Linux sagir-us1.hostever.us 5.14.0-570.51.1.el9_6.x86_64 #1 SMP PREEMPT_DYNAMIC Wed Oct 8 09:41:34 EDT 2025 x86_64
LiteSpeed
Server IP : 104.247.108.91 & Your IP : 216.73.216.105
Domains : 74 Domain
User : georgeto
Terminal
Auto Root
Create File
Create Folder
Localroot Suggester
Backdoor Destroyer
Readme
/
usr /
local /
apps /
include /
unicode /
Delete
Unzip
Name
Size
Permission
Date
Action
alphaindex.h
26.25
KB
-rw-r--r--
2019-10-10 05:03
appendable.h
8.38
KB
-rw-r--r--
2019-10-10 05:03
basictz.h
8.85
KB
-rw-r--r--
2019-10-10 05:03
brkiter.h
27.62
KB
-rw-r--r--
2019-10-10 05:03
bytestream.h
9.28
KB
-rw-r--r--
2019-10-10 05:03
bytestrie.h
19.27
KB
-rw-r--r--
2019-10-10 05:03
bytestriebuilder.h
7.11
KB
-rw-r--r--
2019-10-10 05:03
calendar.h
105.38
KB
-rw-r--r--
2019-10-10 05:03
caniter.h
7.36
KB
-rw-r--r--
2019-10-10 05:03
chariter.h
23.83
KB
-rw-r--r--
2019-10-10 05:03
choicfmt.h
23.86
KB
-rw-r--r--
2019-10-10 05:03
coleitr.h
13.69
KB
-rw-r--r--
2019-10-10 05:03
coll.h
55.59
KB
-rw-r--r--
2019-10-10 05:03
compactdecimalformat.h
16.59
KB
-rw-r--r--
2019-10-10 05:03
curramt.h
3.69
KB
-rw-r--r--
2019-10-10 05:03
currpinf.h
7.09
KB
-rw-r--r--
2019-10-10 05:03
currunit.h
2.93
KB
-rw-r--r--
2019-10-10 05:03
datefmt.h
40.19
KB
-rw-r--r--
2019-10-10 05:03
dbbi.h
1.12
KB
-rw-r--r--
2019-10-10 05:03
dcfmtsym.h
16.58
KB
-rw-r--r--
2019-10-10 05:03
decimfmt.h
90.64
KB
-rw-r--r--
2019-10-10 05:03
docmain.h
6.42
KB
-rw-r--r--
2019-10-10 05:03
dtfmtsym.h
37.61
KB
-rw-r--r--
2019-10-10 05:03
dtintrv.h
3.77
KB
-rw-r--r--
2019-10-10 05:03
dtitvfmt.h
42.88
KB
-rw-r--r--
2019-10-10 05:03
dtitvinf.h
18.63
KB
-rw-r--r--
2019-10-10 05:03
dtptngen.h
23.54
KB
-rw-r--r--
2019-10-10 05:03
dtrule.h
8.63
KB
-rw-r--r--
2019-10-10 05:03
enumset.h
2.07
KB
-rw-r--r--
2019-10-10 05:03
errorcode.h
4.79
KB
-rw-r--r--
2019-10-10 05:03
fieldpos.h
8.64
KB
-rw-r--r--
2019-10-10 05:03
filteredbrk.h
4.29
KB
-rw-r--r--
2019-10-10 05:03
fmtable.h
24.4
KB
-rw-r--r--
2019-10-10 05:03
format.h
12.45
KB
-rw-r--r--
2019-10-10 05:03
fpositer.h
2.99
KB
-rw-r--r--
2019-10-10 05:03
gender.h
3.19
KB
-rw-r--r--
2019-10-10 05:03
gregocal.h
31.58
KB
-rw-r--r--
2019-10-10 05:03
icudataver.h
1.04
KB
-rw-r--r--
2019-10-10 05:03
icuplug.h
11.87
KB
-rw-r--r--
2019-10-10 05:03
idna.h
12.65
KB
-rw-r--r--
2019-10-10 05:03
listformatter.h
4.85
KB
-rw-r--r--
2019-10-10 05:03
localpointer.h
18.19
KB
-rw-r--r--
2019-10-10 05:03
locdspnm.h
7.04
KB
-rw-r--r--
2019-10-10 05:03
locid.h
31.42
KB
-rw-r--r--
2019-10-10 05:03
measfmt.h
11.31
KB
-rw-r--r--
2019-10-10 05:03
measunit.h
38.43
KB
-rw-r--r--
2019-10-10 05:03
measure.h
4.27
KB
-rw-r--r--
2019-10-10 05:03
messagepattern.h
33.66
KB
-rw-r--r--
2019-10-10 05:03
msgfmt.h
43.17
KB
-rw-r--r--
2019-10-10 05:03
normalizer2.h
28.29
KB
-rw-r--r--
2019-10-10 05:03
normlzr.h
30.72
KB
-rw-r--r--
2019-10-10 05:03
numfmt.h
45.98
KB
-rw-r--r--
2019-10-10 05:03
numsys.h
6.74
KB
-rw-r--r--
2019-10-10 05:03
parseerr.h
3.09
KB
-rw-r--r--
2019-10-10 05:03
parsepos.h
5.46
KB
-rw-r--r--
2019-10-10 05:03
platform.h
27.25
KB
-rw-r--r--
2019-10-10 05:03
plurfmt.h
25.67
KB
-rw-r--r--
2019-10-10 05:03
plurrule.h
17.28
KB
-rw-r--r--
2019-10-10 05:03
ptypes.h
3.48
KB
-rw-r--r--
2019-10-10 05:03
putil.h
6.35
KB
-rw-r--r--
2019-10-10 05:03
rbbi.h
27.78
KB
-rw-r--r--
2019-10-10 05:03
rbnf.h
47.01
KB
-rw-r--r--
2019-10-10 05:03
rbtz.h
15.48
KB
-rw-r--r--
2019-10-10 05:03
regex.h
84.88
KB
-rw-r--r--
2019-10-10 05:03
region.h
9.13
KB
-rw-r--r--
2019-10-10 05:03
reldatefmt.h
14.34
KB
-rw-r--r--
2019-10-10 05:03
rep.h
9.5
KB
-rw-r--r--
2019-10-10 05:03
resbund.h
18.01
KB
-rw-r--r--
2019-10-10 05:03
schriter.h
6.27
KB
-rw-r--r--
2019-10-10 05:03
scientificnumberformatter.h
6.6
KB
-rw-r--r--
2019-10-10 05:03
search.h
22.22
KB
-rw-r--r--
2019-10-10 05:03
selfmt.h
14.25
KB
-rw-r--r--
2019-10-10 05:03
simpleformatter.h
11.26
KB
-rw-r--r--
2019-10-10 05:03
simpletz.h
45.25
KB
-rw-r--r--
2019-10-10 05:03
smpdtfmt.h
69.73
KB
-rw-r--r--
2019-10-10 05:03
sortkey.h
11.13
KB
-rw-r--r--
2019-10-10 05:03
std_string.h
1
KB
-rw-r--r--
2019-10-10 05:03
strenum.h
10.04
KB
-rw-r--r--
2019-10-10 05:03
stringpiece.h
6.49
KB
-rw-r--r--
2019-10-10 05:03
stringtriebuilder.h
15.17
KB
-rw-r--r--
2019-10-10 05:03
stsearch.h
21.3
KB
-rw-r--r--
2019-10-10 05:03
symtable.h
4.22
KB
-rw-r--r--
2019-10-10 05:03
tblcoll.h
36.37
KB
-rw-r--r--
2019-10-10 05:03
timezone.h
40.83
KB
-rw-r--r--
2019-10-10 05:03
tmunit.h
3.3
KB
-rw-r--r--
2019-10-10 05:03
tmutamt.h
4.81
KB
-rw-r--r--
2019-10-10 05:03
tmutfmt.h
7.8
KB
-rw-r--r--
2019-10-10 05:03
translit.h
54.53
KB
-rw-r--r--
2019-10-10 05:03
tzfmt.h
42.81
KB
-rw-r--r--
2019-10-10 05:03
tznames.h
16.86
KB
-rw-r--r--
2019-10-10 05:03
tzrule.h
35.33
KB
-rw-r--r--
2019-10-10 05:03
tztrans.h
6.08
KB
-rw-r--r--
2019-10-10 05:03
ubidi.h
89.14
KB
-rw-r--r--
2019-10-10 05:03
ubiditransform.h
12.76
KB
-rw-r--r--
2019-10-10 05:03
ubrk.h
20.74
KB
-rw-r--r--
2019-10-10 05:03
ucal.h
54.55
KB
-rw-r--r--
2019-10-10 05:03
ucasemap.h
16.57
KB
-rw-r--r--
2019-10-10 05:03
ucat.h
5.37
KB
-rw-r--r--
2019-10-10 05:03
uchar.h
130.91
KB
-rw-r--r--
2019-10-10 05:03
ucharstrie.h
20.92
KB
-rw-r--r--
2019-10-10 05:03
ucharstriebuilder.h
7.13
KB
-rw-r--r--
2019-10-10 05:03
uchriter.h
12.95
KB
-rw-r--r--
2019-10-10 05:03
uclean.h
11.15
KB
-rw-r--r--
2019-10-10 05:03
ucnv.h
83.07
KB
-rw-r--r--
2019-10-10 05:03
ucnv_cb.h
6.6
KB
-rw-r--r--
2019-10-10 05:03
ucnv_err.h
21
KB
-rw-r--r--
2019-10-10 05:03
ucnvsel.h
6.15
KB
-rw-r--r--
2019-10-10 05:03
ucol.h
61.36
KB
-rw-r--r--
2019-10-10 05:03
ucoleitr.h
9.46
KB
-rw-r--r--
2019-10-10 05:03
uconfig.h
11.88
KB
-rw-r--r--
2019-10-10 05:03
ucsdet.h
14.47
KB
-rw-r--r--
2019-10-10 05:03
ucurr.h
15.16
KB
-rw-r--r--
2019-10-10 05:03
udat.h
60.25
KB
-rw-r--r--
2019-10-10 05:03
udata.h
15.55
KB
-rw-r--r--
2019-10-10 05:03
udateintervalformat.h
6.8
KB
-rw-r--r--
2019-10-10 05:03
udatpg.h
24.12
KB
-rw-r--r--
2019-10-10 05:03
udisplaycontext.h
6
KB
-rw-r--r--
2019-10-10 05:03
uenum.h
7.88
KB
-rw-r--r--
2019-10-10 05:03
ufieldpositer.h
4.37
KB
-rw-r--r--
2019-10-10 05:03
uformattable.h
10.94
KB
-rw-r--r--
2019-10-10 05:03
ugender.h
2.01
KB
-rw-r--r--
2019-10-10 05:03
uidna.h
33.38
KB
-rw-r--r--
2019-10-10 05:03
uiter.h
22.79
KB
-rw-r--r--
2019-10-10 05:03
uldnames.h
10.46
KB
-rw-r--r--
2019-10-10 05:03
ulistformatter.h
4.55
KB
-rw-r--r--
2019-10-10 05:03
uloc.h
50.07
KB
-rw-r--r--
2019-10-10 05:03
ulocdata.h
11.27
KB
-rw-r--r--
2019-10-10 05:03
umachine.h
11.56
KB
-rw-r--r--
2019-10-10 05:03
umisc.h
1.35
KB
-rw-r--r--
2019-10-10 05:03
umsg.h
24.24
KB
-rw-r--r--
2019-10-10 05:03
unifilt.h
3.64
KB
-rw-r--r--
2019-10-10 05:03
unifunct.h
3.99
KB
-rw-r--r--
2019-10-10 05:03
unimatch.h
6.05
KB
-rw-r--r--
2019-10-10 05:03
unirepl.h
3.33
KB
-rw-r--r--
2019-10-10 05:03
uniset.h
63.99
KB
-rw-r--r--
2019-10-10 05:03
unistr.h
169.99
KB
-rw-r--r--
2019-10-10 05:03
unorm.h
20.43
KB
-rw-r--r--
2019-10-10 05:03
unorm2.h
25.23
KB
-rw-r--r--
2019-10-10 05:03
unum.h
50.01
KB
-rw-r--r--
2019-10-10 05:03
unumsys.h
7.15
KB
-rw-r--r--
2019-10-10 05:03
uobject.h
10.73
KB
-rw-r--r--
2019-10-10 05:03
upluralrules.h
4.74
KB
-rw-r--r--
2019-10-10 05:03
uregex.h
72.07
KB
-rw-r--r--
2019-10-10 05:03
uregion.h
9.84
KB
-rw-r--r--
2019-10-10 05:03
ureldatefmt.h
12.31
KB
-rw-r--r--
2019-10-10 05:03
urename.h
124.22
KB
-rw-r--r--
2019-10-10 05:03
urep.h
5.39
KB
-rw-r--r--
2019-10-10 05:03
ures.h
36.53
KB
-rw-r--r--
2019-10-10 05:03
uscript.h
25.66
KB
-rw-r--r--
2019-10-10 05:03
usearch.h
38.15
KB
-rw-r--r--
2019-10-10 05:03
uset.h
39.92
KB
-rw-r--r--
2019-10-10 05:03
usetiter.h
9.5
KB
-rw-r--r--
2019-10-10 05:03
ushape.h
18.01
KB
-rw-r--r--
2019-10-10 05:03
uspoof.h
65.05
KB
-rw-r--r--
2019-10-10 05:03
usprep.h
8.14
KB
-rw-r--r--
2019-10-10 05:03
ustdio.h
38.55
KB
-rw-r--r--
2019-10-10 05:03
ustream.h
1.92
KB
-rw-r--r--
2019-10-10 05:03
ustring.h
72.81
KB
-rw-r--r--
2019-10-10 05:03
ustringtrie.h
3.16
KB
-rw-r--r--
2019-10-10 05:03
utext.h
58.09
KB
-rw-r--r--
2019-10-10 05:03
utf.h
7.76
KB
-rw-r--r--
2019-10-10 05:03
utf16.h
18.75
KB
-rw-r--r--
2019-10-10 05:03
utf32.h
777
B
-rw-r--r--
2019-10-10 05:03
utf8.h
27.74
KB
-rw-r--r--
2019-10-10 05:03
utf_old.h
42.62
KB
-rw-r--r--
2019-10-10 05:03
utmscale.h
13.49
KB
-rw-r--r--
2019-10-10 05:03
utrace.h
13.9
KB
-rw-r--r--
2019-10-10 05:03
utrans.h
25.54
KB
-rw-r--r--
2019-10-10 05:03
utypes.h
29.92
KB
-rw-r--r--
2019-10-10 05:03
uvernum.h
5.71
KB
-rw-r--r--
2019-10-10 05:03
uversion.h
6.42
KB
-rw-r--r--
2019-10-10 05:03
vtzone.h
20.18
KB
-rw-r--r--
2019-10-10 05:03
Save
Rename
// Copyright (C) 2016 and later: Unicode, Inc. and others. // License & terms of use: http://www.unicode.org/copyright.html /* *************************************************************************** * Copyright (C) 1999-2016 International Business Machines Corporation * * and others. All rights reserved. * *************************************************************************** ********************************************************************** * Date Name Description * 10/22/99 alan Creation. * 11/11/99 rgillam Complete port from Java. ********************************************************************** */ #ifndef RBBI_H #define RBBI_H #include "unicode/utypes.h" /** * \file * \brief C++ API: Rule Based Break Iterator */ #if !UCONFIG_NO_BREAK_ITERATION #include "unicode/brkiter.h" #include "unicode/udata.h" #include "unicode/parseerr.h" #include "unicode/schriter.h" #include "unicode/uchriter.h" struct UTrie; U_NAMESPACE_BEGIN /** @internal */ struct RBBIDataHeader; class RuleBasedBreakIteratorTables; class BreakIterator; class RBBIDataWrapper; class UStack; class LanguageBreakEngine; class UnhandledEngine; struct RBBIStateTable; /** * * A subclass of BreakIterator whose behavior is specified using a list of rules. * <p>Instances of this class are most commonly created by the factory methods of * BreakIterator::createWordInstance(), BreakIterator::createLineInstance(), etc., * and then used via the abstract API in class BreakIterator</p> * * <p>See the ICU User Guide for information on Break Iterator Rules.</p> * * <p>This class is not intended to be subclassed.</p> */ class U_COMMON_API RuleBasedBreakIterator /*U_FINAL*/ : public BreakIterator { private: /** * The UText through which this BreakIterator accesses the text * @internal */ UText *fText; /** * A character iterator that refers to the same text as the UText, above. * Only included for compatibility with old API, which was based on CharacterIterators. * Value may be adopted from outside, or one of fSCharIter or fDCharIter, below. */ CharacterIterator *fCharIter; /** * When the input text is provided by a UnicodeString, this will point to * a characterIterator that wraps that data. Needed only for the * implementation of getText(), a backwards compatibility issue. */ StringCharacterIterator *fSCharIter; /** * When the input text is provided by a UText, this * dummy CharacterIterator over an empty string will * be returned from getText() */ UCharCharacterIterator *fDCharIter; /** * The rule data for this BreakIterator instance * @internal */ RBBIDataWrapper *fData; /** Index of the Rule {tag} values for the most recent match. * @internal */ int32_t fLastRuleStatusIndex; /** * Rule tag value valid flag. * Some iterator operations don't intrinsically set the correct tag value. * This flag lets us lazily compute the value if we are ever asked for it. * @internal */ UBool fLastStatusIndexValid; /** * Counter for the number of characters encountered with the "dictionary" * flag set. * @internal */ uint32_t fDictionaryCharCount; /** * When a range of characters is divided up using the dictionary, the break * positions that are discovered are stored here, preventing us from having * to use either the dictionary or the state table again until the iterator * leaves this range of text. Has the most impact for line breaking. * @internal */ int32_t* fCachedBreakPositions; /** * The number of elements in fCachedBreakPositions * @internal */ int32_t fNumCachedBreakPositions; /** * if fCachedBreakPositions is not null, this indicates which item in the * cache the current iteration position refers to * @internal */ int32_t fPositionInCache; /** * * If present, UStack of LanguageBreakEngine objects that might handle * dictionary characters. Searched from top to bottom to find an object to * handle a given character. * @internal */ UStack *fLanguageBreakEngines; /** * * If present, the special LanguageBreakEngine used for handling * characters that are in the dictionary set, but not handled by any * LangugageBreakEngine. * @internal */ UnhandledEngine *fUnhandledBreakEngine; /** * * The type of the break iterator, or -1 if it has not been set. * @internal */ int32_t fBreakType; //======================================================================= // constructors //======================================================================= /** * Constructor from a flattened set of RBBI data in malloced memory. * RulesBasedBreakIterators built from a custom set of rules * are created via this constructor; the rules are compiled * into memory, then the break iterator is constructed here. * * The break iterator adopts the memory, and will * free it when done. * @internal */ RuleBasedBreakIterator(RBBIDataHeader* data, UErrorCode &status); friend class RBBIRuleBuilder; /** @internal */ friend class BreakIterator; public: /** Default constructor. Creates an empty shell of an iterator, with no * rules or text to iterate over. Object can subsequently be assigned to. * @stable ICU 2.2 */ RuleBasedBreakIterator(); /** * Copy constructor. Will produce a break iterator with the same behavior, * and which iterates over the same text, as the one passed in. * @param that The RuleBasedBreakIterator passed to be copied * @stable ICU 2.0 */ RuleBasedBreakIterator(const RuleBasedBreakIterator& that); /** * Construct a RuleBasedBreakIterator from a set of rules supplied as a string. * @param rules The break rules to be used. * @param parseError In the event of a syntax error in the rules, provides the location * within the rules of the problem. * @param status Information on any errors encountered. * @stable ICU 2.2 */ RuleBasedBreakIterator( const UnicodeString &rules, UParseError &parseError, UErrorCode &status); /** * Contruct a RuleBasedBreakIterator from a set of precompiled binary rules. * Binary rules are obtained from RulesBasedBreakIterator::getBinaryRules(). * Construction of a break iterator in this way is substantially faster than * constuction from source rules. * * Ownership of the storage containing the compiled rules remains with the * caller of this function. The compiled rules must not be modified or * deleted during the life of the break iterator. * * The compiled rules are not compatible across different major versions of ICU. * The compiled rules are comaptible only between machines with the same * byte ordering (little or big endian) and the same base character set family * (ASCII or EBCDIC). * * @see #getBinaryRules * @param compiledRules A pointer to the compiled break rules to be used. * @param ruleLength The length of the compiled break rules, in bytes. This * corresponds to the length value produced by getBinaryRules(). * @param status Information on any errors encountered, including invalid * binary rules. * @stable ICU 4.8 */ RuleBasedBreakIterator(const uint8_t *compiledRules, uint32_t ruleLength, UErrorCode &status); /** * This constructor uses the udata interface to create a BreakIterator * whose internal tables live in a memory-mapped file. "image" is an * ICU UDataMemory handle for the pre-compiled break iterator tables. * @param image handle to the memory image for the break iterator data. * Ownership of the UDataMemory handle passes to the Break Iterator, * which will be responsible for closing it when it is no longer needed. * @param status Information on any errors encountered. * @see udata_open * @see #getBinaryRules * @stable ICU 2.8 */ RuleBasedBreakIterator(UDataMemory* image, UErrorCode &status); /** * Destructor * @stable ICU 2.0 */ virtual ~RuleBasedBreakIterator(); /** * Assignment operator. Sets this iterator to have the same behavior, * and iterate over the same text, as the one passed in. * @param that The RuleBasedBreakItertor passed in * @return the newly created RuleBasedBreakIterator * @stable ICU 2.0 */ RuleBasedBreakIterator& operator=(const RuleBasedBreakIterator& that); /** * Equality operator. Returns TRUE if both BreakIterators are of the * same class, have the same behavior, and iterate over the same text. * @param that The BreakIterator to be compared for equality * @return TRUE if both BreakIterators are of the * same class, have the same behavior, and iterate over the same text. * @stable ICU 2.0 */ virtual UBool operator==(const BreakIterator& that) const; /** * Not-equal operator. If operator== returns TRUE, this returns FALSE, * and vice versa. * @param that The BreakIterator to be compared for inequality * @return TRUE if both BreakIterators are not same. * @stable ICU 2.0 */ UBool operator!=(const BreakIterator& that) const; /** * Returns a newly-constructed RuleBasedBreakIterator with the same * behavior, and iterating over the same text, as this one. * Differs from the copy constructor in that it is polymorphic, and * will correctly clone (copy) a derived class. * clone() is thread safe. Multiple threads may simultaeneously * clone the same source break iterator. * @return a newly-constructed RuleBasedBreakIterator * @stable ICU 2.0 */ virtual BreakIterator* clone() const; /** * Compute a hash code for this BreakIterator * @return A hash code * @stable ICU 2.0 */ virtual int32_t hashCode(void) const; /** * Returns the description used to create this iterator * @return the description used to create this iterator * @stable ICU 2.0 */ virtual const UnicodeString& getRules(void) const; //======================================================================= // BreakIterator overrides //======================================================================= /** * <p> * Return a CharacterIterator over the text being analyzed. * The returned character iterator is owned by the break iterator, and must * not be deleted by the caller. Repeated calls to this function may * return the same CharacterIterator. * </p> * <p> * The returned character iterator must not be used concurrently with * the break iterator. If concurrent operation is needed, clone the * returned character iterator first and operate on the clone. * </p> * <p> * When the break iterator is operating on text supplied via a UText, * this function will fail. Lacking any way to signal failures, it * returns an CharacterIterator containing no text. * The function getUText() provides similar functionality, * is reliable, and is more efficient. * </p> * * TODO: deprecate this function? * * @return An iterator over the text being analyzed. * @stable ICU 2.0 */ virtual CharacterIterator& getText(void) const; /** * Get a UText for the text being analyzed. * The returned UText is a shallow clone of the UText used internally * by the break iterator implementation. It can safely be used to * access the text without impacting any break iterator operations, * but the underlying text itself must not be altered. * * @param fillIn A UText to be filled in. If NULL, a new UText will be * allocated to hold the result. * @param status receives any error codes. * @return The current UText for this break iterator. If an input * UText was provided, it will always be returned. * @stable ICU 3.4 */ virtual UText *getUText(UText *fillIn, UErrorCode &status) const; /** * Set the iterator to analyze a new piece of text. This function resets * the current iteration position to the beginning of the text. * @param newText An iterator over the text to analyze. The BreakIterator * takes ownership of the character iterator. The caller MUST NOT delete it! * @stable ICU 2.0 */ virtual void adoptText(CharacterIterator* newText); /** * Set the iterator to analyze a new piece of text. This function resets * the current iteration position to the beginning of the text. * * The BreakIterator will retain a reference to the supplied string. * The caller must not modify or delete the text while the BreakIterator * retains the reference. * * @param newText The text to analyze. * @stable ICU 2.0 */ virtual void setText(const UnicodeString& newText); /** * Reset the break iterator to operate over the text represented by * the UText. The iterator position is reset to the start. * * This function makes a shallow clone of the supplied UText. This means * that the caller is free to immediately close or otherwise reuse the * Utext that was passed as a parameter, but that the underlying text itself * must not be altered while being referenced by the break iterator. * * @param text The UText used to change the text. * @param status Receives any error codes. * @stable ICU 3.4 */ virtual void setText(UText *text, UErrorCode &status); /** * Sets the current iteration position to the beginning of the text, position zero. * @return The offset of the beginning of the text, zero. * @stable ICU 2.0 */ virtual int32_t first(void); /** * Sets the current iteration position to the end of the text. * @return The text's past-the-end offset. * @stable ICU 2.0 */ virtual int32_t last(void); /** * Advances the iterator either forward or backward the specified number of steps. * Negative values move backward, and positive values move forward. This is * equivalent to repeatedly calling next() or previous(). * @param n The number of steps to move. The sign indicates the direction * (negative is backwards, and positive is forwards). * @return The character offset of the boundary position n boundaries away from * the current one. * @stable ICU 2.0 */ virtual int32_t next(int32_t n); /** * Advances the iterator to the next boundary position. * @return The position of the first boundary after this one. * @stable ICU 2.0 */ virtual int32_t next(void); /** * Moves the iterator backwards, to the last boundary preceding this one. * @return The position of the last boundary position preceding this one. * @stable ICU 2.0 */ virtual int32_t previous(void); /** * Sets the iterator to refer to the first boundary position following * the specified position. * @param offset The position from which to begin searching for a break position. * @return The position of the first break after the current position. * @stable ICU 2.0 */ virtual int32_t following(int32_t offset); /** * Sets the iterator to refer to the last boundary position before the * specified position. * @param offset The position to begin searching for a break from. * @return The position of the last boundary before the starting position. * @stable ICU 2.0 */ virtual int32_t preceding(int32_t offset); /** * Returns true if the specfied position is a boundary position. As a side * effect, leaves the iterator pointing to the first boundary position at * or after "offset". * @param offset the offset to check. * @return True if "offset" is a boundary position. * @stable ICU 2.0 */ virtual UBool isBoundary(int32_t offset); /** * Returns the current iteration position. * @return The current iteration position. * @stable ICU 2.0 */ virtual int32_t current(void) const; /** * Return the status tag from the break rule that determined the most recently * returned break position. For break rules that do not specify a * status, a default value of 0 is returned. If more than one break rule * would cause a boundary to be located at some position in the text, * the numerically largest of the applicable status values is returned. * <p> * Of the standard types of ICU break iterators, only word break and * line break provide status values. The values are defined in * the header file ubrk.h. For Word breaks, the status allows distinguishing between words * that contain alphabetic letters, "words" that appear to be numbers, * punctuation and spaces, words containing ideographic characters, and * more. For Line Break, the status distinguishes between hard (mandatory) breaks * and soft (potential) break positions. * <p> * <code>getRuleStatus()</code> can be called after obtaining a boundary * position from <code>next()</code>, <code>previous()</code>, or * any other break iterator functions that returns a boundary position. * <p> * When creating custom break rules, one is free to define whatever * status values may be convenient for the application. * <p> * Note: this function is not thread safe. It should not have been * declared const, and the const remains only for compatibility * reasons. (The function is logically const, but not bit-wise const). * <p> * @return the status from the break rule that determined the most recently * returned break position. * * @see UWordBreak * @stable ICU 2.2 */ virtual int32_t getRuleStatus() const; /** * Get the status (tag) values from the break rule(s) that determined the most * recently returned break position. * <p> * The returned status value(s) are stored into an array provided by the caller. * The values are stored in sorted (ascending) order. * If the capacity of the output array is insufficient to hold the data, * the output will be truncated to the available length, and a * U_BUFFER_OVERFLOW_ERROR will be signaled. * * @param fillInVec an array to be filled in with the status values. * @param capacity the length of the supplied vector. A length of zero causes * the function to return the number of status values, in the * normal way, without attemtping to store any values. * @param status receives error codes. * @return The number of rule status values from rules that determined * the most recent boundary returned by the break iterator. * In the event of a U_BUFFER_OVERFLOW_ERROR, the return value * is the total number of status values that were available, * not the reduced number that were actually returned. * @see getRuleStatus * @stable ICU 3.0 */ virtual int32_t getRuleStatusVec(int32_t *fillInVec, int32_t capacity, UErrorCode &status); /** * Returns a unique class ID POLYMORPHICALLY. Pure virtual override. * This method is to implement a simple version of RTTI, since not all * C++ compilers support genuine RTTI. Polymorphic operator==() and * clone() methods call this method. * * @return The class ID for this object. All objects of a * given class have the same class ID. Objects of * other classes have different class IDs. * @stable ICU 2.0 */ virtual UClassID getDynamicClassID(void) const; /** * Returns the class ID for this class. This is useful only for * comparing to a return value from getDynamicClassID(). For example: * * Base* polymorphic_pointer = createPolymorphicObject(); * if (polymorphic_pointer->getDynamicClassID() == * Derived::getStaticClassID()) ... * * @return The class ID for all objects of this class. * @stable ICU 2.0 */ static UClassID U_EXPORT2 getStaticClassID(void); /** * Deprecated functionality. Use clone() instead. * * Create a clone (copy) of this break iterator in memory provided * by the caller. The idea is to increase performance by avoiding * a storage allocation. Use of this functoin is NOT RECOMMENDED. * Performance gains are minimal, and correct buffer management is * tricky. Use clone() instead. * * @param stackBuffer The pointer to the memory into which the cloned object * should be placed. If NULL, allocate heap memory * for the cloned object. * @param BufferSize The size of the buffer. If zero, return the required * buffer size, but do not clone the object. If the * size was too small (but not zero), allocate heap * storage for the cloned object. * * @param status Error status. U_SAFECLONE_ALLOCATED_WARNING will be * returned if the the provided buffer was too small, and * the clone was therefore put on the heap. * * @return Pointer to the clone object. This may differ from the stackBuffer * address if the byte alignment of the stack buffer was not suitable * or if the stackBuffer was too small to hold the clone. * @deprecated ICU 52. Use clone() instead. */ virtual BreakIterator * createBufferClone(void *stackBuffer, int32_t &BufferSize, UErrorCode &status); /** * Return the binary form of compiled break rules, * which can then be used to create a new break iterator at some * time in the future. Creating a break iterator from pre-compiled rules * is much faster than building one from the source form of the * break rules. * * The binary data can only be used with the same version of ICU * and on the same platform type (processor endian-ness) * * @param length Returns the length of the binary data. (Out paramter.) * * @return A pointer to the binary (compiled) rule data. The storage * belongs to the RulesBasedBreakIterator object, not the * caller, and must not be modified or deleted. * @stable ICU 4.8 */ virtual const uint8_t *getBinaryRules(uint32_t &length); /** * Set the subject text string upon which the break iterator is operating * without changing any other aspect of the matching state. * The new and previous text strings must have the same content. * * This function is intended for use in environments where ICU is operating on * strings that may move around in memory. It provides a mechanism for notifying * ICU that the string has been relocated, and providing a new UText to access the * string in its new position. * * Note that the break iterator implementation never copies the underlying text * of a string being processed, but always operates directly on the original text * provided by the user. Refreshing simply drops the references to the old text * and replaces them with references to the new. * * Caution: this function is normally used only by very specialized, * system-level code. One example use case is with garbage collection that moves * the text in memory. * * @param input The new (moved) text string. * @param status Receives errors detected by this function. * @return *this * * @stable ICU 49 */ virtual RuleBasedBreakIterator &refreshInputText(UText *input, UErrorCode &status); private: //======================================================================= // implementation //======================================================================= /** * Dumps caches and performs other actions associated with a complete change * in text or iteration position. * @internal */ void reset(void); /** * Set the type of the break iterator. * @internal */ void setBreakType(int32_t type); /** * Common initialization function, used by constructors and bufferClone. * @internal */ void init(); /** * This method backs the iterator back up to a "safe position" in the text. * This is a position that we know, without any context, must be a break position. * The various calling methods then iterate forward from this safe position to * the appropriate position to return. (For more information, see the description * of buildBackwardsStateTable() in RuleBasedBreakIterator.Builder.) * @param statetable state table used of moving backwards * @internal */ int32_t handlePrevious(const RBBIStateTable *statetable); /** * This method is the actual implementation of the next() method. All iteration * vectors through here. This method initializes the state machine to state 1 * and advances through the text character by character until we reach the end * of the text or the state machine transitions to state 0. We update our return * value every time the state machine passes through a possible end state. * @param statetable state table used of moving forwards * @internal */ int32_t handleNext(const RBBIStateTable *statetable); /** * This is the function that actually implements dictionary-based * breaking. Covering at least the range from startPos to endPos, * it checks for dictionary characters, and if it finds them determines * the appropriate object to deal with them. It may cache found breaks in * fCachedBreakPositions as it goes. It may well also look at text outside * the range startPos to endPos. * If going forward, endPos is the normal Unicode break result, and * if goind in reverse, startPos is the normal Unicode break result * @param startPos The start position of a range of text * @param endPos The end position of a range of text * @param reverse The call is for the reverse direction * @internal */ int32_t checkDictionary(int32_t startPos, int32_t endPos, UBool reverse); /** * This function returns the appropriate LanguageBreakEngine for a * given character c. * @param c A character in the dictionary set * @internal */ const LanguageBreakEngine *getLanguageBreakEngine(UChar32 c); /** * @internal */ void makeRuleStatusValid(); }; //------------------------------------------------------------------------------ // // Inline Functions Definitions ... // //------------------------------------------------------------------------------ inline UBool RuleBasedBreakIterator::operator!=(const BreakIterator& that) const { return !operator==(that); } U_NAMESPACE_END #endif /* #if !UCONFIG_NO_BREAK_ITERATION */ #endif