trunk/doc/pcre16.3

.TH PCRE 3
.SH NAME
PCRE - Perl-compatible regular expressions
.sp
.B #include <pcre.h>
.
.
.SH "PCRE 16-BIT API BASIC FUNCTIONS"
.rs
.sp
.SM
.B pcre16 *pcre16_compile(PCRE_SPTR16 \fIpattern\fP, int \fIoptions\fP,
.ti +5n
.B const char **\fIerrptr\fP, int *\fIerroffset\fP,
.ti +5n
.B const unsigned char *\fItableptr\fP);
.PP
.B pcre16 *pcre16_compile2(PCRE_SPTR16 \fIpattern\fP, int \fIoptions\fP,
.ti +5n
.B int *\fIerrorcodeptr\fP,
.ti +5n
.B const char **\fIerrptr\fP, int *\fIerroffset\fP,
.ti +5n
.B const unsigned char *\fItableptr\fP);
.PP
.B pcre16_extra *pcre16_study(const pcre16 *\fIcode\fP, int \fIoptions\fP,
.ti +5n
.B const char **\fIerrptr\fP);
.PP
.B void pcre16_free_study(pcre16_extra *\fIextra\fP);
.PP
.B int pcre16_exec(const pcre16 *\fIcode\fP, "const pcre16_extra *\fIextra\fP,"
.ti +5n
.B "PCRE_SPTR16 \fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
.ti +5n
.B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
.PP
.B int pcre16_dfa_exec(const pcre16 *\fIcode\fP, "const pcre16_extra *\fIextra\fP,"
.ti +5n
.B "PCRE_SPTR16 \fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
.ti +5n
.B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
.ti +5n
.B int *\fIworkspace\fP, int \fIwscount\fP);
.
.
.SH "PCRE 16-BIT API STRING EXTRACTION FUNCTIONS"
.rs
.sp
.B int pcre16_copy_named_substring(const pcre16 *\fIcode\fP,
.ti +5n
.B PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
.ti +5n
.B int \fIstringcount\fP, PCRE_SPTR16 \fIstringname\fP,
.ti +5n
.B PCRE_SCHAR16 *\fIbuffer\fP, int \fIbuffersize\fP);
.PP
.B int pcre16_copy_substring(PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
.ti +5n
.B int \fIstringcount\fP, int \fIstringnumber\fP, PCRE_SCHAR16 *\fIbuffer\fP,
.ti +5n
.B int \fIbuffersize\fP);
.PP
.B int pcre16_get_named_substring(const pcre16 *\fIcode\fP,
.ti +5n
.B PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
.ti +5n
.B int \fIstringcount\fP, PCRE_SPTR16 \fIstringname\fP,
.ti +5n
.B PCRE_SPTR16 *\fIstringptr\fP);
.PP
.B int pcre16_get_stringnumber(const pcre16 *\fIcode\fP,
.ti +5n
.B PCRE_SPTR16 \fIname\fP);
.PP
.B int pcre16_get_stringtable_entries(const pcre16 *\fIcode\fP,
.ti +5n
.B PCRE_SPTR16 \fIname\fP, PCRE_SCHAR16 **\fIfirst\fP, PCRE_SCHAR16 **\fIlast\fP);
.PP
.B int pcre16_get_substring(PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
.ti +5n
.B int \fIstringcount\fP, int \fIstringnumber\fP,
.ti +5n
.B PCRE_SPTR16 *\fIstringptr\fP);
.PP
.B int pcre16_get_substring_list(PCRE_SPTR16 \fIsubject\fP,
.ti +5n
.B int *\fIovector\fP, int \fIstringcount\fP, "PCRE_SPTR16 **\fIlistptr\fP);"
.PP
.B void pcre16_free_substring(PCRE_SPTR16 \fIstringptr\fP);
.PP
.B void pcre16_free_substring_list(PCRE_SPTR16 *\fIstringptr\fP);
.
.
.SH "PCRE 16-BIT API AUXILIARY FUNCTIONS"
.rs
.sp
.B pcre16_jit_stack *pcre16_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
.PP
.B void pcre16_jit_stack_free(pcre16_jit_stack *\fIstack\fP);
.PP
.B void pcre16_assign_jit_stack(pcre16_extra *\fIextra\fP,
.ti +5n
.B pcre16_jit_callback \fIcallback\fP, void *\fIdata\fP);
.PP
.B const unsigned char *pcre16_maketables(void);
.PP
.B int pcre16_fullinfo(const pcre16 *\fIcode\fP, "const pcre16_extra *\fIextra\fP,"
.ti +5n
.B int \fIwhat\fP, void *\fIwhere\fP);
.PP
.B int pcre16_refcount(pcre16 *\fIcode\fP, int \fIadjust\fP);
.PP
.B int pcre16_config(int \fIwhat\fP, void *\fIwhere\fP);
.PP
.B const char *pcre16_version(void);
.PP
.B int pcre16_pattern_to_host_byte_order(pcre16 *\fIcode\fP,
.ti +5n
.B pcre16_extra *\fIextra\fP, const unsigned char *\fItables\fP);
.
.
.SH "PCRE 16-BIT API INDIRECTED FUNCTIONS"
.rs
.sp
.B void *(*pcre16_malloc)(size_t);
.PP
.B void (*pcre16_free)(void *);
.PP
.B void *(*pcre16_stack_malloc)(size_t);
.PP
.B void (*pcre16_stack_free)(void *);
.PP
.B int (*pcre16_callout)(pcre16_callout_block *);
.
.
.SH "PCRE 16-BIT API 16-BIT-ONLY FUNCTION"
.rs
.sp
.B int pcre16_utf16_to_host_byte_order(PCRE_SCHAR16 *\fIoutput\fP,
.ti +5n
.B PCRE_SPTR16 \fIinput\fP, int \fIlength\fP, int *\fIbyte_order\fP, 
.ti +5n
.B int \fIkeep_boms\fP);
.
.
.SH "THE PCRE 16-BIT LIBRARY"
.rs
.sp
Starting with release 8.30, it is possible to compile a PCRE library that
supports 16-bit character strings, including UTF-16 strings, as well as or
instead of the original 8-bit library. The majority of the work to make this
possible was done by Zoltan Herczeg. The two libraries contain identical sets
of functions, used in exactly the same way. Only the names of the functions and
the data types of their arguments and results are different. To avoid
over-complication and reduce the documentation maintenance load, most of the
PCRE documentation describes the 8-bit library, with only occasional references
to the 16-bit library. This page describes what is different when you use the
16-bit library.
.P
WARNING: A single application can be linked with both libraries, but you must 
take care when processing any particular pattern to use functions from just one 
library. For example, if you want to study a pattern that was compiled with
\fBpcre16_compile()\fP, you must do so with \fBpcre16_study()\fP, not
\fBpcre_study()\fP, and you must free the study data with
\fBpcre16_free_study()\fP.
.
.
.SH "THE HEADER FILE"
.rs
.sp
There is only one header file, \fBpcre.h\fP. It contains prototypes for all the 
functions in both libraries, as well as definitions of flags, structures, error
codes, etc.
.
.
.SH "THE LIBRARY NAME"
.rs
.sp
In Unix-like systems, the 16-bit library is called \fBlibpcre16\fP, and can 
normally be accesss by adding \fB-lpcre16\fP to the command for linking an 
application that uses PCRE.
.
.
.SH "STRING TYPES"
.rs
.sp
In the 8-bit library, strings are passed to PCRE library functions as vectors 
of bytes with the C type "char *". In the 16-bit library, strings are passed as 
vectors of unsigned 16-bit quantities. The macro PCRE_SCHAR16 specifies an 
appropriate data type, and PCRE_SPTR16 is defined as "const PCRE_SCHAR16 *". In 
very many environments, "short int" is a 16-bit data type. When PCRE is built, 
it defines PCRE_SCHAR16 as "short int", but checks that it really is a 16-bit 
data type. If it is not, the build fails with an error message telling the 
maintainer to modify the definition appropriately.
.
.
.SH "STRUCTURE TYPES"
.rs
.sp
The types of the opaque structures that are used for compiled 16-bit patterns 
and JIT stacks are \fBpcre16\fP and \fBpcre16_jit_stack\fP respectively. The 
type of the user-accessible structure that is returned by \fBpcre16_study()\fP 
is \fBpcre16_extra\fP, and the type of the structure that is used for passing
data to a callout function is \fBpcre16_callout_block\fP. These structures 
contain the same fields, with the same names, as their 8-bit counterparts. The 
only difference is that pointers to character strings are 16-bit instead of 
8-bit types.
.
.
.SH "16-BIT FUNCTIONS"
.rs
.sp
For every function in the 8-bit library there is a corresponding function in
the 16-bit library with a name that starts with \fBpcre16_\fP instead of 
\fBpcre_\fP. The prototypes are listed above. In addition, there is one extra
function, \fBpcre16_utf16_to_host_byte_order()\fP. This is a utility function 
that converts a UTF-16 character string to host byte order if necessary. The 
other 16-bit functions expect the strings they are passed to be in host byte 
order. 
.P
The \fIlength\fP argument of \fBpcre16_utf16_to_host_byte_order()\fP specifies
the number of 16-bit data units in the input string; a negative value specifies
a zero-terminated string.
.P
If \fIbyte_order\fP is NULL, it is assumed that the string starts off in host 
byte order. This may be changed by byte-order marks (BOMs) anywhere in the
string (commonly as the first character).
.P
If \fIbyte_order\fP is not NULL, a non-zero value of the integer to which it 
points means that the input starts off in host byte order, otherwise the 
opposite order is assumed. Again, BOMs in the string can change this. The final 
byte order is passed back at the end of processing.
.P
If \fIkeep_boms\fP is non zero, byte-order mark characters (0xfeff) are copied 
into the output string. Otherwise they are discarded.
.P
The result of the function is the number of 16-bit units placed into the output
buffer, including the zero terminator if the string was zero-terminated.
.
.
.SH "SUBJECT STRING OFFSETS"
.rs
.sp
The offsets within subject strings that are returned by the matching functions 
are in 16-bit units rather than bytes.
.
.
.SH "NAMED SUBPATTERNS"
.rs
.sp
The name-to-number translation table that is maintained for named subpatterns 
uses 16-bit characters. The \fBpcre16_get_stringtable_entries()\fP function 
returns the length of each entry in the table as the number of 16-bit data 
units.
.
.
.SH "OPTION NAMES"
.rs
.sp
There are two new general option names, PCRE_UTF16 and PCRE_NO_UTF16_CHECK,
which correspond to PCRE_UTF8 and PCRE_NO_UTF8_CHECK in the 8-bit library. In
fact, these new options define the same bits in the options word.
.P
For the \fBpcre16_config()\fP function there is an option PCRE_CONFIG_UTF16 
that returns 1 if UTF-16 support is configured, otherwise 0. If this option is
given to \fBpcre_config()\fP, or if the PCRE_CONFIG_UTF8 option is given to
\fBpcre16_config()\fP, the result is the PCRE_ERROR_BADOPTION error.
.
.
.SH "CHARACTER CODES"
.rs
.sp
In 16-bit mode, when PCRE_UTF16 is not set, character values are treated in the 
same way as in 8-bit, non UTF-8 mode, except, of course, that they can range 
from 0 to 0xffff instead of 0 to 0xff. Character types for characters less than 
0xff can therefore be influenced by the locale in the same way as before. 
Characters greater than 0xff have only one case, and no "type" (such as letter 
or digit).
.P
In UTF-16 mode, the character code is Unicode, in the range 0 to 0x10ffff, with 
the exception of values in the range 0xd800 to 0xdfff because those are 
"surrogate" values that are used in pairs to encode values greater than 0xffff.
.P
A UTF-16 string can indicate its endianness by special code knows as a 
byte-order mark (BOM). The PCRE functions do not handle this, expecting strings
to be in host byte order. A utility function called
\fBpcre16_utf16_to_host_byte_order()\fP is provided to help with this (see
above).
.
.
.SH "ERROR NAMES"
.rs
.sp
The errors PCRE_ERROR_BADUTF16_OFFSET and PCRE_ERROR_SHORTUTF16 correspond to 
their 8-bit counterparts. The error PCRE_ERROR_BADMODE is given when a compiled
pattern is passed to a function that processes patterns in the other
mode, for example, if a pattern compiled with \fBpcre_compile()\fP is passed to 
\fBpcre16_exec()\fP.
.P
There are new error codes whose names begin with PCRE_UTF16_ERR for invalid
UTF-16 strings, corresponding to the PCRE_UTF8_ERR codes for UTF-8 strings that 
are described in the section entitled
.\" HTML <a href="pcreapi.html#badutf8reasons">
.\" </a>
"Reason codes for invalid UTF-8 strings"
.\"
in the main 
.\" HREF
\fBpcreapi\fP
.\"
page. The UTF-16 errors are:
.sp
  PCRE_UTF16_ERR1  Missing low surrogate at end of string
  PCRE_UTF16_ERR2  Invalid low surrogate follows high surrogate
  PCRE_UTF16_ERR3  Isolated low surrogate
  PCRE_UTF16_ERR4  Invalid character 0xfffe
.
.
.SH "ERROR TEXTS"
.rs
.sp
If there is an error while compiling a pattern, the error text that is passed 
back by \fBpcre16_compile()\fP or \fBpcre16_compile2()\fP is still an 8-bit 
character string, zero-terminated.
.
.
.SH "CALLOUTS"
.rs
.sp
The \fIsubject\fP and \fImark\fP fields in the callout block that is passed to
a callout function point to 16-bit vectors.
.
.
.SH "TESTING"
.rs
.sp
The \fBpcretest\fP program continues to operate with 8-bit input and output 
files, but it can be used for testing the 16-bit library. If it is run with the 
command line option \fB-16\fP, patterns and subject strings are converted from 
8-bit to 16-bit before being passed to PCRE, and the 16-bit library functions 
are used instead of the 8-bit ones. Returned 16-bit strings are converted to 
8-bit for output. If the 8-bit library was not compiled, \fBpcretest\fP
defaults to 16-bit and the \fB-16\fP option is ignored.
.P
When PCRE is being built, the \fBRunTest\fP script that is called by "make 
check" uses the \fBpcretest\fP \fB-C\fP option to discover which of the 8-bit
and 16-bit libraries has been built, and runs the tests appropriately.
.
.
.SH "NOT SUPPORTED IN 16-BIT MODE"
.rs
.sp
Not all the features of the 8-bit library are available with the 16-bit 
library. The C++ and POSIX wrapper functions support only the 8-bit library, 
and the \fBpcregrep\fP program is at present 8-bit only.
.
.
.SH AUTHOR
.rs
.sp
.nf
Philip Hazel
University Computing Service
Cambridge CB2 3QH, England.
.fi
.
.
.SH REVISION
.rs
.sp
.nf
Last updated: 07 January 2012
Copyright (c) 1997-2012 University of Cambridge.
.fi
1	ph10	846	.TH PCRE 3
2			.SH NAME
3			PCRE - Perl-compatible regular expressions
4	ph10	856	.sp
5			.B #include <pcre.h>
6			.
7			.
8	ph10	855	.SH "PCRE 16-BIT API BASIC FUNCTIONS"
9			.rs
10			.sp
11			.SM
12			.B pcre16 *pcre16_compile(PCRE_SPTR16 \fIpattern\fP, int \fIoptions\fP,
13			.ti +5n
14			.B const char *\fIerrptr\fP, int \fIerroffset\fP,
15			.ti +5n
16			.B const unsigned char *\fItableptr\fP);
17			.PP
18			.B pcre16 *pcre16_compile2(PCRE_SPTR16 \fIpattern\fP, int \fIoptions\fP,
19			.ti +5n
20			.B int *\fIerrorcodeptr\fP,
21			.ti +5n
22			.B const char *\fIerrptr\fP, int \fIerroffset\fP,
23			.ti +5n
24			.B const unsigned char *\fItableptr\fP);
25			.PP
26			.B pcre16_extra pcre16_study(const pcre16 \fIcode\fP, int \fIoptions\fP,
27			.ti +5n
28			.B const char **\fIerrptr\fP);
29			.PP
30			.B void pcre16_free_study(pcre16_extra *\fIextra\fP);
31			.PP
32			.B int pcre16_exec(const pcre16 \fIcode\fP, "const pcre16_extra \fIextra\fP,"
33			.ti +5n
34			.B "PCRE_SPTR16 \fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
35			.ti +5n
36			.B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
37			.PP
38			.B int pcre16_dfa_exec(const pcre16 \fIcode\fP, "const pcre16_extra \fIextra\fP,"
39			.ti +5n
40			.B "PCRE_SPTR16 \fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
41			.ti +5n
42			.B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
43			.ti +5n
44			.B int *\fIworkspace\fP, int \fIwscount\fP);
45	ph10	856	.
46			.
47			.SH "PCRE 16-BIT API STRING EXTRACTION FUNCTIONS"
48			.rs
49			.sp
50	ph10	855	.B int pcre16_copy_named_substring(const pcre16 *\fIcode\fP,
51			.ti +5n
52			.B PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
53			.ti +5n
54			.B int \fIstringcount\fP, PCRE_SPTR16 \fIstringname\fP,
55			.ti +5n
56			.B PCRE_SCHAR16 *\fIbuffer\fP, int \fIbuffersize\fP);
57			.PP
58			.B int pcre16_copy_substring(PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
59			.ti +5n
60			.B int \fIstringcount\fP, int \fIstringnumber\fP, PCRE_SCHAR16 *\fIbuffer\fP,
61			.ti +5n
62			.B int \fIbuffersize\fP);
63			.PP
64			.B int pcre16_get_named_substring(const pcre16 *\fIcode\fP,
65			.ti +5n
66			.B PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
67			.ti +5n
68			.B int \fIstringcount\fP, PCRE_SPTR16 \fIstringname\fP,
69			.ti +5n
70			.B PCRE_SPTR16 *\fIstringptr\fP);
71			.PP
72			.B int pcre16_get_stringnumber(const pcre16 *\fIcode\fP,
73			.ti +5n
74			.B PCRE_SPTR16 \fIname\fP);
75			.PP
76			.B int pcre16_get_stringtable_entries(const pcre16 *\fIcode\fP,
77			.ti +5n
78			.B PCRE_SPTR16 \fIname\fP, PCRE_SCHAR16 \fIfirst\fP, PCRE_SCHAR16 \fIlast\fP);
79			.PP
80			.B int pcre16_get_substring(PCRE_SPTR16 \fIsubject\fP, int *\fIovector\fP,
81			.ti +5n
82			.B int \fIstringcount\fP, int \fIstringnumber\fP,
83			.ti +5n
84			.B PCRE_SPTR16 *\fIstringptr\fP);
85			.PP
86			.B int pcre16_get_substring_list(PCRE_SPTR16 \fIsubject\fP,
87			.ti +5n
88			.B int \fIovector\fP, int \fIstringcount\fP, "PCRE_SPTR16 *\fIlistptr\fP);"
89			.PP
90			.B void pcre16_free_substring(PCRE_SPTR16 \fIstringptr\fP);
91			.PP
92			.B void pcre16_free_substring_list(PCRE_SPTR16 *\fIstringptr\fP);
93	ph10	856	.
94			.
95			.SH "PCRE 16-BIT API AUXILIARY FUNCTIONS"
96			.rs
97			.sp
98			.B pcre16_jit_stack *pcre16_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
99	ph10	855	.PP
100	ph10	856	.B void pcre16_jit_stack_free(pcre16_jit_stack *\fIstack\fP);
101			.PP
102			.B void pcre16_assign_jit_stack(pcre16_extra *\fIextra\fP,
103			.ti +5n
104			.B pcre16_jit_callback \fIcallback\fP, void *\fIdata\fP);
105			.PP
106	ph10	855	.B const unsigned char *pcre16_maketables(void);
107			.PP
108			.B int pcre16_fullinfo(const pcre16 \fIcode\fP, "const pcre16_extra \fIextra\fP,"
109			.ti +5n
110			.B int \fIwhat\fP, void *\fIwhere\fP);
111			.PP
112			.B int pcre16_refcount(pcre16 *\fIcode\fP, int \fIadjust\fP);
113			.PP
114			.B int pcre16_config(int \fIwhat\fP, void *\fIwhere\fP);
115			.PP
116			.B const char *pcre16_version(void);
117	ph10	856	.PP
118			.B int pcre16_pattern_to_host_byte_order(pcre16 *\fIcode\fP,
119			.ti +5n
120			.B pcre16_extra \fIextra\fP, const unsigned char \fItables\fP);
121	ph10	855	.
122			.
123			.SH "PCRE 16-BIT API INDIRECTED FUNCTIONS"
124			.rs
125			.sp
126			.B void (pcre16_malloc)(size_t);
127			.PP
128			.B void (pcre16_free)(void );
129			.PP
130			.B void (pcre16_stack_malloc)(size_t);
131			.PP
132			.B void (pcre16_stack_free)(void );
133			.PP
134			.B int (pcre16_callout)(pcre16_callout_block );
135			.
136			.
137			.SH "PCRE 16-BIT API 16-BIT-ONLY FUNCTION"
138			.rs
139			.sp
140			.B int pcre16_utf16_to_host_byte_order(PCRE_SCHAR16 *\fIoutput\fP,
141			.ti +5n
142			.B PCRE_SPTR16 \fIinput\fP, int \fIlength\fP, int *\fIbyte_order\fP,
143			.ti +5n
144			.B int \fIkeep_boms\fP);
145			.
146			.
147	ph10	846	.SH "THE PCRE 16-BIT LIBRARY"
148			.rs
149			.sp
150			Starting with release 8.30, it is possible to compile a PCRE library that
151			supports 16-bit character strings, including UTF-16 strings, as well as or
152			instead of the original 8-bit library. The majority of the work to make this
153			possible was done by Zoltan Herczeg. The two libraries contain identical sets
154			of functions, used in exactly the same way. Only the names of the functions and
155	ph10	855	the data types of their arguments and results are different. To avoid
156	ph10	846	over-complication and reduce the documentation maintenance load, most of the
157	ph10	855	PCRE documentation describes the 8-bit library, with only occasional references
158			to the 16-bit library. This page describes what is different when you use the
159	ph10	846	16-bit library.
160			.P
161			WARNING: A single application can be linked with both libraries, but you must
162			take care when processing any particular pattern to use functions from just one
163			library. For example, if you want to study a pattern that was compiled with
164			\fBpcre16_compile()\fP, you must do so with \fBpcre16_study()\fP, not
165			\fBpcre_study()\fP, and you must free the study data with
166			\fBpcre16_free_study()\fP.
167			.
168			.
169			.SH "THE HEADER FILE"
170			.rs
171			.sp
172			There is only one header file, \fBpcre.h\fP. It contains prototypes for all the
173	ph10	855	functions in both libraries, as well as definitions of flags, structures, error
174			codes, etc.
175	ph10	846	.
176			.
177	ph10	855	.SH "THE LIBRARY NAME"
178			.rs
179			.sp
180			In Unix-like systems, the 16-bit library is called \fBlibpcre16\fP, and can
181			normally be accesss by adding \fB-lpcre16\fP to the command for linking an
182			application that uses PCRE.
183			.
184			.
185	ph10	846	.SH "STRING TYPES"
186			.rs
187			.sp
188			In the 8-bit library, strings are passed to PCRE library functions as vectors
189			of bytes with the C type "char *". In the 16-bit library, strings are passed as
190			vectors of unsigned 16-bit quantities. The macro PCRE_SCHAR16 specifies an
191			appropriate data type, and PCRE_SPTR16 is defined as "const PCRE_SCHAR16 *". In
192			very many environments, "short int" is a 16-bit data type. When PCRE is built,
193			it defines PCRE_SCHAR16 as "short int", but checks that it really is a 16-bit
194			data type. If it is not, the build fails with an error message telling the
195			maintainer to modify the definition appropriately.
196			.
197			.
198	ph10	855	.SH "STRUCTURE TYPES"
199	ph10	846	.rs
200			.sp
201	ph10	855	The types of the opaque structures that are used for compiled 16-bit patterns
202			and JIT stacks are \fBpcre16\fP and \fBpcre16_jit_stack\fP respectively. The
203			type of the user-accessible structure that is returned by \fBpcre16_study()\fP
204			is \fBpcre16_extra\fP, and the type of the structure that is used for passing
205			data to a callout function is \fBpcre16_callout_block\fP. These structures
206			contain the same fields, with the same names, as their 8-bit counterparts. The
207			only difference is that pointers to character strings are 16-bit instead of
208			8-bit types.
209			.
210			.
211			.SH "16-BIT FUNCTIONS"
212			.rs
213			.sp
214	ph10	846	For every function in the 8-bit library there is a corresponding function in
215			the 16-bit library with a name that starts with \fBpcre16_\fP instead of
216	ph10	855	\fBpcre_\fP. The prototypes are listed above. In addition, there is one extra
217			function, \fBpcre16_utf16_to_host_byte_order()\fP. This is a utility function
218			that converts a UTF-16 character string to host byte order if necessary. The
219			other 16-bit functions expect the strings they are passed to be in host byte
220			order.
221			.P
222	ph10	856	The \fIlength\fP argument of \fBpcre16_utf16_to_host_byte_order()\fP specifies
223			the number of 16-bit data units in the input string; a negative value specifies
224			a zero-terminated string.
225	ph10	855	.P
226			If \fIbyte_order\fP is NULL, it is assumed that the string starts off in host
227			byte order. This may be changed by byte-order marks (BOMs) anywhere in the
228			string (commonly as the first character).
229			.P
230			If \fIbyte_order\fP is not NULL, a non-zero value of the integer to which it
231			points means that the input starts off in host byte order, otherwise the
232			opposite order is assumed. Again, BOMs in the string can change this. The final
233			byte order is passed back at the end of processing.
234			.P
235			If \fIkeep_boms\fP is non zero, byte-order mark characters (0xfeff) are copied
236			into the output string. Otherwise they are discarded.
237			.P
238			The result of the function is the number of 16-bit units placed into the output
239			buffer, including the zero terminator if the string was zero-terminated.
240	ph10	846	.
241			.
242			.SH "SUBJECT STRING OFFSETS"
243			.rs
244			.sp
245			The offsets within subject strings that are returned by the matching functions
246			are in 16-bit units rather than bytes.
247			.
248			.
249			.SH "NAMED SUBPATTERNS"
250			.rs
251			.sp
252			The name-to-number translation table that is maintained for named subpatterns
253			uses 16-bit characters. The \fBpcre16_get_stringtable_entries()\fP function
254			returns the length of each entry in the table as the number of 16-bit data
255	ph10	855	units.
256	ph10	846	.
257			.
258			.SH "OPTION NAMES"
259			.rs
260			.sp
261			There are two new general option names, PCRE_UTF16 and PCRE_NO_UTF16_CHECK,
262			which correspond to PCRE_UTF8 and PCRE_NO_UTF8_CHECK in the 8-bit library. In
263			fact, these new options define the same bits in the options word.
264			.P
265			For the \fBpcre16_config()\fP function there is an option PCRE_CONFIG_UTF16
266			that returns 1 if UTF-16 support is configured, otherwise 0. If this option is
267			given to \fBpcre_config()\fP, or if the PCRE_CONFIG_UTF8 option is given to
268			\fBpcre16_config()\fP, the result is the PCRE_ERROR_BADOPTION error.
269			.
270			.
271			.SH "CHARACTER CODES"
272			.rs
273			.sp
274			In 16-bit mode, when PCRE_UTF16 is not set, character values are treated in the
275			same way as in 8-bit, non UTF-8 mode, except, of course, that they can range
276	ph10	855	from 0 to 0xffff instead of 0 to 0xff. Character types for characters less than
277			0xff can therefore be influenced by the locale in the same way as before.
278			Characters greater than 0xff have only one case, and no "type" (such as letter
279	ph10	846	or digit).
280			.P
281	ph10	855	In UTF-16 mode, the character code is Unicode, in the range 0 to 0x10ffff, with
282			the exception of values in the range 0xd800 to 0xdfff because those are
283			"surrogate" values that are used in pairs to encode values greater than 0xffff.
284	ph10	846	.P
285	ph10	855	A UTF-16 string can indicate its endianness by special code knows as a
286			byte-order mark (BOM). The PCRE functions do not handle this, expecting strings
287			to be in host byte order. A utility function called
288			\fBpcre16_utf16_to_host_byte_order()\fP is provided to help with this (see
289			above).
290	ph10	846	.
291			.
292			.SH "ERROR NAMES"
293			.rs
294			.sp
295			The errors PCRE_ERROR_BADUTF16_OFFSET and PCRE_ERROR_SHORTUTF16 correspond to
296			their 8-bit counterparts. The error PCRE_ERROR_BADMODE is given when a compiled
297			pattern is passed to a function that processes patterns in the other
298			mode, for example, if a pattern compiled with \fBpcre_compile()\fP is passed to
299			\fBpcre16_exec()\fP.
300			.P
301			There are new error codes whose names begin with PCRE_UTF16_ERR for invalid
302	ph10	855	UTF-16 strings, corresponding to the PCRE_UTF8_ERR codes for UTF-8 strings that
303			are described in the section entitled
304			.\" HTML <a href="pcreapi.html#badutf8reasons">
305			.\" </a>
306			"Reason codes for invalid UTF-8 strings"
307			.\"
308			in the main
309	ph10	846	.\" HREF
310			\fBpcreapi\fP
311			.\"
312	ph10	855	page. The UTF-16 errors are:
313			.sp
314			PCRE_UTF16_ERR1 Missing low surrogate at end of string
315			PCRE_UTF16_ERR2 Invalid low surrogate follows high surrogate
316			PCRE_UTF16_ERR3 Isolated low surrogate
317			PCRE_UTF16_ERR4 Invalid character 0xfffe
318	ph10	846	.
319			.
320			.SH "ERROR TEXTS"
321			.rs
322			.sp
323			If there is an error while compiling a pattern, the error text that is passed
324			back by \fBpcre16_compile()\fP or \fBpcre16_compile2()\fP is still an 8-bit
325			character string, zero-terminated.
326			.
327			.
328			.SH "CALLOUTS"
329			.rs
330			.sp
331			The \fIsubject\fP and \fImark\fP fields in the callout block that is passed to
332			a callout function point to 16-bit vectors.
333			.
334			.
335			.SH "TESTING"
336			.rs
337			.sp
338			The \fBpcretest\fP program continues to operate with 8-bit input and output
339			files, but it can be used for testing the 16-bit library. If it is run with the
340			command line option \fB-16\fP, patterns and subject strings are converted from
341			8-bit to 16-bit before being passed to PCRE, and the 16-bit library functions
342			are used instead of the 8-bit ones. Returned 16-bit strings are converted to
343			8-bit for output. If the 8-bit library was not compiled, \fBpcretest\fP
344			defaults to 16-bit and the \fB-16\fP option is ignored.
345			.P
346			When PCRE is being built, the \fBRunTest\fP script that is called by "make
347			check" uses the \fBpcretest\fP \fB-C\fP option to discover which of the 8-bit
348			and 16-bit libraries has been built, and runs the tests appropriately.
349			.
350			.
351			.SH "NOT SUPPORTED IN 16-BIT MODE"
352			.rs
353			.sp
354			Not all the features of the 8-bit library are available with the 16-bit
355			library. The C++ and POSIX wrapper functions support only the 8-bit library,
356			and the \fBpcregrep\fP program is at present 8-bit only.
357			.
358			.
359			.SH AUTHOR
360			.rs
361			.sp
362			.nf
363			Philip Hazel
364			University Computing Service
365			Cambridge CB2 3QH, England.
366			.fi
367			.
368			.
369			.SH REVISION
370			.rs
371			.sp
372			.nf
373	ph10	856	Last updated: 07 January 2012
374	ph10	846	Copyright (c) 1997-2012 University of Cambridge.
375			.fi