#! /bin/sh
#
# tagusage
#
# count up tags between <text> and </text> in an SGML document
# provide output in <tagusage> tags sorted by element name.
# 
# Original script by Heiki-Jaan Kaalep (hkaalep@psych.ut.ee) 
# that counted the tag usage was modified by Greg Priest-Dorman 
# (priestdo@cs.vassar.edu) to give output in a format that 
# could be pasted into a tagsdecl element.  Developed in the 
# context of the MULTEXT and MULTEXT-EAST projects.
# 10/95
#

#
# requires gnu sed and possibly gnu grep
# 
tr '\012' ' ' |\
sed 's/^..*<text/<text/g' | \
sed 's/</@</g' | \
sed 's/>/>@/g' | tr '@' '\012' | \
grep "<[^<>/]*>" | \
sed 's/ [^>]*>/>/g' | \
sort| uniq -c | \
sed 's/^ *//'| \
sed 's/^\([0-9]*\)\(.*\)$/\2 \1/' | \
#
# the next grep will remove the text tag from the count
# grep -v "<text>" |
#
# the next grep will remove the comments from the count
grep -v "<\!--" |
sort | \
sed 's/^.</<tagusage gi=/'| \
sed 's/> \([0-9]*\)/ occurs=\1><\/tagusage>/'
#