Download channels from Discord and save as a webarchive for offline reading

discord-dl.sh

#!/bin/bash
#
# Make sure to have a .env with `DISCORD_TOKEN=your_token_here` in the same directory as this script.
# Get your token as per: https://github.com/Tyrrrz/DiscordChatExporter/blob/master/.docs/Token-and-IDs.md

dce() {
  docker run -it --rm --env-file .env --volume "$(pwd)/out":/out tyrrrz/discordchatexporter:latest "$@"
}

ffmpeg() {
  docker run --rm --volume "$(pwd)/out":/out lscr.io/linuxserver/ffmpeg:latest "$@"
}

# Ensure all files/tools exist
if [ ! -e .env ] || ! command -v docker >/dev/null 2>&1 || ! command -v webarchiver >/dev/null 2>&1; then
  echo "You need a .env file with DISCORD_TOKEN in it, and you must have docker + webarchiver installed"
  exit 1
fi

# Input for what to download
echo "Getting guilds"
dce guilds
echo "Which guild id would you like to download from?"; read -r guild_id
dce channels --guild "$guild_id"
echo "Which channel id?"; read -r channel_id
echo "Download from what timestamp/messageid (0 for beginning)?"; read -r start_id
echo "What name should we assign to the download (no spaces)?"; read -r name

# Download
dce export --media --reuse-media --media-dir media --markdown True --format HtmlDark --locale Pacific \
  --channel "$channel_id" --after "$start_id" --output "$name.html"

echo "Converting videos to animated gifs"  # Needed for webarchive for iOS/iPadOS
videos=$(grep -oE 'src=media/[^"]*.mp4' "out/$name.html")
errors=0
if [[ -n "$videos" ]]; then
  while IFS= read -r match; do
    mp4_filename_with_media=$(echo "$match" | cut -d'=' -f2)
    gif_filename_with_media="${mp4_filename_with_media%.*}.gif"
    if ! ffmpeg -nostdin -hide_banner -loglevel error -y -i "/out/$mp4_filename_with_media" -vf "scale=320:-1,fps=10" "/out/$gif_filename_with_media"; then errors=1; fi
    sed -i "" "s|<video.*$match.*</video>|<img class=chatlog__embed-generic-gifv src=\"$gif_filename_with_media\" \/>|g" "out/$name.html"
  done <<< "$videos"
fi

# Remove lazy loading and generate webarchive files and remove html files
echo "Fixing image urls and converting to webarchive"
sed -i "" 's/loading="*lazy"*//g' "out/$name.html"                      # Disable lazyloading so webarchiver downloads everything
sed -i "" -e ':a' -e 's/\(src=[^ >]*\)%/\1@!@25/g; t a' "out/$name.html"  # dce has a bug when % is in the url
sed -i "" -e ':a' -e 's/\(src=[^ >]*\)@!@/\1%/g; t a' "out/$name.html"    # replace the intermediary @!@ with % to prevent infinite loops

echo "Converting to webarchive"
if ! webarchiver -url "out/$name.html" -output "./out/$name.webarchive"; then errors=1; fi
[ "$errors" -eq 0 ] && rm "out/$name.html" || echo "There were errors, not deleting html file out/$name.html"

echo "Completed"