SoulStream/crates/metadata/examples/mb_query_merge_probe.rs

//! Go/no-go experiment for merging the identity pass's two release-group
//! queries into one (theory 5 of the 2026-07-22 load audit).
//!
//! Current scheme (2 queries):
//!   STUDIO: arid:X AND primarytype:album AND -secondarytype:* AND status:official   (1 page)
//!   HINTS:  arid:X AND primarytype:album AND secondarytype:*                        (≤5 pages)
//!
//! Candidate merged scheme (1 query):
//!   ALL-OFFICIAL: arid:X AND primarytype:album AND status:official                  (≤5 pages)
//!   → client-side split: no secondary types ⇒ studio, some ⇒ hint.
//!
//! The merge is adoptable ONLY if, for every probed artist, the split
//! reproduces BOTH current sets exactly. The known hazard: the hints query
//! has no status filter today, so status:official could drop bootleg-only
//! compilation groups from the hint set — a classification change.
//!
//! Usage: cargo run -p metadata --example mb_query_merge_probe -- "Artist" ...

use std::collections::BTreeSet;
use std::time::Duration;

const UA: &str = "SoulSeekPlus/0.1 (dev build)";

#[tokio::main]
async fn main() {
    let artists: Vec<String> = std::env::args().skip(1).collect();
    if artists.is_empty() {
        eprintln!("usage: mb_query_merge_probe -- \"Artist\" ...");
        return;
    }
    let http = reqwest::Client::builder()
        .user_agent(UA)
        .timeout(Duration::from_secs(15))
        .build()
        .expect("client");
    let meta = metadata::MetadataClient::new();

    let mut all_ok = true;
    for artist in &artists {
        let search = meta.catalog_search(artist).await;
        let Some(deezer_id) = search
            .artists
            .iter()
            .find(|a| norm(&a.name) == norm(artist))
            .map(|a| a.id)
        else {
            println!("== {artist}: no exact Deezer artist, skipped ==");
            continue;
        };
        let Some(mbid) = linked_mbid(&http, deezer_id).await else {
            println!("== {artist}: no linked MBID, skipped ==");
            continue;
        };

        let studio_q =
            format!("arid:{mbid} AND primarytype:album AND -secondarytype:* AND status:official");
        let hints_q = format!("arid:{mbid} AND primarytype:album AND secondarytype:*");
        let merged_q = format!("arid:{mbid} AND primarytype:album AND status:official");

        // Current studio behavior is a SINGLE page.
        let studio_now: BTreeSet<String> = fetch_groups(&http, &studio_q, 1)
            .await
            .into_iter()
            .filter(|g| g.secondary.is_empty())
            .map(|g| g.title)
            .collect();
        let hints_now: BTreeSet<String> = fetch_groups(&http, &hints_q, 5)
            .await
            .into_iter()
            .map(|g| format!("{} :: {:?}", g.title, g.secondary))
            .collect();

        let merged = fetch_groups(&http, &merged_q, 5).await;
        let studio_merged: BTreeSet<String> = merged
            .iter()
            .filter(|g| g.secondary.is_empty())
            .map(|g| g.title.clone())
            .collect();
        let hints_merged: BTreeSet<String> = merged
            .iter()
            .filter(|g| !g.secondary.is_empty())
            .map(|g| format!("{} :: {:?}", g.title, g.secondary))
            .collect();

        let studio_ok = studio_now == studio_merged;
        let hints_ok = hints_now == hints_merged;
        all_ok &= studio_ok && hints_ok;
        println!(
            "== {artist} == studio {}/{} {}   hints {}/{} {}",
            studio_now.len(),
            studio_merged.len(),
            if studio_ok { "IDENTICAL" } else { "DIFFERS" },
            hints_now.len(),
            hints_merged.len(),
            if hints_ok { "IDENTICAL" } else { "DIFFERS" },
        );
        if !studio_ok {
            for extra in studio_merged.difference(&studio_now) {
                println!("   studio only-in-merged: {extra}");
            }
            for missing in studio_now.difference(&studio_merged) {
                println!("   studio missing-from-merged: {missing}");
            }
        }
        if !hints_ok {
            for extra in hints_merged.difference(&hints_now) {
                println!("   hint only-in-merged: {extra}");
            }
            for missing in hints_now.difference(&hints_merged) {
                println!("   hint missing-from-merged: {missing}");
            }
        }
    }
    println!(
        "\nVERDICT: {}",
        if all_ok {
            "merged query reproduced both sets exactly for every probed artist"
        } else {
            "NOT identical — do not merge"
        }
    );
}

struct Group {
    title: String,
    secondary: Vec<String>,
}

async fn fetch_groups(http: &reqwest::Client, query: &str, max_pages: usize) -> Vec<Group> {
    let mut out = Vec::new();
    for page in 0..max_pages {
        throttle().await;
        let url = format!(
            "https://musicbrainz.org/ws/2/release-group/?query={}&limit=100&offset={}&fmt=json",
            urlencoding(query),
            page * 100
        );
        let Ok(resp) = http.get(&url).send().await else {
            break;
        };
        let Ok(body) = resp.json::<serde_json::Value>().await else {
            break;
        };
        let Some(groups) = body.get("release-groups").and_then(|v| v.as_array()) else {
            break;
        };
        if groups.is_empty() {
            break;
        }
        let page_len = groups.len();
        for g in groups {
            let Some(title) = g.get("title").and_then(|v| v.as_str()) else {
                continue;
            };
            let secondary = g
                .get("secondary-types")
                .and_then(|v| v.as_array())
                .map(|a| {
                    let mut s: Vec<String> = a
                        .iter()
                        .filter_map(|v| v.as_str().map(String::from))
                        .collect();
                    s.sort();
                    s
                })
                .unwrap_or_default();
            out.push(Group {
                title: title.to_string(),
                secondary,
            });
        }
        if page_len < 100 {
            break;
        }
    }
    out
}

async fn linked_mbid(http: &reqwest::Client, deezer_id: u64) -> Option<String> {
    throttle().await;
    let resource = urlencoding(&format!("https://www.deezer.com/artist/{deezer_id}"));
    let url =
        format!("https://musicbrainz.org/ws/2/url?resource={resource}&inc=artist-rels&fmt=json");
    let body = http
        .get(&url)
        .send()
        .await
        .ok()?
        .json::<serde_json::Value>()
        .await
        .ok()?;
    body.get("relations")?.as_array()?.iter().find_map(|r| {
        r.pointer("/artist/id")
            .and_then(|v| v.as_str())
            .map(String::from)
    })
}

async fn throttle() {
    tokio::time::sleep(Duration::from_millis(1100)).await;
}

fn urlencoding(s: &str) -> String {
    let mut out = String::new();
    for b in s.bytes() {
        match b {
            b'A'..=b'Z' | b'a'..=b'z' | b'0'..=b'9' | b'-' | b'_' | b'.' | b'~' => {
                out.push(b as char)
            }
            _ => out.push_str(&format!("%{b:02X}")),
        }
    }
    out
}

fn norm(s: &str) -> String {
    s.to_lowercase()
        .chars()
        .filter(|c| c.is_alphanumeric())
        .collect()
}