#! /usr/bin/gawk -f
# Last edited on 2003-10-12 12:29:48 by stolfi

BEGIN{ 
  abort = -1;
  usage = ( ARGV[0] " < INFILE > OUTFILE" );
  #
  # Cleans up a list of entries copied from the "Nature" contents alert.
  # Output is a single line per entry, TITLE @ AUTHORS @ URL
  # where the AUTHORS have been shortened and the URL has "${NAT"" instead 
  # of "http://www.nature.com".
  #
  begin_entry();
}

(abort >= 0) { exit abort; }

//{
  gsub(/[ \011]+$/, "", $0);
}

/^$/{ 
  finish_entry();
  begin_entry();
  next;
}

/^http:/ {
  # URL
  if (n == 0) { error("missing author"); }
  if (n == 2) { error("duplicate URL"); }
  n = 2;
  field[n] = (field[n] " " $0);
  next;
}

/^[-A-ZΑΙΣΘά'.,\& ]+$/ {
  # Author(s)
  if (n == 2) { error("garbled entry"); }
  n = 1;
  field[n] = (field[n] " " $0);
  next;
}

/./ {
  if (n == 2) { error("garbled entry"); }
  if (n == 1) { error("garbled entry"); }
  n = 0;
  field[n] = (field[n] " " $0);
  next;
}

END { 
  if (abort >= 0) { exit abort; }
  finish_entry();
}

function error(msg)
{
  printf "*** line %d: %s\n", FNR, msg;
  abort = 1;
  exit abort;
}
  
funtion finish_entry(  tit,aut,url)
{
  tit = field[0];
  aut = field[1];
  url = field[2];
  aut = gensub(/^([A-ZΑΙΣΘά][A-ZΑΙΣΘά' ]*[A-ZΑΙΣΘά])[ ]*[,\&].*$/, "\\1 ...", "g", aut);
  gsub(/http[:][\/][\/]www.nature.com/, "${NAT}", url);
  printf "%s @ %s @ %s\n", tit, aut, url;
}

function begin_entry()
{
  split("", field);
  n = 0;
}