Add playlist dedup methods

This commit is contained in:
Polochon-street 2021-08-23 19:37:51 +02:00
parent a3fcccbf2a
commit 23d4d36cb4
3 changed files with 154 additions and 2 deletions

2
Cargo.lock generated
View file

@ -77,7 +77,7 @@ checksum = "cf1de2fe8c75bc145a2f577add951f8134889b4795d47466a54a5c846d691693"
[[package]]
name = "bliss-audio"
version = "0.3.5"
version = "0.4.0"
dependencies = [
"bliss-audio-aubio-rs",
"crossbeam",

View file

@ -1,6 +1,6 @@
[package]
name = "bliss-audio"
version = "0.3.5"
version = "0.4.0"
authors = ["Polochon-street <polochonstreet@gmx.fr>"]
edition = "2018"
license = "GPL-3.0-only"

View file

@ -73,6 +73,50 @@ pub fn song_to_song(first_song: &Song, songs: &mut Vec<Song>, distance: impl Dis
*songs = new_songs;
}
/// Remove duplicate songs from a playlist, in place.
///
/// Two songs are considered duplicates if they either have the same,
/// non-empty title and artist name, or if they are close enough in terms
/// of distance.
///
/// # Arguments
///
/// * `songs`: The playlist to remove duplicates from.
/// * `distance_threshold`: The distance threshold under which two songs are
/// considered identical. If `None`, a default value of 0.05 will be used.
pub fn dedup_playlist(songs: &mut Vec<Song>, distance_threshold: Option<f32>) {
dedup_playlist_custom_distance(songs, distance_threshold, euclidean_distance);
}
/// Remove duplicate songs from a playlist, in place, using a custom distance
/// metric.
///
/// Two songs are considered duplicates if they either have the same,
/// non-empty title and artist name, or if they are close enough in terms
/// of distance.
///
/// # Arguments
///
/// * `songs`: The playlist to remove duplicates from.
/// * `distance_threshold`: The distance threshold under which two songs are
/// considered identical. If `None`, a default value of 0.05 will be used.
/// * `distance`: A custom distance metric.
pub fn dedup_playlist_custom_distance(
songs: &mut Vec<Song>,
distance_threshold: Option<f32>,
distance: impl DistanceMetric,
) {
songs.dedup_by(|s1, s2| {
n32(s1.custom_distance(&s2, &distance)) < distance_threshold.unwrap_or(0.05)
|| (s1.title.is_some()
&& s2.title.is_some()
&& s1.artist.is_some()
&& s2.artist.is_some()
&& s1.title == s2.title
&& s1.artist == s2.artist)
});
}
#[cfg(test)]
mod test {
use super::*;
@ -80,6 +124,114 @@ mod test {
use ndarray::arr1;
use std::path::Path;
#[test]
fn test_dedup_playlist_custom_distance() {
let first_song = Song {
path: Path::new("path-to-first").to_path_buf(),
analysis: Analysis::new([
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
]),
..Default::default()
};
let first_song_dupe = Song {
path: Path::new("path-to-dupe").to_path_buf(),
analysis: Analysis::new([
1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
]),
..Default::default()
};
let second_song = Song {
path: Path::new("path-to-second").to_path_buf(),
analysis: Analysis::new([
2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 1.9, 1., 1., 1.,
]),
title: Some(String::from("dupe-title")),
artist: Some(String::from("dupe-artist")),
..Default::default()
};
let third_song = Song {
path: Path::new("path-to-third").to_path_buf(),
title: Some(String::from("dupe-title")),
artist: Some(String::from("dupe-artist")),
analysis: Analysis::new([
2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2.5, 1., 1., 1.,
]),
..Default::default()
};
let fourth_song = Song {
path: Path::new("path-to-fourth").to_path_buf(),
artist: Some(String::from("no-dupe-artist")),
title: Some(String::from("dupe-title")),
analysis: Analysis::new([
2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 0., 1., 1., 1.,
]),
..Default::default()
};
let fifth_song = Song {
path: Path::new("path-to-fourth").to_path_buf(),
analysis: Analysis::new([
2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 2., 0.001, 1., 1., 1.,
]),
..Default::default()
};
let mut playlist = vec![
first_song.to_owned(),
first_song_dupe.to_owned(),
second_song.to_owned(),
third_song.to_owned(),
fourth_song.to_owned(),
fifth_song.to_owned(),
];
dedup_playlist_custom_distance(&mut playlist, None, euclidean_distance);
assert_eq!(
playlist,
vec![
first_song.to_owned(),
second_song.to_owned(),
fourth_song.to_owned(),
],
);
let mut playlist = vec![
first_song.to_owned(),
first_song_dupe.to_owned(),
second_song.to_owned(),
third_song.to_owned(),
fourth_song.to_owned(),
fifth_song.to_owned(),
];
dedup_playlist_custom_distance(&mut playlist, Some(20.), cosine_distance);
assert_eq!(playlist, vec![first_song.to_owned()]);
let mut playlist = vec![
first_song.to_owned(),
first_song_dupe.to_owned(),
second_song.to_owned(),
third_song.to_owned(),
fourth_song.to_owned(),
fifth_song.to_owned(),
];
dedup_playlist(&mut playlist, Some(20.));
assert_eq!(playlist, vec![first_song.to_owned()]);
let mut playlist = vec![
first_song.to_owned(),
first_song_dupe.to_owned(),
second_song.to_owned(),
third_song.to_owned(),
fourth_song.to_owned(),
fifth_song.to_owned(),
];
dedup_playlist(&mut playlist, None);
assert_eq!(
playlist,
vec![
first_song.to_owned(),
second_song.to_owned(),
fourth_song.to_owned()
]
);
}
#[test]
fn test_song_to_song() {
let first_song = Song {