You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Martin Kraemer <Ma...@mch.sni.de> on 1999/04/20 20:56:33 UTC
Re: cvs commit: apache-1.3/src/modules/standard mod_mime.c
Dean writes,...
> dgaudet 99/04/20 10:27:51
>
> Modified: src CHANGES
> src/modules/standard mod_mime.c
> Log:
> islower() returns true for stuff outside a-z when locale isn't C
>...
> Index: mod_mime.c
> ===================================================================
> RCS file: /home/cvs/apache-1.3/src/modules/standard/mod_mime.c,v
> retrieving revision 1.48
> retrieving revision 1.49
> diff -u -r1.48 -r1.49
> --- mod_mime.c 1999/01/01 19:05:11 1.48
> +++ mod_mime.c 1999/04/20 17:27:51 1.49
> @@ -233,13 +233,8 @@
> +#define MIME_HASHSIZE (32)
> +#define hash(i) (ap_tolower(i) % MIME_HASHSIZE)
Oops! While is makes sense to delimit the possible hash index by using
modulo, I expect better hash results for EBCDIC by using a slightly larger
prime (EBCDIC('z') - EBCDIC('a') + 1 + 1 = 42! What a coincidence ;-)
Maybe I should put an #ifdef CHARSET_EBCDIC around it.
Martin
--
<Ma...@MchP.Siemens.De> | Siemens Information and
Phone: +49-89-636-46021 | Communication Products
FAX: +49-89-636-47816 | 81730 Munich, Germany
Re: cvs commit: apache-1.3/src/modules/standard mod_mime.c
Posted by Dean Gaudet <dg...@arctic.org>.
On Tue, 20 Apr 1999, Martin Kraemer wrote:
> Dean writes,...
> > dgaudet 99/04/20 10:27:51
> >
> > Modified: src CHANGES
> > src/modules/standard mod_mime.c
> > Log:
> > islower() returns true for stuff outside a-z when locale isn't C
> >...
> > Index: mod_mime.c
> > ===================================================================
> > RCS file: /home/cvs/apache-1.3/src/modules/standard/mod_mime.c,v
> > retrieving revision 1.48
> > retrieving revision 1.49
> > diff -u -r1.48 -r1.49
> > --- mod_mime.c 1999/01/01 19:05:11 1.48
> > +++ mod_mime.c 1999/04/20 17:27:51 1.49
> > @@ -233,13 +233,8 @@
> > +#define MIME_HASHSIZE (32)
> > +#define hash(i) (ap_tolower(i) % MIME_HASHSIZE)
>
> Oops! While is makes sense to delimit the possible hash index by using
> modulo, I expect better hash results for EBCDIC by using a slightly larger
> prime (EBCDIC('z') - EBCDIC('a') + 1 + 1 = 42! What a coincidence ;-)
> Maybe I should put an #ifdef CHARSET_EBCDIC around it.
It's actually a pretty bogus hash regardless... and it's only there due to
laziness. Given that the complete set of extensions is known at
configuration time someone could have easily built a binary search table
(or something better). I don't think it really matters much...
Dean