Archive Synthwave Mixes from Astral Throb

Astral-Throb_archiver.sh

## Download the excellent synthwave mixes created by Astral Throb, split the
## audio into individual songs, and label them with metadata, including
## excellent cover art. Ideal for offline listening.

## Dependencies:
##   - youtube-dl
##   - jq
##   - ffmpeg
##   - eyeD3

## TODO
## 1. Catch errors where YouTube-DL fails to grab the initial playlist.
## 2. Don't allow the infofile, album, title, &c to be blank when parsing —
##      treat as an error and skip that album while downloading.
## 3. If there have been errors, don't try to move the final file.

ALBUM_ARTIST="Astral Throb"
GENRE="Electronic"

ARTIST_HOMEPAGE="https://www.youtube.com/channel/UCpbH_7H71IPKq4eH7CD5spg"
PLAYLIST_URL="https://www.youtube.com/playlist?list=UUpbH_7H71IPKq4eH7CD5spg"

TARGET_DIR="$HOME/media/music/library/$ALBUM_ARTIST"
ARTIST_LOGO="$TARGET_DIR/astral-throb.jpg" # This file is not created, it should already exist in the target directory

## Colors for output
RED='\033[0;31m'
LIGHT_PURPLE='\033[1;35m'
NC='\033[0m' # No Color

echo "Archiving $ALBUM_ARTIST to ${LIGHT_PURPLE}$TARGET_DIR${NC}."

## Download and parse the list of URLs in the playlist
echo "Fetching work queue…"
PLAYLIST_LIST=$(youtube-dl --dump-json --flat-playlist "$PLAYLIST_URL" | jq -r '.id' | sed 's_^_https://youtube.com/v/_')
#PLAYLIST_LIST="https://www.youtube.com/watch?v=isIj3tuQTDY&t=2015s" # For testing with individual albums which encounter errors

for URL in $PLAYLIST_LIST
do

  ## Skip SIMULAKRUM LAB, which isn't actually a mix
  [[ "$URL" == "https://youtube.com/v/6DaQbh45OxI" ]] && echo "Skipping SIMULAKRUM LAB…" && continue

  ## Download the metadata file for the album
  INFOFILE=$(youtube-dl --skip-download --write-info-json "$URL" | grep "Writing video description metadata as JSON to" | cut -c 55- ) #; echo "INFOFILE: $INFOFILE"
  [[ "$INFOFILE" == "" ]] && echo "Metadata download failed, skipping…" && continue
  PREFIX=$(echo "$INFOFILE" | rev | cut -c 11- | rev)

  ## Parse basic album metadata from the file
  ALBUM_TITLE=$(cat "$INFOFILE" | jq .fulltitle | cut -d ' ' -f 1 | cut -c 2-) #; echo "ALBUM_TITLE: $ALBUM_TITLE" # The short all-caps title
  ALBUM_YEAR=$(cat "$INFOFILE" | jq ".upload_date" | cut -c 2-5) #; echo "ALBUM_YEAR: $ALBUM_YEAR" # Four-digit album year
  ALBUM_DIRECTORY="$ALBUM_ARTIST - $ALBUM_TITLE ($ALBUM_YEAR)" #; echo "ALBUM_DIRECTORY: $ALBUM_DIRECTORY"

  ## Check to see if this album already existst in the target directory, otherwise create it here
  if [ -d "$TARGET_DIR/$ALBUM_DIRECTORY" ]
  then
    rm "$INFOFILE"
    echo "$ALBUM_TITLE has already been downloaded."
    rm "$PREFIX"*
    continue
  else
    mkdir "$ALBUM_DIRECTORY"
  fi

  ALBUM_FULLTITLE=$(cat "$INFOFILE" | jq --raw-output .fulltitle)
  echo "Processing $ALBUM_FULLTITLE …"

  ## Download the audio and the thumbnail from YouTube
  echo "  Downloading $ALBUM_TITLE combined mix & art…"
  ALBUM_AUDIO=$(youtube-dl --format bestaudio --extract-audio --write-thumbnail "$URL" | grep Destination | grep ffmpeg | cut -c 23-)

  ## Parse the remaining album metadata
  RELEASE_DATE=$(cat "$INFOFILE" | jq ".upload_date" | cut -c 2-5)-$(cat "$INFOFILE" | jq ".upload_date" | cut -c 6-7)-$(cat "$INFOFILE" | jq ".upload_date" | cut -c 8-9)
  TRACKS=$(cat "$INFOFILE" | jq ".chapters | length")

  ## Move the album art
  echo "  Exporting $ALBUM_TITLE cover art…"
  ALBUM_ART="$ALBUM_DIRECTORY/$ALBUM_FULLTITLE.jpg"
  mv "$PREFIX.jpg" "$ALBUM_ART"


  ## Extract the individual audio files from the combined album mix
  for INDEX in $(seq 1 $TRACKS)
  do

    ## Parse this audio track's metadata
    TRACK_ARTIST="$(cat "$INFOFILE" | jq --raw-output ".chapters["$[$INDEX-1]"].title" | cut -d - -f 1 | rev | cut -c 2- | rev | sed 's/^\///;s/\// /g' )"
    TRACK_TITLE="$(cat "$INFOFILE" | jq --raw-output ".chapters["$[$INDEX-1]"].title" | cut -d - -f 2 | cut -c 2- | sed 's/^\///;s/\// /g' )"
    TRACK_START="$(cat "$INFOFILE" | jq ".chapters["$[$INDEX-1]"].start_time")"
    TRACK_END="$(cat "$INFOFILE" | jq ".chapters["$[$INDEX-1]"].end_time")"

    ## Single-digit track numbers are zero-prefixed in the filename
    if (( $INDEX < 10 )) ; then
      FILENAME="$ALBUM_DIRECTORY/0$INDEX. $TRACK_ARTIST - $TRACK_TITLE.mp3"
    else
      FILENAME="$ALBUM_DIRECTORY/$INDEX. $TRACK_ARTIST - $TRACK_TITLE.mp3"
    fi

    ## The heavy lifting of audio extraction for this track
    echo "  Extracting $ALBUM_TITLE track $INDEX/$TRACKS $TRACK_ARTIST - $TRACK_TITLE …"
    ffmpeg -loglevel warning -y -i "$ALBUM_AUDIO" -ss "$TRACK_START" -to "$TRACK_END" "$FILENAME" && echo "    Track extraction successful." || echo "    ${RED}Track extraction failed!${NC}"

    ## Write the metadata to the file
    echo "    Writing tack metadata for $TRACK_ARTIST - $TRACK_TITLE …"
    eyeD3 \
      --artist "$TRACK_ARTIST" \
      --album "$ALBUM_FULLTITLE" \
      --album-artist "$ALBUM_ARTIST" \
      --title "$TRACK_TITLE" \
      --release-year "$ALBUM_YEAR" \
      --track "$INDEX" \
      --track-total "$TRACKS" \
      --genre "$GENRE" \
      --release-date "$RELEASE_DATE" \
      --url-frame "WOAS:$URL" \
      --url-frame "WOAR:$ARTIST_HOMEPAGE" \
      --add-image "$ARTIST_LOGO:BAND_LOGO" \
      --add-image "$ALBUM_ART:FRONT_COVER" \
      "$FILENAME" >/dev/null

  done

  ## Move the complete albim to the target directory
  echo "  Extraction complete. Moving $ALBUM_TITLE to library…"
  rsync --recursive --remove-source-files --compress "$ALBUM_DIRECTORY" "$TARGET_DIR"

  ## Remove the original audio, thumbnail, and metadata file
  echo "  $ALBUM_TITLE archived, cleaning up…"
  rm "$PREFIX"*
  rm -r "$ALBUM_DIRECTORY"

done

exit 0