I’m sure I’ll be modifying this more over time. But, it reads from a text file (podcasts.txt) in the given DOWNLOAD_PATH–this file contains the RSS URL(s) for the podcast(s). It organizes the podcasts into directories, if those directories don’t already exist. It uses eyeD3 to add metadata (including cover art). Works fairly well, though I’m still stomping bugs.

Maybe next I’ll put together a nice little frontend for it–maybe convert the whole thing to a Blazor app. But, for now, it does the thing and it’s fun being a script kiddie 🙂

#!/bin/bash

create_path() {
    TITLE_PATH="${DOWNLOAD_PATH}/${1}"

    if [ ! -d "${TITLE_PATH}" ]; then
       mkdir "${TITLE_PATH}"
    fi
}

sanitize_filename() {
  local filename="$1"
  # Replace unfriendly characters with underscore
  sanitized=$(echo "$filename" | sed 's/[\/:*?"<>|]/_/g')
  echo "$sanitized"
}

dl_podcast() {
    local RSS_FEED_URL=$1

    curl -s "$RSS_FEED_URL" > /tmp/rss_feed.xml

    title=$(get_podcast_title "${RSS_FEED_URL}")

    create_path "${title}"

    # Initialize arrays to store episode titles and enclosure URLs
    titles=()
    urls=()

    # Extract episode titles and enclosure URLs
    mapfile -t titles < <(xmllint --xpath '//item/title/text()' /tmp/rss_feed.xml)
    mapfile -t urls < <(xmllint --xpath '//item/enclosure/@url' /tmp/rss_feed.xml)

    # Print the episode titles and enclosure URLs
    j=${#titles[@]}
    padtowidth=${#j}
    for ((i=0; i<${#titles[@]}; i++)); do
	ep_title=$(echo "${titles[$i]}" | sed -e 's/<!\[CDATA\[//' -e 's/\]\]>//')
        url=`echo ${urls[$i]} | sed "s/url=//g"`
        url=`echo ${url} | sed "s/\"//g"`
        filenum=`printf "%0*d\n" $padtowidth $j`
        file="${filenum} $ep_title.mp3"
        file=$(sanitize_filename "${file}")

        if [ ! -f "${TITLE_PATH}/${file}" ]; then
            echo "Downloading: ${file} (${url})"
            wget -qO "${TITLE_PATH}/${file}" "${url}"
	    /usr/bin/eyeD3 --track "${filenum}" --disc-num "1" "${TITLE_PATH}/${file}"
	    /usr/bin/eyeD3 --add-image="/tmp/cover.jpg":FRONT_COVER "${TITLE_PATH}/${file}"
        fi
        j=`expr ${j} - 1`
    done

    unset ${TITLE_PATH}
}

get_podcast_title() {
    local url="${1}"
    local title="$(curl -s "$url" | grep -oP '<title>\K[^<]+' | sed -n '2p')"
    echo "${title}"
}

get_podcast_cover() {
    local url=$1
    local cover_url=$(curl -s "$url" | grep -oP '<itunes:image\s+href="\K[^"]+' | head -n 1)

    curl -s -o /tmp/cover.jpg ${cover_url}
}

cleanup() {
    rm /tmp/cover.jpg
    rm /tmp/rss_feed.xml
}

read_lines_into_array() {
    local file=$1

    # Check if file exists
    if [ ! -f "$file" ]; then
        echo "Error: File $file not found."
        exit 1
    fi

    # Read the file line by line and append each line to the array
    mapfile -t lines_array < "${file}"

    # Return the array
    echo "${lines_array[@]}"
}

process_podcasts() {
    local -n podcasts=$1

    for link in "${podcasts[@]}"; do
        get_podcast_cover "$link"

        title=$(get_podcast_title "$link")

        dl_podcast $link
    done
}

main()
{
    DOWNLOAD_PATH="<PODCAST DIRECTORY GOES HERE>"
    TITLE_PATH=${DOWNLOAD_PATH}

    if [ $# -ne 1 ]; then
        echo "Usage: $0 <rss url>"
        exit 1
    fi

    file=$1

    # Call the function and store the result in an array
    declare -A links
    links=$(read_lines_into_array "$file")

    # Print the contents of the resulting array
    process_podcasts links
    cleanup
}

main "$@"

Been working on a podcast downloader for use with Plex, since Plex decided to take away podcasts–basically I keep the ones I wanna retain long-term in a library collection now. The downloader is a simple BASH script. Seems to be nearly perfected, even adds metadata (track #, cover art, etc) to the downloaded mp3.

I think my last little issue is that I need to have it rename all the files whenever episodes hit a factor of 10 (e.g., if there are 10 episodes, episode one should be “01 – Title”, when it hits 100, episode one should be renamed to “001 – Title”). It’s a little thing–I just like my files sorting properly by name, no matter what OS I’m listing them in. And Linux sorts in lexicographical by default, as opposed to natural order

Some guys work in the garage on Saturday mornings–I work on code 🤣