OXIESEC PANEL
- Current Dir:
/
/
usr
/
include
/
unicode
Server IP: 139.59.38.164
Upload:
Create Dir:
Name
Size
Modified
Perms
📁
..
-
10/28/2024 06:50:42 AM
rwxr-xr-x
📄
alphaindex.h
26.27 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
appendable.h
8.43 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
basictz.h
8.84 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
brkiter.h
27.54 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
bytestream.h
9.59 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
bytestrie.h
19.26 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
bytestriebuilder.h
7.1 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
calendar.h
105.35 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
caniter.h
7.35 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
casemap.h
25.31 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
char16ptr.h
7.29 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
chariter.h
24 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
choicfmt.h
23.85 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
coleitr.h
13.7 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
coll.h
55.61 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
compactdecimalformat.h
16.58 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
curramt.h
3.69 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
currpinf.h
7.08 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
currunit.h
3.49 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
datefmt.h
40.18 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dbbi.h
1.11 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dcfmtsym.h
17.66 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
decimfmt.h
89.73 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
docmain.h
6.56 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dtfmtsym.h
37.64 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dtintrv.h
3.76 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dtitvfmt.h
42.87 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dtitvinf.h
18.46 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dtptngen.h
23.72 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
dtrule.h
8.62 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
edits.h
15.54 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
enumset.h
2.05 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
errorcode.h
4.78 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
fieldpos.h
8.63 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
filteredbrk.h
5.46 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
fmtable.h
24.39 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
format.h
12.44 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
fpositer.h
3.14 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
gender.h
3.18 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
gregocal.h
31.57 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
icudataver.h
1.03 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
icuplug.h
11.86 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
idna.h
12.63 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
listformatter.h
4.98 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
localpointer.h
18.19 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
locdspnm.h
7.05 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
locid.h
31.4 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
measfmt.h
11.25 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
measunit.h
37.74 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
measure.h
4.26 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
messagepattern.h
33.64 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
msgfmt.h
43.16 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
normalizer2.h
33.97 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
normlzr.h
30.74 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
nounit.h
2.6 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
numberformatter.h
64.98 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
numfmt.h
47.76 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
numsys.h
6.73 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
parseerr.h
3.08 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
parsepos.h
5.45 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
platform.h
27.71 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
plurfmt.h
25.66 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
plurrule.h
18.34 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ptypes.h
3.47 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
putil.h
6.34 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
rbbi.h
27.09 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
rbnf.h
49.76 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
rbtz.h
15.47 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
regex.h
84.91 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
region.h
9.12 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
reldatefmt.h
14.3 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
rep.h
9.5 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
resbund.h
18.01 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
schriter.h
6.26 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
scientificnumberformatter.h
6.59 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
search.h
22.21 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
selfmt.h
14.24 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
simpleformatter.h
11.43 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
simpletz.h
45.24 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
smpdtfmt.h
70.09 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
sortkey.h
11.12 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
std_string.h
1015 bytes
10/19/2021 07:44:28 PM
rw-r--r--
📄
strenum.h
10.04 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
stringoptions.h
5.89 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
stringpiece.h
6.46 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
stringtriebuilder.h
15.2 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
stsearch.h
21.29 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
symtable.h
4.21 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tblcoll.h
36.4 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
timezone.h
40.83 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tmunit.h
3.29 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tmutamt.h
4.8 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tmutfmt.h
7.79 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
translit.h
54.53 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tzfmt.h
42.81 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tznames.h
16.79 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tzrule.h
35.32 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
tztrans.h
6.07 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ubidi.h
89.28 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ubiditransform.h
12.71 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ubrk.h
24.08 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucal.h
54.56 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucasemap.h
15.18 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucat.h
5.36 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uchar.h
131.61 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucharstrie.h
21.06 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucharstriebuilder.h
7.14 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uchriter.h
13.14 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uclean.h
11.24 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucnv.h
83.06 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucnv_cb.h
6.59 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucnv_err.h
20.98 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucnvsel.h
6.14 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucol.h
61.36 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucoleitr.h
9.46 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uconfig.h
11.91 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucsdet.h
14.67 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ucurr.h
15.15 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
udat.h
60.13 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
udata.h
15.53 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
udateintervalformat.h
6.79 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
udatpg.h
24.11 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
udisplaycontext.h
5.89 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uenum.h
7.9 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ufieldpositer.h
4.36 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uformattable.h
10.94 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ugender.h
2 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uidna.h
33.37 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uiter.h
22.77 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uldnames.h
10.45 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ulistformatter.h
4.54 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uloc.h
50.75 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ulocdata.h
11.26 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
umachine.h
13.1 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
umisc.h
1.33 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
umsg.h
24.23 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unifilt.h
3.63 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unifunct.h
3.98 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unimatch.h
6.04 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unirepl.h
3.32 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uniset.h
63.98 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unistr.h
175.65 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unorm.h
20.45 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unorm2.h
24.66 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unum.h
52.49 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
unumsys.h
7.14 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uobject.h
10.72 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
upluralrules.h
6.77 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uregex.h
72.05 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uregion.h
9.84 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ureldatefmt.h
12.22 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
urename.h
124.49 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
urep.h
5.38 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ures.h
36.52 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uscript.h
25.91 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
usearch.h
38.14 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uset.h
39.91 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
usetiter.h
9.49 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ushape.h
18 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uspoof.h
64.9 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
usprep.h
8.13 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ustdio.h
38.54 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ustream.h
1.8 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ustring.h
72.52 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
ustringtrie.h
3.15 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utext.h
58.11 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utf.h
7.86 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utf16.h
22.49 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utf32.h
763 bytes
10/19/2021 07:44:28 PM
rw-r--r--
📄
utf8.h
28.16 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utf_old.h
43.78 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utmscale.h
13.78 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utrace.h
13.89 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utrans.h
25.53 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
utypes.h
29.92 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uvernum.h
5.69 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
uversion.h
6.4 KB
10/19/2021 07:44:28 PM
rw-r--r--
📄
vtzone.h
20.17 KB
10/19/2021 07:44:28 PM
rw-r--r--
Editing: chariter.h
Close
// © 2016 and later: Unicode, Inc. and others. // License & terms of use: http://www.unicode.org/copyright.html /* ******************************************************************** * * Copyright (C) 1997-2011, International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************** */ #ifndef CHARITER_H #define CHARITER_H #include "unicode/utypes.h" #include "unicode/uobject.h" #include "unicode/unistr.h" /** * \file * \brief C++ API: Character Iterator */ U_NAMESPACE_BEGIN /** * Abstract class that defines an API for forward-only iteration * on text objects. * This is a minimal interface for iteration without random access * or backwards iteration. It is especially useful for wrapping * streams with converters into an object for collation or * normalization. * * <p>Characters can be accessed in two ways: as code units or as * code points. * Unicode code points are 21-bit integers and are the scalar values * of Unicode characters. ICU uses the type UChar32 for them. * Unicode code units are the storage units of a given * Unicode/UCS Transformation Format (a character encoding scheme). * With UTF-16, all code points can be represented with either one * or two code units ("surrogates"). * String storage is typically based on code units, while properties * of characters are typically determined using code point values. * Some processes may be designed to work with sequences of code units, * or it may be known that all characters that are important to an * algorithm can be represented with single code units. * Other processes will need to use the code point access functions.</p> * * <p>ForwardCharacterIterator provides nextPostInc() to access * a code unit and advance an internal position into the text object, * similar to a <code>return text[position++]</code>.<br> * It provides next32PostInc() to access a code point and advance an internal * position.</p> * * <p>next32PostInc() assumes that the current position is that of * the beginning of a code point, i.e., of its first code unit. * After next32PostInc(), this will be true again. * In general, access to code units and code points in the same * iteration loop should not be mixed. In UTF-16, if the current position * is on a second code unit (Low Surrogate), then only that code unit * is returned even by next32PostInc().</p> * * <p>For iteration with either function, there are two ways to * check for the end of the iteration. When there are no more * characters in the text object: * <ul> * <li>The hasNext() function returns FALSE.</li> * <li>nextPostInc() and next32PostInc() return DONE * when one attempts to read beyond the end of the text object.</li> * </ul> * * Example: * \code * void function1(ForwardCharacterIterator &it) { * UChar32 c; * while(it.hasNext()) { * c=it.next32PostInc(); * // use c * } * } * * void function1(ForwardCharacterIterator &it) { * char16_t c; * while((c=it.nextPostInc())!=ForwardCharacterIterator::DONE) { * // use c * } * } * \endcode * </p> * * @stable ICU 2.0 */ class U_COMMON_API ForwardCharacterIterator : public UObject { public: /** * Value returned by most of ForwardCharacterIterator's functions * when the iterator has reached the limits of its iteration. * @stable ICU 2.0 */ enum { DONE = 0xffff }; /** * Destructor. * @stable ICU 2.0 */ virtual ~ForwardCharacterIterator(); /** * Returns true when both iterators refer to the same * character in the same character-storage object. * @param that The ForwardCharacterIterator to be compared for equality * @return true when both iterators refer to the same * character in the same character-storage object * @stable ICU 2.0 */ virtual UBool operator==(const ForwardCharacterIterator& that) const = 0; /** * Returns true when the iterators refer to different * text-storage objects, or to different characters in the * same text-storage object. * @param that The ForwardCharacterIterator to be compared for inequality * @return true when the iterators refer to different * text-storage objects, or to different characters in the * same text-storage object * @stable ICU 2.0 */ inline UBool operator!=(const ForwardCharacterIterator& that) const; /** * Generates a hash code for this iterator. * @return the hash code. * @stable ICU 2.0 */ virtual int32_t hashCode(void) const = 0; /** * Returns a UClassID for this ForwardCharacterIterator ("poor man's * RTTI").<P> Despite the fact that this function is public, * DO NOT CONSIDER IT PART OF CHARACTERITERATOR'S API! * @return a UClassID for this ForwardCharacterIterator * @stable ICU 2.0 */ virtual UClassID getDynamicClassID(void) const = 0; /** * Gets the current code unit for returning and advances to the next code unit * in the iteration range * (toward endIndex()). If there are * no more code units to return, returns DONE. * @return the current code unit. * @stable ICU 2.0 */ virtual char16_t nextPostInc(void) = 0; /** * Gets the current code point for returning and advances to the next code point * in the iteration range * (toward endIndex()). If there are * no more code points to return, returns DONE. * @return the current code point. * @stable ICU 2.0 */ virtual UChar32 next32PostInc(void) = 0; /** * Returns FALSE if there are no more code units or code points * at or after the current position in the iteration range. * This is used with nextPostInc() or next32PostInc() in forward * iteration. * @returns FALSE if there are no more code units or code points * at or after the current position in the iteration range. * @stable ICU 2.0 */ virtual UBool hasNext() = 0; protected: /** Default constructor to be overridden in the implementing class. @stable ICU 2.0*/ ForwardCharacterIterator(); /** Copy constructor to be overridden in the implementing class. @stable ICU 2.0*/ ForwardCharacterIterator(const ForwardCharacterIterator &other); /** * Assignment operator to be overridden in the implementing class. * @stable ICU 2.0 */ ForwardCharacterIterator &operator=(const ForwardCharacterIterator&) { return *this; } }; /** * Abstract class that defines an API for iteration * on text objects. * This is an interface for forward and backward iteration * and random access into a text object. * * <p>The API provides backward compatibility to the Java and older ICU * CharacterIterator classes but extends them significantly: * <ol> * <li>CharacterIterator is now a subclass of ForwardCharacterIterator.</li> * <li>While the old API functions provided forward iteration with * "pre-increment" semantics, the new one also provides functions * with "post-increment" semantics. They are more efficient and should * be the preferred iterator functions for new implementations. * The backward iteration always had "pre-decrement" semantics, which * are efficient.</li> * <li>Just like ForwardCharacterIterator, it provides access to * both code units and code points. Code point access versions are available * for the old and the new iteration semantics.</li> * <li>There are new functions for setting and moving the current position * without returning a character, for efficiency.</li> * </ol> * * See ForwardCharacterIterator for examples for using the new forward iteration * functions. For backward iteration, there is also a hasPrevious() function * that can be used analogously to hasNext(). * The old functions work as before and are shown below.</p> * * <p>Examples for some of the new functions:</p> * * Forward iteration with hasNext(): * \code * void forward1(CharacterIterator &it) { * UChar32 c; * for(it.setToStart(); it.hasNext();) { * c=it.next32PostInc(); * // use c * } * } * \endcode * Forward iteration more similar to loops with the old forward iteration, * showing a way to convert simple for() loops: * \code * void forward2(CharacterIterator &it) { * char16_t c; * for(c=it.firstPostInc(); c!=CharacterIterator::DONE; c=it.nextPostInc()) { * // use c * } * } * \endcode * Backward iteration with setToEnd() and hasPrevious(): * \code * void backward1(CharacterIterator &it) { * UChar32 c; * for(it.setToEnd(); it.hasPrevious();) { * c=it.previous32(); * // use c * } * } * \endcode * Backward iteration with a more traditional for() loop: * \code * void backward2(CharacterIterator &it) { * char16_t c; * for(c=it.last(); c!=CharacterIterator::DONE; c=it.previous()) { * // use c * } * } * \endcode * * Example for random access: * \code * void random(CharacterIterator &it) { * // set to the third code point from the beginning * it.move32(3, CharacterIterator::kStart); * // get a code point from here without moving the position * UChar32 c=it.current32(); * // get the position * int32_t pos=it.getIndex(); * // get the previous code unit * char16_t u=it.previous(); * // move back one more code unit * it.move(-1, CharacterIterator::kCurrent); * // set the position back to where it was * // and read the same code point c and move beyond it * it.setIndex(pos); * if(c!=it.next32PostInc()) { * exit(1); // CharacterIterator inconsistent * } * } * \endcode * * <p>Examples, especially for the old API:</p> * * Function processing characters, in this example simple output * <pre> * \code * void processChar( char16_t c ) * { * cout << " " << c; * } * \endcode * </pre> * Traverse the text from start to finish * <pre> * \code * void traverseForward(CharacterIterator& iter) * { * for(char16_t c = iter.first(); c != CharacterIterator.DONE; c = iter.next()) { * processChar(c); * } * } * \endcode * </pre> * Traverse the text backwards, from end to start * <pre> * \code * void traverseBackward(CharacterIterator& iter) * { * for(char16_t c = iter.last(); c != CharacterIterator.DONE; c = iter.previous()) { * processChar(c); * } * } * \endcode * </pre> * Traverse both forward and backward from a given position in the text. * Calls to notBoundary() in this example represents some additional stopping criteria. * <pre> * \code * void traverseOut(CharacterIterator& iter, int32_t pos) * { * char16_t c; * for (c = iter.setIndex(pos); * c != CharacterIterator.DONE && (Unicode::isLetter(c) || Unicode::isDigit(c)); * c = iter.next()) {} * int32_t end = iter.getIndex(); * for (c = iter.setIndex(pos); * c != CharacterIterator.DONE && (Unicode::isLetter(c) || Unicode::isDigit(c)); * c = iter.previous()) {} * int32_t start = iter.getIndex() + 1; * * cout << "start: " << start << " end: " << end << endl; * for (c = iter.setIndex(start); iter.getIndex() < end; c = iter.next() ) { * processChar(c); * } * } * \endcode * </pre> * Creating a StringCharacterIterator and calling the test functions * <pre> * \code * void CharacterIterator_Example( void ) * { * cout << endl << "===== CharacterIterator_Example: =====" << endl; * UnicodeString text("Ein kleiner Satz."); * StringCharacterIterator iterator(text); * cout << "----- traverseForward: -----------" << endl; * traverseForward( iterator ); * cout << endl << endl << "----- traverseBackward: ----------" << endl; * traverseBackward( iterator ); * cout << endl << endl << "----- traverseOut: ---------------" << endl; * traverseOut( iterator, 7 ); * cout << endl << endl << "-----" << endl; * } * \endcode * </pre> * * @stable ICU 2.0 */ class U_COMMON_API CharacterIterator : public ForwardCharacterIterator { public: /** * Origin enumeration for the move() and move32() functions. * @stable ICU 2.0 */ enum EOrigin { kStart, kCurrent, kEnd }; /** * Destructor. * @stable ICU 2.0 */ virtual ~CharacterIterator(); /** * Returns a pointer to a new CharacterIterator of the same * concrete class as this one, and referring to the same * character in the same text-storage object as this one. The * caller is responsible for deleting the new clone. * @return a pointer to a new CharacterIterator * @stable ICU 2.0 */ virtual CharacterIterator* clone(void) const = 0; /** * Sets the iterator to refer to the first code unit in its * iteration range, and returns that code unit. * This can be used to begin an iteration with next(). * @return the first code unit in its iteration range. * @stable ICU 2.0 */ virtual char16_t first(void) = 0; /** * Sets the iterator to refer to the first code unit in its * iteration range, returns that code unit, and moves the position * to the second code unit. This is an alternative to setToStart() * for forward iteration with nextPostInc(). * @return the first code unit in its iteration range. * @stable ICU 2.0 */ virtual char16_t firstPostInc(void); /** * Sets the iterator to refer to the first code point in its * iteration range, and returns that code unit, * This can be used to begin an iteration with next32(). * Note that an iteration with next32PostInc(), beginning with, * e.g., setToStart() or firstPostInc(), is more efficient. * @return the first code point in its iteration range. * @stable ICU 2.0 */ virtual UChar32 first32(void) = 0; /** * Sets the iterator to refer to the first code point in its * iteration range, returns that code point, and moves the position * to the second code point. This is an alternative to setToStart() * for forward iteration with next32PostInc(). * @return the first code point in its iteration range. * @stable ICU 2.0 */ virtual UChar32 first32PostInc(void); /** * Sets the iterator to refer to the first code unit or code point in its * iteration range. This can be used to begin a forward * iteration with nextPostInc() or next32PostInc(). * @return the start position of the iteration range * @stable ICU 2.0 */ inline int32_t setToStart(); /** * Sets the iterator to refer to the last code unit in its * iteration range, and returns that code unit. * This can be used to begin an iteration with previous(). * @return the last code unit. * @stable ICU 2.0 */ virtual char16_t last(void) = 0; /** * Sets the iterator to refer to the last code point in its * iteration range, and returns that code unit. * This can be used to begin an iteration with previous32(). * @return the last code point. * @stable ICU 2.0 */ virtual UChar32 last32(void) = 0; /** * Sets the iterator to the end of its iteration range, just behind * the last code unit or code point. This can be used to begin a backward * iteration with previous() or previous32(). * @return the end position of the iteration range * @stable ICU 2.0 */ inline int32_t setToEnd(); /** * Sets the iterator to refer to the "position"-th code unit * in the text-storage object the iterator refers to, and * returns that code unit. * @param position the "position"-th code unit in the text-storage object * @return the "position"-th code unit. * @stable ICU 2.0 */ virtual char16_t setIndex(int32_t position) = 0; /** * Sets the iterator to refer to the beginning of the code point * that contains the "position"-th code unit * in the text-storage object the iterator refers to, and * returns that code point. * The current position is adjusted to the beginning of the code point * (its first code unit). * @param position the "position"-th code unit in the text-storage object * @return the "position"-th code point. * @stable ICU 2.0 */ virtual UChar32 setIndex32(int32_t position) = 0; /** * Returns the code unit the iterator currently refers to. * @return the current code unit. * @stable ICU 2.0 */ virtual char16_t current(void) const = 0; /** * Returns the code point the iterator currently refers to. * @return the current code point. * @stable ICU 2.0 */ virtual UChar32 current32(void) const = 0; /** * Advances to the next code unit in the iteration range * (toward endIndex()), and returns that code unit. If there are * no more code units to return, returns DONE. * @return the next code unit. * @stable ICU 2.0 */ virtual char16_t next(void) = 0; /** * Advances to the next code point in the iteration range * (toward endIndex()), and returns that code point. If there are * no more code points to return, returns DONE. * Note that iteration with "pre-increment" semantics is less * efficient than iteration with "post-increment" semantics * that is provided by next32PostInc(). * @return the next code point. * @stable ICU 2.0 */ virtual UChar32 next32(void) = 0; /** * Advances to the previous code unit in the iteration range * (toward startIndex()), and returns that code unit. If there are * no more code units to return, returns DONE. * @return the previous code unit. * @stable ICU 2.0 */ virtual char16_t previous(void) = 0; /** * Advances to the previous code point in the iteration range * (toward startIndex()), and returns that code point. If there are * no more code points to return, returns DONE. * @return the previous code point. * @stable ICU 2.0 */ virtual UChar32 previous32(void) = 0; /** * Returns FALSE if there are no more code units or code points * before the current position in the iteration range. * This is used with previous() or previous32() in backward * iteration. * @return FALSE if there are no more code units or code points * before the current position in the iteration range, return TRUE otherwise. * @stable ICU 2.0 */ virtual UBool hasPrevious() = 0; /** * Returns the numeric index in the underlying text-storage * object of the character returned by first(). Since it's * possible to create an iterator that iterates across only * part of a text-storage object, this number isn't * necessarily 0. * @returns the numeric index in the underlying text-storage * object of the character returned by first(). * @stable ICU 2.0 */ inline int32_t startIndex(void) const; /** * Returns the numeric index in the underlying text-storage * object of the position immediately BEYOND the character * returned by last(). * @return the numeric index in the underlying text-storage * object of the position immediately BEYOND the character * returned by last(). * @stable ICU 2.0 */ inline int32_t endIndex(void) const; /** * Returns the numeric index in the underlying text-storage * object of the character the iterator currently refers to * (i.e., the character returned by current()). * @return the numberic index in the text-storage object of * the character the iterator currently refers to * @stable ICU 2.0 */ inline int32_t getIndex(void) const; /** * Returns the length of the entire text in the underlying * text-storage object. * @return the length of the entire text in the text-storage object * @stable ICU 2.0 */ inline int32_t getLength() const; /** * Moves the current position relative to the start or end of the * iteration range, or relative to the current position itself. * The movement is expressed in numbers of code units forward * or backward by specifying a positive or negative delta. * @param delta the position relative to origin. A positive delta means forward; * a negative delta means backward. * @param origin Origin enumeration {kStart, kCurrent, kEnd} * @return the new position * @stable ICU 2.0 */ virtual int32_t move(int32_t delta, EOrigin origin) = 0; /** * Moves the current position relative to the start or end of the * iteration range, or relative to the current position itself. * The movement is expressed in numbers of code points forward * or backward by specifying a positive or negative delta. * @param delta the position relative to origin. A positive delta means forward; * a negative delta means backward. * @param origin Origin enumeration {kStart, kCurrent, kEnd} * @return the new position * @stable ICU 2.0 */ #ifdef move32 // One of the system headers right now is sometimes defining a conflicting macro we don't use #undef move32 #endif virtual int32_t move32(int32_t delta, EOrigin origin) = 0; /** * Copies the text under iteration into the UnicodeString * referred to by "result". * @param result Receives a copy of the text under iteration. * @stable ICU 2.0 */ virtual void getText(UnicodeString& result) = 0; protected: /** * Empty constructor. * @stable ICU 2.0 */ CharacterIterator(); /** * Constructor, just setting the length field in this base class. * @stable ICU 2.0 */ CharacterIterator(int32_t length); /** * Constructor, just setting the length and position fields in this base class. * @stable ICU 2.0 */ CharacterIterator(int32_t length, int32_t position); /** * Constructor, just setting the length, start, end, and position fields in this base class. * @stable ICU 2.0 */ CharacterIterator(int32_t length, int32_t textBegin, int32_t textEnd, int32_t position); /** * Copy constructor. * * @param that The CharacterIterator to be copied * @stable ICU 2.0 */ CharacterIterator(const CharacterIterator &that); /** * Assignment operator. Sets this CharacterIterator to have the same behavior, * as the one passed in. * @param that The CharacterIterator passed in. * @return the newly set CharacterIterator. * @stable ICU 2.0 */ CharacterIterator &operator=(const CharacterIterator &that); /** * Base class text length field. * Necessary this for correct getText() and hashCode(). * @stable ICU 2.0 */ int32_t textLength; /** * Base class field for the current position. * @stable ICU 2.0 */ int32_t pos; /** * Base class field for the start of the iteration range. * @stable ICU 2.0 */ int32_t begin; /** * Base class field for the end of the iteration range. * @stable ICU 2.0 */ int32_t end; }; inline UBool ForwardCharacterIterator::operator!=(const ForwardCharacterIterator& that) const { return !operator==(that); } inline int32_t CharacterIterator::setToStart() { return move(0, kStart); } inline int32_t CharacterIterator::setToEnd() { return move(0, kEnd); } inline int32_t CharacterIterator::startIndex(void) const { return begin; } inline int32_t CharacterIterator::endIndex(void) const { return end; } inline int32_t CharacterIterator::getIndex(void) const { return pos; } inline int32_t CharacterIterator::getLength(void) const { return textLength; } U_NAMESPACE_END #endif