#! /bin/bash

url="https://www.dialectsarchive.com"

continents="africa
asia
australia-oceania 
caribbean
central-america
europe
middle-east
north-america
south-america"

destination="$1"

for continent in $continents
do
    countries_html="$(wget -q -O - $url/$continent)"

    country_urls="$( \
        echo "$countries_html" | \
        sed -n -e '/Please select a/,/clear/{p;}' | \
        grep -io --color=never "$url/[^\"]*" | \
        uniq
    )"

    for country_url in $country_urls
    do
        dialect_html="$(wget -q -O - $country_url)"

        country="$(basename $country_url)"

        dialect_lines="$( \
            echo "$dialect_html" | \
            grep -iP --color=never "$url/$country-\d+\">[^>]*>.*?(fe)?male" \
        )"

        directory="$destination/$continent/$country"

        mkdir -p "$directory"

        IFS=$'\n'
        for dialect_line in $dialect_lines
        do
            dialect_url="$( \
                echo "$dialect_line" | \
                grep -io --color=never 'https[^"]*' \
            )"

            dialect_info="$( \
                echo "$dialect_line" | \
                grep -Po --color=never "(?<=</a>)[^<]+(?=<)" | \
                sed 's/\W*$//' | \
                sed 's/^\W*//' | \
                sed 's/\//_/g' \
            )"

            filename="$(basename $dialect_url): $dialect_info.mp3"
            filepath="$directory/$filename"

            if [ -f "$filepath" ]
            then
                echo "Skipping existing file $filepath"
                continue
            fi

            dialect_html="$(wget -q -O - $dialect_url)"

            dialect_download="$( \
                echo $dialect_html | \
                grep -io 'https[^"]*\.mp3' | \
                head -1 \
            )"

            echo "Downloading $filename to $directory"
            wget -q -O "$filepath" $dialect_download
        done
    done
done