fix some extractions?

scripts/extract-used-tags.sh

@@ -15,6 +15,7 @@ sed 's/ [0-9A-Za-z/#_-]* ;//'     | # get rid of all contlexes
 cut -d':' -f1                     | # get rid of the surface side
 grep -E '[@+][0-9A-Za-z][0-9A-Za-z@+]' |  # Only keep interesting strings
 sed 's/@@/@€@/g'                  | # insert newline placeholder between flag diacritics
+sed 's/@+/@€+/g'                  | # insert newline placeholder between flag and tag
 tr '€#"' '\n'                     | # insert newlines now, for cleaner data for the next steps
 sed '/^[^@+]/ s/^[^@+]*//'        | # remove non-tag text at beginning of line
 sed 's/^\(@[^@]*@\)/\1€/'         | # insert newline placeholer after initial flag diacritic
@@ -24,5 +25,5 @@ sed '/^\+/ s/\+/€+/g'             | # if begins with +, insert newlines before
 sed '/\+$/ s/\+/\+€/g'            | # if ends with +, insert newlines after + (prefix tags)
 tr '€' '\n'                       | # insert newlines
 grep '[\+@].'                     | # keep only interesting stuff
-sed 's/[-%]*$//'
+sed -e 's/[-%]*$//' -e 's/^[-%]*//'
 # Get rid of final stuff that is odd