Youtube Archiver Script

youtube-archive.sh

#!/bin/bash
# Youtube Archiver v0.1.0
#
# Requires `yt-dlp` and `jq` to be installed and on $PATH
#
# Downloads videos into `./recent` and rechecks them within 30 days because
# it may take YT some time to encode the highest quality version of a video.
# Will move any replaced videos in `./replaced`
# NOTE: Script depends on this specific folder and file naming format, do not change!

set -euo pipefail

SCRIPT_DIR=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
ARCHIVE_DIR="$SCRIPT_DIR/archive"
REPLACED_DIR="$SCRIPT_DIR/replaced"

function yt_dlp_with_args() {
  yt-dlp \
    --verbose \
    --force-ipv4 \
    --sleep-requests 1 \
    --sleep-interval 5 \
    --max-sleep-interval 30 \
    --no-continue \
    --no-overwrites \
    --add-metadata \
    --parse-metadata "%(title)s:%(meta_title)s" \
    --parse-metadata "%(uploader)s:%(meta_artist)s" \
    --write-description \
    --write-info-json \
    --write-annotations \
    --write-thumbnail \
    --embed-thumbnail \
    --all-subs \
    --embed-subs \
    --check-formats \
    --concurrent-fragments 3 \
    --match-filter "!is_live & !live" \
    --output "%(uploader)s/%(uploader)s - %(upload_date)s - %(title)s/%(uploader)s - %(upload_date)s - %(title)s [%(id)s].%(ext)s" \
    --merge-output-format "mkv" \
    --mtime \
    "$@"
}

function archive_channels() {
  pushd "$ARCHIVE_DIR" > /dev/null
  printf "\n[ARCHIVE CHANNELS]\n\n"
  yt_dlp_with_args \
    --download-archive archive.log \
    --batch-file "$SCRIPT_DIR/channels.txt" \
  2>&1 | tee output.log
  popd > /dev/null
}



function refresh_recent_videos() {
  pushd "$ARCHIVE_DIR" > /dev/null
  printf "\n[REFRESH RECENT VIDEOS]\n\n"

  # Find any files downloaded within the last 30 days and created in the last 31 days
  find . -name "*.mkv" -mtime -30 -print0 | 
    while IFS= read -r -d '' recent_vid; do 
      recent_folder=$(dirname "$recent_vid")
      recent_prefix=$(basename "$recent_vid" ".mkv")
      recent_json_file="$recent_folder/$recent_prefix.info.json"

      current_format_id=$(jq -r '.format_id' "$recent_json_file")
      url=$(jq -r '.webpage_url' "$recent_json_file")

      new_format_id=$(yt-dlp --print format_id "$url")

      if [ "$current_format_id" != "$new_format_id" ]; then
        echo "Need to reconcile video at $recent_folder: new format: $new_format_id, old format: $current_format_id"
        replacement_folder="$REPLACED_DIR/$recent_folder-$(date +%Y%m%d)"
        mkdir -p "$replacement_folder"
        mv "$recent_folder" "$replacement_folder"
        yt_dlp_with_args "$url"
      fi
    done

  popd > /dev/null
}

archive_channels_failed="0"
refresh_recent_videos="0"

archive_channels || archive_channels_failed="1"
refresh_recent_videos || refresh_recent_videos="1"

echo "Jobs status:"
echo "archive_channels_failed: $archive_channels_failed"
echo "refresh_recent_videos: $refresh_recent_videos"

if [ "$archive_channels_failed" == "1" ] || [ "$refresh_recent_videos" == "1" ]; then
  exit 1
fi