Contribute a script to convert from YouTube chapters to CD-DA

1 files changed, 143 insertions, 0 deletions

diff --git a/contrib/scripts/yt-chapters-to-cdda.sh b/contrib/scripts/yt-chapters-to-cdda.sh
new file mode 100755
index 000000000..7d4cbc9b9
--- /dev/null
+++ b/contrib/scripts/yt-chapters-to-cdda.sh
@@ -0,0 +1,143 @@
+#!/usr/bin/env bash
+
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# Copyright (C) 2022-2022  kcgen <kcgen@users.noreply.github.com>
+
+# Bash strict-mode
+set -euo pipefail
+IFS=$'\n\t'
+
+SCRIPT=$(basename "$0")
+readonly SCRIPT
+
+print_usage() {
+    echo ""
+    echo "usage: $SCRIPT URI [URI [...]]"
+    echo
+    echo "Convert a YouTube video with chapters into CDDA+CUE format."
+    echo " - Multiple videos can be provided; each will be converted."
+    echo " - URIs can be the full https:// URL or just the identifier."
+    echo ""
+    echo "Depends on:"
+    echo " - yt-dlp (install with pip3 install yt-dlp)"
+    echo " - ffmpeg: install with package manager"
+    echo " - basename: install with package manager"
+    echo ""
+}
+
+main() {
+    case ${1:-} in
+    -h | -help | --help) print_usage ;;
+    *)
+        check-dependencies
+        convert-chapters-to-cdda "$@"
+        ;;
+    esac
+}
+
+check-dependencies() {
+    local missing_deps=0
+    for dep in yt-dlp ffmpeg find sort rm basename; do
+        if ! command -v "$dep" &>/dev/null; then
+            echo "Missing dependency: $dep could not be found in the PATH"
+            ((missing_deps++))
+        fi
+    done
+    # Were any missing?
+    if [[ "$missing_deps" -gt 0 ]]; then
+        echo "Install the above programs and try again"
+        exit 1
+    fi
+}
+
+download-chapters-from-uri() {
+    local uri="$1"
+
+    # YouTube Opus format identifiers in order of quality
+    local webm_formats="338/251/250/249"
+
+    # Directory/##-Track output layout
+    local cdda_ouput="chapter:%(title)s/%(section_number)02d-%(section_title)s.%(ext)s"
+
+    # Fetch the track and split it on chapters
+    yt-dlp \
+        --split-chapters \
+        --restrict-filenames \
+        --format "$webm_formats" \
+        --output "$cdda_ouput" \
+        "$uri"
+}
+
+extract-opus-from-webm() {
+    local webm="$1"
+    local opus="${1%.webm}.opus"
+
+    if [[ -f "$webm" && ! -f "$opus" ]]; then
+        # Extract the Opus stream from the WebM
+        ffmpeg -hide_banner -loglevel error \
+            -i "$webm" \
+            -vn -acodec copy "$opus"
+        # Delete the WebM source if we've got the Opus
+        if [[ -f "$opus" ]]; then
+            rm "$webm"
+        fi
+    fi
+}
+
+add-track-to-cue() {
+    local dir
+    local webm
+    local opus
+    dir=$(basename "${1%/*}")
+    webm=$(basename "$1")
+    opus=$(basename "$webm" webm)opus
+
+    local track_number="$2"
+
+    # Ensure this directory and the track exists
+    if [[ ! -d "$dir" || ! -f "$dir/$opus" ]]; then
+        echo "Problem finding the CDDA directory ($dir) and/or track ($dir/$opus)"
+        exit 1
+    fi
+
+    # Add a CUE entry for the track
+    cue_path="$dir/cdrom.cue"
+    {
+        echo "FILE \"$opus\" OPUS"
+        echo "  TRACK $track_number AUDIO"
+        # Add pre-gap just for the first track
+        if [[ "$track_number" == "1" ]]; then
+            echo "  PREGAP 00:02:00"
+        fi
+        echo "  INDEX 01 00:00:00"
+        echo ""
+    } >>"$cue_path"
+}
+
+convert-local-webm-to-cdda() {
+    local count=1
+    for webm in $(find . -maxdepth 2 -mindepth 2 -name '*.webm' | sort); do
+        extract-opus-from-webm "$webm"
+        add-track-to-cue "$webm" "$count"
+        ((count += 1))
+    done
+}
+
+print-dosbox-imgmount-line() {
+    local cue="$1"
+    if [[ -n "$cue" ]]; then
+        echo "imgmount d \"$cue\" -t cdrom"
+    fi
+}
+
+convert-chapters-to-cdda() {
+    for uri in "$@"; do
+        cue_path=""
+        download-chapters-from-uri "$uri"
+        convert-local-webm-to-cdda
+        print-dosbox-imgmount-line "$cue_path"
+    done
+}
+
+main "$@"