#! /bin/bash
# Cleans the HTML files produced by the OneTab Chrome extension.
# Reada "${name}.html" and the favicons in the directory "${name}_files/".
# Writes to standard output.
name="$1"; shift; # Name of webpage and associated
title="$1"; shift;
name="${name/.html/}" # Just in case.
hfile="${name}.html"
if [[ ! ( -s ${hfile} ) ]]; then echo "** file \"${hfile}\" not found" 1>&2 ; exit 1; fi
hdir="${name}_files"
if [[ ! ( -d ${hdir} ) ]]; then echo "** \"${hdir}\" not found or not a directory" 1>&2 ; exit 1; fi
tmp="/tmp/$$" # Temp file prefix.
printf "\n"
printf "\n"
printf "\n"
printf " \n"
printf " %s\n" "${title}"
printf " \n"
printf "\n"
printf "\n"
printf "%s
\n" "${title}"
printf "\n"
# Copy each line from the file and append it to the end of the img line. Do not leave any space or create new lines. Although the Base64 encoded text in mypicture.txt are separated into many lines, join all of them into a single continuous line with no embedded spaces.
afile="${tmp}_afile.html"
cat ${hfile} \
| recode windows-1252..UTF-8 \
| sed \
-e "s@