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

 *
 *	Compare at most n bytes of utf-8 strings cs and ct.  Both cs
 *	and ct are assumed to be at least n bytes long.
 *
 * Results:
 *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
TclpUtfNcmp2(cs, ct, n)
    CONST char *cs;		/* UTF string to compare to ct. */
    CONST char *ct;		/* UTF string cs is compared to. */
    unsigned long n;		/* Number of *bytes* to compare. */
{
    /*
     * We can't simply call 'memcmp(cs, ct, n);' because we need to check
     * for Tcl's \xC0\x80 non-utf-8 null encoding.
     * Otherwise utf-8 lexes fine in the strcmp manner.
     */
    register int result = 0;

    for ( ; n != 0; n--, cs++, ct++) {
	if (*cs != *ct) {
	    result = UCHAR(*cs) - UCHAR(*ct);
	    break;
	}
    }
    if (n && ((UCHAR(*cs) == 0xC0) || (UCHAR(*ct) == 0xC0))) {
	unsigned char c1, c2;
	c1 = ((UCHAR(*cs) == 0xC0) && (UCHAR(cs[1]) == 0x80)) ? 0 : UCHAR(*cs);
	c2 = ((UCHAR(*ct) == 0xC0) && (UCHAR(ct[1]) == 0x80)) ? 0 : UCHAR(*ct);
	result = (c1 - c2);
    }
    return result;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UtfNcmp --
 *
 *	Compare at most n UTF chars of string cs to string ct.  Both cs
 *	and ct are assumed to be at least n UTF chars long.
 *
 * Results:
 *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UtfNcmp(cs, ct, n)
    CONST char *cs;		/* UTF string to compare to ct. */
    CONST char *ct;		/* UTF string cs is compared to. */
    unsigned long n;		/* Number of UTF chars to compare. */
{
    Tcl_UniChar ch1, ch2;
    /*
     * Cannot use 'memcmp(cs, ct, n);' as byte representation of
     * \u0000 (the pair of bytes 0xc0,0x80) is larger than byte
     * representation of \u0001 (the byte 0x01.)
     */
    while (n-- > 0) {
	/*
	 * n must be interpreted as chars, not bytes.
	 * This should be called only when both strings are of
	 * at least n chars long (no need for \0 check)
	 */
	cs += TclUtfToUniChar(cs, &ch1);
	ct += TclUtfToUniChar(ct, &ch2);
	if (ch1 != ch2) {
	    return (ch1 - ch2);
	}
    }
    return 0;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UtfNcasecmp --
 *
 *	Compare at most n UTF chars of string cs to string ct case
 *	insensitive.  Both cs and ct are assumed to be at least n
 *	UTF chars long.
 *
 * Results:
 *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UtfNcasecmp(cs, ct, n)
    CONST char *cs;		/* UTF string to compare to ct. */
    CONST char *ct;		/* UTF string cs is compared to. */
    unsigned long n;			/* Number of UTF chars to compare. */
{
    Tcl_UniChar ch1, ch2;
    while (n-- > 0) {
	/*
	 * n must be interpreted as chars, not bytes.
	 * This should be called only when both strings are of
	 * at least n chars long (no need for \0 check)
	 */
	cs += TclUtfToUniChar(cs, &ch1);
	ct += TclUtfToUniChar(ct, &ch2);
	if (ch1 != ch2) {
	    ch1 = Tcl_UniCharToLower(ch1);
	    ch2 = Tcl_UniCharToLower(ch2);
	    if (ch1 != ch2) {
		return (ch1 - ch2);
	    }
	}
    }
    return 0;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToUpper --
 *
 *	Compute the uppercase equivalent of the given Unicode character.
 *
 * Results:
 *	Returns the uppercase Unicode character.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

Tcl_UniChar
Tcl_UniCharToUpper(ch)
    int ch;			/* Unicode character to convert. */
{
    int info = GetUniCharInfo(ch);

    if (GetCaseType(info) & 0x04) {
	return (Tcl_UniChar) (ch - GetDelta(info));
    } else {
	return ch;
    }
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToLower --
 *
 *	Compute the lowercase equivalent of the given Unicode character.
 *
 * Results:
 *	Returns the lowercase Unicode character.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

Tcl_UniChar
Tcl_UniCharToLower(ch)
    int ch;			/* Unicode character to convert. */
{
    int info = GetUniCharInfo(ch);

    if (GetCaseType(info) & 0x02) {
	return (Tcl_UniChar) (ch + GetDelta(info));
    } else {
	return ch;
    }
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharToTitle --
 *
 *	Compute the titlecase equivalent of the given Unicode character.
 *
 * Results:
 *	Returns the titlecase Unicode character.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

Tcl_UniChar
Tcl_UniCharToTitle(ch)
    int ch;			/* Unicode character to convert. */
{
    int info = GetUniCharInfo(ch);
    int mode = GetCaseType(info);

    if (mode & 0x1) {
	/*
	 * Subtract or add one depending on the original case.
	 */

	return (Tcl_UniChar) (ch + ((mode & 0x4) ? -1 : 1));
    } else if (mode == 0x4) {
	return (Tcl_UniChar) (ch - GetDelta(info));
    } else {
	return ch;
    }
}

/*
 *----------------------------------------------------------------------
 *
 * 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.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharLen(str)
    CONST Tcl_UniChar *str;	/* Unicode string to find length of. */
{
    int len = 0;
    
    while (*str != '\0') {
	len++;
	str++;
    }
    return len;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharNcmp --
 *
 *	Compare at most n unichars of string cs to string ct.  Both cs
 *	and ct are assumed to be at least n unichars long.
 *
 * Results:
 *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharNcmp(cs, ct, n)
    CONST Tcl_UniChar *cs;		/* Unicode string to compare to ct. */
    CONST Tcl_UniChar *ct;		/* Unicode string cs is compared to. */
    unsigned long n;			/* Number of unichars to compare. */
{
#ifdef WORDS_BIGENDIAN
    /*
     * We are definitely on a big-endian machine; memcmp() is safe
     */
    return memcmp(cs, ct, n*sizeof(Tcl_UniChar));

#else /* !WORDS_BIGENDIAN */
    /*
     * We can't simply call memcmp() because that is not lexically correct.
     */
    for ( ; n != 0; cs++, ct++, n--) {
	if (*cs != *ct) {
	    return (*cs - *ct);
	}
    }
    return 0;
#endif /* WORDS_BIGENDIAN */
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharNcasecmp --
 *
 *	Compare at most n unichars of string cs to string ct case
 *	insensitive.  Both cs and ct are assumed to be at least n
 *	unichars long.
 *
 * Results:
 *	Return <0 if cs < ct, 0 if cs == ct, or >0 if cs > ct.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharNcasecmp(cs, ct, n)
    CONST Tcl_UniChar *cs;		/* Unicode string to compare to ct. */
    CONST Tcl_UniChar *ct;		/* Unicode string cs is compared to. */
    unsigned long n;			/* Number of unichars to compare. */
{
    for ( ; n != 0; n--, cs++, ct++) {
	if ((*cs != *ct) &&
		(Tcl_UniCharToLower(*cs) != Tcl_UniCharToLower(*ct))) {
	    return (*cs - *ct);
	}
    }
    return 0;
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsAlnum --
 *
 *	Test if a character is an alphanumeric Unicode character.
 *
 * Results:
 *	Returns 1 if character is alphanumeric.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharIsAlnum(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);

    return (((ALPHA_BITS | DIGIT_BITS) >> category) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsAlpha --
 *
 *	Test if a character is an alphabetic Unicode character.
 *
 * Results:
 *	Returns 1 if character is alphabetic.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharIsAlpha(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
    return ((ALPHA_BITS >> category) & 1);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsControl --
 *
 *	Test if a character is a Unicode control character.
 *
 * Results:
 *	Returns non-zero if character is a control.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharIsControl(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK) == CONTROL);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsDigit --
 *
 *	Test if a character is a numeric Unicode character.
 *
 * Results:
 *	Returns non-zero if character is a digit.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharIsDigit(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK)
	    == DECIMAL_DIGIT_NUMBER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsGraph --
 *
 *	Test if a character is any Unicode print character except space.
 *
 * Results:
 *	Returns non-zero if character is printable, but not space.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharIsGraph(ch)
    int ch;			/* Unicode character to test. */
{
    register int category = (GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK);
    return (((PRINT_BITS >> category) & 1) && ((unsigned char) ch != ' '));
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsLower --
 *
 *	Test if a character is a lowercase Unicode character.
 *
 * Results:
 *	Returns non-zero if character is lowercase.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int
Tcl_UniCharIsLower(ch)
    int ch;			/* Unicode character to test. */
{
    return ((GetUniCharInfo(ch) & UNICODE_CATEGORY_MASK) == LOWERCASE_LETTER);
}

/*
 *----------------------------------------------------------------------
 *
 * Tcl_UniCharIsPrint --
 *
 *	Test if a character is a Unicode print character.
 *
 * Results:
 *	Returns non-zero if character is printable.
 *
 * Side effects:
 *	None.
 *
 *----------------------------------------------------------------------
 */

int