Tcl Source Code

Check-in [1841bf54d1]
Login
Bounty program for improvements to Tcl and certain Tcl packages.
Tcl 2019 Conference, Houston/TX, US, Nov 4-8
Send your abstracts to [email protected]
or submit via the online form by Sep 9.

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:Deduplicate code in INST_STR_CMP, StringCmpCmd, and StringEqualCmd.
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | core-8-6-branch
Files: files | file ages | folders
SHA3-256: 1841bf54d1b3257e14be24a6205a56ade3fc86b62a4fba25d86235888942933e
User & Date: pooryorick 2018-05-07 07:43:00
Context
2018-05-08
09:56
merge fix for [92564326a98b5510] from 8.5 check-in: b6d467a174 user: sebres tags: core-8-6-branch
2018-05-07
14:33
merge core-8-6-branch check-in: cb43f6d65a user: pooryorick tags: core-8-branch
07:43
Deduplicate code in INST_STR_CMP, StringCmpCmd, and StringEqualCmd. check-in: 1841bf54d1 user: pooryorick tags: core-8-6-branch
2018-05-06
18:13
Factor options handling out of StringCmpCmd. check-in: 88e60c5289 user: pooryorick tags: core-8-6-branch
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to generic/tclCmdMZ.c.

2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
....
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
....
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772

2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839



2840
2841
2842
2843
2844

2845
2846
2847
2848

2849
2850
2851
2852
2853

2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873





2874
2875
2876
2877
2878
2879
2880
....
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905

2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939





2940
2941









2942




2943
2944
2945













2946
2947

2948
2949
2950
2951
2952
2953
2954

2955
2956
2957

2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971

2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
{
    /*
     * Remember to keep code here in some sync with the byte-compiled versions
     * in tclExecute.c (INST_STR_EQ, INST_STR_NEQ and INST_STR_CMP as well as
     * the expr string comparison in INST_EQ/INST_NEQ/INST_LT/...).
     */

    const char *string1, *string2;
    int length1, length2, i, match, length, nocase = 0, reqlength = -1;
    typedef int (*strCmpFn_t)(const char *, const char *, unsigned int);
    strCmpFn_t strCmpFn;

    if (objc < 3 || objc > 6) {
    str_cmp_args:
	Tcl_WrongNumArgs(interp, 1, objv,
		"?-nocase? ?-length int? string1 string2");
	return TCL_ERROR;
    }
................................................................................
    /*
     * From now on, we only access the two objects at the end of the argument
     * array.
     */

    objv += objc-2;

    if ((reqlength == 0) || (objv[0] == objv[1])) {
	/*
	 * Always match at 0 chars of if it is the same obj.
	 */

	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(1));
	return TCL_OK;
    }

    if (!nocase && TclIsPureByteArray(objv[0]) &&
	    TclIsPureByteArray(objv[1])) {
	/*
	 * Use binary versions of comparisons since that won't cause undue
	 * type conversions and it is much faster. Only do this if we're
	 * case-sensitive (which is all that really makes sense with byte
	 * arrays anyway, and we have no memcasecmp() for some reason... :^)
	 */

	string1 = (char *) Tcl_GetByteArrayFromObj(objv[0], &length1);
	string2 = (char *) Tcl_GetByteArrayFromObj(objv[1], &length2);
	strCmpFn = (strCmpFn_t) memcmp;
    } else if ((objv[0]->typePtr == &tclStringType)
	    && (objv[1]->typePtr == &tclStringType)) {
	/*
	 * Do a unicode-specific comparison if both of the args are of String
	 * type. In benchmark testing this proved the most efficient check
	 * between the unicode and string comparison operations.
	 */

	string1 = (char *) Tcl_GetUnicodeFromObj(objv[0], &length1);
	string2 = (char *) Tcl_GetUnicodeFromObj(objv[1], &length2);
	strCmpFn = (strCmpFn_t)
		(nocase ? Tcl_UniCharNcasecmp : Tcl_UniCharNcmp);
    } else {
	/*
	 * As a catch-all we will work with UTF-8. We cannot use memcmp() as
	 * that is unsafe with any string containing NUL (\xC0\x80 in Tcl's
	 * utf rep). We can use the more efficient TclpUtfNcmp2 if we are
	 * case-sensitive and no specific length was requested.
	 */

	string1 = (char *) TclGetStringFromObj(objv[0], &length1);
	string2 = (char *) TclGetStringFromObj(objv[1], &length2);
	if ((reqlength < 0) && !nocase) {
	    strCmpFn = (strCmpFn_t) TclpUtfNcmp2;
	} else {
	    length1 = Tcl_NumUtfChars(string1, length1);
	    length2 = Tcl_NumUtfChars(string2, length2);
	    strCmpFn = (strCmpFn_t) (nocase ? Tcl_UtfNcasecmp : Tcl_UtfNcmp);
	}
    }

    if ((reqlength < 0) && (length1 != length2)) {
	match = 1;		/* This will be reversed below. */
    } else {
	length = (length1 < length2) ? length1 : length2;
	if (reqlength > 0 && reqlength < length) {
	    length = reqlength;
	} else if (reqlength < 0) {
	    /*
	     * The requested length is negative, so we ignore it by setting it
	     * to length + 1 so we correct the match var.
	     */

	    reqlength = length + 1;
	}

	match = strCmpFn(string1, string2, (unsigned) length);
	if ((match == 0) && (reqlength > length)) {
	    match = length1 - length2;
	}
    }

    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(match ? 0 : 1));
    return TCL_OK;
}
 
/*
 *----------------------------------------------------------------------
................................................................................
{
    /*
     * Remember to keep code here in some sync with the byte-compiled versions
     * in tclExecute.c (INST_STR_EQ, INST_STR_NEQ and INST_STR_CMP as well as
     * the expr string comparison in INST_EQ/INST_NEQ/INST_LT/...).
     */

    const char *string1, *string2;
    int length1, length2, match, length, nocase, reqlength, status;
    typedef int (*strCmpFn_t)(const char *, const char *, unsigned int);
    strCmpFn_t strCmpFn;

    if ((status = TclStringCmpOpts(interp, objc, objv, &reqlength,
	(char **)&string2, &length2, &nocase)) != TCL_OK){
	return status;
    }

    /*
     * From now on, we only access the two objects at the end of the argument
     * array.
     */
    objv += objc-2;

    if ((reqlength == 0) || (objv[0] == objv[1])) {
	/*
	 * Always match at 0 chars of if it is the same obj.
	 */


	Tcl_SetObjResult(interp, Tcl_NewBooleanObj(0));
	return TCL_OK;
    }

    if (!nocase && TclIsPureByteArray(objv[0]) &&
	    TclIsPureByteArray(objv[1])) {

	string1 = (char *) Tcl_GetByteArrayFromObj(objv[0], &length1);
	string2 = (char *) Tcl_GetByteArrayFromObj(objv[1], &length2);
	strCmpFn = (strCmpFn_t) memcmp;
    } else if ((objv[0]->typePtr == &tclStringType)
	    && (objv[1]->typePtr == &tclStringType)) {
	/*
	 * Do a unicode-specific comparison if both of the args are of String
	 * type. In benchmark testing this proved the most efficient check
	 * between the unicode and string comparison operations.
	 */

	string1 = (char *) Tcl_GetUnicodeFromObj(objv[0], &length1);
	string2 = (char *) Tcl_GetUnicodeFromObj(objv[1], &length2);
	strCmpFn = (strCmpFn_t)
		(nocase ? Tcl_UniCharNcasecmp : Tcl_UniCharNcmp);
    } else {
	/*
	 * As a catch-all we will work with UTF-8. We cannot use memcmp() as
	 * that is unsafe with any string containing NUL (\xC0\x80 in Tcl's
	 * utf rep). We can use the more efficient TclpUtfNcmp2 if we are
	 * case-sensitive and no specific length was requested.
	 */

	string1 = (char *) TclGetStringFromObj(objv[0], &length1);
	string2 = (char *) TclGetStringFromObj(objv[1], &length2);
	if ((reqlength < 0) && !nocase) {
	    strCmpFn = (strCmpFn_t) TclpUtfNcmp2;
	} else {
	    length1 = Tcl_NumUtfChars(string1, length1);
	    length2 = Tcl_NumUtfChars(string2, length2);
	    strCmpFn = (strCmpFn_t) (nocase ? Tcl_UtfNcasecmp : Tcl_UtfNcmp);
	}
    }

    length = (length1 < length2) ? length1 : length2;
    if (reqlength > 0 && reqlength < length) {
	length = reqlength;
    } else if (reqlength < 0) {
	/*
	 * The requested length is negative, so we ignore it by setting it to
	 * length + 1 so we correct the match var.
	 */

	reqlength = length + 1;
    }

    match = strCmpFn(string1, string2, (unsigned) length);
    if ((match == 0) && (reqlength > length)) {
	match = length1 - length2;
    }

    Tcl_SetObjResult(interp,
	    Tcl_NewIntObj((match > 0) ? 1 : (match < 0) ? -1 : 0));
    return TCL_OK;
}
 
int TclStringCmp (
	Tcl_Obj *value1Ptr,
	Tcl_Obj *value2Ptr,
	int checkEq



) {
	char *s1, *s2;
	int empty, match, s1len, s2len;
	memCmpFn_t memCmpFn;


	/*
	 * When we have equal-length we can check only for (in)equality.
	 * We can use memcmp in all (n)eq cases because we
	 * don't need to worry about lexical LE/BE variance.

	 */

	if (value1Ptr == value2Ptr) {
	    match = 0;
	} else {

	    if (TclIsPureByteArray(value1Ptr)
		    && TclIsPureByteArray(value2Ptr)) {
		/*
		 * Use binary versions of comparisons since that won't cause undue
		 * type conversions and it is much faster. Only do this if we're
		 * case-sensitive (which is all that really makes sense with byte
		 * arrays anyway, and we have no memcasecmp() for some reason... :^)
		 */
		s1 = (char *) Tcl_GetByteArrayFromObj(value1Ptr, &s1len);
		s2 = (char *) Tcl_GetByteArrayFromObj(value2Ptr, &s2len);
		memCmpFn = memcmp;
	    } else if ((value1Ptr->typePtr == &tclStringType)
		    && (value2Ptr->typePtr == &tclStringType)) {
		/*
		 * Do a unicode-specific comparison if both of the args are of
		 * String type. If the char length == byte length, we can do a
		 * memcmp. In benchmark testing this proved the most efficient
		 * check between the unicode and string comparison operations.
		 */






		s1len = Tcl_GetCharLength(value1Ptr);
		s2len = Tcl_GetCharLength(value2Ptr);
		if ((s1len == value1Ptr->length)
			&& (value1Ptr->bytes != NULL)
			&& (s2len == value2Ptr->length)
			&& (value2Ptr->bytes != NULL)) {
		    s1 = value1Ptr->bytes;
................................................................................
			memCmpFn = memcmp;
			s1len *= sizeof(Tcl_UniChar);
			s2len *= sizeof(Tcl_UniChar);
		    } else {
			memCmpFn = (memCmpFn_t) Tcl_UniCharNcmp;
		    }
		}
	    } else {
		/*
		 * In order to handle the special Tcl \xC0\x80 null encoding
		 * for utf-8, strcmp can't do a simple memcmp.
		 */


		if ((empty = TclCheckEmptyString(value1Ptr)) > 0) {
		    switch (TclCheckEmptyString(value2Ptr)) {
			case -1:
			s1 = "";
			s1len = 0;
			s2 = TclGetStringFromObj(value2Ptr, &s2len);
			break;
			case 0:
			match = -1;
			goto matchdone;
			case 1:
			match = 0;
			goto matchdone;
		    }
		} else if (TclCheckEmptyString(value2Ptr) > 0) {
		    switch (empty) {
			case -1:
			s2 = "";
			s2len = 0;
			s1 = TclGetStringFromObj(value1Ptr, &s1len);
			break;
			case 0:
			match = 1;
			goto matchdone;
			case 1:
			match = 0;
			goto matchdone;
		    }
		} else {
		    s1 = TclGetStringFromObj(value1Ptr, &s1len);
		    s2 = TclGetStringFromObj(value2Ptr, &s2len);
		}

		if (checkEq) {





		    memCmpFn = memcmp;
		} else {









		    memCmpFn = (memCmpFn_t) TclpUtfNcmp2;




		}
	    }














	    if (checkEq && (s1len != s2len)) {
		match = 1;

	    }  else {
		/*
		 * The comparison function should compare up to the minimum
		 * byte length only.
		 */
		match = memCmpFn(s1, s2,
			(size_t) ((s1len < s2len) ? s1len : s2len));

		if (match == 0) {
		    match = s1len - s2len;
		}

	    }
	}
    matchdone:
    return match;
}
 
int TclStringCmpOpts (
    Tcl_Interp *interp,		/* Current interpreter. */
    int objc,			/* Number of arguments. */
    Tcl_Obj *const objv[],	/* Argument objects. */
    int *reqlength,
    char **stringPtr,
    int *length,
    int *nocase


)
{
    int i;
    const char *string = *stringPtr;

    *reqlength = -1;
    *nocase = 0;
    if (objc < 3 || objc > 6) {
    str_cmp_args:
	Tcl_WrongNumArgs(interp, 1, objv,
		"?-nocase? ?-length int? string1 string2");
	return TCL_ERROR;
    }

    for (i = 1; i < objc-2; i++) {
	string = TclGetStringFromObj(objv[i], length);
	if ((*length > 1) && !strncmp(string, "-nocase", (size_t)*length)) {
	    *nocase = 1;
	} else if ((*length > 1)
		&& !strncmp(string, "-length", (size_t)*length)) {
	    if (i+1 >= objc-2) {
		goto str_cmp_args;
	    }
	    i++;
	    if (TclGetIntFromObj(interp, objv[i], reqlength) != TCL_OK) {
		return TCL_ERROR;
	    }






|
|
<
<







 







|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<







 







|
|
|
|
|
<
<



<
<
<
<

<
<
<
<
<
<
>
|
|
|
|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<



<
>
>
>

|
|
|

>

<
<
<
>

<
<
|
|
>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|

>
>
>
>
>







 







<
<
<
<
<
|
>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<
|
>
>
>
>
>
|
|
>
>
>
>
>
>
>
>
>

>
>
>
>


|
>
>
>
>
>
>
>
>
>
>
>
>
>
|
<
>
|
|
|
|
|
|
<
>
|
|
|
>
|
<








<
<
<
|
>
|
<

|
|











|
|

|
|







2595
2596
2597
2598
2599
2600
2601
2602
2603


2604
2605
2606
2607
2608
2609
2610
....
2635
2636
2637
2638
2639
2640
2641
2642







































































2643
2644
2645
2646
2647
2648
2649
....
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683


2684
2685
2686




2687






2688
2689
2690
2691
2692



























































2693
2694
2695

2696
2697
2698
2699
2700
2701
2702
2703
2704
2705



2706
2707


2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
....
2755
2756
2757
2758
2759
2760
2761





2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795

2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834

2835
2836
2837
2838
2839
2840
2841

2842
2843
2844
2845
2846
2847

2848
2849
2850
2851
2852
2853
2854
2855



2856
2857
2858

2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
{
    /*
     * Remember to keep code here in some sync with the byte-compiled versions
     * in tclExecute.c (INST_STR_EQ, INST_STR_NEQ and INST_STR_CMP as well as
     * the expr string comparison in INST_EQ/INST_NEQ/INST_LT/...).
     */

    const char *string2;
    int length2, i, match, nocase = 0, reqlength = -1;



    if (objc < 3 || objc > 6) {
    str_cmp_args:
	Tcl_WrongNumArgs(interp, 1, objv,
		"?-nocase? ?-length int? string1 string2");
	return TCL_ERROR;
    }
................................................................................
    /*
     * From now on, we only access the two objects at the end of the argument
     * array.
     */

    objv += objc-2;

    match = TclStringCmp (objv[0], objv[1], 0, nocase, reqlength);








































































    Tcl_SetObjResult(interp, Tcl_NewBooleanObj(match ? 0 : 1));
    return TCL_OK;
}
 
/*
 *----------------------------------------------------------------------
................................................................................
{
    /*
     * Remember to keep code here in some sync with the byte-compiled versions
     * in tclExecute.c (INST_STR_EQ, INST_STR_NEQ and INST_STR_CMP as well as
     * the expr string comparison in INST_EQ/INST_NEQ/INST_LT/...).
     */

    int match, nocase, reqlength, status;

    if ((status = TclStringCmpOpts(interp, objc, objv, &nocase, &reqlength))
	!= TCL_OK) {



	return status;
    }





    objv += objc-2;






    match = TclStringCmp (objv[0], objv[1], 0, nocase, reqlength);
    Tcl_SetObjResult(interp, Tcl_NewIntObj(match));
    return TCL_OK;
}
 



























































int TclStringCmp (
	Tcl_Obj *value1Ptr,
	Tcl_Obj *value2Ptr,

	int checkEq,        /* comparison is only for equality */
	int nocase,	    /* comparison is not case sensitive */
	int reqlength	    /* requested length */
) {
    char *s1, *s2;
    int empty, length, match, s1len, s2len;
    memCmpFn_t memCmpFn;

    if ((reqlength == 0) || (value1Ptr == value2Ptr)) {
	/*



	 * Always match at 0 chars of if it is the same obj.
	 */


	match = 0;
    } else {

	if (!nocase && TclIsPureByteArray(value1Ptr)
		&& TclIsPureByteArray(value2Ptr)) {
	    /*
	     * Use binary versions of comparisons since that won't cause undue
	     * type conversions and it is much faster. Only do this if we're
	     * case-sensitive (which is all that really makes sense with byte
	     * arrays anyway, and we have no memcasecmp() for some reason... :^)
	     */
	    s1 = (char *) Tcl_GetByteArrayFromObj(value1Ptr, &s1len);
	    s2 = (char *) Tcl_GetByteArrayFromObj(value2Ptr, &s2len);
	    memCmpFn = memcmp;
	} else if ((value1Ptr->typePtr == &tclStringType)
		&& (value2Ptr->typePtr == &tclStringType)) {
	    /*
	     * Do a unicode-specific comparison if both of the args are of
	     * String type. If the char length == byte length, we can do a
	     * memcmp. In benchmark testing this proved the most efficient
	     * check between the unicode and string comparison operations.
	     */

	    if (nocase) {
		s1 = (char *) Tcl_GetUnicodeFromObj(value1Ptr, &s1len);
		s2 = (char *) Tcl_GetUnicodeFromObj(value2Ptr, &s2len);
		memCmpFn = (memCmpFn_t)Tcl_UniCharNcasecmp;
	    } else {
		s1len = Tcl_GetCharLength(value1Ptr);
		s2len = Tcl_GetCharLength(value2Ptr);
		if ((s1len == value1Ptr->length)
			&& (value1Ptr->bytes != NULL)
			&& (s2len == value2Ptr->length)
			&& (value2Ptr->bytes != NULL)) {
		    s1 = value1Ptr->bytes;
................................................................................
			memCmpFn = memcmp;
			s1len *= sizeof(Tcl_UniChar);
			s2len *= sizeof(Tcl_UniChar);
		    } else {
			memCmpFn = (memCmpFn_t) Tcl_UniCharNcmp;
		    }
		}





	    }
	} else {
	    if ((empty = TclCheckEmptyString(value1Ptr)) > 0) {
		switch (TclCheckEmptyString(value2Ptr)) {
		    case -1:
		    s1 = "";
		    s1len = 0;
		    s2 = TclGetStringFromObj(value2Ptr, &s2len);
		    break;
		    case 0:
		    match = -1;
		    goto matchdone;
		    case 1:
		    match = 0;
		    goto matchdone;
		}
	    } else if (TclCheckEmptyString(value2Ptr) > 0) {
		switch (empty) {
		    case -1:
		    s2 = "";
		    s2len = 0;
		    s1 = TclGetStringFromObj(value1Ptr, &s1len);
		    break;
		    case 0:
		    match = 1;
		    goto matchdone;
		    case 1:
		    match = 0;
		    goto matchdone;
		}
	    } else {
		s1 = TclGetStringFromObj(value1Ptr, &s1len);
		s2 = TclGetStringFromObj(value2Ptr, &s2len);
	    }

	    if (!nocase && checkEq) {
		/*
		 * When we have equal-length we can check only for (in)equality.
		 * We can use memcmp in all (n)eq cases because we
		 * don't need to worry about lexical LE/BE variance.
		 */
		memCmpFn = memcmp;
	    } else {

		/*
		 * As a catch-all we will work with UTF-8. We cannot use memcmp() as
		 * that is unsafe with any string containing NUL (\xC0\x80 in Tcl's
		 * utf rep). We can use the more efficient TclpUtfNcmp2 if we are
		 * case-sensitive and no specific length was requested.
		 */

		if ((reqlength < 0) && !nocase) {
		    memCmpFn = (memCmpFn_t) TclpUtfNcmp2;
		} else {
		    s1len = Tcl_NumUtfChars(s1, s1len);
		    s2len = Tcl_NumUtfChars(s2, s2len);
		    memCmpFn = (memCmpFn_t) (nocase ? Tcl_UtfNcasecmp : Tcl_UtfNcmp);
		}
	    }
	}

	length = (s1len < s2len) ? s1len : s2len;
	if (reqlength > 0 && reqlength < length) {
	    length = reqlength;
	} else if (reqlength < 0) {
	    /*
	     * The requested length is negative, so we ignore it by setting it to
	     * length + 1 so we correct the match var.
	     */

	    reqlength = length + 1;
	}

	if (checkEq && (s1len != s2len)) {

	    match = 1;		/* This will be reversed below. */
	}  else {
	    /*
	     * The comparison function should compare up to the minimum
	     * byte length only.
	     */
	    match = memCmpFn(s1, s2, (size_t) length);

	}
	if ((match == 0) && (reqlength > length)) {
	    match = s1len - s2len;
	}
	match = (match > 0) ? 1 : (match < 0) ? -1 : 0;
    }

    matchdone:
    return match;
}
 
int TclStringCmpOpts (
    Tcl_Interp *interp,		/* Current interpreter. */
    int objc,			/* Number of arguments. */
    Tcl_Obj *const objv[],	/* Argument objects. */



    int *nocase,
    int *reqlength
)

{
    int i, length;
    const char *string;

    *reqlength = -1;
    *nocase = 0;
    if (objc < 3 || objc > 6) {
    str_cmp_args:
	Tcl_WrongNumArgs(interp, 1, objv,
		"?-nocase? ?-length int? string1 string2");
	return TCL_ERROR;
    }

    for (i = 1; i < objc-2; i++) {
	string = TclGetStringFromObj(objv[i], &length);
	if ((length > 1) && !strncmp(string, "-nocase", (size_t)length)) {
	    *nocase = 1;
	} else if ((length > 1)
		&& !strncmp(string, "-length", (size_t)length)) {
	    if (i+1 >= objc-2) {
		goto str_cmp_args;
	    }
	    i++;
	    if (TclGetIntFromObj(interp, objv[i], reqlength) != TCL_OK) {
		return TCL_ERROR;
	    }

Changes to generic/tclExecute.c.

5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
    stringCompare:
	value2Ptr = OBJ_AT_TOS;
	valuePtr = OBJ_UNDER_TOS;

	{
	    int checkEq = ((*pc == INST_EQ) || (*pc == INST_NEQ)
		    || (*pc == INST_STR_EQ) || (*pc == INST_STR_NEQ));
	    match = TclStringCmp(valuePtr, value2Ptr, checkEq);
	}

	/*
	 * Make sure only -1,0,1 is returned
	 * TODO: consider peephole opt.
	 */







|







5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
    stringCompare:
	value2Ptr = OBJ_AT_TOS;
	valuePtr = OBJ_UNDER_TOS;

	{
	    int checkEq = ((*pc == INST_EQ) || (*pc == INST_NEQ)
		    || (*pc == INST_STR_EQ) || (*pc == INST_STR_NEQ));
	    match = TclStringCmp(valuePtr, value2Ptr, checkEq, 0, -1);
	}

	/*
	 * Make sure only -1,0,1 is returned
	 * TODO: consider peephole opt.
	 */

Changes to generic/tclInt.h.

3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
			    Tcl_Obj *const *objv, int objc, int subIdx,
			    Tcl_Obj *bad, Tcl_Obj *fix);
MODULE_SCOPE void *	TclStackRealloc(Tcl_Interp *interp, void *ptr,
			    int numBytes);

typedef int (*memCmpFn_t)(const void*, const void*, size_t);
MODULE_SCOPE int	TclStringCmp (Tcl_Obj *value1Ptr, Tcl_Obj *value2Ptr,
			    int checkEq);
MODULE_SCOPE int	TclStringCmpOpts (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[],
			    int *reqlength, char **stringPtr, int *length2, int *nocase);
MODULE_SCOPE int	TclStringMatch(const char *str, int strLen,
			    const char *pattern, int ptnLen, int flags);
MODULE_SCOPE int	TclStringMatchObj(Tcl_Obj *stringObj,
			    Tcl_Obj *patternObj, int flags);
MODULE_SCOPE Tcl_Obj *	TclStringObjReverse(Tcl_Obj *objPtr);
MODULE_SCOPE void	TclSubstCompile(Tcl_Interp *interp, const char *bytes,
			    int numBytes, int flags, int line,






|

|







3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
			    Tcl_Obj *const *objv, int objc, int subIdx,
			    Tcl_Obj *bad, Tcl_Obj *fix);
MODULE_SCOPE void *	TclStackRealloc(Tcl_Interp *interp, void *ptr,
			    int numBytes);

typedef int (*memCmpFn_t)(const void*, const void*, size_t);
MODULE_SCOPE int	TclStringCmp (Tcl_Obj *value1Ptr, Tcl_Obj *value2Ptr,
			    int checkEq, int nocase, int reqlength);
MODULE_SCOPE int	TclStringCmpOpts (Tcl_Interp *interp, int objc, Tcl_Obj *const objv[],
			    int *nocase, int *reqlength);
MODULE_SCOPE int	TclStringMatch(const char *str, int strLen,
			    const char *pattern, int ptnLen, int flags);
MODULE_SCOPE int	TclStringMatchObj(Tcl_Obj *stringObj,
			    Tcl_Obj *patternObj, int flags);
MODULE_SCOPE Tcl_Obj *	TclStringObjReverse(Tcl_Obj *objPtr);
MODULE_SCOPE void	TclSubstCompile(Tcl_Interp *interp, const char *bytes,
			    int numBytes, int flags, int line,