Tcl Source Code

'\"
'\" Copyright (c) 1997 Sun Microsystems, Inc.
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH Utf 3 "8.1" Tcl "Tcl Library Procedures"
.so man.macros
.BS
.SH NAME
Tcl_UniChar, Tcl_UniCharToUtf, Tcl_UtfToUniChar, Tcl_UtfToChar16, Tcl_UtfToWChar, Tcl_UniCharToUtfDString, Tcl_UtfToUniCharDString, Tcl_Char16ToUtfDString, Tcl_UtfToWCharDString, Tcl_UtfToChar16DString, Tcl_UniCharLen, Tcl_UniCharNcmp, Tcl_UniCharNcasecmp, Tcl_UniCharCaseMatch, Tcl_UtfNcmp, Tcl_UtfNcasecmp, Tcl_UtfCharComplete, Tcl_NumUtfChars, Tcl_UtfFindFirst, Tcl_UtfFindLast, Tcl_UtfNext, Tcl_UtfPrev, Tcl_UniCharAtIndex, Tcl_UtfAtIndex, Tcl_UtfBackslash \- routines for manipulating UTF-8 strings
Tcl_UniChar, Tcl_UniCharToUtf, Tcl_UtfToUniChar, Tcl_UtfToChar16, Tcl_UtfToWChar, Tcl_UniCharToUtfDString, Tcl_UtfToUniCharDString, Tcl_Char16ToUtfDString, Tcl_UtfToWCharDString, Tcl_UtfToChar16DString, Tcl_WCharLen, Tcl_Char16Len, Tcl_UniCharLen, Tcl_UniCharNcmp, Tcl_UniCharNcasecmp, Tcl_UniCharCaseMatch, Tcl_UtfNcmp, Tcl_UtfNcasecmp, Tcl_UtfCharComplete, Tcl_NumUtfChars, Tcl_UtfFindFirst, Tcl_UtfFindLast, Tcl_UtfNext, Tcl_UtfPrev, Tcl_UniCharAtIndex, Tcl_UtfAtIndex, Tcl_UtfBackslash \- routines for manipulating UTF-8 strings
.SH SYNOPSIS
.nf
\fB#include <tcl.h>\fR
.sp
typedef ... \fBTcl_UniChar\fR;
.sp
int

.sp
unsigned short *
\fBTcl_UtfToChar16DString\fR(\fIsrc, length, dsPtr\fR)
.sp
wchar_t *
\fBTcl_UtfToWCharDString\fR(\fIsrc, length, dsPtr\fR)
.sp
int
\fBTcl_Char16Len\fR(\fIuniStr\fR)
.sp
int
\fBTcl_WCharLen\fR(\fIuniStr\fR)
.sp
int
\fBTcl_UniCharLen\fR(\fIuniStr\fR)
.sp
int
\fBTcl_UniCharNcmp\fR(\fIucs, uct, numChars\fR)
.sp
int

.QW \-1 ,
in which case \fBTcl_UtfToUniCharDString\fR uses \fBstrlen\fR to
calculate the length.  The return value is a pointer to the Unicode
representation of the UTF-8 string.  Storage for the return value
is appended to the end of the \fBTcl_DString\fR.  The Unicode string
is terminated with a Unicode null character.
.PP
\fBTcl_Char16Len\fR corresponds to \fBstrlen\fR for UTF-16
characters.  It accepts a null-terminated Unicode string and returns
the number of Unicode characters (not bytes) in that string.
.PP
\fBTcl_WCharLen\fR corresponds to \fBstrlen\fR for wchar_t
characters.  It accepts a null-terminated Unicode string and returns
the number of Unicode characters (not bytes) in that string.
.PP
\fBTcl_UniCharLen\fR corresponds to \fBstrlen\fR for Unicode
characters.  It accepts a null-terminated Unicode string and returns
the number of Unicode characters (not bytes) in that string.
.PP
\fBTcl_UniCharNcmp\fR and \fBTcl_UniCharNcasecmp\fR correspond to
\fBstrncmp\fR and \fBstrncasecmp\fR, respectively, for Unicode characters.
They accept two null-terminated Unicode strings and the number of characters

}
declare 350 {
    int Tcl_UniCharIsUpper(int ch)
}
declare 351 {
    int Tcl_UniCharIsWordChar(int ch)
}
declare 352 {deprecated {Use Tcl_GetCharLength}} {
    int Tcl_UniCharLen(const Tcl_UniChar *uniStr)
declare 352 {
    int Tcl_Char16Len(const unsigned short *uniStr)
}
declare 353 {deprecated {Use Tcl_UtfNcmp}} {
    int Tcl_UniCharNcmp(const Tcl_UniChar *ucs, const Tcl_UniChar *uct,
	    unsigned long numChars)
}
declare 354 {
    char *Tcl_Char16ToUtfDString(const unsigned short *uniStr,

    int Tcl_UniCharIsUnicode(int ch)
}

# TIP #511
declare 660 {
    int Tcl_AsyncMarkFromSignal(Tcl_AsyncHandler async, int sigNumber)
}

# TIP #617
declare 668 {
    int Tcl_UniCharLen(const int *uniStr)
}


# ----- BASELINE -- FOR -- 8.7.0 ----- #

##############################################################################

# Define the platform specific public Tcl interface. These functions are only
# available on the designated platform.

/* 349 */
EXTERN int		Tcl_UniCharIsSpace(int ch);
/* 350 */
EXTERN int		Tcl_UniCharIsUpper(int ch);
/* 351 */
EXTERN int		Tcl_UniCharIsWordChar(int ch);
/* 352 */
TCL_DEPRECATED("Use Tcl_GetCharLength")
int			Tcl_UniCharLen(const Tcl_UniChar *uniStr);
EXTERN int		Tcl_Char16Len(const unsigned short *uniStr);
/* 353 */
TCL_DEPRECATED("Use Tcl_UtfNcmp")
int			Tcl_UniCharNcmp(const Tcl_UniChar *ucs,
				const Tcl_UniChar *uct,
				unsigned long numChars);
/* 354 */
EXTERN char *		Tcl_Char16ToUtfDString(const unsigned short *uniStr,

/* 657 */
EXTERN int		Tcl_UniCharIsUnicode(int ch);
/* Slot 658 is reserved */
/* Slot 659 is reserved */
/* 660 */
EXTERN int		Tcl_AsyncMarkFromSignal(Tcl_AsyncHandler async,
				int sigNumber);
/* Slot 661 is reserved */
/* Slot 662 is reserved */
/* Slot 663 is reserved */
/* Slot 664 is reserved */
/* Slot 665 is reserved */
/* Slot 666 is reserved */
/* Slot 667 is reserved */
/* 668 */
EXTERN int		Tcl_UniCharLen(const int *uniStr);

typedef struct {
    const struct TclPlatStubs *tclPlatStubs;
    const struct TclIntStubs *tclIntStubs;
    const struct TclIntPlatStubs *tclIntPlatStubs;
} TclStubHooks;

    int (*tcl_UniCharIsAlnum) (int ch); /* 345 */
    int (*tcl_UniCharIsAlpha) (int ch); /* 346 */
    int (*tcl_UniCharIsDigit) (int ch); /* 347 */
    int (*tcl_UniCharIsLower) (int ch); /* 348 */
    int (*tcl_UniCharIsSpace) (int ch); /* 349 */
    int (*tcl_UniCharIsUpper) (int ch); /* 350 */
    int (*tcl_UniCharIsWordChar) (int ch); /* 351 */
    TCL_DEPRECATED_API("Use Tcl_GetCharLength") int (*tcl_UniCharLen) (const Tcl_UniChar *uniStr); /* 352 */
    int (*tcl_Char16Len) (const unsigned short *uniStr); /* 352 */
    TCL_DEPRECATED_API("Use Tcl_UtfNcmp") int (*tcl_UniCharNcmp) (const Tcl_UniChar *ucs, const Tcl_UniChar *uct, unsigned long numChars); /* 353 */
    char * (*tcl_Char16ToUtfDString) (const unsigned short *uniStr, int uniLength, Tcl_DString *dsPtr); /* 354 */
    unsigned short * (*tcl_UtfToChar16DString) (const char *src, int length, Tcl_DString *dsPtr); /* 355 */
    Tcl_RegExp (*tcl_GetRegExpFromObj) (Tcl_Interp *interp, Tcl_Obj *patObj, int flags); /* 356 */
    TCL_DEPRECATED_API("Use Tcl_EvalTokensStandard") Tcl_Obj * (*tcl_EvalTokens) (Tcl_Interp *interp, Tcl_Token *tokenPtr, int count); /* 357 */
    void (*tcl_FreeParse) (Tcl_Parse *parsePtr); /* 358 */
    void (*tcl_LogCommandInfo) (Tcl_Interp *interp, const char *script, const char *command, int length); /* 359 */

    int (*tcl_UtfCharComplete) (const char *src, int length); /* 654 */
    const char * (*tcl_UtfNext) (const char *src); /* 655 */
    const char * (*tcl_UtfPrev) (const char *src, const char *start); /* 656 */
    int (*tcl_UniCharIsUnicode) (int ch); /* 657 */
    void (*reserved658)(void);
    void (*reserved659)(void);
    int (*tcl_AsyncMarkFromSignal) (Tcl_AsyncHandler async, int sigNumber); /* 660 */
    void (*reserved661)(void);
    void (*reserved662)(void);
    void (*reserved663)(void);
    void (*reserved664)(void);
    void (*reserved665)(void);
    void (*reserved666)(void);
    void (*reserved667)(void);
    int (*tcl_UniCharLen) (const int *uniStr); /* 668 */
} TclStubs;

extern const TclStubs *tclStubsPtr;

#ifdef __cplusplus
}
#endif

	(tclStubsPtr->tcl_UniCharIsLower) /* 348 */
#define Tcl_UniCharIsSpace \
	(tclStubsPtr->tcl_UniCharIsSpace) /* 349 */
#define Tcl_UniCharIsUpper \
	(tclStubsPtr->tcl_UniCharIsUpper) /* 350 */
#define Tcl_UniCharIsWordChar \
	(tclStubsPtr->tcl_UniCharIsWordChar) /* 351 */
#define Tcl_UniCharLen \
	(tclStubsPtr->tcl_UniCharLen) /* 352 */
#define Tcl_Char16Len \
	(tclStubsPtr->tcl_Char16Len) /* 352 */
#define Tcl_UniCharNcmp \
	(tclStubsPtr->tcl_UniCharNcmp) /* 353 */
#define Tcl_Char16ToUtfDString \
	(tclStubsPtr->tcl_Char16ToUtfDString) /* 354 */
#define Tcl_UtfToChar16DString \
	(tclStubsPtr->tcl_UtfToChar16DString) /* 355 */
#define Tcl_GetRegExpFromObj \

	(tclStubsPtr->tcl_UtfPrev) /* 656 */
#define Tcl_UniCharIsUnicode \
	(tclStubsPtr->tcl_UniCharIsUnicode) /* 657 */
/* Slot 658 is reserved */
/* Slot 659 is reserved */
#define Tcl_AsyncMarkFromSignal \
	(tclStubsPtr->tcl_AsyncMarkFromSignal) /* 660 */
/* Slot 661 is reserved */
/* Slot 662 is reserved */
/* Slot 663 is reserved */
/* Slot 664 is reserved */
/* Slot 665 is reserved */
/* Slot 666 is reserved */
/* Slot 667 is reserved */
#define Tcl_UniCharLen \
	(tclStubsPtr->tcl_UniCharLen) /* 668 */

#endif /* defined(USE_TCL_STUBS) */

/* !END!: Do not edit above this line. */

#undef TclUnusedStubEntry
#if defined(USE_TCL_STUBS)

#if TCL_UTF_MAX <= 3
#   undef Tcl_UniCharToUtfDString
#   define Tcl_UniCharToUtfDString Tcl_Char16ToUtfDString
#   undef Tcl_UtfToUniCharDString
#   define Tcl_UtfToUniCharDString Tcl_UtfToChar16DString
#   undef Tcl_UtfToUniChar
#   define Tcl_UtfToUniChar Tcl_UtfToChar16
#   undef Tcl_UniCharLen
#   define Tcl_UniCharLen Tcl_Char16Len
#endif
#if defined(USE_TCL_STUBS)
#   define Tcl_WCharToUtfDString (sizeof(wchar_t) != sizeof(short) \
		? (char *(*)(const wchar_t *, int, Tcl_DString *))tclStubsPtr->tcl_UniCharToUtfDString \
		: (char *(*)(const wchar_t *, int, Tcl_DString *))Tcl_Char16ToUtfDString)
#   define Tcl_UtfToWCharDString (sizeof(wchar_t) != sizeof(short) \
		? (wchar_t *(*)(const char *, int, Tcl_DString *))tclStubsPtr->tcl_UtfToUniCharDString \
		: (wchar_t *(*)(const char *, int, Tcl_DString *))Tcl_UtfToChar16DString)
#   define Tcl_UtfToWChar (sizeof(wchar_t) != sizeof(short) \
		? (int (*)(const char *, wchar_t *))tclStubsPtr->tcl_UtfToUniChar \
		: (int (*)(const char *, wchar_t *))Tcl_UtfToChar16)
#   define Tcl_WCharLen (sizeof(wchar_t) != sizeof(short) \
		? (int (*)(wchar_t *))tclStubsPtr->tcl_UniCharLen \
		: (int (*)(wchar_t *))Tcl_Char16Len)
#else
#   define Tcl_WCharToUtfDString (sizeof(wchar_t) != sizeof(short) \
		? (char *(*)(const wchar_t *, int, Tcl_DString *))Tcl_UniCharToUtfDString \
		: (char *(*)(const wchar_t *, int, Tcl_DString *))Tcl_Char16ToUtfDString)
#   define Tcl_UtfToWCharDString (sizeof(wchar_t) != sizeof(short) \
		? (wchar_t *(*)(const char *, int, Tcl_DString *))Tcl_UtfToUniCharDString \
		: (wchar_t *(*)(const char *, int, Tcl_DString *))Tcl_UtfToChar16DString)
#   define Tcl_UtfToWChar (sizeof(wchar_t) != sizeof(short) \
		? (int (*)(const char *, wchar_t *))Tcl_UtfToUniChar \
		: (int (*)(const char *, wchar_t *))Tcl_UtfToChar16)
#   define Tcl_WCharLen (sizeof(wchar_t) != sizeof(short) \
		? (int (*)(wchar_t *))Tcl_UniCharLen \
		: (int (*)(wchar_t *))Tcl_Char16Len)
#endif

/*
 * Deprecated Tcl procedures:
 */

#undef Tcl_EvalObj

    Tcl_UniCharIsAlnum, /* 345 */
    Tcl_UniCharIsAlpha, /* 346 */
    Tcl_UniCharIsDigit, /* 347 */
    Tcl_UniCharIsLower, /* 348 */
    Tcl_UniCharIsSpace, /* 349 */
    Tcl_UniCharIsUpper, /* 350 */
    Tcl_UniCharIsWordChar, /* 351 */
    Tcl_UniCharLen, /* 352 */
    Tcl_Char16Len, /* 352 */
    Tcl_UniCharNcmp, /* 353 */
    Tcl_Char16ToUtfDString, /* 354 */
    Tcl_UtfToChar16DString, /* 355 */
    Tcl_GetRegExpFromObj, /* 356 */
    Tcl_EvalTokens, /* 357 */
    Tcl_FreeParse, /* 358 */
    Tcl_LogCommandInfo, /* 359 */

    Tcl_UtfCharComplete, /* 654 */
    Tcl_UtfNext, /* 655 */
    Tcl_UtfPrev, /* 656 */
    Tcl_UniCharIsUnicode, /* 657 */
    0, /* 658 */
    0, /* 659 */
    Tcl_AsyncMarkFromSignal, /* 660 */
    0, /* 661 */
    0, /* 662 */
    0, /* 663 */
    0, /* 664 */
    0, /* 665 */
    0, /* 666 */
    0, /* 667 */
    Tcl_UniCharLen, /* 668 */
};

/* !END!: Do not edit above this line. */

    /* Clear away extension bits, if any */
    return ch & 0x1FFFFF;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_Char16Len --
 *
 *	Find the length of a UniChar string. The str input must be null
 *	terminated.
 *
 * Results:
 *	Returns the length of str in UniChars (not bytes).
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_Char16Len(
    const unsigned short *uniStr)	/* Unicode string to find length of. */
{
    int len = 0;

    while (*uniStr != '\0') {
	len++;
	uniStr++;
    }
    return len;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharLen --
 *
 *	Find the length of a UniChar string. The str input must be null
 *	terminated.
 *
 * Results:
 *	Returns the length of str in UniChars (not bytes).
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

#undef Tcl_UniCharLen
int
Tcl_UniCharLen(
    const Tcl_UniChar *uniStr)	/* Unicode string to find length of. */
    const int *uniStr)	/* Unicode string to find length of. */
{
    int len = 0;

    while (*uniStr != '\0') {
	len++;
	uniStr++;
    }