INFZ=words.num.zip
INF=words.num
OUTF=dict-ru.xml
INURL=http://bokrcorpora.narod.ru/frqlist/$INFZ
set -e
TDIR=$(mktemp -d)
trap "rm -rf $TDIR" EXIT
wget -P $TDIR $INURL
(cd $TDIR && unzip $INFZ)
test -f $TDIR/$INF
cat $TDIR/$INF | iconv -f cp1251 -t utf8 | awk '
BEGIN {
	print "<?xml version=\"1.0\" encoding=\"UTF-8\" ?>";
	print "<wordlist>"
}
{
	printf " <w f=\"%d\">%s</w>\n", int($2), $3
}

END {
	print "</wordlist>"
}
' > $OUTF
ls -l $OUTF
echo "$OUTF generation succesful"