#!/bin/sh

# this script generates a variety of lists / databases.

# artist.txt: a list of all artist in the torrent, including sub-units
# album.txt: a list of all releases in the torrent
# lossy.txt: a list of all lossy releases in the torrent
# db.csv: a database of specific tags. this is required for other scripts to run properly.

check() {
	[ ! -d "$dir" ] && echo "invalid directory!" && exit
	[ ! -d "$(dirname "$db")" ] && mkdir "$(dirname "$db")"
	[ ! -d "$(dirname "$log")" ] && mkdir "$(dirname "$log")"
}

artist() {
	printf "\033[1mgenerating artist list...\033[0m\n"

	[ -f "$art" ] && mv -f "$art" "$art".bak
	find "$dir" -mindepth 1 -maxdepth 2 -type d | grep -vE '(Lossless|Lossy)$' |
	sed "s|$dir/||;s|^.*/\[||;s|\]$||" | LC_COLLATE=C sort -f > "$art"
}

album() {
	printf "\033[1mgenerating album list...\033[0m\n"

	[ -f "$alb" ] && mv -f "$alb" "$alb".bak
	find "$dir" -type d -links 2 | sed "s|.*/||" | LC_COLLATE=C sort -f > "$alb"
}

lossy() {
	printf "\033[1mgenerating lossy album list...\033[0m\n"

	[ -f "$los" ] && mv -f "$los" "$los".bak
	find "$dir" -type d -links 2 | grep Lossy | sed "s|.*/||" | LC_COLLATE=C sort -f > "$los"
}

csv() {
	printf "\033[1mgenerating csv database...\033[0m\n"

	[ -f "$db" ] && mv -f "$db" "$db".bak
	echo "#TAG,ALBUMARTIST,ORGANIZATION,CONTENTGROUP" > "$db"

	# get artist list
	find "$dir" -mindepth 1 -maxdepth 2 -type d | grep -vE '(Lossless|Lossy)$' | while read -r parent; do
		# prefer using the most recent lossless korean release
		child="$(find "$parent" -mindepth 3 -maxdepth 3 -type d | grep '/Lossless/Korean/' | sort -n | tail -n 1)"

		# if there is none, use the most recent release of any type
		if [ -z "$child" ]; then
			child="$(find "$parent" -mindepth 3 -maxdepth 3 -type d | sort -n | tail -n 1)"
		fi

		if [ -n "$child" ]; then
			# prefer using a track without features
			file="$(find "$child" -type f -name '*.flac' | grep -ivE feat | head -n 1)"

			if [ "$(find "$child" -type f -name '*.flac' | wc -l)" -eq 1 ] || [ -z "$file" ]; then
				file="$(find "$child" -type f -name '*.flac' | head -n 1)"
			fi

			# tag sub-units
			echo "$parent" | grep -q '\[.*\]$' && tag="U" || tag=""

			artist="$(metaflac --show-tag=ARTIST "$file" 2>/dev/null | cut -d '=' -f 2)"
			org="$(metaflac --show-tag=ORGANIZATION "$file" 2>/dev/null | cut -d '=' -f 2)"
			group="$(metaflac --show-tag=CONTENTGROUP "$file" 2>/dev/null | cut -d '=' -f 2)"

			# only skip adding entry to the database if artist isn't found or goes over more than one line
			if [ -z "$artist" ] || [ "$(echo "$artist" | wc -l)" -gt 1 ]; then
				echo "$(basename "$parent"): cound not get artist tag. skipping..." >> "$log" && continue
			fi

			# print a warning for other missing tags
			[ -z "$org" ] && echo "$(basename "$parent"): could not get organization tag" >> "$log"
			[ -z "$group" ] && echo "$(basename "$parent"): could not get content group tag" >> "$log"

			printf "\n%s\t%s\t%s\t%s" "$tag" "$artist" "$org" "$group" >> "$db"
		else
			echo "$(basename "$parent"): error finding files. skipping..." && continue
		fi
	done
}

while getopts "d:arlch" o; do case "${o}" in
	d) dir="$OPTARG";;
	a) artist=1;;
	r) album=1;;
	l) lossy=1;;
	c) csv=1;;
	h|*) printf "usage: ./%s [OPTION]...\n\noptions:
 -d specify directory to scan
 -a generate artist list
 -r generate complete release list
 -l generate lossy release list
 -c generate csv database\n" "$(basename "$0")" && exit 1
esac done

# set file locations
log="../logs/list.log"
art="../lists/artists.txt"
alb="../lists/albums.txt"
los="../lists/lossy.txt"
db="../lists/db.csv"

# check variables.
check

# begin new log.
printf "\nscript run - %s\n" "$(date)" >> "$log"

# generate lists
[ -n "$artist" ] && artist
[ -n "$album" ] && album
[ -n "$lossy" ] && lossy
[ -n "$csv" ] && csv

printf "\n\033[1mscript completed! remember to check log for errors.\033[0m\n"