Re: hhctrl.ocx: Parse HTML entities in the table of contents.

[Alexandre Julliard]

Robert Shearman  writes:
> -
>            /* hexadecimal entity */
> -
>            while ((*p >= '0' && *p <= '9') || (*p >= 'a' && *p <= 'f') ||
> -
>                   (*p >= 'A' && *p <= 'F') || (*p == ';'))
> -
>                p++;

You should exit the loop at the first ';'. Also you have to increment
p first.
> -
>        while (isalpha(*p))
> -
>            p++;

You can't use ASCII ctype functions on Unicode chars.
> -
>        for (i = 0; i < sizeof(char_refs)/sizeof(char_refs[0]); i++)
> -
>            if (p - start - 1 <= sizeof(char_refs[i].name))
> -
>            {
> -
>                if (!strncmpW(char_refs[i].name, start + 1, p - start - 1))
> -
>                    break;
> -
>            }

The sizeof check doesn't make much sense, especially inside the
loop. What you need is to check that you reached the end of the string
if the strncmpW succeeded.

hhctrl.ocx: Parse HTML entities in the table of contents.

[Robert Shearman]

Alexandre Julliard wrote:
> Robert Shearman  writes:
> > -
> >            /* hexadecimal entity */
> > -
> >            while ((*p >= '0' && *p <= '9') || (*p >= 'a' && *p <= 'f') ||
> > -
> >                   (*p >= 'A' && *p <= 'F') || (*p == ';'))
> > -
> >                p++;
>
> You should exit the loop at the first ';'. Also you have to increment
> p first.
> > -
> >        while (isalpha(*p))
> > -
> >            p++;
>
> You can't use ASCII ctype functions on Unicode chars.

Thanks, I'll send an updated patch which should address these comments.
> > -
> >        for (i = 0; i < sizeof(char_refs)/sizeof(char_refs[0]); i++)
> > -
> >            if (p - start - 1 <= sizeof(char_refs[i].name))
> > -
> >            {
> > -
> >                if (!strncmpW(char_refs[i].name, start + 1, p - start - 1))
> > -
> >                    break;
> > -
> >            }
>
> The sizeof check doesn't make much sense, especially inside the
> loop. What you need is to check that you reached the end of the string
> if the strncmpW succeeded.

Again, I'm not sure what you mean here. I know (start + 1) -> p will not 
have any nul's in it, therefore strncmpW will return non-zero if 
char_refs[i].name has a nul in it before (p - start - 1) characters. I 
wrote a small test program to prove this:
    static const WCHAR str1[] = {'t','e','s','t',0};
    static const WCHAR str2[] = {'t','e','s','t','s','t','r','i','n','g',0};
    printf("strncmpW(str1, str2, sizeof(str2)/sizeof(str2[0])-1) = 
%d\n", strncmpW(str1, str2, sizeof(str2)/sizeof(str2[0])-1));
It outputs:
strncmpW(str1, str2, sizeof(str2)/sizeof(str2[0])-1) = -115

So I'm what other cases I need to handle where strncmpW succeeds when it 
shouldn't.