Re: Issues with gendocs and images

[Akim Demaille]

Hi all,

I don't understand too well some bits of gendocs:

  opt="--html -o $PACKAGE.html $commonarg $htmlarg"
  cmd="$SETLANG $MAKEINFO $opt \"$srcfile\""
  echo "Generating html by node... ($cmd)"
  eval "$cmd"
  split_html_dir=$PACKAGE.html
  (
   cd ${split_html_dir} || exit 1
   tar -czf "$abs_outdir/${PACKAGE}.html_node.tar.gz" -- *.html
  )
  html_node_tgz_size=`calcsize "$outdir/${PACKAGE}.html_node.tar.gz"`
  rm -f "$outdir"/html_node/*.html
  mkdir -p "$outdir/html_node/"
  mv ${split_html_dir}/*.html "$outdir/html_node/"
  rmdir ${split_html_dir}

I don't understand why it seems to take that many precautions
to get only the *html files that are generated by makeinfo
--html, why no simply take everything that is in the directory?
Also, at the end, we empty a directory, and then mv *html files
from one to the other, and kill the empty directory.  Why didn't
we simply rename that directory?  Actually, why would we rename it
instead of simply passing to -o the wanted directory, instead of
creating this directory?

Thanks in advance.  My questions are related to changes I am
making to import the needed images (which, of course, don't
match *.html).