Sylpheed/crates/sylpheed-formats/tests/ui_opt_link_disc.rs

//! What does the `.rat` record's `opt ` link actually point at?
//!
//! `ui-rat-layout.md` reads it as *normal state → focused state*, from one
//! example: `pgpbtn00.rat` links to `pgpbtn00f.rat`. The backlog carries the
//! opposite verdict — "refuted as focus; unexplained otherwise" — with no number
//! behind either. This settles it by classifying **every** `opt ` link on the
//! disc, so the answer is a distribution rather than an anecdote.
//!
//! Four questions per link, in order of how much they would explain:
//!   1. is the target `<source stem>f.<ext>` — the focus pattern?
//!   2. is the target another element of the SAME build?
//!   3. is it a RATC child of the bundle (a resource rather than an element)?
//!   4. otherwise: what do the leftovers look like?

use std::collections::{HashMap, HashSet};
use std::path::{Path, PathBuf};

use sylpheed_formats::{pak::PakArchive, ratc, ui_layout};

mod common;
use common::disc_root;

fn for_each_build(root: &Path, mut f: impl FnMut(&str, &[u8])) {
    let mut paks: Vec<PathBuf> = std::fs::read_dir(root.join("dat"))
        .expect("dat/")
        .flatten()
        .map(|e| e.path())
        .filter(|p| p.extension().and_then(|s| s.to_str()) == Some("pak"))
        .collect();
    paks.sort();
    for p in &paks {
        let name = p.file_name().unwrap().to_string_lossy().to_string();
        let Ok(arc) = PakArchive::open(p) else {
            continue;
        };
        for e in arc.entries() {
            let Ok(bytes) = arc.read(e) else { continue };
            if ratc::is_ratc(&bytes) {
                f(&name, &bytes);
            }
        }
    }
}

#[test]
fn every_opt_link_on_the_disc_is_classified() {
    let Some(root) = disc_root() else {
        eprintln!("SKIP: extracted disc not found (set SYLPHEED_DISC to enable)");
        return;
    };

    let (mut links, mut focus_pattern, mut same_build, mut ratc_child) = (0usize, 0, 0, 0);
    // `opt_link` returns the FIRST `opt ` tag in a record. If records can carry
    // several, every count here understates - so count the tags themselves in
    // the raw bundle rather than trusting the parser's one-per-record view.
    let mut opt_tags = 0usize;
    let mut self_link = 0usize;
    let mut ext_pairs: HashMap<(String, String), usize> = HashMap::new();
    let mut leftovers: Vec<String> = Vec::new();

    for_each_build(&root, |pak, bytes| {
        let Some(build) = ui_layout::parse_build(bytes) else {
            return;
        };
        if build.from_fallback {
            return;
        }
        let names: HashSet<String> = build
            .elements
            .iter()
            .map(|e| e.name.to_ascii_lowercase())
            .collect();
        let kids: HashSet<String> = ratc::parse(bytes)
            .unwrap_or_default()
            .into_iter()
            .map(|c| c.name.to_ascii_lowercase())
            .collect();
        opt_tags += bytes.windows(4).filter(|w| *w == b"opt ").count();
        for el in &build.elements {
            let Some(target) = el.focus_link.as_ref() else {
                continue;
            };
            links += 1;
            let src = el.name.to_ascii_lowercase();
            let tgt = target.to_ascii_lowercase();
            let se = src
                .rsplit_once('.')
                .map(|(_, e)| e.to_string())
                .unwrap_or_default();
            let te = tgt
                .rsplit_once('.')
                .map(|(_, e)| e.to_string())
                .unwrap_or_default();
            *ext_pairs.entry((se, te)).or_default() += 1;
            if tgt == src {
                self_link += 1;
            }
            let is_focus = src
                .rsplit_once('.')
                .map(|(stem, ext)| format!("{stem}f.{ext}") == tgt)
                .unwrap_or(false);
            if is_focus {
                focus_pattern += 1;
            }
            if names.contains(&tgt) {
                same_build += 1;
            }
            if kids.contains(&tgt) {
                ratc_child += 1;
            }
            if !is_focus && leftovers.len() < 14 {
                leftovers.push(format!(
                    "{pak}: {} -> {}{}",
                    el.name,
                    target,
                    if names.contains(&tgt) {
                        "  [declared element]"
                    } else {
                        ""
                    }
                ));
            }
        }
    });

    eprintln!("opt links on the disc: {links}   raw `opt ` tags in the bundles: {opt_tags}");
    eprintln!("  target is <stem>f.<ext> (the focus pattern): {focus_pattern}");
    eprintln!("  target is another element of the same build: {same_build}");
    eprintln!("  target is a RATC child of the bundle:        {ratc_child}");
    eprintln!("  target is the element ITSELF:                {self_link}");
    let mut ep: Vec<_> = ext_pairs.iter().collect();
    ep.sort_by_key(|(_, n)| std::cmp::Reverse(**n));
    eprintln!("  source-ext -> target-ext: {:?}", &ep[..ep.len().min(8)]);
    for l in &leftovers {
        eprintln!("  NOT the f-pattern: {l}");
    }

    assert!(links > 0, "no opt links found — the sweep is broken");

    // MEASURED 2026-08-24 and asserted, because both prior readings were wrong
    // in different directions: `opt ` is not "the focused state" (that fits 73 %
    // of links, not all of them) and it is not "unexplained" either — EVERY
    // link resolves to a RATC child of its own bundle, and every one is
    // .rat -> .rat.
    assert_eq!(links, 1467, "the disc has 1467 opt links");
    assert_eq!(
        ratc_child, links,
        "an opt link does not resolve to a RATC child"
    );
    assert_eq!(
        ext_pairs
            .get(&("rat".to_string(), "rat".to_string()))
            .copied(),
        Some(links),
        "an opt link points at something other than a .rat record"
    );
    assert_eq!(self_link, 0, "a record links to itself");
    assert_eq!(focus_pattern, 1076, "the f-pattern share moved");
}