2
**********************************************************************
3
* Copyright (C) 1999-2006, International Business Machines
4
* Corporation and others. All Rights Reserved.
5
**********************************************************************
6
* Date Name Description
7
* 11/17/99 aliu Creation.
8
**********************************************************************
13
#include "unicode/utypes.h"
17
* \brief C++ API: Tranforms text from one format to another.
20
#if !UCONFIG_NO_TRANSLITERATION
22
#include "unicode/uobject.h"
23
#include "unicode/unistr.h"
24
#include "unicode/parseerr.h"
25
#include "unicode/utrans.h" // UTransPosition, UTransDirection
26
#include "unicode/strenum.h"
32
class CompoundTransliterator;
33
class TransliteratorParser;
34
class NormalizationTransliterator;
35
class TransliteratorIDParser;
39
* <code>Transliterator</code> is an abstract class that
40
* transliterates text from one format to another. The most common
41
* kind of transliterator is a script, or alphabet, transliterator.
42
* For example, a Russian to Latin transliterator changes Russian text
43
* written in Cyrillic characters to phonetically equivalent Latin
44
* characters. It does not <em>translate</em> Russian to English!
45
* Transliteration, unlike translation, operates on characters, without
46
* reference to the meanings of words and sentences.
48
* <p>Although script conversion is its most common use, a
49
* transliterator can actually perform a more general class of tasks.
50
* In fact, <code>Transliterator</code> defines a very general API
51
* which specifies only that a segment of the input text is replaced
52
* by new text. The particulars of this conversion are determined
53
* entirely by subclasses of <code>Transliterator</code>.
55
* <p><b>Transliterators are stateless</b>
57
* <p><code>Transliterator</code> objects are <em>stateless</em>; they
58
* retain no information between calls to
59
* <code>transliterate()</code>. (However, this does <em>not</em>
60
* mean that threads may share transliterators without synchronizing
61
* them. Transliterators are not immutable, so they must be
62
* synchronized when shared between threads.) This1 might seem to
63
* limit the complexity of the transliteration operation. In
64
* practice, subclasses perform complex transliterations by delaying
65
* the replacement of text until it is known that no other
66
* replacements are possible. In other words, although the
67
* <code>Transliterator</code> objects are stateless, the source text
68
* itself embodies all the needed information, and delayed operation
69
* allows arbitrary complexity.
71
* <p><b>Batch transliteration</b>
73
* <p>The simplest way to perform transliteration is all at once, on a
74
* string of existing text. This is referred to as <em>batch</em>
75
* transliteration. For example, given a string <code>input</code>
76
* and a transliterator <code>t</code>, the call
78
* \htmlonly<blockquote>\endhtmlonly<code>String result = t.transliterate(input);
79
* </code>\htmlonly</blockquote>\endhtmlonly
81
* will transliterate it and return the result. Other methods allow
82
* the client to specify a substring to be transliterated and to use
83
* {@link Replaceable } objects instead of strings, in order to
84
* preserve out-of-band information (such as text styles).
86
* <p><b>Keyboard transliteration</b>
88
* <p>Somewhat more involved is <em>keyboard</em>, or incremental
89
* transliteration. This is the transliteration of text that is
90
* arriving from some source (typically the user's keyboard) one
91
* character at a time, or in some other piecemeal fashion.
93
* <p>In keyboard transliteration, a <code>Replaceable</code> buffer
94
* stores the text. As text is inserted, as much as possible is
95
* transliterated on the fly. This means a GUI that displays the
96
* contents of the buffer may show text being modified as each new
99
* <p>Consider the simple <code>RuleBasedTransliterator</code>:
101
* \htmlonly<blockquote>\endhtmlonly<code>
104
* </code>\htmlonly</blockquote>\endhtmlonly
106
* When the user types 't', nothing will happen, since the
107
* transliterator is waiting to see if the next character is 'h'. To
108
* remedy this, we introduce the notion of a cursor, marked by a '|'
109
* in the output string:
111
* \htmlonly<blockquote>\endhtmlonly<code>
114
* </code>\htmlonly</blockquote>\endhtmlonly
116
* Now when the user types 't', tau appears, and if the next character
117
* is 'h', the tau changes to a theta. This is accomplished by
118
* maintaining a cursor position (independent of the insertion point,
119
* and invisible in the GUI) across calls to
120
* <code>transliterate()</code>. Typically, the cursor will
121
* be coincident with the insertion point, but in a case like the one
122
* above, it will precede the insertion point.
124
* <p>Keyboard transliteration methods maintain a set of three indices
125
* that are updated with each call to
126
* <code>transliterate()</code>, including the cursor, start,
127
* and limit. Since these indices are changed by the method, they are
128
* passed in an <code>int[]</code> array. The <code>START</code> index
129
* marks the beginning of the substring that the transliterator will
130
* look at. It is advanced as text becomes committed (but it is not
131
* the committed index; that's the <code>CURSOR</code>). The
132
* <code>CURSOR</code> index, described above, marks the point at
133
* which the transliterator last stopped, either because it reached
134
* the end, or because it required more characters to disambiguate
135
* between possible inputs. The <code>CURSOR</code> can also be
136
* explicitly set by rules in a <code>RuleBasedTransliterator</code>.
137
* Any characters before the <code>CURSOR</code> index are frozen;
138
* future keyboard transliteration calls within this input sequence
139
* will not change them. New text is inserted at the
140
* <code>LIMIT</code> index, which marks the end of the substring that
141
* the transliterator looks at.
143
* <p>Because keyboard transliteration assumes that more characters
144
* are to arrive, it is conservative in its operation. It only
145
* transliterates when it can do so unambiguously. Otherwise it waits
146
* for more characters to arrive. When the client code knows that no
147
* more characters are forthcoming, perhaps because the user has
148
* performed some input termination operation, then it should call
149
* <code>finishTransliteration()</code> to complete any
150
* pending transliterations.
154
* <p>Pairs of transliterators may be inverses of one another. For
155
* example, if transliterator <b>A</b> transliterates characters by
156
* incrementing their Unicode value (so "abc" -> "def"), and
157
* transliterator <b>B</b> decrements character values, then <b>A</b>
158
* is an inverse of <b>B</b> and vice versa. If we compose <b>A</b>
159
* with <b>B</b> in a compound transliterator, the result is the
160
* indentity transliterator, that is, a transliterator that does not
161
* change its input text.
163
* The <code>Transliterator</code> method <code>getInverse()</code>
164
* returns a transliterator's inverse, if one exists, or
165
* <code>null</code> otherwise. However, the result of
166
* <code>getInverse()</code> usually will <em>not</em> be a true
167
* mathematical inverse. This is because true inverse transliterators
168
* are difficult to formulate. For example, consider two
169
* transliterators: <b>AB</b>, which transliterates the character 'A'
170
* to 'B', and <b>BA</b>, which transliterates 'B' to 'A'. It might
171
* seem that these are exact inverses, since
173
* \htmlonly<blockquote>\endhtmlonly"A" x <b>AB</b> -> "B"<br>
174
* "B" x <b>BA</b> -> "A"\htmlonly</blockquote>\endhtmlonly
176
* where 'x' represents transliteration. However,
178
* \htmlonly<blockquote>\endhtmlonly"ABCD" x <b>AB</b> -> "BBCD"<br>
179
* "BBCD" x <b>BA</b> -> "AACD"\htmlonly</blockquote>\endhtmlonly
181
* so <b>AB</b> composed with <b>BA</b> is not the
182
* identity. Nonetheless, <b>BA</b> may be usefully considered to be
183
* <b>AB</b>'s inverse, and it is on this basis that
184
* <b>AB</b><code>.getInverse()</code> could legitimately return
187
* <p><b>IDs and display names</b>
189
* <p>A transliterator is designated by a short identifier string or
190
* <em>ID</em>. IDs follow the format <em>source-destination</em>,
191
* where <em>source</em> describes the entity being replaced, and
192
* <em>destination</em> describes the entity replacing
193
* <em>source</em>. The entities may be the names of scripts,
194
* particular sequences of characters, or whatever else it is that the
195
* transliterator converts to or from. For example, a transliterator
196
* from Russian to Latin might be named "Russian-Latin". A
197
* transliterator from keyboard escape sequences to Latin-1 characters
198
* might be named "KeyboardEscape-Latin1". By convention, system
199
* entity names are in English, with the initial letters of words
200
* capitalized; user entity names may follow any format so long as
201
* they do not contain dashes.
203
* <p>In addition to programmatic IDs, transliterator objects have
204
* display names for presentation in user interfaces, returned by
205
* {@link #getDisplayName }.
207
* <p><b>Factory methods and registration</b>
209
* <p>In general, client code should use the factory method
210
* {@link #createInstance } to obtain an instance of a
211
* transliterator given its ID. Valid IDs may be enumerated using
212
* <code>getAvailableIDs()</code>. Since transliterators are mutable,
213
* multiple calls to {@link #createInstance } with the same ID will
214
* return distinct objects.
216
* <p>In addition to the system transliterators registered at startup,
217
* user transliterators may be registered by calling
218
* <code>registerInstance()</code> at run time. A registered instance
219
* acts a template; future calls to {@link #createInstance } with the ID
220
* of the registered object return clones of that object. Thus any
221
* object passed to <tt>registerInstance()</tt> must implement
222
* <tt>clone()</tt> propertly. To register a transliterator subclass
223
* without instantiating it (until it is needed), users may call
224
* {@link #registerFactory }. In this case, the objects are
225
* instantiated by invoking the zero-argument public constructor of
228
* <p><b>Subclassing</b>
230
* Subclasses must implement the abstract method
231
* <code>handleTransliterate()</code>. <p>Subclasses should override
232
* the <code>transliterate()</code> method taking a
233
* <code>Replaceable</code> and the <code>transliterate()</code>
234
* method taking a <code>String</code> and <code>StringBuffer</code>
235
* if the performance of these methods can be improved over the
236
* performance obtained by the default implementations in this class.
241
class U_I18N_API Transliterator : public UObject {
246
* Programmatic name, e.g., "Latin-Arabic".
251
* This transliterator's filter. Any character for which
252
* <tt>filter.contains()</tt> returns <tt>false</tt> will not be
253
* altered by this transliterator. If <tt>filter</tt> is
254
* <tt>null</tt> then no filtering is applied.
256
UnicodeFilter* filter;
258
int32_t maximumContextLength;
263
* A context integer or pointer for a factory function, passed by
269
* This token, interpreted as a 32-bit integer.
274
* This token, interpreted as a native pointer.
281
* Return a token containing an integer.
282
* @return a token containing an integer.
285
inline static Token integerToken(int32_t);
288
* Return a token containing a pointer.
289
* @return a token containing a pointer.
292
inline static Token pointerToken(void*);
295
* A function that creates and returns a Transliterator. When
296
* invoked, it will be passed the ID string that is being
297
* instantiated, together with the context pointer that was passed
298
* in when the factory function was first registered. Many
299
* factory functions will ignore both parameters, however,
300
* functions that are registered to more than one ID may use the
301
* ID or the context parameter to parameterize the transliterator
303
* @param ID the string identifier for this transliterator
304
* @param context a context pointer that will be stored and
305
* later passed to the factory function when an ID matching
306
* the registration ID is being instantiated with this factory.
309
typedef Transliterator* (U_EXPORT2 *Factory)(const UnicodeString& ID, Token context);
314
* Default constructor.
315
* @param ID the string identifier for this transliterator
316
* @param adoptedFilter the filter. Any character for which
317
* <tt>filter.contains()</tt> returns <tt>false</tt> will not be
318
* altered by this transliterator. If <tt>filter</tt> is
319
* <tt>null</tt> then no filtering is applied.
322
Transliterator(const UnicodeString& ID, UnicodeFilter* adoptedFilter);
328
Transliterator(const Transliterator&);
331
* Assignment operator.
334
Transliterator& operator=(const Transliterator&);
337
* Create a transliterator from a basic ID. This is an ID
338
* containing only the forward direction source, target, and
340
* @param id a basic ID of the form S-T or S-T/V.
341
* @param canon canonical ID to assign to the object, or
342
* NULL to leave the ID unchanged
343
* @return a newly created Transliterator or null if the ID is
347
static Transliterator* createBasicInstance(const UnicodeString& id,
348
const UnicodeString* canon);
350
friend class TransliteratorParser; // for parseID()
351
friend class TransliteratorIDParser; // for createBasicInstance()
352
friend class TransliteratorAlias; // for setID()
360
virtual ~Transliterator();
363
* Implements Cloneable.
364
* All subclasses are encouraged to implement this method if it is
365
* possible and reasonable to do so. Subclasses that are to be
366
* registered with the system using <tt>registerInstance()</tt>
367
* are required to implement this method. If a subclass does not
368
* implement clone() properly and is registered with the system
369
* using registerInstance(), then the default clone() implementation
370
* will return null, and calls to createInstance() will fail.
372
* @return a copy of the object.
373
* @see #registerInstance
376
virtual Transliterator* clone() const;
379
* Transliterates a segment of a string, with optional filtering.
381
* @param text the string to be transliterated
382
* @param start the beginning index, inclusive; <code>0 <= start
384
* @param limit the ending index, exclusive; <code>start <= limit
385
* <= text.length()</code>.
386
* @return The new limit index. The text previously occupying <code>[start,
387
* limit)</code> has been transliterated, possibly to a string of a different
388
* length, at <code>[start, </code><em>new-limit</em><code>)</code>, where
389
* <em>new-limit</em> is the return value. If the input offsets are out of bounds,
390
* the returned value is -1 and the input string remains unchanged.
393
virtual int32_t transliterate(Replaceable& text,
394
int32_t start, int32_t limit) const;
397
* Transliterates an entire string in place. Convenience method.
398
* @param text the string to be transliterated
401
virtual void transliterate(Replaceable& text) const;
404
* Transliterates the portion of the text buffer that can be
405
* transliterated unambiguosly after new text has been inserted,
406
* typically as a result of a keyboard event. The new text in
407
* <code>insertion</code> will be inserted into <code>text</code>
408
* at <code>index.limit</code>, advancing
409
* <code>index.limit</code> by <code>insertion.length()</code>.
410
* Then the transliterator will try to transliterate characters of
411
* <code>text</code> between <code>index.cursor</code> and
412
* <code>index.limit</code>. Characters before
413
* <code>index.cursor</code> will not be changed.
415
* <p>Upon return, values in <code>index</code> will be updated.
416
* <code>index.start</code> will be advanced to the first
417
* character that future calls to this method will read.
418
* <code>index.cursor</code> and <code>index.limit</code> will
419
* be adjusted to delimit the range of text that future calls to
420
* this method may change.
422
* <p>Typical usage of this method begins with an initial call
423
* with <code>index.start</code> and <code>index.limit</code>
424
* set to indicate the portion of <code>text</code> to be
425
* transliterated, and <code>index.cursor == index.start</code>.
426
* Thereafter, <code>index</code> can be used without
427
* modification in future calls, provided that all changes to
428
* <code>text</code> are made via this method.
430
* <p>This method assumes that future calls may be made that will
431
* insert new text into the buffer. As a result, it only performs
432
* unambiguous transliterations. After the last call to this
433
* method, there may be untransliterated text that is waiting for
434
* more input to resolve an ambiguity. In order to perform these
435
* pending transliterations, clients should call {@link
436
* #finishTransliteration } after the last call to this
437
* method has been made.
439
* @param text the buffer holding transliterated and untransliterated text
440
* @param index an array of three integers.
442
* <ul><li><code>index.start</code>: the beginning index,
443
* inclusive; <code>0 <= index.start <= index.limit</code>.
445
* <li><code>index.limit</code>: the ending index, exclusive;
446
* <code>index.start <= index.limit <= text.length()</code>.
447
* <code>insertion</code> is inserted at
448
* <code>index.limit</code>.
450
* <li><code>index.cursor</code>: the next character to be
451
* considered for transliteration; <code>index.start <=
452
* index.cursor <= index.limit</code>. Characters before
453
* <code>index.cursor</code> will not be changed by future calls
454
* to this method.</ul>
456
* @param insertion text to be inserted and possibly
457
* transliterated into the translation buffer at
458
* <code>index.limit</code>. If <code>null</code> then no text
460
* @param status Output param to filled in with a success or an error.
461
* @see #handleTransliterate
462
* @exception IllegalArgumentException if <code>index</code>
464
* @see UTransPosition
467
virtual void transliterate(Replaceable& text, UTransPosition& index,
468
const UnicodeString& insertion,
469
UErrorCode& status) const;
472
* Transliterates the portion of the text buffer that can be
473
* transliterated unambiguosly after a new character has been
474
* inserted, typically as a result of a keyboard event. This is a
475
* convenience method; see {@link
476
* #transliterate(Replaceable&, UTransPosition&, const UnicodeString&, UErrorCode&) const}
478
* @param text the buffer holding transliterated and
479
* untransliterated text
480
* @param index an array of three integers. See {@link
481
* #transliterate(Replaceable&, UTransPosition&, const UnicodeString&, UErrorCode&) const }.
482
* @param insertion text to be inserted and possibly
483
* transliterated into the translation buffer at
484
* <code>index.limit</code>.
485
* @param status Output param to filled in with a success or an error.
486
* @see #transliterate(Replaceable&, UTransPosition&, const UnicodeString&, UErrorCode&) const
489
virtual void transliterate(Replaceable& text, UTransPosition& index,
491
UErrorCode& status) const;
494
* Transliterates the portion of the text buffer that can be
495
* transliterated unambiguosly. This is a convenience method; see
497
* #transliterate(Replaceable&, UTransPosition&, const UnicodeString&, UErrorCode&) const }
499
* @param text the buffer holding transliterated and
500
* untransliterated text
501
* @param index an array of three integers. See {@link
502
* #transliterate(Replaceable&, UTransPosition&, const UnicodeString&, UErrorCode&) const }.
503
* @param status Output param to filled in with a success or an error.
504
* @see #transliterate(Replaceable, int[], String)
507
virtual void transliterate(Replaceable& text, UTransPosition& index,
508
UErrorCode& status) const;
511
* Finishes any pending transliterations that were waiting for
512
* more characters. Clients should call this method as the last
513
* call after a sequence of one or more calls to
514
* <code>transliterate()</code>.
515
* @param text the buffer holding transliterated and
516
* untransliterated text.
517
* @param index the array of indices previously passed to {@link
521
virtual void finishTransliteration(Replaceable& text,
522
UTransPosition& index) const;
527
* This internal method does incremental transliteration. If the
528
* 'insertion' is non-null then we append it to 'text' before
529
* proceeding. This method calls through to the pure virtual
530
* framework method handleTransliterate() to do the actual
532
* @param text the buffer holding transliterated and
533
* untransliterated text
534
* @param index an array of three integers. See {@link
535
* #transliterate(Replaceable, int[], String)}.
536
* @param insertion text to be inserted and possibly
537
* transliterated into the translation buffer at
538
* <code>index.limit</code>.
539
* @param status Output param to filled in with a success or an error.
541
void _transliterate(Replaceable& text,
542
UTransPosition& index,
543
const UnicodeString* insertion,
544
UErrorCode &status) const;
549
* Abstract method that concrete subclasses define to implement
550
* their transliteration algorithm. This method handles both
551
* incremental and non-incremental transliteration. Let
552
* <code>originalStart</code> refer to the value of
553
* <code>pos.start</code> upon entry.
556
* <li>If <code>incremental</code> is false, then this method
557
* should transliterate all characters between
558
* <code>pos.start</code> and <code>pos.limit</code>. Upon return
559
* <code>pos.start</code> must == <code> pos.limit</code>.</li>
561
* <li>If <code>incremental</code> is true, then this method
562
* should transliterate all characters between
563
* <code>pos.start</code> and <code>pos.limit</code> that can be
564
* unambiguously transliterated, regardless of future insertions
565
* of text at <code>pos.limit</code>. Upon return,
566
* <code>pos.start</code> should be in the range
567
* [<code>originalStart</code>, <code>pos.limit</code>).
568
* <code>pos.start</code> should be positioned such that
569
* characters [<code>originalStart</code>, <code>
570
* pos.start</code>) will not be changed in the future by this
571
* transliterator and characters [<code>pos.start</code>,
572
* <code>pos.limit</code>) are unchanged.</li>
575
* <p>Implementations of this method should also obey the
576
* following invariants:</p>
579
* <li> <code>pos.limit</code> and <code>pos.contextLimit</code>
580
* should be updated to reflect changes in length of the text
581
* between <code>pos.start</code> and <code>pos.limit</code>. The
582
* difference <code> pos.contextLimit - pos.limit</code> should
585
* <li><code>pos.contextStart</code> should not change.</li>
587
* <li>Upon return, neither <code>pos.start</code> nor
588
* <code>pos.limit</code> should be less than
589
* <code>originalStart</code>.</li>
591
* <li>Text before <code>originalStart</code> and text after
592
* <code>pos.limit</code> should not change.</li>
594
* <li>Text before <code>pos.contextStart</code> and text after
595
* <code> pos.contextLimit</code> should be ignored.</li>
598
* <p>Subclasses may safely assume that all characters in
599
* [<code>pos.start</code>, <code>pos.limit</code>) are filtered.
600
* In other words, the filter has already been applied by the time
601
* this method is called. See
602
* <code>filteredTransliterate()</code>.
604
* <p>This method is <b>not</b> for public consumption. Calling
605
* this method directly will transliterate
606
* [<code>pos.start</code>, <code>pos.limit</code>) without
607
* applying the filter. End user code should call <code>
608
* transliterate()</code> instead of this method. Subclass code
609
* and wrapping transliterators should call
610
* <code>filteredTransliterate()</code> instead of this method.<p>
612
* @param text the buffer holding transliterated and
613
* untransliterated text
615
* @param pos the indices indicating the start, limit, context
616
* start, and context limit of the text.
618
* @param incremental if true, assume more text may be inserted at
619
* <code>pos.limit</code> and act accordingly. Otherwise,
620
* transliterate all text between <code>pos.start</code> and
621
* <code>pos.limit</code> and move <code>pos.start</code> up to
622
* <code>pos.limit</code>.
624
* @see #transliterate
627
virtual void handleTransliterate(Replaceable& text,
629
UBool incremental) const = 0;
633
* Transliterate a substring of text, as specified by index, taking filters
634
* into account. This method is for subclasses that need to delegate to
635
* another transliterator, such as CompoundTransliterator.
636
* @param text the text to be transliterated
637
* @param index the position indices
638
* @param incremental if TRUE, then assume more characters may be inserted
639
* at index.limit, and postpone processing to accomodate future incoming
643
virtual void filteredTransliterate(Replaceable& text,
644
UTransPosition& index,
645
UBool incremental) const;
650
* Top-level transliteration method, handling filtering, incremental and
651
* non-incremental transliteration, and rollback. All transliteration
652
* public API methods eventually call this method with a rollback argument
653
* of TRUE. Other entities may call this method but rollback should be
656
* <p>If this transliterator has a filter, break up the input text into runs
657
* of unfiltered characters. Pass each run to
658
* <subclass>.handleTransliterate().
660
* <p>In incremental mode, if rollback is TRUE, perform a special
661
* incremental procedure in which several passes are made over the input
662
* text, adding one character at a time, and committing successful
663
* transliterations as they occur. Unsuccessful transliterations are rolled
664
* back and retried with additional characters to give correct results.
666
* @param text the text to be transliterated
667
* @param index the position indices
668
* @param incremental if TRUE, then assume more characters may be inserted
669
* at index.limit, and postpone processing to accomodate future incoming
671
* @param rollback if TRUE and if incremental is TRUE, then perform special
672
* incremental processing, as described above, and undo partial
673
* transliterations where necessary. If incremental is FALSE then this
674
* parameter is ignored.
676
virtual void filteredTransliterate(Replaceable& text,
677
UTransPosition& index,
679
UBool rollback) const;
684
* Returns the length of the longest context required by this transliterator.
685
* This is <em>preceding</em> context. The default implementation supplied
686
* by <code>Transliterator</code> returns zero; subclasses
687
* that use preceding context should override this method to return the
688
* correct value. For example, if a transliterator translates "ddd" (where
689
* d is any digit) to "555" when preceded by "(ddd)", then the preceding
690
* context length is 5, the length of "(ddd)".
692
* @return The maximum number of preceding context characters this
693
* transliterator needs to examine
696
int32_t getMaximumContextLength(void) const;
701
* Method for subclasses to use to set the maximum context length.
702
* @param maxContextLength the new value to be set.
703
* @see #getMaximumContextLength
706
void setMaximumContextLength(int32_t maxContextLength);
711
* Returns a programmatic identifier for this transliterator.
712
* If this identifier is passed to <code>createInstance()</code>, it
713
* will return this object, if it has been registered.
714
* @return a programmatic identifier for this transliterator.
715
* @see #registerInstance
716
* @see #registerFactory
717
* @see #getAvailableIDs
720
virtual const UnicodeString& getID(void) const;
723
* Returns a name for this transliterator that is appropriate for
724
* display to the user in the default locale. See {@link
725
* #getDisplayName } for details.
726
* @param ID the string identifier for this transliterator
727
* @param result Output param to receive the display name
728
* @return A reference to 'result'.
731
static UnicodeString& U_EXPORT2 getDisplayName(const UnicodeString& ID,
732
UnicodeString& result);
735
* Returns a name for this transliterator that is appropriate for
736
* display to the user in the given locale. This name is taken
737
* from the locale resource data in the standard manner of the
738
* <code>java.text</code> package.
740
* <p>If no localized names exist in the system resource bundles,
741
* a name is synthesized using a localized
742
* <code>MessageFormat</code> pattern from the resource data. The
743
* arguments to this pattern are an integer followed by one or two
744
* strings. The integer is the number of strings, either 1 or 2.
745
* The strings are formed by splitting the ID for this
746
* transliterator at the first '-'. If there is no '-', then the
747
* entire ID forms the only string.
748
* @param ID the string identifier for this transliterator
749
* @param inLocale the Locale in which the display name should be
751
* @param result Output param to receive the display name
752
* @return A reference to 'result'.
755
static UnicodeString& U_EXPORT2 getDisplayName(const UnicodeString& ID,
756
const Locale& inLocale,
757
UnicodeString& result);
760
* Returns the filter used by this transliterator, or <tt>NULL</tt>
761
* if this transliterator uses no filter.
762
* @return the filter used by this transliterator, or <tt>NULL</tt>
763
* if this transliterator uses no filter.
766
const UnicodeFilter* getFilter(void) const;
769
* Returns the filter used by this transliterator, or <tt>NULL</tt> if this
770
* transliterator uses no filter. The caller must eventually delete the
771
* result. After this call, this transliterator's filter is set to
773
* @return the filter used by this transliterator, or <tt>NULL</tt> if this
774
* transliterator uses no filter.
777
UnicodeFilter* orphanFilter(void);
780
* Changes the filter used by this transliterator. If the filter
781
* is set to <tt>null</tt> then no filtering will occur.
783
* <p>Callers must take care if a transliterator is in use by
784
* multiple threads. The filter should not be changed by one
785
* thread while another thread may be transliterating.
786
* @param adoptedFilter the new filter to be adopted.
789
void adoptFilter(UnicodeFilter* adoptedFilter);
792
* Returns this transliterator's inverse. See the class
793
* documentation for details. This implementation simply inverts
794
* the two entities in the ID and attempts to retrieve the
795
* resulting transliterator. That is, if <code>getID()</code>
796
* returns "A-B", then this method will return the result of
797
* <code>createInstance("B-A")</code>, or <code>null</code> if that
800
* <p>Subclasses with knowledge of their inverse may wish to
801
* override this method.
803
* @param status Output param to filled in with a success or an error.
804
* @return a transliterator that is an inverse, not necessarily
805
* exact, of this transliterator, or <code>null</code> if no such
806
* transliterator is registered.
807
* @see #registerInstance
810
Transliterator* createInverse(UErrorCode& status) const;
813
* Returns a <code>Transliterator</code> object given its ID.
814
* The ID must be either a system transliterator ID or a ID registered
815
* using <code>registerInstance()</code>.
817
* @param ID a valid ID, as enumerated by <code>getAvailableIDs()</code>
818
* @param dir either FORWARD or REVERSE.
819
* @param parseError Struct to recieve information on position
820
* of error if an error is encountered
821
* @param status Output param to filled in with a success or an error.
822
* @return A <code>Transliterator</code> object with the given ID
823
* @see #registerInstance
824
* @see #getAvailableIDs
828
static Transliterator* U_EXPORT2 createInstance(const UnicodeString& ID,
830
UParseError& parseError,
834
* Returns a <code>Transliterator</code> object given its ID.
835
* The ID must be either a system transliterator ID or a ID registered
836
* using <code>registerInstance()</code>.
837
* @param ID a valid ID, as enumerated by <code>getAvailableIDs()</code>
838
* @param dir either FORWARD or REVERSE.
839
* @param status Output param to filled in with a success or an error.
840
* @return A <code>Transliterator</code> object with the given ID
843
static Transliterator* U_EXPORT2 createInstance(const UnicodeString& ID,
848
* Returns a <code>Transliterator</code> object constructed from
849
* the given rule string. This will be a RuleBasedTransliterator,
850
* if the rule string contains only rules, or a
851
* CompoundTransliterator, if it contains ID blocks, or a
852
* NullTransliterator, if it contains ID blocks which parse as
853
* empty for the given direction.
854
* @param ID the id for the transliterator.
855
* @param rules rules, separated by ';'
856
* @param dir either FORWARD or REVERSE.
857
* @param parseError Struct to recieve information on position
858
* of error if an error is encountered
859
* @param status Output param set to success/failure code.
862
static Transliterator* U_EXPORT2 createFromRules(const UnicodeString& ID,
863
const UnicodeString& rules,
865
UParseError& parseError,
869
* Create a rule string that can be passed to createFromRules()
870
* to recreate this transliterator.
871
* @param result the string to receive the rules. Previous
872
* contents will be deleted.
873
* @param escapeUnprintable if TRUE then convert unprintable
874
* character to their hex escape representations, \\uxxxx or
875
* \\Uxxxxxxxx. Unprintable characters are those other than
876
* U+000A, U+0020..U+007E.
879
virtual UnicodeString& toRules(UnicodeString& result,
880
UBool escapeUnprintable) const;
883
* Return the number of elements that make up this transliterator.
884
* For example, if the transliterator "NFD;Jamo-Latin;Latin-Greek"
885
* were created, the return value of this method would be 3.
887
* <p>If this transliterator is not composed of other
888
* transliterators, then this method returns 1.
889
* @return the number of transliterators that compose this
890
* transliterator, or 1 if this transliterator is not composed of
891
* multiple transliterators
894
int32_t countElements() const;
897
* Return an element that makes up this transliterator. For
898
* example, if the transliterator "NFD;Jamo-Latin;Latin-Greek"
899
* were created, the return value of this method would be one
900
* of the three transliterator objects that make up that
901
* transliterator: [NFD, Jamo-Latin, Latin-Greek].
903
* <p>If this transliterator is not composed of other
904
* transliterators, then this method will return a reference to
905
* this transliterator when given the index 0.
906
* @param index a value from 0..countElements()-1 indicating the
907
* transliterator to return
908
* @param ec input-output error code
909
* @return one of the transliterators that makes up this
910
* transliterator, if this transliterator is made up of multiple
911
* transliterators, otherwise a reference to this object if given
915
const Transliterator& getElement(int32_t index, UErrorCode& ec) const;
918
* Returns the set of all characters that may be modified in the
919
* input text by this Transliterator. This incorporates this
920
* object's current filter; if the filter is changed, the return
921
* value of this function will change. The default implementation
922
* returns an empty set. Some subclasses may override {@link
923
* #handleGetSourceSet } to return a more precise result. The
924
* return result is approximate in any case and is intended for
925
* use by tests, tools, or utilities.
926
* @param result receives result set; previous contents lost
927
* @return a reference to result
929
* @see #handleGetSourceSet
932
UnicodeSet& getSourceSet(UnicodeSet& result) const;
935
* Framework method that returns the set of all characters that
936
* may be modified in the input text by this Transliterator,
937
* ignoring the effect of this object's filter. The base class
938
* implementation returns the empty set. Subclasses that wish to
939
* implement this should override this method.
940
* @return the set of characters that this transliterator may
941
* modify. The set may be modified, so subclasses should return a
942
* newly-created object.
943
* @param result receives result set; previous contents lost
948
virtual void handleGetSourceSet(UnicodeSet& result) const;
951
* Returns the set of all characters that may be generated as
952
* replacement text by this transliterator. The default
953
* implementation returns the empty set. Some subclasses may
954
* override this method to return a more precise result. The
955
* return result is approximate in any case and is intended for
956
* use by tests, tools, or utilities requiring such
958
* @param result receives result set; previous contents lost
959
* @return a reference to result
963
virtual UnicodeSet& getTargetSet(UnicodeSet& result) const;
968
* Registers a factory function that creates transliterators of
970
* @param id the ID being registered
971
* @param factory a function pointer that will be copied and
972
* called later when the given ID is passed to createInstance()
973
* @param context a context pointer that will be stored and
974
* later passed to the factory function when an ID matching
975
* the registration ID is being instantiated with this factory.
978
static void U_EXPORT2 registerFactory(const UnicodeString& id,
983
* Registers an instance <tt>obj</tt> of a subclass of
984
* <code>Transliterator</code> with the system. When
985
* <tt>createInstance()</tt> is called with an ID string that is
986
* equal to <tt>obj->getID()</tt>, then <tt>obj->clone()</tt> is
989
* After this call the Transliterator class owns the adoptedObj
990
* and will delete it.
992
* @param adoptedObj an instance of subclass of
993
* <code>Transliterator</code> that defines <tt>clone()</tt>
994
* @see #createInstance
995
* @see #registerFactory
999
static void U_EXPORT2 registerInstance(Transliterator* adoptedObj);
1002
* Registers an ID string as an alias of another ID string.
1003
* That is, after calling this function, <tt>createInstance(aliasID)</tt>
1004
* will return the same thing as <tt>createInstance(realID)</tt>.
1005
* This is generally used to create shorter, more mnemonic aliases
1006
* for long compound IDs.
1008
* @param aliasID The new ID being registered.
1009
* @param realID The ID that the new ID is to be an alias for.
1010
* This can be a compound ID and can include filters and should
1011
* refer to transliterators that have already been registered with
1012
* the framework, although this isn't checked.
1015
static void U_EXPORT2 registerAlias(const UnicodeString& aliasID,
1016
const UnicodeString& realID);
1022
* @param id the ID being registered
1023
* @param factory a function pointer that will be copied and
1024
* called later when the given ID is passed to createInstance()
1025
* @param context a context pointer that will be stored and
1026
* later passed to the factory function when an ID matching
1027
* the registration ID is being instantiated with this factory.
1029
static void _registerFactory(const UnicodeString& id,
1036
static void _registerInstance(Transliterator* adoptedObj);
1041
static void _registerAlias(const UnicodeString& aliasID, const UnicodeString& realID);
1044
* Register two targets as being inverses of one another. For
1045
* example, calling registerSpecialInverse("NFC", "NFD", true) causes
1046
* Transliterator to form the following inverse relationships:
1049
* Any-NFC => Any-NFD
1051
* Any-NFD => Any-NFC</pre>
1053
* (Without the special inverse registration, the inverse of NFC
1054
* would be NFC-Any.) Note that NFD is shorthand for Any-NFD, but
1055
* that the presence or absence of "Any-" is preserved.
1057
* <p>The relationship is symmetrical; registering (a, b) is
1058
* equivalent to registering (b, a).
1060
* <p>The relevant IDs must still be registered separately as
1061
* factories or classes.
1063
* <p>Only the targets are specified. Special inverses always
1064
* have the form Any-Target1 <=> Any-Target2. The target should
1065
* have canonical casing (the casing desired to be produced when
1066
* an inverse is formed) and should contain no whitespace or other
1067
* extraneous characters.
1069
* @param target the target against which to register the inverse
1070
* @param inverseTarget the inverse of target, that is
1071
* Any-target.getInverse() => Any-inverseTarget
1072
* @param bidirectional if true, register the reverse relation
1073
* as well, that is, Any-inverseTarget.getInverse() => Any-target
1076
static void _registerSpecialInverse(const UnicodeString& target,
1077
const UnicodeString& inverseTarget,
1078
UBool bidirectional);
1083
* Unregisters a transliterator or class. This may be either
1084
* a system transliterator or a user transliterator or class.
1085
* Any attempt to construct an unregistered transliterator based
1086
* on its ID will fail.
1088
* @param ID the ID of the transliterator or class
1089
* @return the <code>Object</code> that was registered with
1090
* <code>ID</code>, or <code>null</code> if none was
1091
* @see #registerInstance
1092
* @see #registerFactory
1095
static void U_EXPORT2 unregister(const UnicodeString& ID);
1100
* Return a StringEnumeration over the IDs available at the time of the
1101
* call, including user-registered IDs.
1102
* @param ec input-output error code
1103
* @return a newly-created StringEnumeration over the transliterators
1104
* available at the time of the call. The caller should delete this object
1105
* when done using it.
1108
static StringEnumeration* U_EXPORT2 getAvailableIDs(UErrorCode& ec);
1111
* Return the number of registered source specifiers.
1112
* @return the number of registered source specifiers.
1115
static int32_t U_EXPORT2 countAvailableSources(void);
1118
* Return a registered source specifier.
1119
* @param index which specifier to return, from 0 to n-1, where
1120
* n = countAvailableSources()
1121
* @param result fill-in paramter to receive the source specifier.
1122
* If index is out of range, result will be empty.
1123
* @return reference to result
1126
static UnicodeString& U_EXPORT2 getAvailableSource(int32_t index,
1127
UnicodeString& result);
1130
* Return the number of registered target specifiers for a given
1132
* @param source the given source specifier.
1133
* @return the number of registered target specifiers for a given
1137
static int32_t U_EXPORT2 countAvailableTargets(const UnicodeString& source);
1140
* Return a registered target specifier for a given source.
1141
* @param index which specifier to return, from 0 to n-1, where
1142
* n = countAvailableTargets(source)
1143
* @param source the source specifier
1144
* @param result fill-in paramter to receive the target specifier.
1145
* If source is invalid or if index is out of range, result will
1147
* @return reference to result
1150
static UnicodeString& U_EXPORT2 getAvailableTarget(int32_t index,
1151
const UnicodeString& source,
1152
UnicodeString& result);
1155
* Return the number of registered variant specifiers for a given
1156
* source-target pair.
1157
* @param source the source specifiers.
1158
* @param target the target specifiers.
1161
static int32_t U_EXPORT2 countAvailableVariants(const UnicodeString& source,
1162
const UnicodeString& target);
1165
* Return a registered variant specifier for a given source-target
1167
* @param index which specifier to return, from 0 to n-1, where
1168
* n = countAvailableVariants(source, target)
1169
* @param source the source specifier
1170
* @param target the target specifier
1171
* @param result fill-in paramter to receive the variant
1172
* specifier. If source is invalid or if target is invalid or if
1173
* index is out of range, result will be empty.
1174
* @return reference to result
1177
static UnicodeString& U_EXPORT2 getAvailableVariant(int32_t index,
1178
const UnicodeString& source,
1179
const UnicodeString& target,
1180
UnicodeString& result);
1185
* Non-mutexed internal method
1188
static int32_t _countAvailableSources(void);
1191
* Non-mutexed internal method
1194
static UnicodeString& _getAvailableSource(int32_t index,
1195
UnicodeString& result);
1198
* Non-mutexed internal method
1201
static int32_t _countAvailableTargets(const UnicodeString& source);
1204
* Non-mutexed internal method
1207
static UnicodeString& _getAvailableTarget(int32_t index,
1208
const UnicodeString& source,
1209
UnicodeString& result);
1212
* Non-mutexed internal method
1215
static int32_t _countAvailableVariants(const UnicodeString& source,
1216
const UnicodeString& target);
1219
* Non-mutexed internal method
1222
static UnicodeString& _getAvailableVariant(int32_t index,
1223
const UnicodeString& source,
1224
const UnicodeString& target,
1225
UnicodeString& result);
1230
* Set the ID of this transliterators. Subclasses shouldn't do
1231
* this, unless the underlying script behavior has changed.
1232
* @param id the new id t to be set.
1235
void setID(const UnicodeString& id);
1240
* Return the class ID for this class. This is useful only for
1241
* comparing to a return value from getDynamicClassID().
1242
* Note that Transliterator is an abstract base class, and therefor
1243
* no fully constructed object will have a dynamic
1244
* UCLassID that equals the UClassID returned from
1245
* TRansliterator::getStaticClassID().
1246
* @return The class ID for class Transliterator.
1249
static UClassID U_EXPORT2 getStaticClassID(void);
1252
* Returns a unique class ID <b>polymorphically</b>. This method
1253
* is to implement a simple version of RTTI, since not all C++
1254
* compilers support genuine RTTI. Polymorphic operator==() and
1255
* clone() methods call this method.
1257
* <p>Concrete subclasses of Transliterator must use the
1258
* UOBJECT_DEFINE_RTTI_IMPLEMENTATION macro from
1259
* uobject.h to provide the RTTI functions.
1261
* @return The class ID for this object. All objects of a given
1262
* class have the same class ID. Objects of other classes have
1263
* different class IDs.
1266
virtual UClassID getDynamicClassID(void) const = 0;
1269
static UBool initializeRegistry(void);
1273
* Return the number of IDs currently registered with the system.
1274
* To retrieve the actual IDs, call getAvailableID(i) with
1275
* i from 0 to countAvailableIDs() - 1.
1276
* @return the number of IDs currently registered with the system.
1277
* @obsolete ICU 3.4 use getAvailableIDs() instead
1279
static int32_t U_EXPORT2 countAvailableIDs(void);
1282
* Return the index-th available ID. index must be between 0
1283
* and countAvailableIDs() - 1, inclusive. If index is out of
1284
* range, the result of getAvailableID(0) is returned.
1285
* @param index the given ID index.
1286
* @return the index-th available ID. index must be between 0
1287
* and countAvailableIDs() - 1, inclusive. If index is out of
1288
* range, the result of getAvailableID(0) is returned.
1289
* @obsolete ICU 3.4 use getAvailableIDs() instead; this function
1290
* is not thread safe, since it returns a reference to storage that
1291
* may become invalid if another thread calls unregister
1293
static const UnicodeString& U_EXPORT2 getAvailableID(int32_t index);
1296
inline int32_t Transliterator::getMaximumContextLength(void) const {
1297
return maximumContextLength;
1300
inline void Transliterator::setID(const UnicodeString& id) {
1302
// NUL-terminate the ID string, which is a non-aliased copy.
1303
ID.append((UChar)0);
1304
ID.truncate(ID.length()-1);
1307
inline Transliterator::Token Transliterator::integerToken(int32_t i) {
1313
inline Transliterator::Token Transliterator::pointerToken(void* p) {
1321
#endif /* #if !UCONFIG_NO_TRANSLITERATION */