experimental messing about with nzh source ;)

h2txt

#!/bin/zsh
[ $1 ]||{echo "Usage: $0 <nzherald-url>"&&exit 1;}
page=`mktemp herald.htm.XXXXX`
curl -s $1 > $page
echo "Extracting to: $page.txt"
xmllint --html --format $page --nowarning --xpath "//p" 2>/dev/null \
  | perl -pe 's|<p.*?>||g;s|<span.*?>||g;s|</.*?>|\n|g;s|\n\n         ||g' \
  | fmt -p -s | perl -pe 's|<strong>||ig;s|&amp\;|&|g;s|         ||g' \
  | fmt -s | tee $page.txt | less
#