OXIESEC PANEL
- Current Dir:
/
/
opt
/
alt
/
libicu71
/
usr
/
include
/
unicode
Server IP: 2a02:4780:11:1594:0:ef5:22d7:a
Upload:
Create Dir:
Name
Size
Modified
Perms
📁
..
-
05/14/2024 03:54:45 PM
rwxr-xr-x
📄
alphaindex.h
26.52 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
appendable.h
8.53 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
basictz.h
9.86 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
brkiter.h
27.81 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
bytestream.h
10.75 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
bytestrie.h
20.77 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
bytestriebuilder.h
7.46 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
calendar.h
105.88 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
caniter.h
7.44 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
casemap.h
25.33 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
char16ptr.h
7.22 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
chariter.h
24.06 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
choicfmt.h
23.97 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
coleitr.h
13.77 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
coll.h
56.26 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
compactdecimalformat.h
6.88 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
curramt.h
3.78 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
currpinf.h
7.3 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
currunit.h
4.02 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
datefmt.h
40.71 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
dbbi.h
1.19 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
dcfmtsym.h
20.61 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
decimfmt.h
87.57 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
docmain.h
7.21 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
dtfmtsym.h
38.21 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
dtintrv.h
3.84 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
dtitvfmt.h
49.27 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
dtitvinf.h
18.63 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
dtptngen.h
28.84 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
dtrule.h
8.69 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
edits.h
20.74 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
enumset.h
2.08 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
errorcode.h
4.84 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
fieldpos.h
8.7 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
filteredbrk.h
5.37 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
fmtable.h
24.43 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
format.h
12.5 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
formattedvalue.h
9.75 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
fpositer.h
3.03 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
gender.h
3.33 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
gregocal.h
31.88 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
icudataver.h
1.02 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
icuplug.h
11.83 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
idna.h
12.7 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
listformatter.h
8.79 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
localebuilder.h
11.08 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
localematcher.h
26.84 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
localpointer.h
19.36 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
locdspnm.h
7.12 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
locid.h
47.66 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
measfmt.h
11.42 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
measunit.h
104.97 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
measure.h
4.33 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
messagepattern.h
33.71 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
msgfmt.h
44.22 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
normalizer2.h
33.66 KB
12/01/2022 11:12:32 AM
rw-r--r--
📄
normlzr.h
30.95 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
nounit.h
2.25 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
numberformatter.h
94.15 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
numberrangeformatter.h
25.19 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
numfmt.h
49.84 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
numsys.h
7.33 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
parseerr.h
3.08 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
parsepos.h
5.57 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
platform.h
28.09 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
plurfmt.h
25.24 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
plurrule.h
20.82 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ptypes.h
3.49 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
putil.h
6.32 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
rbbi.h
28.64 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
rbnf.h
48.85 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
rbtz.h
15.77 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
regex.h
84.36 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
region.h
9.18 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
reldatefmt.h
22.22 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
rep.h
9.37 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
resbund.h
18.08 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
schriter.h
6.36 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
scientificnumberformatter.h
6.43 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
search.h
22.22 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
selfmt.h
14.34 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
simpleformatter.h
12.59 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
simpletz.h
45.64 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
smpdtfmt.h
71.07 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
sortkey.h
11.18 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
std_string.h
1.05 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
strenum.h
9.92 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
stringoptions.h
5.79 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
stringpiece.h
10.05 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
stringtriebuilder.h
15.47 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
stsearch.h
21.41 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
symtable.h
4.27 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tblcoll.h
36.92 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
timezone.h
43.81 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tmunit.h
3.4 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tmutamt.h
4.91 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tmutfmt.h
7.42 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
translit.h
65.82 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tzfmt.h
42.93 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tznames.h
16.85 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tzrule.h
35.6 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
tztrans.h
6.13 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ubidi.h
89.61 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ubiditransform.h
12.7 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ubrk.h
24.43 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucal.h
60.68 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucasemap.h
15.21 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucat.h
5.35 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uchar.h
144.64 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucharstrie.h
22.53 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucharstriebuilder.h
7.47 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uchriter.h
13.42 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uclean.h
11.2 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucnv.h
83.46 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucnv_cb.h
6.58 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucnv_err.h
20.98 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucnvsel.h
6.19 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucol.h
61.95 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucoleitr.h
9.82 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uconfig.h
12.07 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucpmap.h
5.53 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucptrie.h
22.5 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucsdet.h
14.69 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ucurr.h
16.72 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
udat.h
62.4 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
udata.h
15.63 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
udateintervalformat.h
11.93 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
udatpg.h
30.18 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
udisplaycontext.h
5.94 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uenum.h
7.79 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ufieldpositer.h
4.41 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uformattable.h
10.97 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uformattedvalue.h
12.3 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ugender.h
2.06 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uidna.h
33.43 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uiter.h
22.75 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uldnames.h
10.48 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ulistformatter.h
10.78 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uloc.h
54.63 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ulocdata.h
11.3 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
umachine.h
16.05 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
umisc.h
1.33 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
umsg.h
24.25 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
umutablecptrie.h
8.29 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unifilt.h
4 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unifunct.h
4.05 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unimatch.h
6.1 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unirepl.h
3.38 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uniset.h
66.46 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unistr.h
170.53 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unorm.h
20.55 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unorm2.h
24.68 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unounclass.h
735 bytes
12/01/2022 11:12:33 AM
rw-r--r--
📄
unum.h
55.31 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unumberformatter.h
30.28 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unumberrangeformatter.h
15.37 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
unumsys.h
7.26 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uobject.h
10.68 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
upluralrules.h
8.79 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uregex.h
71.99 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uregion.h
9.81 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ureldatefmt.h
17.04 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
urename.h
135 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
urep.h
5.38 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ures.h
36.54 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uscript.h
27.63 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
usearch.h
39.21 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uset.h
44.29 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
usetiter.h
9.68 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ushape.h
18 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uspoof.h
65.84 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
usprep.h
8.19 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ustdio.h
38.56 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ustream.h
1.89 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ustring.h
72.36 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
ustringtrie.h
3.15 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utext.h
58.08 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utf.h
7.87 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utf16.h
23.35 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utf32.h
763 bytes
12/01/2022 11:12:33 AM
rw-r--r--
📄
utf8.h
30.83 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utf_old.h
45.83 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utmscale.h
13.78 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utrace.h
17.18 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utrans.h
25.54 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
utypes.h
31.06 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uvernum.h
6.68 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
uversion.h
5.99 KB
12/01/2022 11:12:33 AM
rw-r--r--
📄
vtzone.h
20.74 KB
12/01/2022 11:12:33 AM
rw-r--r--
Editing: utf16.h
Close
// © 2016 and later: Unicode, Inc. and others. // License & terms of use: http://www.unicode.org/copyright.html /* ******************************************************************************* * * Copyright (C) 1999-2012, International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************************* * file name: utf16.h * encoding: UTF-8 * tab size: 8 (not used) * indentation:4 * * created on: 1999sep09 * created by: Markus W. Scherer */ /** * \file * \brief C API: 16-bit Unicode handling macros * * This file defines macros to deal with 16-bit Unicode (UTF-16) code units and strings. * * For more information see utf.h and the ICU User Guide Strings chapter * (https://unicode-org.github.io/icu/userguide/strings). * * <em>Usage:</em> * ICU coding guidelines for if() statements should be followed when using these macros. * Compound statements (curly braces {}) must be used for if-else-while... * bodies and all macro statements should be terminated with semicolon. */ #ifndef __UTF16_H__ #define __UTF16_H__ #include <stdbool.h> #include "unicode/umachine.h" #ifndef __UTF_H__ # include "unicode/utf.h" #endif /* single-code point definitions -------------------------------------------- */ /** * Does this code unit alone encode a code point (BMP, not a surrogate)? * @param c 16-bit code unit * @return true or false * @stable ICU 2.4 */ #define U16_IS_SINGLE(c) !U_IS_SURROGATE(c) /** * Is this code unit a lead surrogate (U+d800..U+dbff)? * @param c 16-bit code unit * @return true or false * @stable ICU 2.4 */ #define U16_IS_LEAD(c) (((c)&0xfffffc00)==0xd800) /** * Is this code unit a trail surrogate (U+dc00..U+dfff)? * @param c 16-bit code unit * @return true or false * @stable ICU 2.4 */ #define U16_IS_TRAIL(c) (((c)&0xfffffc00)==0xdc00) /** * Is this code unit a surrogate (U+d800..U+dfff)? * @param c 16-bit code unit * @return true or false * @stable ICU 2.4 */ #define U16_IS_SURROGATE(c) U_IS_SURROGATE(c) /** * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)), * is it a lead surrogate? * @param c 16-bit code unit * @return true or false * @stable ICU 2.4 */ #define U16_IS_SURROGATE_LEAD(c) (((c)&0x400)==0) /** * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)), * is it a trail surrogate? * @param c 16-bit code unit * @return true or false * @stable ICU 4.2 */ #define U16_IS_SURROGATE_TRAIL(c) (((c)&0x400)!=0) /** * Helper constant for U16_GET_SUPPLEMENTARY. * @internal */ #define U16_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) /** * Get a supplementary code point value (U+10000..U+10ffff) * from its lead and trail surrogates. * The result is undefined if the input values are not * lead and trail surrogates. * * @param lead lead surrogate (U+d800..U+dbff) * @param trail trail surrogate (U+dc00..U+dfff) * @return supplementary code point (U+10000..U+10ffff) * @stable ICU 2.4 */ #define U16_GET_SUPPLEMENTARY(lead, trail) \ (((UChar32)(lead)<<10UL)+(UChar32)(trail)-U16_SURROGATE_OFFSET) /** * Get the lead surrogate (0xd800..0xdbff) for a * supplementary code point (0x10000..0x10ffff). * @param supplementary 32-bit code point (U+10000..U+10ffff) * @return lead surrogate (U+d800..U+dbff) for supplementary * @stable ICU 2.4 */ #define U16_LEAD(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) /** * Get the trail surrogate (0xdc00..0xdfff) for a * supplementary code point (0x10000..0x10ffff). * @param supplementary 32-bit code point (U+10000..U+10ffff) * @return trail surrogate (U+dc00..U+dfff) for supplementary * @stable ICU 2.4 */ #define U16_TRAIL(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00) /** * How many 16-bit code units are used to encode this Unicode code point? (1 or 2) * The result is not defined if c is not a Unicode code point (U+0000..U+10ffff). * @param c 32-bit code point * @return 1 or 2 * @stable ICU 2.4 */ #define U16_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) /** * The maximum number of 16-bit code units per Unicode code point (U+0000..U+10ffff). * @return 2 * @stable ICU 2.4 */ #define U16_MAX_LENGTH 2 /** * Get a code point from a string at a random-access offset, * without changing the offset. * "Unsafe" macro, assumes well-formed UTF-16. * * The offset may point to either the lead or trail surrogate unit * for a supplementary code point, in which case the macro will read * the adjacent matching surrogate as well. * The result is undefined if the offset points to a single, unpaired surrogate. * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT. * * @param s const UChar * string * @param i string offset * @param c output UChar32 variable * @see U16_GET * @stable ICU 2.4 */ #define U16_GET_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[i]; \ if(U16_IS_SURROGATE(c)) { \ if(U16_IS_SURROGATE_LEAD(c)) { \ (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)+1]); \ } else { \ (c)=U16_GET_SUPPLEMENTARY((s)[(i)-1], (c)); \ } \ } \ } UPRV_BLOCK_MACRO_END /** * Get a code point from a string at a random-access offset, * without changing the offset. * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The offset may point to either the lead or trail surrogate unit * for a supplementary code point, in which case the macro will read * the adjacent matching surrogate as well. * * The length can be negative for a NUL-terminated string. * * If the offset points to a single, unpaired surrogate, then * c is set to that unpaired surrogate. * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT. * * @param s const UChar * string * @param start starting string offset (usually 0) * @param i string offset, must be start<=i<length * @param length string length * @param c output UChar32 variable * @see U16_GET_UNSAFE * @stable ICU 2.4 */ #define U16_GET(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[i]; \ if(U16_IS_SURROGATE(c)) { \ uint16_t __c2; \ if(U16_IS_SURROGATE_LEAD(c)) { \ if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \ (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ } \ } else { \ if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ } \ } \ } \ } UPRV_BLOCK_MACRO_END /** * Get a code point from a string at a random-access offset, * without changing the offset. * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The offset may point to either the lead or trail surrogate unit * for a supplementary code point, in which case the macro will read * the adjacent matching surrogate as well. * * The length can be negative for a NUL-terminated string. * * If the offset points to a single, unpaired surrogate, then * c is set to U+FFFD. * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT_OR_FFFD. * * @param s const UChar * string * @param start starting string offset (usually 0) * @param i string offset, must be start<=i<length * @param length string length * @param c output UChar32 variable * @see U16_GET_UNSAFE * @stable ICU 60 */ #define U16_GET_OR_FFFD(s, start, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[i]; \ if(U16_IS_SURROGATE(c)) { \ uint16_t __c2; \ if(U16_IS_SURROGATE_LEAD(c)) { \ if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \ (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ } else { \ (c)=0xfffd; \ } \ } else { \ if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ } else { \ (c)=0xfffd; \ } \ } \ } \ } UPRV_BLOCK_MACRO_END /* definitions with forward iteration --------------------------------------- */ /** * Get a code point from a string at a code point boundary offset, * and advance the offset to the next code point boundary. * (Post-incrementing forward iteration.) * "Unsafe" macro, assumes well-formed UTF-16. * * The offset may point to the lead surrogate unit * for a supplementary code point, in which case the macro will read * the following trail surrogate as well. * If the offset points to a trail surrogate, then that itself * will be returned as the code point. * The result is undefined if the offset points to a single, unpaired lead surrogate. * * @param s const UChar * string * @param i string offset * @param c output UChar32 variable * @see U16_NEXT * @stable ICU 2.4 */ #define U16_NEXT_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[(i)++]; \ if(U16_IS_LEAD(c)) { \ (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)++]); \ } \ } UPRV_BLOCK_MACRO_END /** * Get a code point from a string at a code point boundary offset, * and advance the offset to the next code point boundary. * (Post-incrementing forward iteration.) * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The length can be negative for a NUL-terminated string. * * The offset may point to the lead surrogate unit * for a supplementary code point, in which case the macro will read * the following trail surrogate as well. * If the offset points to a trail surrogate or * to a single, unpaired lead surrogate, then c is set to that unpaired surrogate. * * @param s const UChar * string * @param i string offset, must be i<length * @param length string length * @param c output UChar32 variable * @see U16_NEXT_UNSAFE * @stable ICU 2.4 */ #define U16_NEXT(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[(i)++]; \ if(U16_IS_LEAD(c)) { \ uint16_t __c2; \ if((i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \ ++(i); \ (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ } \ } \ } UPRV_BLOCK_MACRO_END /** * Get a code point from a string at a code point boundary offset, * and advance the offset to the next code point boundary. * (Post-incrementing forward iteration.) * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The length can be negative for a NUL-terminated string. * * The offset may point to the lead surrogate unit * for a supplementary code point, in which case the macro will read * the following trail surrogate as well. * If the offset points to a trail surrogate or * to a single, unpaired lead surrogate, then c is set to U+FFFD. * * @param s const UChar * string * @param i string offset, must be i<length * @param length string length * @param c output UChar32 variable * @see U16_NEXT_UNSAFE * @stable ICU 60 */ #define U16_NEXT_OR_FFFD(s, i, length, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[(i)++]; \ if(U16_IS_SURROGATE(c)) { \ uint16_t __c2; \ if(U16_IS_SURROGATE_LEAD(c) && (i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \ ++(i); \ (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ } else { \ (c)=0xfffd; \ } \ } \ } UPRV_BLOCK_MACRO_END /** * Append a code point to a string, overwriting 1 or 2 code units. * The offset points to the current end of the string contents * and is advanced (post-increment). * "Unsafe" macro, assumes a valid code point and sufficient space in the string. * Otherwise, the result is undefined. * * @param s const UChar * string buffer * @param i string offset * @param c code point to append * @see U16_APPEND * @stable ICU 2.4 */ #define U16_APPEND_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ if((uint32_t)(c)<=0xffff) { \ (s)[(i)++]=(uint16_t)(c); \ } else { \ (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ } \ } UPRV_BLOCK_MACRO_END /** * Append a code point to a string, overwriting 1 or 2 code units. * The offset points to the current end of the string contents * and is advanced (post-increment). * "Safe" macro, checks for a valid code point. * If a surrogate pair is written, checks for sufficient space in the string. * If the code point is not valid or a trail surrogate does not fit, * then isError is set to true. * * @param s const UChar * string buffer * @param i string offset, must be i<capacity * @param capacity size of the string buffer * @param c code point to append * @param isError output UBool set to true if an error occurs, otherwise not modified * @see U16_APPEND_UNSAFE * @stable ICU 2.4 */ #define U16_APPEND(s, i, capacity, c, isError) UPRV_BLOCK_MACRO_BEGIN { \ if((uint32_t)(c)<=0xffff) { \ (s)[(i)++]=(uint16_t)(c); \ } else if((uint32_t)(c)<=0x10ffff && (i)+1<(capacity)) { \ (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ } else /* c>0x10ffff or not enough space */ { \ (isError)=true; \ } \ } UPRV_BLOCK_MACRO_END /** * Advance the string offset from one code point boundary to the next. * (Post-incrementing iteration.) * "Unsafe" macro, assumes well-formed UTF-16. * * @param s const UChar * string * @param i string offset * @see U16_FWD_1 * @stable ICU 2.4 */ #define U16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_LEAD((s)[(i)++])) { \ ++(i); \ } \ } UPRV_BLOCK_MACRO_END /** * Advance the string offset from one code point boundary to the next. * (Post-incrementing iteration.) * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The length can be negative for a NUL-terminated string. * * @param s const UChar * string * @param i string offset, must be i<length * @param length string length * @see U16_FWD_1_UNSAFE * @stable ICU 2.4 */ #define U16_FWD_1(s, i, length) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_LEAD((s)[(i)++]) && (i)!=(length) && U16_IS_TRAIL((s)[i])) { \ ++(i); \ } \ } UPRV_BLOCK_MACRO_END /** * Advance the string offset from one code point boundary to the n-th next one, * i.e., move forward by n code points. * (Post-incrementing iteration.) * "Unsafe" macro, assumes well-formed UTF-16. * * @param s const UChar * string * @param i string offset * @param n number of code points to skip * @see U16_FWD_N * @stable ICU 2.4 */ #define U16_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ int32_t __N=(n); \ while(__N>0) { \ U16_FWD_1_UNSAFE(s, i); \ --__N; \ } \ } UPRV_BLOCK_MACRO_END /** * Advance the string offset from one code point boundary to the n-th next one, * i.e., move forward by n code points. * (Post-incrementing iteration.) * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The length can be negative for a NUL-terminated string. * * @param s const UChar * string * @param i int32_t string offset, must be i<length * @param length int32_t string length * @param n number of code points to skip * @see U16_FWD_N_UNSAFE * @stable ICU 2.4 */ #define U16_FWD_N(s, i, length, n) UPRV_BLOCK_MACRO_BEGIN { \ int32_t __N=(n); \ while(__N>0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \ U16_FWD_1(s, i, length); \ --__N; \ } \ } UPRV_BLOCK_MACRO_END /** * Adjust a random-access offset to a code point boundary * at the start of a code point. * If the offset points to the trail surrogate of a surrogate pair, * then the offset is decremented. * Otherwise, it is not modified. * "Unsafe" macro, assumes well-formed UTF-16. * * @param s const UChar * string * @param i string offset * @see U16_SET_CP_START * @stable ICU 2.4 */ #define U16_SET_CP_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_TRAIL((s)[i])) { \ --(i); \ } \ } UPRV_BLOCK_MACRO_END /** * Adjust a random-access offset to a code point boundary * at the start of a code point. * If the offset points to the trail surrogate of a surrogate pair, * then the offset is decremented. * Otherwise, it is not modified. * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * @param s const UChar * string * @param start starting string offset (usually 0) * @param i string offset, must be start<=i * @see U16_SET_CP_START_UNSAFE * @stable ICU 2.4 */ #define U16_SET_CP_START(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_TRAIL((s)[i]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \ --(i); \ } \ } UPRV_BLOCK_MACRO_END /* definitions with backward iteration -------------------------------------- */ /** * Move the string offset from one code point boundary to the previous one * and get the code point between them. * (Pre-decrementing backward iteration.) * "Unsafe" macro, assumes well-formed UTF-16. * * The input offset may be the same as the string length. * If the offset is behind a trail surrogate unit * for a supplementary code point, then the macro will read * the preceding lead surrogate as well. * If the offset is behind a lead surrogate, then that itself * will be returned as the code point. * The result is undefined if the offset is behind a single, unpaired trail surrogate. * * @param s const UChar * string * @param i string offset * @param c output UChar32 variable * @see U16_PREV * @stable ICU 2.4 */ #define U16_PREV_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[--(i)]; \ if(U16_IS_TRAIL(c)) { \ (c)=U16_GET_SUPPLEMENTARY((s)[--(i)], (c)); \ } \ } UPRV_BLOCK_MACRO_END /** * Move the string offset from one code point boundary to the previous one * and get the code point between them. * (Pre-decrementing backward iteration.) * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The input offset may be the same as the string length. * If the offset is behind a trail surrogate unit * for a supplementary code point, then the macro will read * the preceding lead surrogate as well. * If the offset is behind a lead surrogate or behind a single, unpaired * trail surrogate, then c is set to that unpaired surrogate. * * @param s const UChar * string * @param start starting string offset (usually 0) * @param i string offset, must be start<i * @param c output UChar32 variable * @see U16_PREV_UNSAFE * @stable ICU 2.4 */ #define U16_PREV(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[--(i)]; \ if(U16_IS_TRAIL(c)) { \ uint16_t __c2; \ if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ --(i); \ (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ } \ } \ } UPRV_BLOCK_MACRO_END /** * Move the string offset from one code point boundary to the previous one * and get the code point between them. * (Pre-decrementing backward iteration.) * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The input offset may be the same as the string length. * If the offset is behind a trail surrogate unit * for a supplementary code point, then the macro will read * the preceding lead surrogate as well. * If the offset is behind a lead surrogate or behind a single, unpaired * trail surrogate, then c is set to U+FFFD. * * @param s const UChar * string * @param start starting string offset (usually 0) * @param i string offset, must be start<i * @param c output UChar32 variable * @see U16_PREV_UNSAFE * @stable ICU 60 */ #define U16_PREV_OR_FFFD(s, start, i, c) UPRV_BLOCK_MACRO_BEGIN { \ (c)=(s)[--(i)]; \ if(U16_IS_SURROGATE(c)) { \ uint16_t __c2; \ if(U16_IS_SURROGATE_TRAIL(c) && (i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ --(i); \ (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ } else { \ (c)=0xfffd; \ } \ } \ } UPRV_BLOCK_MACRO_END /** * Move the string offset from one code point boundary to the previous one. * (Pre-decrementing backward iteration.) * The input offset may be the same as the string length. * "Unsafe" macro, assumes well-formed UTF-16. * * @param s const UChar * string * @param i string offset * @see U16_BACK_1 * @stable ICU 2.4 */ #define U16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_TRAIL((s)[--(i)])) { \ --(i); \ } \ } UPRV_BLOCK_MACRO_END /** * Move the string offset from one code point boundary to the previous one. * (Pre-decrementing backward iteration.) * The input offset may be the same as the string length. * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * @param s const UChar * string * @param start starting string offset (usually 0) * @param i string offset, must be start<i * @see U16_BACK_1_UNSAFE * @stable ICU 2.4 */ #define U16_BACK_1(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_TRAIL((s)[--(i)]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \ --(i); \ } \ } UPRV_BLOCK_MACRO_END /** * Move the string offset from one code point boundary to the n-th one before it, * i.e., move backward by n code points. * (Pre-decrementing backward iteration.) * The input offset may be the same as the string length. * "Unsafe" macro, assumes well-formed UTF-16. * * @param s const UChar * string * @param i string offset * @param n number of code points to skip * @see U16_BACK_N * @stable ICU 2.4 */ #define U16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \ int32_t __N=(n); \ while(__N>0) { \ U16_BACK_1_UNSAFE(s, i); \ --__N; \ } \ } UPRV_BLOCK_MACRO_END /** * Move the string offset from one code point boundary to the n-th one before it, * i.e., move backward by n code points. * (Pre-decrementing backward iteration.) * The input offset may be the same as the string length. * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * @param s const UChar * string * @param start start of string * @param i string offset, must be start<i * @param n number of code points to skip * @see U16_BACK_N_UNSAFE * @stable ICU 2.4 */ #define U16_BACK_N(s, start, i, n) UPRV_BLOCK_MACRO_BEGIN { \ int32_t __N=(n); \ while(__N>0 && (i)>(start)) { \ U16_BACK_1(s, start, i); \ --__N; \ } \ } UPRV_BLOCK_MACRO_END /** * Adjust a random-access offset to a code point boundary after a code point. * If the offset is behind the lead surrogate of a surrogate pair, * then the offset is incremented. * Otherwise, it is not modified. * The input offset may be the same as the string length. * "Unsafe" macro, assumes well-formed UTF-16. * * @param s const UChar * string * @param i string offset * @see U16_SET_CP_LIMIT * @stable ICU 2.4 */ #define U16_SET_CP_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \ if(U16_IS_LEAD((s)[(i)-1])) { \ ++(i); \ } \ } UPRV_BLOCK_MACRO_END /** * Adjust a random-access offset to a code point boundary after a code point. * If the offset is behind the lead surrogate of a surrogate pair, * then the offset is incremented. * Otherwise, it is not modified. * The input offset may be the same as the string length. * "Safe" macro, handles unpaired surrogates and checks for string boundaries. * * The length can be negative for a NUL-terminated string. * * @param s const UChar * string * @param start int32_t starting string offset (usually 0) * @param i int32_t string offset, start<=i<=length * @param length int32_t string length * @see U16_SET_CP_LIMIT_UNSAFE * @stable ICU 2.4 */ #define U16_SET_CP_LIMIT(s, start, i, length) UPRV_BLOCK_MACRO_BEGIN { \ if((start)<(i) && ((i)<(length) || (length)<0) && U16_IS_LEAD((s)[(i)-1]) && U16_IS_TRAIL((s)[i])) { \ ++(i); \ } \ } UPRV_BLOCK_MACRO_END #endif