1 files changed, 127 insertions, 0 deletions
diff --git a/import/src/plugins/vgmdb.rs b/import/src/plugins/vgmdb.rs
new file mode 100644
index 0000000..402fd90
--- /dev/null
+++ b/import/src/plugins/vgmdb.rs
@@ -0,0 +1,127 @@
+/*
+    This file is part of jellything (https://codeberg.org/metamuffin/jellything)
+    which is licensed under the GNU Affero General Public License (version 3); see /COPYING.
+    Copyright (C) 2025 metamuffin <metamuffin.org>
+*/
+
+use crate::USER_AGENT;
+use anyhow::{Context, Result};
+use jellycache::{cache, cache_store, HashKey};
+use jellycommon::Asset;
+use log::info;
+use regex::Regex;
+use reqwest::{
+    header::{HeaderMap, HeaderName, HeaderValue},
+    Client, ClientBuilder,
+};
+use std::{
+    sync::{Arc, LazyLock},
+    time::Duration,
+};
+use tokio::{
+    runtime::Handle,
+    sync::Semaphore,
+    time::{sleep_until, Instant},
+};
+
+pub struct Vgmdb {
+    client: Client,
+    rate_limit: Arc<Semaphore>,
+}
+
+static RE_IMAGE_URL_FROM_HTML: LazyLock<Regex> = LazyLock::new(|| {
+    Regex::new(r#"href='(?<url>https://media.vgm.io/artists/[-/\w\.]+)'"#).unwrap()
+});
+
+impl Default for Vgmdb {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl Vgmdb {
+    pub fn new() -> Self {
+        let client = ClientBuilder::new()
+            .default_headers(HeaderMap::from_iter([
+                (
+                    HeaderName::from_static("user-agent"),
+                    HeaderValue::from_static(USER_AGENT),
+                ),
+                (
+                    HeaderName::from_static("x-comment"),
+                    HeaderValue::from_static("Please add an API, thanks!"),
+                ),
+            ]))
+            .build()
+            .unwrap();
+        Self {
+            client,
+            rate_limit: Arc::new(Semaphore::new(3)),
+        }
+    }
+
+    pub fn get_artist_image(&self, id: u64, rt: &Handle) -> Result<Option<Asset>> {
+        if let Some(url) = self.get_artist_image_url(id, rt)? {
+            cache_store(
+                format!("ext/vgmdb/artist-image/{}.image", HashKey(&url)),
+                move || {
+                    rt.block_on(async {
+                        info!("downloading image {url:?}");
+                        Ok(self
+                            .client
+                            .get(url)
+                            .send()
+                            .await?
+                            .error_for_status()?
+                            .bytes()
+                            .await?
+                            .to_vec())
+                    })
+                },
+            )
+            .context("vgmdb media download")
+            .map(Asset)
+            .map(Some)
+        } else {
+            Ok(None)
+        }
+    }
+
+    pub fn get_artist_image_url(&self, id: u64, rt: &Handle) -> Result<Option<String>> {
+        let html = self.scrape_artist_page(id, rt)?;
+        if let Some(cap) = RE_IMAGE_URL_FROM_HTML.captures(&str::from_utf8(&html).unwrap()) {
+            if let Some(url) = cap.name("url").map(|m| m.as_str()) {
+                return Ok(Some(url.to_string()));
+            }
+        }
+        Ok(None)
+    }
+
+    pub fn scrape_artist_page(&self, id: u64, rt: &Handle) -> Result<Vec<u8>> {
+        cache(&format!("ext/vgmdb/artist-page/{id}.html"), move || {
+            rt.block_on(async {
+                let _permit = self.rate_limit.clone().acquire_owned().await?;
+                let permit_drop_ts = Instant::now() + Duration::from_secs(1);
+                info!("scrape artist: {id}");
+
+                let resp = self
+                    .client
+                    .get(format!("https://vgmdb.net/artist/{id}"))
+                    .send()
+                    .await?
+                    .error_for_status()?
+                    .bytes()
+                    .await?
+                    .to_vec();
+
+                tokio::task::spawn(async move {
+                    sleep_until(permit_drop_ts).await;
+                    drop(_permit);
+                });
+
+                Ok(resp)
+            })
+        })
+        .context("vgmdb artist page scrape")
+    }
+}