format_report_for_wiki

#!/bin/bash

# This script will grab the files that will be created after a scan,
# process then to MediaWiki format and then strip out html markup that
# breaks the formatting when viewing on the wiki.
# Also takes name of the file as the name for the individual reports.

# NAMING CONVENTION FOR FILES: machine-name/item-name.
# Instead of spaces(!), use hyphens in the name of the scan target.
# The script searches for the first underscore and takes everything
# before it as the scan target, adding in the name to the Summary section
# in the report.


# Creation date: 9th September 2011
# Amended 26th September 2011
# Author: Gordon Thomson
# Version 1.0

# Fixes:
# 4/10/2011 - formatting was askew in the wiki file - sub headers were given the same prominence as
# the main header per scan target. Script now gives correct prominence to sub headers.

# get the passed in directory
directory=$1
# and the current directory
current_dir=$2
# grab the end location from params
date_location=$3
# Final filename
wiki_name=$4
# grab the date
date_run=`date +%d_%m_%Y`
year=`date +%Y`
month=`date +%B`
# create the files
touch /tmp/$date_run".html"
echo -e "\n"
for files in `find $directory -maxdepth 2 -mindepth 1 -type f`
do
    if [ "$(file -b $files | grep 'HTML document')" ];
    then
        # grab the first section of the filename (everything before the '_')
        filename=$(basename "$files")
        # this would get the extension. Not sure if it copes with .tar.gz though.
        extension=${filename# # *.}
        file_name_ext=${filename%.*}.$extension
        name_of_file=$(echo $file_name_ext | cut -d'_' -f1)
        # create a temp file to avoid overwriting the original
        cat $files > $files.temp
        # remove top nine lines (including the <h1>Summary</h1> part)
        sed -i '1,8d' $files.temp
        # remove two bottom lines (and spaces between)
        head -n -5 $files.temp > $files.temp2
        # replace the summary part, including the filename
        sed -i -e '1c \<h1>Summary for '$name_of_file'</h1>' $files.temp2
        # Also need to cater for every other machine/item scanned, the above replaces only the first.
        # sed -i 's/">Host/">Host '$name_of_file'-/g' $files
        # Put the temp file into the overall HTML file
        cat $files.temp2 >> /tmp/$date_run".html"
        # Couple of line breaks to separate content.
        echo '<br /><br />' >> /tmp/$date_run".html"
        echo $files" has been processed."
        # Clean up
        rm $files.temp
        rm $files.temp2
        continue
    elif [ "$(file -b $files | grep 'ASCII English text, with very long lines')" ];
    then
        continue
    else
        continue
    fi
done
echo "Converting to MediaWiki format..."
html2wiki --strip-tags --dialect MediaWiki --no-wrap-in-html --no-escape-entities /tmp/$date_run".html" > /tmp/$date_run".wiki"
echo "Correcting indentation structure..."
sed -i 's/=Results per Host=/==Results per Host==/g' /tmp/$date_run".wiki"
echo "Done."
rm /tmp/$date_run".html"