Usenet 1994 October/usenetsourcesnewsgroupsinfomagicoctober1994disk2.iso/misc/volume11/uudec/part01

Text File | 1990-04-06 | 6KB | 141 lines

Newsgroups: comp.sources.misc
subject: v11i097: latest uudec.cc for the world
From: [e-mail].P1 (Barry Gorman )
Sender: [e-mail] (Brandon S. Allbery - comp.sources.misc)

Posting-number: Volume 11, Issue 97
Submitted-by: [e-mail].P1 (Barry Gorman )
Archive-name: uudec/part01

Some little time ago, I asked a good friend on a unix site to send me a copy
of a uudecoder program. I was sent the version by scott@max (see below).
Needless to say I had a fiddle with it, and I append the result of this.
I would claim that the transformed program is more "structured" (what is this?)
and certainly shorter. I have used my personal layout rules (indent=4*{'s)
which will no doubt be controversial. Please feel free to use, distribute etc.

Please do not attempt to reply via the uucp gateway @ uk.ac.ukc; but earn-relay
and nsfnet-relay are ok. Alternatively send via [e-mail] for forwarding

Barry
.....
----------------------------------- cut here -----------------------------------
#! /bin/sh
# This file was wrapped with "dummyshar".  "sh" this file to extract.
# Contents:  uudec.c
echo extracting 'uudec.c'
if test -f 'uudec.c' -a -z "$1"; then echo Not overwriting 'uudec.c'; else
sed 's/^X//' << \EOF > 'uudec.c'
X/********** From: [e-mail] - 20 Feb 90 21:13:03 GMT *********/
X/* For the those may not have access to a UUDECODER, you may not
X * be able to take benefit of like UX-Maze file exchange
X * which require the files to be uuencoded.
X * If that is the case may be just may be the following posts
X * will be of help. This post contains the source code in
X * C of a UUDECODER tested on msc 3.0 and PCDOS 3.1.
X * Perhaps if you have access to an IBM/CLONE you could do
X * the file uudecoding on the IBM before the final transfer
X * to the C64 or C128.
X * Or perhaps if you are knowledgeable with programing you
X * could convert this program for the C64 or perhaps to another
X * popular language like Pascal or Fortran that is supported
X * by your system.
X
X/* source code reduced in size whilst maintaining the exact operation
X * of the original program: [e-mail].P1 March 19th 1990.
X * Probably favours machines capable of 32 bit working; but should
X * work on others. The "register unsigned long word" variable is used
X * for some long (16 place) shifts. This may or may not be faster than
X * the use of division, depending on you processor design. I have left
X * it in this form as I feel it is in the spirit of C and illustrates
X * my technique. (bring back 24 bit machines with cyclic shifts!)
X
X/* usage for this file is as follows:
X * uudecode <input file name (with extension)>
X *     file is placed in the file name contained on the
X *     first line of the uuencoded file ...
X *
X * The program expects this to look like begin XXX <FN>
X * where XXX are the Unix(tm) file permissions
X *   and <fn> is the Unix file name.  If this name is
X *       too long for DOS, edit it to fit.  It has been
X *       tested with msc v3.0 and PC-DOS 3.1.  the companion
X *       file is uuencode.c ...
X
X/*************************** single character decode **************************/
X
X#define DEC(c) (((c) - ' ') & 077)
X#include <stdio.h>
X
X/************************************ MAIN ************************************/
Xmain (argc,argv) int argc; char *argv[];      /* expects one (or no) argument */
X/******************************************************************************/
X
X   {FILE *in, *out; char buf[80];
X
X    switch(argc)                     /* sort out number of arguments supplied */
X
X       {default: fprintf(stderr,"Usage: uudec [file]n"); exit(2); /* 2+ args */
X
X        case 1: in=stdin; break;                    /* take source from stdin */
X
X        case 2: if(!(in=fopen(argv[1],"r"))) {perror(argv[1]); exit(1);}}
X
X/***** now that the source has been identified, scan for the "begin" line *****/
X
X    do if(!(fgets(buf,sizeof buf,in)))             /* read input line by line */
X
X       {fprintf(stderr, "No "begin" linen"); exit(3);}  /* complain if end */
X
X    while(strncmp(buf,"begin ",6));        /* until the "begin" line is found */
X
X/* the begin line must be followed by three (octal) digits and the file name **/
X
X       {char dest[32]; int mode;    /* these variables are needed transiently */
X
X        sscanf(buf,"begin %o %s",&mode,dest);   /* extract mode and file name */
X
X        if(!(out=fopen(dest,"w"))) {perror(dest); exit(4);}} /* open outfile? */
X
X/** having opened the file and read the "begin" line, decode it line by line **/
X
X        {char *p; int n;                  /* line pointer and character count */
X
X         while((p=fgets(buf,sizeof buf,in)) ? (n=DEC(buf[0]))    /* read line */
X
X                : (fprintf(stderr,"Short filen"), exit(10)))  /* end of file */
X
X           {register unsigned long word; int k;            /* local workspace */
X
X            do {k=4; do word=word<<6|DEC(*++p); while(--k);  /* decode 4 to 3 */
X
X                k=3; do {putc(word>>16,out); word<<=8;} while(--n && --k);}}
X
X            while(n);}}                 /* keep going until all decoding done */
X
X/* the decode should now be complete, we have read the zero length byte count */
X
X    if(!(fgets(buf,sizeof buf,in)) || strncmp(buf,"end",3))  /* check for end */
X
X       {fprintf(stderr,"No "end" linen"); exit(5);}}   /* complain if none */
X
X/******************************** end of uudec ********************************/
EOF
chars=`wc -c < 'uudec.c'`
if test $chars !=     4181; then echo 'uudec.c' is $chars characters, should be     4181 characters!; fi
fi
exit 0


---
===============================================================================
John Cullen                     || JANET : [e-mail]
System Support Group            || ARPA  : [e-mail]
Informatics Department          || BITNET: jjsc@ukacrl (I think!)
Rutherford Appleton Laboratory  || UUCP  : {...!mcvax}!ukc!rlinf!jjsc
Chilton, Didcot, Oxon. OX11 0QX || VOICE : [phone] ext 6555
===============================================================================
Your fortune cookie says:
Birth: The first and direst of all disasters.