set, all strings or
1396 * characters 'b' such that foldCase(a) == foldCase(b) are added
1397 * to this set.
1398 *
1399 * 2. For each string 'e' in the resulting set, if e !=
1400 * foldCase(e), 'e' will be removed.
1401 *
1402 * Example: [aq\\u00DF{Bc}{bC}{Fi}] => [aAqQ\\u00DF\\uFB01{ss}{bc}{fi}]
1403 *
1404 * (Here foldCase(x) refers to the operation u_strFoldCase, and a
1405 * == b denotes that the contents are the same, not pointer
1406 * comparison.)
1407 *
1408 * A frozen set will not be modified.
1409 *
1410 * @param attribute bitmask for attributes to close over.
1411 * Currently only the USET_CASE bit is supported. Any undefined bits
1412 * are ignored.
1413 * @return a reference to this set.
1414 * @stable ICU 4.2
1415 */
1416 UnicodeSet& closeOver(int32_t attribute);
1417
1418 /**
1419 * Remove all strings from this set.
1420 *
1421 * @return a reference to this set.
1422 * @stable ICU 4.2
1423 */
1424 virtual UnicodeSet &removeAllStrings();
1425
1426 /**
1427 * Iteration method that returns the number of ranges contained in
1428 * this set.
1429 * @see #getRangeStart
1430 * @see #getRangeEnd
1431 * @stable ICU 2.4
1432 */
1433 virtual int32_t getRangeCount(void) const;
1434
1435 /**
1436 * Iteration method that returns the first character in the
1437 * specified range of this set.
1438 * @see #getRangeCount
1439 * @see #getRangeEnd
1440 * @stable ICU 2.4
1441 */
1442 virtual UChar32 getRangeStart(int32_t index) const;
1443
1444 /**
1445 * Iteration method that returns the last character in the
1446 * specified range of this set.
1447 * @see #getRangeStart
1448 * @see #getRangeEnd
1449 * @stable ICU 2.4
1450 */
1451 virtual UChar32 getRangeEnd(int32_t index) const;
1452
1453 /**
1454 * Serializes this set into an array of 16-bit integers. Serialization
1455 * (currently) only records the characters in the set; multicharacter
1456 * strings are ignored.
1457 *
1458 * The array has following format (each line is one 16-bit
1459 * integer):
1460 *
1461 * length = (n+2*m) | (m!=0?0x8000:0)
1462 * bmpLength = n; present if m!=0
1463 * bmp[0]
1464 * bmp[1]
1465 * ...
1466 * bmp[n-1]
1467 * supp-high[0]
1468 * supp-low[0]
1469 * supp-high[1]
1470 * supp-low[1]
1471 * ...
1472 * supp-high[m-1]
1473 * supp-low[m-1]
1474 *
1475 * The array starts with a header. After the header are n bmp
1476 * code points, then m supplementary code points. Either n or m
1477 * or both may be zero. n+2*m is always <= 0x7FFF.
1478 *
1479 * If there are no supplementary characters (if m==0) then the
1480 * header is one 16-bit integer, 'length', with value n.
1481 *
1482 * If there are supplementary characters (if m!=0) then the header
1483 * is two 16-bit integers. The first, 'length', has value
1484 * (n+2*m)|0x8000. The second, 'bmpLength', has value n.
1485 *
1486 * After the header the code points are stored in ascending order.
1487 * Supplementary code points are stored as most significant 16
1488 * bits followed by least significant 16 bits.
1489 *
1490 * @param dest pointer to buffer of destCapacity 16-bit integers.
1491 * May be NULL only if destCapacity is zero.
1492 * @param destCapacity size of dest, or zero. Must not be negative.
1493 * @param ec error code. Will be set to U_INDEX_OUTOFBOUNDS_ERROR
1494 * if n+2*m > 0x7FFF. Will be set to U_BUFFER_OVERFLOW_ERROR if
1495 * n+2*m+(m!=0?2:1) > destCapacity.
1496 * @return the total length of the serialized format, including
1497 * the header, that is, n+2*m+(m!=0?2:1), or 0 on error other
1498 * than U_BUFFER_OVERFLOW_ERROR.
1499 * @stable ICU 2.4
1500 */
1501 int32_t serialize(uint16_t *dest, int32_t destCapacity, UErrorCode& ec) const;
1502
1503 /**
1504 * Reallocate this objects internal structures to take up the least
1505 * possible space, without changing this object's value.
1506 * A frozen set will not be modified.
1507 * @stable ICU 2.4
1508 */
1509 virtual UnicodeSet& compact();
1510
1511 /**
1512 * Return the class ID for this class. This is useful only for
1513 * comparing to a return value from getDynamicClassID(). For example:
1514 * <pre>
1515 * . Base* polymorphic_pointer = createPolymorphicObject();
1516 * . if (polymorphic_pointer->getDynamicClassID() ==
1517 * . Derived::getStaticClassID()) ...
1518 * </pre>
1519 * @return The class ID for all objects of this class.
1520 * @stable ICU 2.0
1521 */
1522 static UClassID U_EXPORT2 getStaticClassID(void);
1523
1524 /**
1525 * Implement UnicodeFunctor API.
1526 *
1527 * @return The class ID for this object. All objects of a given
1528 * class have the same class ID. Objects of other classes have
1529 * different class IDs.
1530 * @stable ICU 2.4
1531 */
1532 virtual UClassID getDynamicClassID(void) const override;
1533
1534 private:
1535
1536 // Private API for the USet API
1537
1538 friend class USetAccess;
1539
1540 const UnicodeString* getString(int32_t index) const;
1541
1542 //----------------------------------------------------------------
1543 // RuleBasedTransliterator support
1544 //----------------------------------------------------------------
1545
1546 private:
1547
1548 /**
1549 * Returns <tt>true</tt> if this set contains any character whose low byte
1550 * is the given value. This is used by <tt>RuleBasedTransliterator</tt> for
1551 * indexing.
1552 */
1553 virtual UBool matchesIndexValue(uint8_t v) const override;
1554
1555 private:
1556 friend class RBBIRuleScanner;
1557
1558 //----------------------------------------------------------------
1559 // Implementation: Clone as thawed (see ICU4J Freezable)
1560 //----------------------------------------------------------------
1561
1562 UnicodeSet(const UnicodeSet& o, UBool /* asThawed */);
1563 UnicodeSet& copyFrom(const UnicodeSet& o, UBool asThawed);
1564
1565 //----------------------------------------------------------------
1566 // Implementation: Pattern parsing
1567 //----------------------------------------------------------------
1568
1569 void applyPatternIgnoreSpace(const UnicodeString& pattern,
1570 ParsePosition& pos,
1571 const SymbolTable* symbols,
1572 UErrorCode& status);
1573
1574 void applyPattern(RuleCharacterIterator& chars,
1575 const SymbolTable* symbols,
1576 UnicodeString& rebuiltPat,
1577 uint32_t options,
1578 UnicodeSet& (UnicodeSet::*caseClosure)(int32_t attribute),
1579 int32_t depth,
1580 UErrorCode& ec);
1581
1582 //----------------------------------------------------------------
1583 // Implementation: Utility methods
1584 //----------------------------------------------------------------
1585
1586 static int32_t nextCapacity(int32_t minCapacity);
1587
1588 bool ensureCapacity(int32_t newLen);
1589
1590 bool ensureBufferCapacity(int32_t newLen);
1591
1592 void swapBuffers(void);
1593
1594 UBool allocateStrings(UErrorCode &status);
1595 int32_t stringsSize() const;
1596 UBool stringsContains(const UnicodeString &s) const;
1597
1598 UnicodeString& _toPattern(UnicodeString& result,
1599 UBool escapeUnprintable) const;
1600
1601 UnicodeString& _generatePattern(UnicodeString& result,
1602 UBool escapeUnprintable) const;
1603
1604 static void _appendToPat(UnicodeString& buf, const UnicodeString& s, UBool escapeUnprintable);
1605
1606 static void _appendToPat(UnicodeString& buf, UChar32 c, UBool escapeUnprintable);
1607
1608 static void _appendToPat(UnicodeString &result, UChar32 start, UChar32 end,
1609 UBool escapeUnprintable);
1610
1611 //----------------------------------------------------------------
1612 // Implementation: Fundamental operators
1613 //----------------------------------------------------------------
1614
1615 void exclusiveOr(const UChar32* other, int32_t otherLen, int8_t polarity);
1616
1617 void add(const UChar32* other, int32_t otherLen, int8_t polarity);
1618
1619 void retain(const UChar32* other, int32_t otherLen, int8_t polarity);
1620
1621 /**
1622 * Return true if the given position, in the given pattern, appears
1623 * to be the start of a property set pattern [:foo:], \\p{foo}, or
1624 * \\P{foo}, or \\N{name}.
1625 */
1626 static UBool resemblesPropertyPattern(const UnicodeString& pattern,
1627 int32_t pos);
1628
1629 static UBool resemblesPropertyPattern(RuleCharacterIterator& chars,
1630 int32_t iterOpts);
1631
1632 /**
1633 * Parse the given property pattern at the given parse position
1634 * and set this UnicodeSet to the result.
1635 *
1636 * The original design document is out of date, but still useful.
1637 * Ignore the property and value names:
1638 * https://htmlpreview.github.io/?https://github.com/unicode-org/icu-docs/blob/main/design/unicodeset_properties.html
1639 *
1640 * Recognized syntax:
1641 *
1642 * [:foo:] [:^foo:] - white space not allowed within "[:" or ":]"
1643 * \\p{foo} \\P{foo} - white space not allowed within "\\p" or "\\P"
1644 * \\N{name} - white space not allowed within "\\N"
1645 *
1646 * Other than the above restrictions, Unicode Pattern_White_Space characters are ignored.
1647 * Case is ignored except in "\\p" and "\\P" and "\\N". In 'name' leading
1648 * and trailing space is deleted, and internal runs of whitespace
1649 * are collapsed to a single space.
1650 *
1651 * We support binary properties, enumerated properties, and the
1652 * following non-enumerated properties:
1653 *
1654 * Numeric_Value
1655 * Name
1656 * Unicode_1_Name
1657 *
1658 * @param pattern the pattern string
1659 * @param ppos on entry, the position at which to begin parsing.
1660 * This should be one of the locations marked '^':
1661 *
1662 * [:blah:] \\p{blah} \\P{blah} \\N{name}
1663 * ^ % ^ % ^ % ^ %
1664 *
1665 * On return, the position after the last character parsed, that is,
1666 * the locations marked '%'. If the parse fails, ppos is returned
1667 * unchanged.
1668 * @param ec status
1669 * @return a reference to this.
1670 */
1671 UnicodeSet& applyPropertyPattern(const UnicodeString& pattern,
1672 ParsePosition& ppos,
1673 UErrorCode &ec);
1674
1675 void applyPropertyPattern(RuleCharacterIterator& chars,
1676 UnicodeString& rebuiltPat,
1677 UErrorCode& ec);
1678
1679 /**
1680 * A filter that returns true if the given code point should be
1681 * included in the UnicodeSet being constructed.
1682 */
1683 typedef UBool (*Filter)(UChar32 codePoint, void* context);
1684
1685 /**
1686 * Given a filter, set this UnicodeSet to the code points
1687 * contained by that filter. The filter MUST be
1688 * property-conformant. That is, if it returns value v for one
1689 * code point, then it must return v for all affiliated code
1690 * points, as defined by the inclusions list. See
1691 * getInclusions().
1692 * src is a UPropertySource value.
1693 */
1694 void applyFilter(Filter filter,
1695 void* context,
1696 const UnicodeSet* inclusions,
1697 UErrorCode &status);
1698
1699 /**
1700 * Set the new pattern to cache.
1701 */
setPattern(const UnicodeString & newPat)1702 void setPattern(const UnicodeString& newPat) {
1703 setPattern(newPat.getBuffer(), newPat.length());
1704 }
1705 void setPattern(const char16_t *newPat, int32_t newPatLen);
1706 /**
1707 * Release existing cached pattern.
1708 */
1709 void releasePattern();
1710
1711 friend class UnicodeSetIterator;
1712 };
1713
1714
1715
1716 inline bool UnicodeSet::operator!=(const UnicodeSet& o) const {
1717 return !operator==(o);
1718 }
1719
isFrozen()1720 inline UBool UnicodeSet::isFrozen() const {
1721 return (UBool)(bmpSet!=NULL || stringSpan!=NULL);
1722 }
1723
containsSome(UChar32 start,UChar32 end)1724 inline UBool UnicodeSet::containsSome(UChar32 start, UChar32 end) const {
1725 return !containsNone(start, end);
1726 }
1727
containsSome(const UnicodeSet & s)1728 inline UBool UnicodeSet::containsSome(const UnicodeSet& s) const {
1729 return !containsNone(s);
1730 }
1731
containsSome(const UnicodeString & s)1732 inline UBool UnicodeSet::containsSome(const UnicodeString& s) const {
1733 return !containsNone(s);
1734 }
1735
isBogus()1736 inline UBool UnicodeSet::isBogus() const {
1737 return (UBool)(fFlags & kIsBogus);
1738 }
1739
fromUSet(USet * uset)1740 inline UnicodeSet *UnicodeSet::fromUSet(USet *uset) {
1741 return reinterpret_cast<UnicodeSet *>(uset);
1742 }
1743
fromUSet(const USet * uset)1744 inline const UnicodeSet *UnicodeSet::fromUSet(const USet *uset) {
1745 return reinterpret_cast<const UnicodeSet *>(uset);
1746 }
1747
toUSet()1748 inline USet *UnicodeSet::toUSet() {
1749 return reinterpret_cast<USet *>(this);
1750 }
1751
toUSet()1752 inline const USet *UnicodeSet::toUSet() const {
1753 return reinterpret_cast<const USet *>(this);
1754 }
1755
span(const UnicodeString & s,int32_t start,USetSpanCondition spanCondition)1756 inline int32_t UnicodeSet::span(const UnicodeString &s, int32_t start, USetSpanCondition spanCondition) const {
1757 int32_t sLength=s.length();
1758 if(start<0) {
1759 start=0;
1760 } else if(start>sLength) {
1761 start=sLength;
1762 }
1763 return start+span(s.getBuffer()+start, sLength-start, spanCondition);
1764 }
1765
spanBack(const UnicodeString & s,int32_t limit,USetSpanCondition spanCondition)1766 inline int32_t UnicodeSet::spanBack(const UnicodeString &s, int32_t limit, USetSpanCondition spanCondition) const {
1767 int32_t sLength=s.length();
1768 if(limit<0) {
1769 limit=0;
1770 } else if(limit>sLength) {
1771 limit=sLength;
1772 }
1773 return spanBack(s.getBuffer(), limit, spanCondition);
1774 }
1775
1776 U_NAMESPACE_END
1777
1778 #endif /* U_SHOW_CPLUSPLUS_API */
1779
1780 #endif
1781