#!/bin/bash
## Check websites for illegal HTML-comments containing "--"  by Eugene Reimer 2010-04-25;
##
## PREREQ:  find-dirs  fullname -- from http://ereimer.net/programs/general-purpose-scripts.htm
##
## Copyright © 2010 Eugene Reimer;  can be used, modified, copied, and distributed or sold under the terms of either the LGPL or the GPL (your choice);
## see http://www.gnu.org/licenses for the details of these terms.

WEBSITES="/debwendon/website  /er/website  /noci/website"				##--REVISE AS NEEDED--
shopt -s nullglob									##set nullglob, so for-loop works within DIR lacking *htm files (eg cgi-bin)

WEBCOMMENTCHECK_LL () {									##WEBCOMMENTCHECK_LL: function to check in one dir
TMP=/tmp/WEBCOMMENT-$(fullnameNOSLASH "$1")
x01=$'\x01'; x02=$'\x02'
cat "$1"									|
sed "s|<!--|$x01|g; s|-->|$x02|g"						|	##replace comment-starter with x01, comment-ender with x02
sed ":a \$!{N;ba}; s|^[^$x01]*||; s|\($x01[^$x02]*$x02\)[^$x01]*|\n\1|g"	|	##treating file as one line, discard everything but comments
cat >$TMP										##comments-only copy of input-file to tmpfile, in case needed later
if V=$(grep -- '--' $TMP); [[ $V ]];then
  echo;echo "===$(fullname $1) contains double-dash within HTML-comments==="
  echo "$V" |sed "s|$x01|<!--|g; s|$x02|-->|g"						##echo the faulty lines, replacing x01,x02 with comment-delimeters (undoing)
else rm -f $TMP										##discard tmpfile for correct HTML-file (keep for faulty ones)
fi
}

for WEBSITE in $WEBSITES;do								##==for each website...
  cd $WEBSITE || exit									##in its HOME-dir
  for DIR in . $(find-dirs);do								##for this dir and each subdir...
    cd $WEBSITE/$DIR || exit								##==work in DIR
    for H in *htm;do WEBCOMMENTCHECK_LL $H; done					##for each HTML file, check its comments
  done
done


exit
I needed this because my habit of using double-dash for em-dash gets me into trouble, and I sometimes mess-up when yanking a chunk of HTML;
when "yanking" a chunk of HTML by enclosing in comment-delimeters, one needs to modify any comments within that chunk:
(1) to ensure a single ending-delimiter, and (2) starting-delimeters are modified to avoid any use of "--" within the comment;
incidentally most browsers ignore the rule about "--" being illegal within a comment, but not all, for example Firefox enforces it;
the actual rule about "--" is complex, but I see nothing gained by stating it as other than simply disallowed.

NOTE: can do this checking with tidy, as in my verify-html-comments:
    WARN=$(tidy -q -o /dev/null "$F" 2>&1 |egrep "comment|--" |tr '\n' ' ')	##msgs such as:  "adjacent hyphens within comment",  "XML comments can't contain --"