#From cs.utexas.edu!uunet!sparky!kent Sat Aug  1 22:18:55 CDT 1992
#Article: 3798 of comp.sources.misc
#Newsgroups: comp.sources.misc
#Path: cse.uta.edu!cs.utexas.edu!uunet!sparky!kent
#From: middle@cse.uta.edu  (William Middleton)
#Subject:  v31i070:  dlhunter - news binary extraction utility v2.02, Part01/01
#Message-ID: <1992Aug2.025304.9353@sparky.imd.sterling.com>
#Followup-To: comp.sources.d
#X-Md4-Signature: 31f01253e66c626d7d097e87421466c0
#Sender: kent@sparky.imd.sterling.com (Kent Landfield)
#Organization: Sterling Software
#Date: Sun, 2 Aug 1992 02:53:04 GMT
#Approved: kent@sparky.imd.sterling.com
#Lines: 1393
#
#Submitted-by: middle@cse.uta.edu  (William Middleton)
#Posting-number: Volume 31, Issue 70
#Archive-name: dlhunter/part01
#Environment: perl, news
#
#dlhunter is an automated news binary extraction utility.  
#
#dlhunter is designed to run as a bckgd process, preferably daily by cron.
#It hunts for uuencoded files in the binary news directories that you specify.
#It uses (preferably) links, in its own queue directory.  When a given picture 
#or tool has all parts posted, it attempts to assemble it, if successful the 
#complete binary is placed in a download directory.  A short blurb, with the 
#info from part 01, is placed in a blurb directory when a binary is successfully
#decoded.  All directories are expired according to your specification. Note 
#that the download directory should have lots of space available.  
#
#This version has a -i flag which starts dlhunter in interactive mode, to allow
#the admin (user) to manually assemble postings with ungrokkable subject lines.
#
#You (admin preferably) MUST edit the configuration area in the script.  You 
#MUST also create the relevant directories. I've tried to make it as easy as 
#possible.  Send me mail if you have trouble.  Alternatively, learn perl.
#
#---
# /bin/sh
# This is a shell archive.  Remove anything before this line, then feed it
# into a shell via "sh file" or similar.  To overwrite existing files,
# type "sh file -c".
# The tool that generated this appeared in the comp.sources.unix newsgroup;
# send mail to comp-sources-unix@uunet.uu.net if you want that tool.
# Contents:  README dlhunter dlhunter.1 myuud.c
# Wrapped by kent@sparky on Sat Aug  1 21:34:33 1992
PATH=/bin:/usr/bin:/usr/ucb ; export PATH
echo If this archive is complete, you will see the following message:
echo '          "shar: End of archive 1 (of 1)."'
if test -f 'README' -a "${1}" != "-c" ; then 
  echo shar: Will not clobber existing file \"'README'\"
else
  echo shar: Extracting \"'README'\" \(1913 characters\)
  sed "s/^X//" >'README' <<'END_OF_FILE'
X
XDLHUNTER - 
X  dlhunter is designed to run as a bckgd process, preferably daily
Xby cron.  It hunts for uuencoded files in the binary news directories
Xthat you specify.  It uses (preferably) links, in its own queue directory.
XWhen a given picture/tool has all parts posted, it attempts to
Xassemble it, if successful the complete binary is placed in a download
Xdirectory.  A short blurb, with the info from part 01, is placed in a
Xblurb directory when a binary is successfully decoded.  All directories
Xare expired according to your specification. Note that the download
Xdirectory should have lots of space available.  And watch the errors
Xfiles, they grow.
X  This second version has a -i flag which starts dlhunter in interactive 
Xmode, to allow the admin (user) to manually assemble postings with un-
Xgrokkable subject lines.
X  You (admin preferably) MUST edit the configuration area in the script.  
XYou MUST also create the relevant directories. I've tried
Xto make it as easy as possible.  Send me mail if you have trouble.
XAlternatively, learn perl.
X
X
X Update : July 17,1992
X    You must now specify your own decoder in the configuration area.
X
X Update : July 22, 1992
X    myuud.c is now bundled with the release.  
X
XTODO: 
X*  someone has already started to hack this thing to work across NNTP
X
X*  noone has beat my regexp match for parsing subject lines, yet.
X   Of course, that'd save more time in interactive mode.
X
X*  do an analogous thing for the news sources directories, unpacking and
X   tarzeeing them.  volunteers?  I'm gettin kinda bored.
X
X  Enjoy, and save disk space.  Hack at will, but let me know.
X  If you do hack you gotta share.
X
XThanks to:  
X     6o25@sdf.lonestar.org  ( Stephen Jones, the man with the idea )
X     sbw@naucse.cse.nau.edu ( Steve Wampler, turned me on to perl )
X            
X
XbILL
Xwjm@sdf.lonestar.org, middle@cse.uta.edu, wjmi@lightnin.lonestar.org
Xperl, what else?  hmmm?
X
END_OF_FILE
  if test 1913 -ne `wc -c <'README'`; then
    echo shar: \"'README'\" unpacked with wrong size!
  fi
  # end of 'README'
fi
if test -f 'dlhunter' -a "${1}" != "-c" ; then 
  echo shar: Will not clobber existing file \"'dlhunter'\"
else
  echo shar: Extracting \"'dlhunter'\" \(17539 characters\)
  sed "s/^X//" >'dlhunter' <<'END_OF_FILE'
X#!/usr/local/bin/perl
X#eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
X#if 0;
X#
X# July 4, 1992 - First release ( version 1.00 ) 
X#
X# July 7, 1992  -  First revision, ( version 1.01 ) 
X# repair an expire bug that deleted all
X# the files once the script completed.
X#
X# July 13, 1992 -  Second Revision, ( version 2.00 ) 
X# added several new features, including interactive mode
X# to allow the admin ( or his DA ) to easily grok through the
X# remaining links in the queue, assembling binaries with 
X# un-grokkable subject lines, or other things the daemon doesn't like. 
X# You can also (optionally) use a subdirectory structure for downloads.
X# Further, dlhunter has a better decoder now. 
X#  
X# July 16, 1992  - third revision,  (version 2.01)
X# well, i'm sick of try_to_decode screwing up, so i've decided to
X# use a user supplied routine as the exec.  you must now edit the 
X# configuration area to specify the path to the uudecoder of your preference.
X# dlhunter will pass all parts thereto, and use the return value of 0
X# as an indication of success.  there are many decoders out there, this
X# version is released with a call to uud.  ( maman@cosinus.inria.fr )
X# which works pretty good. The decoder must be of the type that will
X# strip all necessary junk.
X# 
X# July 22, 1992  - fourth revision  (version 2.02)
X# now includes a man page and README.  small bugs taken out of
X# sub hunt (interactive mode), which printed zero-length blurbs.
X# and died on expired links.  Now includes myuud, by permission.
X# added one more regexp to try to grok subject lines, and took
X# a little rigor out of the parts verification.
X# 
X#
X# 
X# --------------------------------------------------------------------- 
X# General Info:
X# dlhunter is designed to run as a bckgd process, by cron. 
X#
X# But it now has an interactive mode, invoked as dlhunter -i.  
X# you will see the older links in the queue printed on the screen,
X# with their corresponding subject lines.  These are the ones that dlhunter,
X# or try_to_decode couldnt handle.  You must give filenumbers to the prompt
X# after the listing.  It will then try to do its regular thing with them.
X# If successful, the links will be removed from the queue. 
X#
X# You must edit the following variables.   MUST!
X# You must create relevant directories if necessary.  MUST!
X# Your queue must be on the same fs as news. 
X# if you insist on using symlinks, clear the queue before each run.
X#
X# dlhunter will find binaries in any dir you care to give it.
X# 
X#***************************  Edit these   ***********************************
X#******************* and create relevant directories!!! ***********************
X#
X#  Which decoder do you wanna use?  It's got to be the kind that will
X# strip off all the junk!  the myuud bundled is good, use the -v option.
X$usr_decode = '/usr/local/bin/myuud  -v ';
X
X#  where are the binaries directories, the news directories where
X#  binary files are posted 
X@tooldirs =(
X         '/var/spool/news/alt/binaries/pictures',
X         '/var/spool/news/alt/binaries/pictures/erotica',
X         '/var/spool/news/alt/binaries/pictures/misc',
X         '/var/spool/news/alt/binaries/pictures/utilities',
X         '/var/spool/news/comp/binaries/os2',
X         '/var/spool/news/comp/binaries/ibm/pc'
X           );
X
X# we need an extension for each tooldir, to assure uniqueness in the queue
X@exts = ('.pc1','.ero','.misc','.util','.os2','.ibm');
X
X# Next, by special request, we include an optional associative array to
X# allow dlhunter to put the binaries in a sub directory structure below  
X# the actual download directory.  If you wish to use this feature, you must
X# uncomment and edit the following lines, which define the associative array.
X# Then, create the download dirs corresponding to the exts.  dlhunter will then
X# automatically dump the binaries in the associated subdir.  Remember to also
X# create the blurb directories, beneath each respective download directory.
X# To edit this array declaration, be sure to associate your download/blurb dir
X# with its corresponding extension (key). Separate the download dir from its
X# corresponding blurb dir with a colon.  Send me mail at cse with questions.
X#                     
X# %dldirs = (
X#         '.pc1' , '/tmp/downloads/pictures:/tmp/downloads/pictures/blurbs' ,
X#          '.ero' , '/tmp/downloads/erot:/tmp/downloads/erot/blurbs' ,
X#          '.misc' , '/tmp/downloads/misc:/tmp/downloads/misc/blurbs' ,
X#          '.util' , '/tmp/downloads/util:/tmp/downloads/util/blurbs' ,
X#          '.os2' , '/tmp/downloads/os2:/tmp/downloads/os2/blurbs' ,
X#          '.ibm' , '/tmp/downloads/ibm:/tmp/downloads/ibm/blurbs'
X#           );
X#
X# The queue directory, for links to uu postings, until all parts have arrived.
X#  IMPORTANT!!  This directory MUST be on the same filesystem as the news
X# directories. (@tooldirs)  For a number of good reasons...
X$quedir = '/var/spool/news/dl_hunter/queue';
X
X# Now, where are the assembled tools/pictures to go?  This is 
X# still the default location for the extracted binaries.
X$dloaddir = '/tmp/downloads';
X
X# We also need an associated blurb directory.  This is also the default.
X$blurbdir = '/tmp/downloads/blurbs';
X
X# Expire times correspond to number of days since created, for the queue,
X# and number of days since last accessed, for the downloads (binaries).
X$qkill = 3;    # queue
X$dlkill = 3;  # downloads
X
X###################  END OF CONFIGURATION   ##############################
X
X#  First, if we are running in interactive mode, 
X#  just call the hunt routine and exit.
X#
Xif (@ARGV[0] eq "-i"){
X&hunt();
Xexit(0);
X}
X
X# Next, link (copy) the new postings that are (probably) uuencoded to the queue
X# Pass 1 
X
XT1: for (0 .. $#tooldirs){          
X      $i = $_;
X      opendir(TOOLS,$tooldirs[$i]);
X      @uufiles =grep(!/^\.\.?$/ , readdir(TOOLS)); # get all current postings 
X      closedir(TOOLS);
XM1:   foreach $file (@uufiles){
X        open(CURR,"$tooldirs[$i]/$file")||next M1;      
X        $count = 0;
XL1:       while(<CURR>){    #  check to see if its a uuencoded file
X          if($count > 3){ 
X            close CURR; 
X
X# IMPORTANT : use link here, ( DONT USE SYMLINKS ) otherwise, you must
X# ^^^^^^^^^   hack a copy in here if your queue is not on the same fs as news 
X#             if you can only use symlinks, clear the queue before each run
X
X             if( !(-e "$quedir/$file$exts[$i]")){
X              link ("$tooldirs[$i]/$file","$quedir/$file$exts[$i]") ||
X                die " You cant use hard links in this queue directory.\n";}
X            next M1;}
X            next L1 if /[a-z]/;
X            next L1 unless /(^M.*$)|(^X.*$)/;# if there are 3 lines that beginh
X            $count++;                 # with an M or an X, count it as a encode
X          }                           # or shar, respectively
X        next M1;
X         }
X     next T1;
X      } 
X
X# Now hunt through the files in the queue, build an assoc array
X# of filenames in the queue, keyed on title of the tool/picture
X# (*.zip, *.zoo, *.jpg, *.gif, etc.) given in the Subject: line.
X# Now, this is trivial for moderated groups, but not so easy for alt. 
X# Also, obtain the part number, (m of n, m/n, m|n, etc.) and the total
X# number of parts.  
X# Pass 2
X
Xopendir(UUS,$quedir)||die ("cant open directory $dir");
X@allfiles = grep(!/^\.\.?$/ , readdir(UUS));   
Xclosedir(UUS);
XT2: foreach $file (@allfiles) {
X      open(CURR,"$quedir/$file")||warn "cant open file $file";
XM2:  while (<CURR>){
X        study;
X        next M2 unless 
X#  try to find the standard "name.ext" followed
X#  by the word "part" then the numbers.
X# try to beat the number of matches i get from these by twiddlin the regexps.
X# let me know how it turns out.  always use the part keyword in subject!
X
X(/^Subject:.*\s(\S+\.\S+).*[Pp][Aa][Rr][Tt]\D*[0]*([1-9][0-9]?)\D+[0]?([1-9][0-9]?).+$/) 
X
X# maybe we can match the whole line, followed by the part keyword.
X||(/^Subject:(.*)[Pp][Aa][Rr][Tt]\D*[0]*([1-9][0-9]?)\D+[0]?([1-9][0-9]?).+$/) 
X
X# oh well, default match
X|| /^Subject:.*\s(\S+\.\S+)\s*[,]*\D*[0]?([1-9][0-9]?)\D+[0]?([1-9][0-9]?).+$/;
X
X        $name=$1;$part=$2;$total=$3;
X        $titles{$name} .= "$quedir/$file"." "."$part"." "."$total"." ".":";
X        next T2;
X     }
X    }
X 
X# ok, now work the assoc array. For each title (key) in the associative array 
X# that we filled in the last loop, verify all parts are in, then check
X# to see if the file has already been decoded, try to decode and, if successful,
X# move to download directory.  Then add a blurb file to blurb dir, telling
X# about the new tool or picture.  Make the blurb world writable for comments.
X# Pass 3
X
Xopen(TEST,"> tst.out");            # when done, tst.out tells you how it went.
X@uu_tools = sort(keys(%titles));   # i like things sorted.
XT3: foreach $tool (@uu_tools){
X      @order = "";     # clear the ordered array. 
X      @parts = "";
X      @rel_files = split(/:/,$titles{$tool});   # relevant files to each title
X      print TEST "$tool \n";
X      foreach $tmp2 (@rel_files) {print TEST "$tmp2\n";}
X       print TEST "\n\n I chose this order for them....\n";
XM3A:  foreach $part (@rel_files){
X         @tmp = split(/\s/,$part);
X         $parts[$tmp[1]] = $tmp[2];               # save the total parts
X         if(defined $order[$tmp[1]]){           # may be a repost, use new one 
X          if (-M $tmp[0] < -M $order[$tmp[1]]){
X           unlink $order[$tmp[1]];               # and get rid of the old.
X           $order[$tmp[1]] = $tmp[0]; 
X           }} 
X         else{                                
X           $order[$tmp[1]] = $tmp[0];  }
X      }
X# Now test the order array to make sure all parts are available.
X
X      foreach $tmp2 (@order) {print TEST "$tmp2\n";}
X       print TEST "\n\n";
X       $ck = $#order;   
XM3B:  for (1..$#order){
X        next T3 unless defined $order[$_];
X#        next T3 unless $parts[$_] == $ck ;  
X      }
X
X# If we make it this far, we might have a complete encode.  
X# First check to see if the encode is already available,     
X# then try to decode it if not.  If successful, mv and write a blurb.
X
Xif (&dont_have($dloaddir,$order[1])){         # if no got, try to decode
X  if ($new = &try_to_decode(@order)){        # returns the decode's name or 0
X      &write_blurb($new,$blurbdir,$order[1]);# a short blurb written out
X      unlink @order;                        # remove the queue links
X}}                                   
Xelse{
X   unlink @order;}         # unlink the queue links if already decoded.
X
X}    # end of third pass
X
X# Finally, expire the old stuff.  We use last modification dates for the queue,
X# and last-access dates for the download directory, and blurb directory. 
X
Xif(defined (%dldirs)){      # if using the option...
X  foreach $ext (@exts){
X   ($dir1,$dir2) = split(/:/,$dldirs{$ext});
X   &expire($dir1,$dlkill,'-A');
X   &expire($dir2,$dlkill,'-A');
X  }
X   &expire($quedir,$qkill,'-M');
X}
Xelse{                          # the one dir approach.
X&expire($quedir,$qkill,'-M');
X&expire($dloaddir,$dlkill,'-A');  
X&expire($blurbdir,$dlkill,'-M');
X}
Xclose(TEST);
Xexit(0);      # thats it, nice job.
X
X##########################################################################
X
Xsub dont_have{          # checks if the decoded file already exists
X  local ($name) = $_[1];
X  local($dir,$junk) = &which_dir( $name);
X  open (CK, "$name")||die "couldnt open the file $_[1]";
X  while(<CK>){          # find the begin line and name
X    next unless
X     /^begin\s*\d*\s*(\S*)/;
X     if (-e "$dir/$1"){ return 0;}   # already exists
X     else{ return 1;}
X  }
Xreturn 0;        # begin not found
X}
X
X############################################################################
X
Xsub try_to_decode{       # a semi-bulletproof decoder.
X   local($cur) = `pwd`;  # now calls myuud  (Thanks Maman!)
X   local($tmpf) = 'tmp';
X   local(@files) = @_;
X   local($dl,$junk) = &which_dir($files[1]); # for the optional download sub-dir
X   chdir $dl;
X   open(BEGIN,"<$files[1]");
X   open(ENCODE,">$tmpf");
X   $_ = <BEGIN> until ($mode,$file) = /^begin\s*(\d*)\s*(\S*)/i;
X   for (1..$#files){        #build up a temporary file from the parts
X    $val = open(IN,"< $files[$_]");
X    if(! $val){ warn "couldnt open $files[$_]"; return 0;}
X    while(<IN>){ next unless /^\s*$/;last;}  # skip the header.
X    while (<IN>){
X      print ENCODE $_;
X   }}
X  close(ENCODE);
X# implicit child, to handle errors from uudecode, as recommended by lwall.
X  $pid = open(PIPE, "|-");
X   die "Fork failure" unless defined $pid;
X   if(!$pid){
X    exec "$usr_decode $tmpf >> errors  2>> errors";  # spawn it on the child.
X   }
X  close (PIPE);
X  if($?){unlink($tmpf,$file);           # uudecode didnt like it.
X        system "echo $file >>errors";   # a line to the errors file
X        chdir $cur;
X        return 0;} # error in the decoding, rm and return
X  chmod 0644, $file;
X  unlink ($tmpf);
X  chdir $cur;
X  return $file;                        # return the name.
X}                 
X
X#########################################################################
X
Xsub write_blurb{   # makes a file with the info on the tool/picture
X  local($name,$dir,$file) = @_;
X  local($junk) = 0;
X  ($junk,$dir) = &which_dir($file);
X  open(BLURB,"> $dir/$name")||die "couldnt open blurb $name";
X  open(IN, "< $file");  # stuff from part 1 up to the begin line
X   while(<IN>){
X    last if /^begin/;
X    print BLURB $_;}
X close BLURB;   
X chmod 0666, "$dir/$name";   # make it available to comments
X}
X########################################################################## 
X
Xsub expire{                # delete old files from the downloads, blurbs,
X  local($dir,$days,$how) = @_; # and queue directories. $how is the method. 
X  opendir(UL,$dir);
X  local(@allfiles) = grep(!/^\.\.?$/ , readdir(UL));   
X  closedir(UL);
X  local($file) = "";
X  foreach $file (@allfiles){
X  if ( $how eq '-A'){
X   unlink "$dir/$file" unless  (-A "$dir/$file" < $days);}
X  elsif ( $how eq '-M'){
X   unlink "$dir/$file" unless  (-M "$dir/$file" < $days);}
X  elsif ($how eq '-C'){
X   unlink "$dir/$file" unless  (-C "$dir/$file" < $days);}
X  }
X  } 
X ##########################################################################
X
Xsub hunt
X# hunt for and manually assemble the older links in the queue.
X# open the queue directory, and report on the articles
X# that have not been deleted, and are greater than 1 day
X# old. Then give the user the opportunity to assemble
X# uuencoded binaries from the various articles. 
X{
Xlocal($ldir,$file,$ans,$tmp,$subject,$wait,$tmpo) = (0,0,0,0,0,0,0);
Xlocal(@allfiles)=();
Xlocal(@order) = ();
Xlocal(@number) = ();
Xlocal(@forder) = ();
X$= = 22;          # a screenful at a time
XT: while (1) 
X{
X  system "clear";
X  $- = 0;
X  opendir(QUEUE,$quedir);
X  @allfiles = grep(!/^\.\.?$/ , readdir(QUEUE)); # get all the current postings 
X  closedir(QUEUE);
X  $i = 0;
XMID: foreach $file (@allfiles) {
X      next MID unless -M "$quedir/$file" > 1;   # dont bother with the new ones
X      open(CURR,"$quedir/$file");
XLOW:  while (<CURR>){
X        next LOW unless /^Subject:(.*$)/;
X        $subject = $1;
X        $number[$i] = "$quedir/$file";
X        #feed a screenful at a time.  crude.
X        if($- == 0){ print "<CR> to continue...";chop($wait=<STDIN>);write;}
X        else{write;}
X        $i++;
X        next MID; 
X  }
X }    # still in the T loop...
X    print "\nEnter the article numbers which make up the tool/pict\n";
X    print "\n             IN THE CORRECT ORDER \n\n";
X    print "     separated by a space, and followed by a <CR>.\n";
X    print "     Or just enter a carriage return to quit.\n ";
X     chop($tmp =<STDIN>);
X     if ($tmp eq ""){ return;}
X     @order = split(/ /,$tmp);        
X     $i=1;
X     foreach $tmpo (@order){ $forder[$i] = $number[$tmpo];
X                            print $forder[$i],"\n"; $i++;} #whew
X
X     if(!&dont_have(0,$forder[1])){
X       print "already got that one, deleting links\n"; 
X       print "<CR> to continue...";chop($wait=<STDIN>);
X       unlink @forder;
X       next T;}
X     else{          # if the tool/picture is not already present
X       if ($name = &try_to_decode(@forder)){    # try to decode it
X         print " successful decode, new file is named $name\n" ;
X         print " writing a new blurb now\n";
X         print "<CR> to continue...";chop($wait=<STDIN>);
X         &write_blurb($name,$ldir,$forder[1]);
X         unlink @forder;
X         next T;}
X       else{
X         print " decoding unsuccessful, try to find out what the prob is.";
X         print "\n<CR> to continue...";chop($wait=<STDIN>);
X         next T;}
X }
X}
X} 
X###########################################################################
Xsub which_dir{   # determine which dir to dump binaries in.
X                
Xlocal($name,$ext) = (@_[0],0);
X$name =~ /.*(\..*\b).*/;
X$ext = $1;
Xif(defined $dldirs{$ext}){    #  if the option is in use.
X  ($dl,$bl) = split(/:/,$dldirs{$ext});
X  return ($dl,$bl);}
Xelse{
X  return ($dloaddir,$blurbdir);}
X}
X#########################################################################
Xformat STDOUT_TOP =
XThese are the current postings in @<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< 
X$quedir
Xwhich MAY be tools/picts.  You must give me the article 
Xnumbers, in the CORRECT order, for me to feed to uudecode.
XFirst I will list them for you. 
X
XARTICLE 
XNUMBER     SUBJECT
X------   ----------------------------------------------------------------------
X. 
Xformat STDOUT =
X@>>>    @<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
X$i, $subject
X. 
X
X #  Bill Middleton
X #  wjm@sdf.lonestar.org, middle@cse.uta.edu
X #  let me know what you think...
X
X
X
END_OF_FILE
  if test 17539 -ne `wc -c <'dlhunter'`; then
    echo shar: \"'dlhunter'\" unpacked with wrong size!
  fi
  chmod +x 'dlhunter'
  # end of 'dlhunter'
fi
if test -f 'dlhunter.1' -a "${1}" != "-c" ; then 
  echo shar: Will not clobber existing file \"'dlhunter.1'\"
else
  echo shar: Extracting \"'dlhunter.1'\" \(1318 characters\)
  sed "s/^X//" >'dlhunter.1' <<'END_OF_FILE'
X.TH dlhunter 1
X.SH Name
Xdlhunter \- automated news binary extraction utility. 
X.SH Syntax
X.B dlhunter
X[\fB\-i\fR]
X.SH Description
X.PN dlhunter 
Xis designed to run as a background process, by cron.  It will find 
Xand extract uuencoded binaries in the news directories where they
Xare posted. In fact, it will find uuencoded files in any directory, and 
Xassemble them if the Subject lines give the correct part numbers.
XThe command,
X.PN  dlhunter 
X\fB\-i\fR, starts 
Xdlhunter in interactive mode, allowing the user to assemble those postings 
Xwhich had Subject lines which 
X.PN dlhunter 
Xcould not grok.  All directories are expired according to the 
Xuser-defined variables.  The code is
X.PN perl
Xand the script must be edited to suit the individual site.
X.PN dlhunter
Xuses hard links to the files it thinks are uuencoded.  And thus,
Xthe queue directory, where the links are created, 
X.PN must 
Xbe on the
Xsame fs as news.  
X.PP
X.PN dlhunter
Xrequires a SMART uudecoder.  There are many available.
X.PN myuud
Xis bundled with version 2.02.
XWhen you edit the configuration variables, you must specify
Xthe path to one.
X.PN
X.PN  NOTE:  
Xdlhunter is designed to save disk space and time.
XThus, it should be run by 
X.PN ONE 
Xuser, and the binaries placed in a 
X.PN PUBLIC 
Xdirectory.  
X.PN DONT
X.PN BE
X.PN A
X.PN DISK
X.PN HOG!
X.PP
END_OF_FILE
  if test 1318 -ne `wc -c <'dlhunter.1'`; then
    echo shar: \"'dlhunter.1'\" unpacked with wrong size!
  fi
  # end of 'dlhunter.1'
fi
if test -f 'myuud.c' -a "${1}" != "-c" ; then 
  echo shar: Will not clobber existing file \"'myuud.c'\"
else
  echo shar: Extracting \"'myuud.c'\" \(16085 characters\)
  sed "s/^X//" >'myuud.c' <<'END_OF_FILE'
X/*
X * Uud -- decode a uuencoded file back to binary form.
X *
X * From the Berkeley original, modified by MSD, RDR, JPHD & WLS.
X * The Atari GEMDOS version compiled with MWC 2.x.
X * The MSDOS version with TurboC.
X * The Unix version with cc.
X * this version is made: 25 Nov 1988.
X *
X * $Id: myuud.c,v 1.1 92/05/06 22:48:06 maman Exp Locker: maman $
X *
X * Mods :
X * ------ 16 May 1991 ------ maman@cosinus.inria.fr ------
X *    Table is reinitialised at the end of files in order to
X *    avoid a table modification effect for next files.
X * ------ 21 Feb 1991 ------ maman@cosinus.inria.fr ------
X *    Now if the file is unwritable, opens /dev/null instead
X *    to avoid breaking a multiple files run.
X * ------ 14 Feb 1991 ------ maman@cosinus.inria.fr ------
X *    (-c) dochmod added for UNIX.
X * ------ 24 Jan 1991 ------ maman@cosinus.inria.fr ------
X *    No use of ceil. (cdlen[] is already there).
X *    valid_char, valid_length and valid_chars used.
X *    (-v) verify flag added for more tests and more verbose,
X *    exactly the way I like it ;-)
X * ------ 23 Jan 1991 ------ maman@cosinus.inria.fr ------
X *    I've taken this from ftp at sics.se:/mailserver/uudecode.shar
X *    It was said Wrapped by Edwin Kremer <edwin@zlotty> on \
X *    Wed Dec 20 17:06:37 1989
X *
X *    Checks the declared length and the effective length
X *    in order to skip the header/signature of different
X *    files and test the range validity of one character
X *    in the middle of the buffer.
X *    Works almost all the time.
X *    Tolerate one more byte at the end for another count
X *    (should be given as a parameter ?).
X *    UNIX compilation: cc [-O] -o uud uud.c
X */
X
X/*
X * Be sure to have the proper symbol at this point. (GEMDOS, MSDOS, UNIX...)
X */
X/*
X#ifndef GEMDOS
X#define GEMDOS 1
X#endif
X */
X#ifndef UNIX
X#define UNIX 1
X#endif
X/*
X#ifndef MSDOS
X#define MSDOS 1
X#endif
X */
X
X#undef GWMDOS
X#undef MSDOS
X
X#ifdef GEMDOS
X#define SYSNAME "gemdos"
X#define SMALL 1
X#endif
X#ifdef MSDOS
X#define SYSNAME "msdos"
X#define SMALL 1
X#endif
X#ifdef UNIX
X#define SYSNAME "unix"
X#define format printf
X#define TEST_LENGTH
X#endif
X
X#include <stdio.h>
X
X#ifdef GEMDOS
X#include <osbind.h>
X#define Error(n)  { Bconin(2); exit(n); }
X#define WRITE	  "wb"
X#else
X#define Error(n)  exit(n)
X#define WRITE	  "w"
X#endif
X
X#define loop	while (1)
X
Xextern FILE *fopen();
Xextern char *strcpy();
Xextern char *strcat();
X
Xchar *getnword();
X
X#define MAXCHAR 256
X#define LINELEN 256
X#define FILELEN 64
X#define NORMLEN 60	/* allows for 80 encoded chars per line */
X
X#define SEQMAX 'z'
X#define SEQMIN 'a'
Xchar seqc;
Xint first, secnd, check, numl;
X#ifdef TEST_LENGTH
Xint numlbad = -1;
X#endif
X
XFILE *in, *out;
Xchar *pos;
Xchar ifname[FILELEN], ofname[FILELEN];
Xchar *source = NULL, *target = NULL;
Xchar blank, part = '\0';
Xint partn, lens;
Xint debug = 0, nochk = 0, onedone = 0, verify = 0;
X#ifdef UNIX
Xint dochmod = 1;
X#endif
Xint chtbl[MAXCHAR], cdlen[NORMLEN + 3];
X
Xmain(argc, argv) int argc; char *argv[];
X{
X	int mode;
X	register int i, j;
X	char *curarg;
X	char dest[FILELEN], buf[LINELEN];
X
X	if (argc < 2) {
X		format("Almost foolproof uudecode v3.4.2 (%s) 14-Feb-91\n",
X			SYSNAME);
X		format("\n");
X#ifdef UNIX
X		format("Usage: uud [-n] [-d] [-v] [-c] %s\n",
X		       "[-s dir] [-t dir] input-file");
X#else
X		format("Usage: uud [-n] [-d] [-v] %s\n",
X		       "[-s dir] [-t dir] input-file");
X#endif
X		format("\n");
X		format("Option: -n -> No line sequence check\n");
X		format("Option: -d -> Debug/verbose mode\n");
X		format("Option: -v -> Verify all chars\n");
X#ifdef UNIX
X		format("Option: -c -> chmod not performed (umask used)\n");
X#endif
X		format("Option: -s + Source directory for all input files\n");
X		format("  (MUST be terminated by directory separator)\n");
X		format("Option: -t + Target directory for all output files\n");
X		format("  (MUST be terminated by directory separator)\n");
X		format("If input-file is - then stdin is used as input-file\n");
X		Error(1);
X	}
X
X	curarg = argv[1];
X	
X	while (curarg[0] == '-') {
X		if (((curarg[1] == 'd') || (curarg[1] == 'D')) &&
X		    (curarg[2] == '\0')) {
X			debug = 1;
X#ifdef UNIX
X		} else if (((curarg[1] == 'c') || (curarg[1] == 'C')) &&
X			   (curarg[2] == '\0')) {
X			dochmod = 0;
X#endif
X		} else if (((curarg[1] == 'v') || (curarg[1] == 'V')) &&
X			   (curarg[2] == '\0')) {
X			verify = 1;
X		} else if (((curarg[1] == 'n') || (curarg[1] == 'N')) &&
X			   (curarg[2] == '\0')) {
X			nochk = 1;
X		} else if (((curarg[1] == 't') || (curarg[1] == 'T')) &&
X			   (curarg[2] == '\0')) {
X			argv++;
X			argc--;
X			if (argc < 2) {
X				format("uud: Missing target directory.\n");
X				Error(15);
X			}
X			target = argv[1];
X			if (debug||verify)
X				format("Target dir = %s\n",target);
X		} else if (((curarg[1] == 's') || (curarg[1] == 'S')) &&
X			   (curarg[2] == '\0')) {
X			argv++;
X			argc--;
X			if (argc < 2) {
X				format("uud: Missing source directory.\n");
X				Error(15);
X			}
X			source = argv[1];
X			if (debug||verify)
X				format("Source dir = %s\n",source);
X		} else if (curarg[1] != '\0') {
X			format("uud: Unknown option <%s>\n", curarg);
X			Error(15);
X		} else
X			break;
X		argv++;
X		argc--;
X		if (argc < 2) {
X			format("uud: Missing file name.\n");
X			Error(15);
X		}
X		curarg = argv[1];
X	}
X
X	if ((curarg[0] == '-') && (curarg[1] == '\0')) {
X		in = stdin;
X		strcpy(ifname, "<stdin>");
X	} else {
X		if (source != NULL) {
X			strcpy(ifname, source);
X			strcat(ifname, curarg);
X		} else
X			strcpy(ifname, curarg);
X		if ((in = fopen(ifname, "r")) == NULL) {
X			format("uud: Can't open %s\n", ifname);
X			Error(2);
X		}
X		numl = 0;
X	}
X
X	settable();
X
X/*
X * set up the line length table, to avoid computing lotsa * and / ...
X */
X	cdlen[0] = 1;
X	for (i = 1, j = 5; i <= NORMLEN; i += 3, j += 4)
X		cdlen[i] = (cdlen[i + 1] = (cdlen[i + 2] = j));
X/*
X * search for header or translation table line.
X */
X	loop {	/* master loop for multiple decodes in one file */
X		partn = 'a';
X		loop {
X			if (fgets(buf, sizeof buf, in) == NULL) {
X				if (onedone) {
X					if (debug||verify)
X					   format("End of file.\n");
X					exit(0);
X				} else {
X					format("uud: No begin line.\n");
X					Error(3);
X				}
X			}
X			numl++;
X			if (strncmp(buf, "table", 5) == 0 &&
X			    buf[5] == '\n' ) {
X				gettable();
X				continue;
X			}
X			if (strncmp(buf, "begin ", 6) == 0) {
X				break;
X			}
X		}
X		lens = strlen(buf);
X		if (lens) buf[--lens] = '\0';
X#ifdef SMALL
X		if ((pos = getnword(buf, 3))) {
X			strcpy(dest, pos);
X		} else
X#else
X		if(sscanf(buf,"begin%o%s", &mode, dest) != 2)
X#endif
X		{
X			format("uud: Missing filename in begin line.\n");
X			Error(10);
X		}
X
X		if (target != NULL) {
X			strcpy(ofname, target);
X			strcat(ofname, dest);
X		} else
X			strcpy(ofname, dest);
X
X		if((out = fopen(ofname, WRITE)) == NULL) {
X			format("uud: Cannot open output file: %s\n", ofname);
X			format("uud: Sending to /dev/null.\n");
X			strcpy(ofname, "/dev/null");
X			out = fopen(ofname, WRITE);
X		}
X		if (debug||verify) format("Begin uudecoding: %s\n", ofname);
X		seqc = SEQMAX;
X		check = nochk ? 0 : 1;
X		first = 1;
X		secnd = 0;
X		decode();
X		fclose(out);
X#ifdef UNIX
X		if(dochmod) chmod(ofname, mode);
X#endif
X		onedone = 1;
X		if (debug||verify) format("End uudecoding: %s\n", ofname);
X		settable();
X	}	/* master loop for multiple decodes in one file */
X}
X
X/*
X * Bring back a pointer to the start of the nth word.
X */
Xchar *getnword(str, n) register char *str; register int n;
X{
X	while((*str == '\t') || (*str == ' ')) str++;
X	if (! *str) return NULL;
X	while(--n) {
X		while ((*str != '\t') && (*str != ' ') && (*str)) str++;
X		if (! *str) return NULL;
X		while((*str == '\t') || (*str == ' ')) str++;
X		if (! *str) return NULL;
X	}
X	return str;
X}
X
X/*
X * Install the table in memory for later use.
X */
Xgettable()
X{
X	char buf[LINELEN];
X	register int c, n = 0;
X	register char *cpt;
X
X	if(debug) format("Getting new table.\n");
X
X	for (c = 0; c <= MAXCHAR; c++) chtbl[c] = -1;
X
Xagain:	if (fgets(buf, sizeof buf, in) == NULL) {
X		format("uud: EOF while in translation table.\n");
X		Error(5);
X	}
X	numl++;
X	if (strncmp(buf, "begin", 5) == 0) {
X		format("uud: Incomplete translation table.\n");
X		format("uud: Using default one.\n");
X		settable();
X		return;
X	}
X	cpt = buf + strlen(buf) - 1;
X	*cpt = ' ';
X	while (*(cpt) == ' ') {
X		*cpt = 0;
X		cpt--;
X	}
X	cpt = buf;
X	while (c = *cpt) {
X		if (chtbl[c] != -1) {
X			format("uud: Duplicate char in translation table.\n");
X			format("uud: Using default one.\n");
X			settable();
X			return;
X		}
X		if (n == 0) blank = c;
X		chtbl[c] = n++;
X		if (n >= 64)
X		{
X		    if(debug) format("End of getting table.\n");
X		    return;
X		}
X		cpt++;
X	}
X	goto again;
X}
X
X/*
X * Set up the default translation table.
X */
Xsettable()
X{
X    int i, j;
X
X    for (i = 0; i < ' '; i++) chtbl[i] = -1;
X    for (i = ' ', j = 0; i < ' ' + 64; i++, j++) chtbl[i] = j;
X    for (i = ' ' + 64; i < MAXCHAR; i++) chtbl[i] = -1;
X    chtbl['`'] = chtbl[' '];	/* common mutation */
X    chtbl['~'] = chtbl['^'];	/* an other common mutation */
X    blank = ' ';
X}
X
X/*
X * Testing functions
X */
X
X#ifdef TEST_LENGTH
Xvalid_length(l,n)
X     int l,n;
X{
X    int k, resp;
X    k = cdlen[n];
X    /*    k = n % 3;
X     *    k = ( n + (k ? 3-k : 0) )*4/3;
X     */
X    resp = ( (l-k >= 0) && (l-k <= 1) );
X    if(debug && !resp && !verify)
X    format("Bad length at line %d\n", numl);
X    return(resp);
X}
X
Xvalid_char(c)
X     char c;
X{
X    int resp;
X    resp = ( (chtbl[c]<0) ? 0 : 1 );
X    if(debug && !resp && !verify)
X    format("Bad char <%c> at line %d\n", c, numl);
X    return(resp);
X}
X#endif
X
Xvalid_chars(s,l)
X     char *s;
X     int l;
X{
X    int resp, k = 0;
X    while( (k<l) && valid_char(s[k]) ) k++;
X    resp = (k==l);
X    return( resp );
X}
X
X/*
X * copy from in to out, decoding as you go along.
X */
X
Xdecode()
X{
X	char buf[LINELEN], outl[LINELEN];
X	register char *bp, *ut;
X	register int *trtbl = chtbl;
X	register int n, c, rlen;
X	register unsigned int len;
X#ifdef TEST_LENGTH
X	int efflen;
X#endif
X
X	loop {
X		if (fgets(buf, sizeof buf, in) == NULL) {
X			format("uud: EOF before end.\n");
X			fclose(out);
X			Error(8);
X		}
X		numl++;
X		len = strlen(buf);
X		if (len) buf[--len] = '\0';
X/*
X * Is it an unprotected empty line before the end line ?
X */
X		if (len == 0) continue;
X/*
X * Get the binary line length.
X */
X		n = trtbl[buf[0]];
X/*
X * end of uuencoded file ?
X */
X		if (strncmp(buf, "end", 3) == 0) return;
X		if (   (strncmp(buf, "begin ", 6) == 0)
X		    || (strncmp(buf, "table", 5) == 0) )
X		{
X		    format("uud: BEGIN or TABLE before END.\n");
X		    /* puts back the line */
X		    fseek(in,(long) - (strlen(buf)+1) , 1);
X		    return;
X		}
X/*
X * end of current file ? : get next one.
X */
X		if (strncmp(buf, "include", 7) == 0) {
X			getfile(buf);
X			continue;
X		}
X#ifdef TEST_LENGTH
X/*
X * good length ?
X */
X		efflen = len;
X
X		/* Suppress bad chars at the end */
X		while(   (n >= 0)
X		      && (efflen > 0)
X		      && ! valid_char(buf[efflen-1])
X		      ) efflen--;
X
X		if(   (n >= 0)
X		   && (efflen > 0)
X		   && valid_length(efflen,n)
X		   && ((debug||verify)
X		       ? valid_chars(buf,efflen)
X		       : valid_char(buf[efflen/2]))
X		   )
X		{
X		    if (debug||verify) {
X			if((numlbad!=-1) && (numlbad<numl-1))
X			{
X			    format("Bad lines %d--%d\n",
X				   numlbad, numl-1);
X			}
X		    }
X		    numlbad = -1;
X		    goto decod;
X		}
X
X		if(numlbad<0) {
X		    numlbad = numl;
X		    if (debug)
X		    {
X			format("Bad line %d =%s\n", numl, buf);
X			format("n=%d --> %d, len=%d\n", n, cdlen[n], len);
X		    }
X		}
X		continue;
X#else /* !TEST_LENGTH */
X		if (n >= 0) goto decod;
X		format("uud: Bad prefix line %d in file: %s\n",numl, ifname);
X		if (debug) format("Bad line =%s\n",buf);
X		Error(11);
X#endif /* TEST_LENGTH */
X/*
X * Sequence checking ?
X */
Xdecod:		rlen = cdlen[n];
X/*
X * Is it the empty line before the end line ?
X */
X		if (n == 0) continue;
X/*
X * Pad with blanks.
X */
X		for (bp = &buf[c = len];
X			c < rlen; c++, bp++) *bp = blank;
X
X#if  !defined(TEST_LENGTH)
X/*
X * Verify if asked for.
X */
X		if (debug||verify) {
X			for (len = 0, bp = buf; len < rlen; len++) {
X				if (trtbl[*bp] < 0) {
X					format(
X	"Non uuencoded char <%c>, line %d in file: %s\n", *bp, numl, ifname);
X					format("Bad line =%s\n",buf);
X					Error(16);
X				}
X				bp++;
X			}
X		}
X#endif
X
X/*
X * All this just to check for uuencodes that append a 'z' to each line....
X */
X		if (secnd && check) {
X			secnd = 0;
X			if (buf[rlen] == SEQMAX) {
X				check = 0;
X				if (debug) format("Sequence check turned off (2).\n");
X			} else
X				if (debug) format("Sequence check on (2).\n");
X		} else if (first && check) {
X			first = 0;
X			secnd = 1;
X			if (buf[rlen] != SEQMAX) {
X				check = 0;
X				if (debug) format("No sequence check (1).\n");
X			} else
X				if (debug) format("Sequence check on (1).\n");
X		}
X/*
X * There we check.
X */
X		if (check) {
X			if (buf[rlen] != seqc) {
X				format("uud: Wrong sequence line %d in %s\n",
X					numl, ifname);
X				if (debug)
X					format(
X	"Sequence char is <%c> instead of <%c>.\n", buf[rlen], seqc);
X				Error(18);
X			}
X			seqc--;
X			if (seqc < SEQMIN) seqc = SEQMAX;
X		}
X/*
X * output a group of 3 bytes (4 input characters).
X * the input chars are pointed to by p, they are to
X * be output to file f.n is used to tell us not to
X * output all of them at the end of the file.
X */
X		ut = outl;
X		len = n;
X		bp = &buf[1];
X		while (n > 0) {
X			*(ut++) = trtbl[*bp] << 2 | trtbl[bp[1]] >> 4;
X			n--;
X			if (n) {
X				*(ut++) = (trtbl[bp[1]] << 4) |
X					  (trtbl[bp[2]] >> 2);
X				n--;
X			}
X			if (n) {
X				*(ut++) = trtbl[bp[2]] << 6 | trtbl[bp[3]];
X				n--;
X			}
X			bp += 4;
X		}
X		if ((n = fwrite(outl, 1, len, out)) <= 0) {
X			format("uud: Error on writing decoded file.\n");
X			Error(18);
X		}
X	}
X}
X
X/*
X * Find the next needed file, if existing, otherwise try further
X * on next file.
X */
Xgetfile(buf) register char *buf;
X{
X	if ((pos = getnword(buf, 2)) == NULL) {
X		format("uud: Missing include file name.\n");
X		Error(17);
X	} else
X		if (source != NULL) {
X			strcpy(ifname, source);
X			strcat(ifname, pos);
X		} else
X			strcpy(ifname, pos);
X#ifdef GEMDOS
X	if (Fattrib(ifname, 0, 0) < 0)
X#else
X	if (access(ifname, 04))
X#endif
X	{
X		if (debug) {
X			format("Cant find: %s\n", ifname);
X			format("Continuing to read same file.\n");
X		}
X	}
X	else {
X		if (freopen(ifname, "r", in) == in) {
X			numl = 0;
X			if (debug||verify) 
X				format("Reading next section from: %s\n", ifname);
X		} else {
X			format("uud: Freopen abort: %s\n", ifname);
X			Error(9);
X		}
X	}
X	loop {
X		if (fgets(buf, LINELEN, in) == NULL) {
X			format("uud: No begin line after include: %s\n", ifname);
X			Error(12);
X		}
X		numl++;
X		if (strncmp(buf, "table", 5) == 0) {
X			gettable();
X			continue;
X		}
X		if (strncmp(buf, "begin", 5) == 0) break;
X	}
X	lens = strlen(buf);
X	if (lens) buf[--lens] = '\0';
X/*
X * Check the part suffix.
X */
X	if ((pos = getnword(buf, 3)) == NULL ) {
X		format("uud: Missing part name, in included file: %s\n", ifname);
X		Error(13);
X	} else {
X		part = *pos;
X		partn++;
X		if (partn > 'z') partn = 'a';
X		if (part != partn) {
X			format("uud: Part suffix mismatch: <%c> instead of <%c>.\n",
X				part, partn);
X			Error(14);
X		}
X		if (debug||verify) format("Reading part %c\n", *pos);
X	}
X}
X
X#ifndef UNIX
X/*
X * Printf style formatting. (Borrowed from MicroEmacs by Dave Conroy.) 
X * A lot smaller than the full fledged printf.
X */
X/* VARARGS1 */
Xformat(fp, args) char *fp;
X{
X	doprnt(fp, (char *)&args);
X}
X
Xdoprnt(fp, ap)
Xregister char	*fp;
Xregister char	*ap;
X{
X	register int	c, k;
X	register char	*s;
X
X	while ((c = *fp++) != '\0') {
X		if (c != '%')
X			outc(c);
X		else {
X			c = *fp++;
X			switch (c) {
X			case 'd':
X				puti(*(int *)ap, 10);
X				ap += sizeof(int);
X				break;
X
X			case 's':
X				s = *(char **)ap;
X				while ((k = *s++) != '\0')
X					outc(k);
X				ap += sizeof(char *);
X				break;
X
X			case 'c':
X				outc(*(int *)ap);
X				ap += sizeof(int);
X				break;
X
X			default:
X				outc(c);
X			}
X		}
X	}
X}
X
X/*
X * Put integer, in radix "r".
X */
Xputi(i, r)
Xregister unsigned int	i;
Xregister unsigned int	r;
X{
X	register unsigned int	q, s;
X
X	if ((q = i / r) != 0)
X		puti(q, r);
X	s = i % r;
X	if (s <= 9)
X		outc(s + '0');
X	else
X		outc(s - 10 + 'A');
X}
Xoutc(c) register char c;
X{
X#ifdef GEMDOS
X	if (c == '\n') Bconout(2, '\r');
X	Bconout(2, c);
X#else
X	putchar(c);
X#endif
X}
X
X#endif
END_OF_FILE
  if test 16085 -ne `wc -c <'myuud.c'`; then
    echo shar: \"'myuud.c'\" unpacked with wrong size!
  fi
  # end of 'myuud.c'
fi
echo shar: End of archive 1 \(of 1\).
cp /dev/null ark1isdone
MISSING=""
for I in 1 ; do
    if test ! -f ark${I}isdone ; then
	MISSING="${MISSING} ${I}"
    fi
done
if test "${MISSING}" = "" ; then
    echo You have the archive.
    rm -f ark[1-9]isdone
else
    echo You still must unpack the following archives:
    echo "        " ${MISSING}
fi
exit 0
exit 0 # Just in case...