1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
//! # lyric_finder
//!
//! This crate provides a [`Client`] struct for retrieving a song's lyric.
//!
//! It ultilizes the [Genius](https://genius.com) website and its APIs to get lyric data.
//!
//! ## Example
//!
//! ```rust
//! # use anyhow::Result;
//! #
//! # async fn run() -> Result<()> {
//! let client = lyric_finder::Client::new();
//! let result = client.get_lyric("shape of you").await?;
//! match result {
//! lyric_finder::LyricResult::Some {
//! track,
//! artists,
//! lyric,
//! } => {
//! println!("{} by {}'s lyric:\n{}", track, artists, lyric);
//! }
//! lyric_finder::LyricResult::None => {
//! println!("lyric not found!");
//! }
//! }
//! # Ok(())
//! # }
//! ```
const SEARCH_BASE_URL: &str = "https://genius.com/api/search";
pub struct Client {
http: reqwest::Client,
}
#[derive(Debug)]
pub enum LyricResult {
Some {
track: String,
artists: String,
lyric: String,
},
None,
}
impl Client {
pub fn new() -> Self {
Self {
http: reqwest::Client::new(),
}
}
/// Construct a client reusing an existing http client
pub fn from_http_client(http: &reqwest::Client) -> Self {
Self { http: http.clone() }
}
/// Search songs satisfying a given `query`.
pub async fn search_songs(&self, query: &str) -> anyhow::Result<Vec<search::Result>> {
let query = improve_query(query);
log::debug!("search songs: query={query}");
let body = self
.http
.get(format!("{SEARCH_BASE_URL}?q={query}"))
.send()
.await?
.json::<search::Body>()
.await?;
if body.meta.status != 200 {
let message = match body.meta.message {
Some(m) => m,
None => format!("request failed with status code: {}", body.meta.status),
};
anyhow::bail!(message);
}
Ok(body
.response
.map(|r| {
r.hits
.into_iter()
.filter(|hit| hit.ty == "song")
.map(|hit| hit.result)
.collect::<Vec<_>>()
})
.unwrap_or_default())
}
/// Retrieve a song's lyric from a "genius.com" `url`.
pub async fn retrieve_lyric(&self, url: &str) -> anyhow::Result<String> {
let html = self.http.get(url).send().await?.text().await?;
log::debug!("retrieve lyric from url={url}: html={html}");
let lyric = parse::parse(html)?;
Ok(lyric.trim().to_string())
}
/// Process a lyric obtained by crawling the [Genius](https://genius.com) website.
///
/// The lyric received this way may have weird newline spacings between sections (*).
/// The below function tries an ad-hoc method to fix this issue.
///
/// (*): A section often starts with `[`.
fn process_lyric(lyric: String) -> String {
// the below code modifies the `lyric` to make the newline between sections consistent
lyric.replace("\n\n[", "\n[").replace("\n[", "\n\n[")
}
/// Get the lyric of a song satisfying a given `query`.
pub async fn get_lyric(&self, query: &str) -> anyhow::Result<LyricResult> {
// The function first searches songs satisfying the query
// then it retrieves the song's lyric by crawling the "genius.com" website.
let result = {
let mut results = self.search_songs(query).await?;
log::debug!("search results: {results:?}");
if results.is_empty() {
return Ok(LyricResult::None);
}
results.remove(0)
};
let lyric = self.retrieve_lyric(&result.url).await?;
Ok(LyricResult::Some {
track: result.title,
artists: result.artist_names,
lyric: Self::process_lyric(lyric),
})
}
}
impl Default for Client {
fn default() -> Self {
Self::new()
}
}
/// Returns `query` without `remaster` & `remix` information from track/artist query.
/// Returned value is lowercase.
/// These caused wildly invalid lyrics to be found.
/// (try yourself adding remastered 2011 to a song's name when searching in Genius!)
fn improve_query(query: &str) -> String {
// flag for doing something wrong if the song name (after removing remix metadata) is too short.
const SONG_MIN_LENGTH_WO_REMIX_METADATA: usize = 3;
let is_dash = |c: char| c == '-';
// reverse finder for non-filler (space, dashes) chars before an index.
// Acts like a trim to remove undesired spaces and dashes.
let rfind_non_filler = |s: &str, idx: usize| {
let Some(s) = s.get(..idx) else { return idx };
s.char_indices()
.rfind(|(_, c)| !(is_dash(*c) || c.is_whitespace()))
.map_or(idx, |(idx, c)| idx + c.len_utf8())
};
// used to handle longer variants of words: `remixed`, `remastered`, etc.
let end_of_word = |s: &str, idx: usize| {
let Some(s) = s.get(idx..) else { return idx };
s.find(|c: char| !c.is_alphanumeric())
.map_or(idx, |found| found + idx)
};
let mut query = query.to_lowercase();
// remove "xxxx Remaster" from the query
// For example, `{song} xxxx Remastered {artists}` becomes `{song} {artists}`.
if let Some(remaster_start) = query.find("remaster") {
let end = remaster_start + "remaster".len();
let end = end_of_word(&query, end);
let mut start = remaster_start.saturating_sub(1);
let prev = query.get(..remaster_start.saturating_sub(2)).unwrap_or("");
let end_of_prev_word = prev.rfind(' ').unwrap_or(0);
if let Some(year) = query.get(end_of_prev_word + 1..remaster_start.saturating_sub(1)) {
if year.chars().all(|c| c.is_whitespace() || c.is_numeric()) {
start = end_of_prev_word;
}
}
start = rfind_non_filler(&query, start);
query.drain(start..end);
}
// remove "- xxxx yyy remix" from the query
// For example, `{song} - xxxx yyy remix {artists}` becomes `{song} {artists}`.
if let Some(remix_start) = query.find("remix") {
let end = remix_start + "remix".len();
let end = end_of_word(&query, end);
if let Some(metadata_start) = query.rfind(is_dash) {
if metadata_start >= SONG_MIN_LENGTH_WO_REMIX_METADATA {
let start = rfind_non_filler(&query, metadata_start);
query.drain(start..end);
}
}
}
query
}
mod parse {
use html5ever::tendril::TendrilSink;
use html5ever::*;
use markup5ever_rcdom::{Handle, NodeData, RcDom};
const LYRIC_CONTAINER_ATTR: &str = "data-lyrics-container";
/// Parse the HTML content of a "genius.com" lyric page to retrieve the corresponding lyric.
pub fn parse(html: String) -> anyhow::Result<String> {
// parse HTML content into DOM node(s)
let dom = parse_document(RcDom::default(), Default::default())
.from_utf8()
.read_from(&mut (html.as_bytes()))?;
let filter = |data: &NodeData| match data {
NodeData::Element { ref attrs, .. } => attrs
.borrow()
.iter()
.any(|attr| attr.name.local.to_string() == LYRIC_CONTAINER_ATTR),
_ => false,
};
Ok(parse_dom_node(dom.document, &Some(filter), false))
}
/// Parse a dom node and extract the text of children nodes satisfying a requirement.
///
/// The requirement is represented by a `filter` function and a `should_parse` variable.
/// Once a node satisfies a requirement, its children should also satisfy it.
fn parse_dom_node<F>(node: Handle, filter: &Option<F>, mut should_parse: bool) -> String
where
F: Fn(&NodeData) -> bool,
{
log::debug!("parse dom node: node={node:?}, should_parse={should_parse}");
let mut s = String::new();
if !should_parse {
if let Some(f) = filter {
should_parse = f(&node.data);
}
}
match &node.data {
NodeData::Text { contents } => {
if should_parse {
s.push_str(&contents.borrow().to_string());
}
}
NodeData::Element { ref name, .. } => {
if let expanded_name!(html "br") = name.expanded() {
if should_parse {
s.push('\n');
}
}
}
_ => {}
}
node.children.borrow().iter().for_each(|node| {
s.push_str(&parse_dom_node(node.clone(), filter, should_parse));
});
s
}
}
mod search {
use serde::Deserialize;
#[derive(Debug, Deserialize)]
pub struct Body {
pub meta: Metadata,
pub response: Option<Response>,
}
#[derive(Debug, Deserialize)]
pub struct Metadata {
pub status: u16,
pub message: Option<String>,
}
#[derive(Debug, Deserialize)]
pub struct Response {
pub hits: Vec<Hit>,
}
#[derive(Debug, Deserialize)]
pub struct Hit {
#[serde(rename(deserialize = "type"))]
pub ty: String,
pub result: Result,
}
#[derive(Debug, Deserialize)]
pub struct Result {
pub url: String,
pub title: String,
pub artist_names: String,
}
}