process subsubdirs of $1 to extract magazine data

magcon.sh

#!/bin/bash

# exit function
die() {
    echo >&2 "$0 ERROR: $@"; exit 1;
}

[ "$1" ] || die "Argument missing."
[ -d "$1" ] || die "Arg '$1' is not a directory."
cd "$1" || die "Can't access '$1'."

dirs=(*)
[ -d "$dirs" ] || die "No dirs found."

# loop through each subdirectory of $1
for d in "${dirs[@]}";do
    cd "$d"
    dirs2=(*)
    [ -d "$dirs2" ] || die "No dirs2 found."
    # loop through each subsubdirectory
    for d2 in "${dirs2[@]}";do
        # abort if no mag.html
        [ -f "$d2/mag.html" ] || die "No $d2/mag.html"

        # extract title from <title> tag
        title=$(awk -vRS="</title>" '/<title>/{gsub(/.*<title>|\n+/,"");print;exit}' "$d2/mag.html")

        # extract TOC from js array midway in the HTML and convert to json format, remove last comma
        toc=$(awk '/contents in the format/,/\};/' "$d2/mag.html" | tail +2 | tac | sed '1,2 d' | tac | awk '!/\/\//' | sed 's/\[ /\{ "name": /g' | sed -E 's/([0-9]+) \]/"page": "\1" }/g' | perl -00pe 's/,(?!.*,)//s' )

        # slugify title
        slug="$(echo -n "${title}" | sed -e 's/[^[:alnum:]]/-/g' | tr -s '-' | tr A-Z a-z)"

        # create json file
        touch "$d2/mag.json"
        echo -e "{\n  \"title\": \"$title\",\n  \"slug\": \"$slug\",\n  \"toc\": [\n$toc\n  ]\n}" > "$d2/mag.json"

        # rename dir to slug (if doesn't exist already)
        if [ ! -d "$slug" ]; then
            mv "$d2" "$slug"
        fi

        # output for future batch processing
        echo "title=\"$title\", dir=\"$1/$d/$slug\""
    done
    cd ..
done

cd ..
echo "WTF DONE"