arxiv2md

#!/usr/bin/perl
use Getopt::Std;
use Switch 'Perl6'; use Sys::Hostname; use Time::HiRes qw(usleep);#usleep($microseconds)
use File::Basename; use File::Copy; use File::Find; use Cwd;use File::Path;
# --------------------- OPTIONS       --------------------------------
$cdir = cwd;chomp($cdir);
$prog = basename($0);
getopt('Tt', \%opts); #here we add only options with args
foreach $opt (keys %opts){
 given($opt){
  when 'h' {usage();last;}
  default  {usage("Not a valid option: $opt");}
 }
}
# -------------------------------------------------------------------
if($#ARGV < 0){usage();}
$arxiv_url  = "http://arxiv.org/e-print/"; # url where source of papers is
$arxiv_abs  = "http://arxiv.org/abs/";     # url where abstract is
$arxiv_no   = $ARGV[0];

$arxiv_url .= $arxiv_no;
$meta_url   = "${arxiv_abs}/${arxiv_no}";# html file with meta information (title/authors)
# in case the paper number is in the arxiv/000000 format, then redefine $arxiv_no:
$arxiv_ar   = "$arxiv_no";
if( $arxiv_no =~ /(.*?)\/(.*?)$/){ $arxiv_ar = $1; $arxiv_no = $2;}
print "${prog}:arxiv_ar = $arxiv_ar ; arxiv_no = $arxiv_no\n";
$mobi       = "${arxiv_no}.mobi"; # final output file!
#--------------------------------------------------------------------
$tmpdir = "/tmp/tmp-${prog}";
if( -d $tmpdir){
 print "${prog}>rm -rf $tmpdir\n";
 system("rm -rf $tmpdir");
}
mkdir $tmpdir;
chdir $tmpdir;
print "${prog}>wget  --user-agent=Mozilla/1.2 $arxiv_url\n";
system("wget  --user-agent=Mozilla/1.2 $arxiv_url");
print "${prog}>tar xvf $arxiv_no\n";
system("tar xvf $arxiv_no");
$meta_source = "${arxiv_no}.html"; # html file with meta information (title/authors)
print "${prog}>wget  --user-agent=Mozilla/1.2 -O $meta_source $meta_url\n";
system("wget  --user-agent=Mozilla/1.2 -O $meta_source $meta_url");
$latex_source = "";
foreach $f (<*.tex>){
 $results  = `grep '\documentclass' $f`;
 $results .= `grep '\documentstyle' $f`;
 if($results){$latex_source = $f;break;}
}
if( $latex_source){
 print "${prog}:Found LaTeX source: $latex_source\n";
} else{
 die "${prog}:LaTeX source not found!";
}
# Determine title and authors:
$title = "[${arxiv_no}] "; $authors = "";
open(META,$meta_source);
while(<META>){
 if(/citation_title/ ){/content=\"(.*?)\"/; $title   .= $1;}
 if(/citation_author/){/content=\"(.*?)\"/; $authors .= "$1&";}
}
chop $authors; #remove trailing &
print "${prog}:title= $title\n${prog}:authors= $authors\n";

print "--------------------------------------------------------\n";
print "${prog}:Temporary files can be found in $tmpdir\n";
print "${prog}:Title:   $title\n";
print "${prog}:Authors: $authors\n";
print "${prog}:arXiv:   $ARGV[0]\n";
# ----------------------- HELP MESSAGE ------------------------------
sub usage(){
 if(@_){$message  = shift(@_)."\n";}
 $message .= << "EOF";
Usage:   ${prog} <arxiv number>
    e.g. ${prog} 1010.0957
         ${prog} hep-lat/0402031

Downloads source files from arXiv (if they exist), converts them to html and then to the mobi format (Kindle friendly)
Uses htlatex and ebook-convert  (assumes TeX4ht and Calibre exist on your system)
On Ubuntu: sudo apt-get  install texlive-full calibre
K. N. Anagnostopoulos, NTUA 2013
EOF
 print STDERR $message;
 exit(1);
}
sub main::HELP_MESSAGE(){ usage();} #for --help (does not work when default?)
# -------------------------------------------------------------------