You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Rob Hartill <ha...@hyperreal.com> on 1995/11/24 21:38:35 UTC
Re: ICP client/server in Perl (fwd)
No ack sent
Forwarded message:
> From martin@mrrl.lut.ac.uk Thu Nov 23 10:29:03 1995
> Message-Id: <19...@gizmo.lut.ac.uk>
> To: Peter Danzig <da...@catarina.usc.edu>
> cc: cybercache@mrrl.lut.ac.uk, apache-bugs@mail.apache.org
> Subject: Re: ICP client/server in Perl
> X-Mailer: exmh version 1.6.4 10/10/95
> X-URI: <URL:http://www.roads.lut.ac.uk/~martin>
> In-reply-to: Your message of "Wed, 22 Nov 1995 18:14:42 PST."
> <19...@warthog.usc.edu>
> Mime-Version: 1.0
> Content-Type: text/plain
> Date: Thu, 23 Nov 1995 18:26:48 +0000
> From: Martin Hamilton <ma...@mrrl.lut.ac.uk>
>
>
> Peter Danzig writes:
>
> | Currently, the cache would return your MD5 checksum as part
> | of the object's MIME header. While I agree that it would be
> | great for consistency if the cache re-computed the object's MD5,
> | do you think we proxy builders can rely on the browsers to detect
> | corrupted objects and to automatically issue a refresh?
>
> I like the model of browsers using MD5 to tell their users whether
> the object got through OK. Guess we'll need most of the popular
> servers to support it before the browser authors catch on. I
> might have a go at hacking it into the odd one...
>
> Content-MD5 would probably be good for caches even if they avoided
> having to calculate the checksum themselves ? e.g. cache the MD5
> checksum provided by the server when the cached object was
> originally fetched, and it becomes a reliable indicator of the
> object having changed... There's probably also some mileage in
> using MD5s where we currently use URLs ? - cf. Keith Moore's
> "Location Independent Filenames"
>
> Cheerio,
>
> Martin
>
> PS Here's a quick hack which adds support for the "Content-MD5:"
> Header (cf. RFC 1864) to Apache httpd 0.8.14.
>
> e.g. (> indicates client, < indicates server)
>
> >GET /~martin/check.txt HTTP/1.0
> >Accept: */*
> >
> <HTTP/1.0 200 OK
> <Date: Thu, 23 Nov 1995 17:59:06 GMT
> <Server: Apache/0.8.14
> <Content-type: text/plain
> <Content-MD5: g5A4uaZgbiUx5+vMX0eHMQ==
> <Content-length: 19
> <Last-modified: Wed, 22 Nov 1995 18:54:59 GMT
> <
> <Squeamish OSIfrage
>
> To build a Content-MD5 enabled server,
>
> o add -DCONTENT_MD5 to the CFLAGS in src/Configuration,
> o copy md5.h, globals.h, and md5c.c from RSA's MD5 package
> (or from the NCSA httpd 1.5 source directory!)
> o copy the new file md5.c (see below) into the src
> directory
> o add md5.o and md5c.o to the list of objects to compile
> in src/Makefile.tmpl
> o now run ./Configure and make as usual
>
> MD5 checksums are only supplied for static files
>
> A drawback with this approach is that the amount of work
> required to calculate the MD5 checksum of a large object can be
> significant. The obvious next step would be to cache MD5
> checksums and file modification times (perhaps in a DBM style
> database?)- and only recalculate the checksum if the object has
> been modfied since the last visit
>
>
> *** src/http_core.c.FCS Thu Nov 23 18:06:47 1995
> --- src/http_core.c Thu Nov 23 17:32:08 1995
> ***************
> *** 677,682 ****
> --- 677,686 ----
> return FORBIDDEN;
> }
>
> + #ifdef CONTENT_MD5
> + r->content_md5 = (char *)md5digest(f);
> + #endif
> +
> soft_timeout ("send", r);
>
> send_http_header (r);
>
> *** src/http_protocol.c.FCS Thu Nov 23 17:24:33 1995
> --- src/http_protocol.c Thu Nov 23 17:32:34 1995
> ***************
> *** 515,520 ****
> --- 515,525 ----
>
> if (r->content_language)
> fprintf (fd, "Content-language: %s\015\012", r->content_language);
> +
> + #ifdef CONTENT_MD5
> + if (r->content_md5)
> + fprintf (fd, "Content-MD5: %s\015\012", r->content_md5);
> + #endif
>
> for (i = 0; i < hdrs_arr->nelts; ++i) {
> if (!hdrs[i].key) continue;
>
> *** src/httpd.h.FCS Thu Nov 23 17:20:36 1995
> --- src/httpd.h Thu Nov 23 17:32:59 1995
> ***************
> *** 306,311 ****
> --- 306,314 ----
> char *content_type; /* Break these out --- we dispatch on 'em */
> char *content_encoding;
> char *content_language;
> + #ifdef CONTENT_MD5
> + char *content_md5;
> + #endif
>
> int no_cache;
>
>
>
> New file: src/md5.c ...
>
>
> #ifdef CONTENT_MD5
>
> #include <stdio.h>
> #include <string.h>
> #include "global.h"
> #include "md5.h"
>
> /* these portions extracted from mpack, John G. Myers - jgm+@cmu.edu */
>
> /* (C) Copyright 1993,1994 by Carnegie Mellon University
> * All Rights Reserved.
> *
> * Permission to use, copy, modify, distribute, and sell this software
> * and its documentation for any purpose is hereby granted without
> * fee, provided that the above copyright notice appear in all copies
> * and that both that copyright notice and this permission notice
> * appear in supporting documentation, and that the name of Carnegie
> * Mellon University not be used in advertising or publicity
> * pertaining to distribution of the software without specific,
> * written prior permission. Carnegie Mellon University makes no
> * representations about the suitability of this software for any
> * purpose. It is provided "as is" without express or implied
> * warranty.
> *
> * CARNEGIE MELLON UNIVERSITY DISCLAIMS ALL WARRANTIES WITH REGARD TO
> * THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
> * AND FITNESS, IN NO EVENT SHALL CARNEGIE MELLON UNIVERSITY BE LIABLE
> * FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
> * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN
> * AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING
> * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
> * SOFTWARE.
> */
> /*
> Copyright (c) 1991 Bell Communications Research, Inc. (Bellcore)
>
> Permission to use, copy, modify, and distribute this material
> for any purpose and without fee is hereby granted, provided
> that the above copyright notice and this permission notice
> appear in all copies, and that the name of Bellcore not be
> used in advertising or publicity pertaining to this
> material without the specific, prior written permission
> of an authorized representative of Bellcore. BELLCORE
> MAKES NO REPRESENTATIONS ABOUT THE ACCURACY OR SUITABILITY
> OF THIS MATERIAL FOR ANY PURPOSE. IT IS PROVIDED "AS IS",
> WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES. */
>
> static char basis_64[] =
> "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
>
> char *md5contextTo64(context)
> MD5_CTX *context;
> {
> unsigned char digest[18];
> char *encodedDigest;
> int i;
> char *p;
>
> encodedDigest = (char *)malloc(25 * sizeof(char));
>
> MD5Final(digest, context);
> digest[sizeof(digest)-1] = digest[sizeof(digest)-2] = 0;
>
> p = encodedDigest;
> for (i=0; i < sizeof(digest); i+=3) {
> *p++ = basis_64[digest[i]>>2];
> *p++ = basis_64[((digest[i] & 0x3)<<4) | ((digest[i+1] & 0xF0)>>4)];
> *p++ = basis_64[((digest[i+1] & 0xF)<<2) | ((digest[i+2] & 0xC0)>>6)];
> *p++ = basis_64[digest[i+2] & 0x3F];
> }
> *p-- = '\0';
> *p-- = '=';
> *p-- = '=';
> return encodedDigest;
> }
>
>
> char *md5digest(infile)
> FILE *infile;
> {
> MD5_CTX context;
> char buf[1000];
> long length = 0;
> int nbytes;
>
> MD5Init(&context);
> while (nbytes = fread(buf, 1, sizeof(buf), infile)) {
> length += nbytes;
> MD5Update(&context, buf, nbytes);
> }
> rewind(infile);
> return md5contextTo64(&context);
> }
> #endif /* CONTENT_MD5 */
>
>
>