The following script can be used to extract and decode images from a Usenet binary archive created by Newsfetch. Newsfetch is free and runs in Unix. The source code for Newsfetch can be found here: ftp://sunsite.unc.edu/pub/Linux/system/news/readers/newsfetch-1.21.tar.gz
Example setup files for downloading train pictures can be found here: http://www.xmission.com/~georgeps/scripts.html
How could this be modified to deal with binhex and xxencode formatted data? -GPS
No need - UUDeview is a Tcl-based tool (written by Frank Pilhofer) that does all this and more. It's very well-written, thorough, flexible and robust. It includes an easy to use Tk interface. Very fast, too. (I once wrote a similar tool to do binary extraction so I really appreciate how good a job Frank did.) - Don Libes
The URL for UUDeview is: http://www.informatik.uni-frankfurt.de/~fp/uudeview/
#!/usr/local/bin/tclsh8.3
#This should work with tclsh > 8.0.5.
#An earlier version of this had problems dealing with some filenames.
set fp [open ./$argv r]
set op 0
set begin ""
set end ""
set prefix "/home/imagine/Trains/pics/"
while {[eof $fp] != 1} \
{
set data [gets $fp]
if {[regexp {\mbegin ([0-9]+) (.*)} $data all mode filename] == 1} \
{
puts "bef $filename"
regsub -all { } $filename {} filename
puts "aft $filename"
set op [open ./_$filename w]
set begin [tell $fp]
}
if {[regexp {\mend$} $data all]} {set end [tell $fp]}
#Eventually I should check for a ` followed by end, to be more accurate.
if {$end != "" && $begin == ""} {set end ""}
if {$begin != "" && $end != ""} \
{
puts "filename $filename"
puts "begin $begin"
puts "end $end"
seek $fp $begin
puts $op "begin 644 $filename"
while 1 \
{
puts $op [gets $fp]
if {$end < $begin} \
{
puts "Error: The end is before the beginning."
exit
}
if {[tell $fp] >= $end} \
{
set begin ""
set end ""
close $op
puts "$prefix$filename"
catch {exec uudecode -p _$filename > \
$prefix$filename}
catch {exec rm _$filename}
puts "done\n"
break
}
}
}
}See also http://brag.sf.net/ for another Tcl application for downloading and assembling multi-part binaries from newsgroups...