tclutf.c

来自「tcl是工具命令语言」· C语言 代码 · 共 1,949 行 · 第 1/4 页

				 * for strlen(string). */
{
    Tcl_UniChar ch;
    register Tcl_UniChar *chPtr = &ch;
    register int i;

    /*
     * The separate implementations are faster.
     *
     * Since this is a time-sensitive function, we also do the check for
     * the single-byte char case specially.
     */

    i = 0;
    if (len < 0) {
	while (1) {
	    str += TclUtfToUniChar(str, chPtr);
	    if (ch == '\0') {
		break;
	    }
	    i++;
	}
    } else {
	register int n;

	while (len > 0) {
	    if (UCHAR(*str) < 0xC0) {
		len--;
		str++;
	    } else {
		n = Tcl_UtfToUniChar(str, chPtr);
		len -= n;
		str += n;
	    }
	    i++;
	}
    }
    return i;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfFindFirst --
 *
 *	Returns a pointer to the first occurance of the given Tcl_UniChar
 *	in the NULL-terminated UTF-8 string.  The NULL terminator is
 *	considered part of the UTF-8 string.  Equivalent to Plan 9
 *	utfrune().
 *
 * Results:
 *	As above.  If the Tcl_UniChar does not exist in the given string,
 *	the return value is NULL.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
CONST char *
Tcl_UtfFindFirst(string, ch)
    CONST char *string;		/* The UTF-8 string to be searched. */
    int ch;			/* The Tcl_UniChar to search for. */
{
    int len;
    Tcl_UniChar find;
    
    while (1) {
	len = TclUtfToUniChar(string, &find);
	if (find == ch) {
	    return string;
	}
	if (*string == '\0') {
	    return NULL;
	}
	string += len;
    }
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfFindLast --
 *
 *	Returns a pointer to the last occurance of the given Tcl_UniChar
 *	in the NULL-terminated UTF-8 string.  The NULL terminator is
 *	considered part of the UTF-8 string.  Equivalent to Plan 9
 *	utfrrune().
 *
 * Results:
 *	As above.  If the Tcl_UniChar does not exist in the given string,
 *	the return value is NULL.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

CONST char *
Tcl_UtfFindLast(string, ch)
    CONST char *string;		/* The UTF-8 string to be searched. */
    int ch;			/* The Tcl_UniChar to search for. */
{
    int len;
    Tcl_UniChar find;
    CONST char *last;
	
    last = NULL;
    while (1) {
	len = TclUtfToUniChar(string, &find);
	if (find == ch) {
	    last = string;
	}
	if (*string == '\0') {
	    break;
	}
	string += len;
    }
    return last;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfNext --
 *
 *	Given a pointer to some current location in a UTF-8 string,
 *	move forward one character.  The caller must ensure that they
 *	are not asking for the next character after the last character
 *	in the string.
 *
 * Results:
 *	The return value is the pointer to the next character in
 *	the UTF-8 string.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
CONST char *
Tcl_UtfNext(str) 
    CONST char *str;		    /* The current location in the string. */
{
    Tcl_UniChar ch;

    return str + TclUtfToUniChar(str, &ch);
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfPrev --
 *
 *	Given a pointer to some current location in a UTF-8 string,
 *	move backwards one character.  This works correctly when the
 *	pointer is in the middle of a UTF-8 character.
 *
 * Results:
 *	The return value is a pointer to the previous character in the
 *	UTF-8 string.  If the current location was already at the
 *	beginning of the string, the return value will also be a
 *	pointer to the beginning of the string.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

CONST char *
Tcl_UtfPrev(str, start)
    CONST char *str;		    /* The current location in the string. */
    CONST char *start;		    /* Pointer to the beginning of the
				     * string, to avoid going backwards too
				     * far. */
{
    CONST char *look;
    int i, byte;
    
    str--;
    look = str;
    for (i = 0; i < TCL_UTF_MAX; i++) {
	if (look < start) {
	    if (str < start) {
		str = start;
	    }
	    break;
	}
	byte = *((unsigned char *) look);
	if (byte < 0x80) {
	    break;
	}
	if (byte >= 0xC0) {
	    return look;
	}
	look--;
    }
    return str;
}
	
/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UniCharAtIndex --
 *
 *	Returns the Unicode character represented at the specified
 *	character (not byte) position in the UTF-8 string.
 *
 * Results:
 *	As above.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */
 
Tcl_UniChar
Tcl_UniCharAtIndex(src, index)
    register CONST char *src;	/* The UTF-8 string to dereference. */
    register int index;		/* The position of the desired character. */
{
    Tcl_UniChar ch;

    while (index >= 0) {
	index--;
	src += TclUtfToUniChar(src, &ch);
    }
    return ch;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfAtIndex --
 *
 *	Returns a pointer to the specified character (not byte) position
 *	in the UTF-8 string.
 *
 * Results:
 *	As above.
 *
 * Side effects:
 *	None.
 *
 *---------------------------------------------------------------------------
 */

CONST char *
Tcl_UtfAtIndex(src, index)
    register CONST char *src;	/* The UTF-8 string. */
    register int index;		/* The position of the desired character. */
{
    Tcl_UniChar ch;
    
    while (index > 0) {
	index--;
	src += TclUtfToUniChar(src, &ch);
    }
    return src;
}

/*
 *---------------------------------------------------------------------------
 *
 * Tcl_UtfBackslash --
 *
 *	Figure out how to handle a backslash sequence.
 *
 * Results:
 *	Stores the bytes represented by the backslash sequence in dst and
 *	returns the number of bytes written to dst.  At most TCL_UTF_MAX
 *	bytes are written to dst; dst must have been large enough to accept
 *	those bytes.  If readPtr isn't NULL then it is filled in with a
 *	count of the number of bytes in the backslash sequence.  
 *
 * Side effects:
 *	The maximum number of bytes it takes to represent a Unicode
 *	character in UTF-8 is guaranteed to be less than the number of
 *	bytes used to express the backslash sequence that represents
 *	that Unicode character.  If the target buffer into which the
 *	caller is going to store the bytes that represent the Unicode
 *	character is at least as large as the source buffer from which
 *	the backslashed sequence was extracted, no buffer overruns should
 *	occur.
 *
 *---------------------------------------------------------------------------
 */

int
Tcl_UtfBackslash(src, readPtr, dst)
    CONST char *src;		/* Points to the backslash character of
				 * a backslash sequence. */
    int *readPtr;		/* Fill in with number of characters read
				 * from src, unless NULL. */
    char *dst;			/* Filled with the bytes represented by the
				 * backslash sequence. */
{
#define LINE_LENGTH 128
    int numRead;
    int result;

    result = TclParseBackslash(src, LINE_LENGTH, &numRead, dst);
    if (numRead == LINE_LENGTH) {
	/* We ate a whole line.  Pay the price of a strlen() */
	result = TclParseBackslash(src, (int)strlen(src), &numRead, dst);
    }
    if (readPtr != NULL) {
	*readPtr = numRead;
    }
    return result;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UtfToUpper --
 *
 *	Convert lowercase characters to uppercase characters in a UTF
 *	string in place.  The conversion may shrink the UTF string.
 *
 * Results:
 *	Returns the number of bytes in the resulting string
 *	excluding the trailing null.
 *
 * Side effects:
 *	Writes a terminating null after the last converted character.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UtfToUpper(str)
    char *str;			/* String to convert in place. */
{
    Tcl_UniChar ch, upChar;
    char *src, *dst;
    int bytes;

    /*
     * Iterate over the string until we hit the terminating null.
     */

    src = dst = str;
    while (*src) {
        bytes = TclUtfToUniChar(src, &ch);
	upChar = Tcl_UniCharToUpper(ch);

	/*
	 * To keep badly formed Utf strings from getting inflated by
	 * the conversion (thereby causing a segfault), only copy the
	 * upper case char to dst if its size is <= the original char.
	 */
	
	if (bytes < UtfCount(upChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(upChar, dst);
	}
	src += bytes;
    }
    *dst = '\0';
    return (dst - str);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UtfToLower --
 *
 *	Convert uppercase characters to lowercase characters in a UTF
 *	string in place.  The conversion may shrink the UTF string.
 *
 * Results:
 *	Returns the number of bytes in the resulting string
 *	excluding the trailing null.
 *
 * Side effects:
 *	Writes a terminating null after the last converted character.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UtfToLower(str)
    char *str;			/* String to convert in place. */
{
    Tcl_UniChar ch, lowChar;
    char *src, *dst;
    int bytes;
    
    /*
     * Iterate over the string until we hit the terminating null.
     */

    src = dst = str;
    while (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	lowChar = Tcl_UniCharToLower(ch);

	/*
	 * To keep badly formed Utf strings from getting inflated by
	 * the conversion (thereby causing a segfault), only copy the
	 * lower case char to dst if its size is <= the original char.
	 */
	
	if (bytes < UtfCount(lowChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += bytes;
    }
    *dst = '\0';
    return (dst - str);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UtfToTitle --
 *
 *	Changes the first character of a UTF string to title case or
 *	uppercase and the rest of the string to lowercase.  The
 *	conversion happens in place and may shrink the UTF string.
 *
 * Results:
 *	Returns the number of bytes in the resulting string
 *	excluding the trailing null.
 *
 * Side effects:
 *	Writes a terminating null after the last converted character.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UtfToTitle(str)
    char *str;			/* String to convert in place. */
{
    Tcl_UniChar ch, titleChar, lowChar;
    char *src, *dst;
    int bytes;
    
    /*
     * Capitalize the first character and then lowercase the rest of the
     * characters until we get to a null.
     */

    src = dst = str;

    if (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	titleChar = Tcl_UniCharToTitle(ch);

	if (bytes < UtfCount(titleChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(titleChar, dst);
	}
	src += bytes;
    }
    while (*src) {
	bytes = TclUtfToUniChar(src, &ch);
	lowChar = Tcl_UniCharToLower(ch);

	if (bytes < UtfCount(lowChar)) {
	    memcpy(dst, src, (size_t) bytes);
	    dst += bytes;
	} else {
	    dst += Tcl_UniCharToUtf(lowChar, dst);
	}
	src += bytes;
    }
    *dst = '\0';
    return (dst - str);
}

/*
 *----------------------------------------------------------------------
 *
 * TclpUtfNcmp2 --