feat(config): P3 config batch — group 7 (user_htmls pruning)

allToHtml now prunes orphan HTML files (no corresponding wiki source) under
html_path — export_ops::prune_orphan_html walks the output tree, maps each
<rel>.html back to <root>/<rel>.<ext>, and deletes those with no source,
skipping the CSS file and any basename in user_htmls. Gated on
!html_filename_parameterization (slugified names can't be reverse-mapped),
mirroring upstream; the pruned paths are reported in the command result.

hl_headers / hl_cb_checked are NOT implemented as toggles: nuwiki already
highlights headers per-level (@vimwikiHeading.level1..6) and checkboxes
always-on via LSP semantic tokens, which supersedes the upstream opt-in
toggles (default off) — same class as the maxhi divergence. Documented in the
gap doc rather than degrading the existing highlighting.

Test: prune_orphan_html_deletes_sourceless_keeps_protected. Full rust suite +
clippy clean.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

crates/nuwiki-lsp/src/commands.rs

@@ -1282,9 +1282,21 @@ fn export_all(backend: &Backend, args: Vec<Value>, force: bool) -> Result<Option
         }
     }
     let css = crate::commands::export_ops::ensure_css(&cfg).map_err(|e| e.to_string())?;
+    // Prune orphan HTML (no wiki source), honouring user_htmls. Skipped when
+    // filename parameterization is on (slugified names can't reverse-map),
+    // mirroring upstream.
+    let pruned: Vec<serde_json::Value> = if cfg.html.html_filename_parameterization {
+        Vec::new()
+    } else {
+        crate::commands::export_ops::prune_orphan_html(&cfg)
+            .into_iter()
+            .map(|p| serde_json::json!(p))
+            .collect()
+    };
     Ok(Some(serde_json::json!({
         "exported": exported,
         "skipped": skipped,
+        "pruned": pruned,
         "files_seen": files.len(),
         "css_path": css.path,
         "css_created": css.created,
@@ -3788,6 +3800,51 @@ pub mod export_ops {
         )
     }
 
+    /// Delete HTML files under `html_path` that have no corresponding wiki
+    /// source (vimwiki's `allToHtml` cleanup), skipping the CSS file and any
+    /// basename listed in `user_htmls`. Returns the pruned paths. Gated by the
+    /// caller on `!html_filename_parameterization` (slugified names can't be
+    /// reverse-mapped to a source).
+    pub fn prune_orphan_html(cfg: &WikiConfig) -> Vec<PathBuf> {
+        let html_root = &cfg.html.html_path;
+        let css = css_path(&cfg.html);
+        let ext = cfg.file_extension.trim_start_matches('.');
+        let mut pruned = Vec::new();
+        let mut stack = vec![html_root.clone()];
+        while let Some(dir) = stack.pop() {
+            let Ok(rd) = std::fs::read_dir(&dir) else {
+                continue;
+            };
+            for entry in rd.flatten() {
+                let path = entry.path();
+                if path.is_dir() {
+                    stack.push(path);
+                    continue;
+                }
+                if path.extension().and_then(|e| e.to_str()) != Some("html") {
+                    continue;
+                }
+                if path == css {
+                    continue;
+                }
+                let base = path.file_name().and_then(|s| s.to_str()).unwrap_or("");
+                if cfg.html.user_htmls.iter().any(|u| u == base) {
+                    continue;
+                }
+                // Map <html_path>/<rel>.html back to <root>/<rel>.<ext>.
+                let Ok(rel) = path.strip_prefix(html_root) else {
+                    continue;
+                };
+                let mut src = cfg.root.join(rel);
+                src.set_extension(ext);
+                if !src.exists() && std::fs::remove_file(&path).is_ok() {
+                    pruned.push(path);
+                }
+            }
+        }
+        pruned
+    }
+
     fn rss_path(cfg: &WikiConfig) -> PathBuf {
         let name = if cfg.html.rss_name.trim().is_empty() {
             "rss.xml"

crates/nuwiki-lsp/tests/html_export.rs

@@ -631,6 +631,32 @@ fn write_rss_uses_base_url_for_public_links() {
     assert!(!xml.contains("file:///tmp/diary"));
 }
 
+#[test]
+fn prune_orphan_html_deletes_sourceless_keeps_protected() {
+    let root = temp_root("prune");
+    let mut c = WikiConfig::from_root(root.clone());
+    c.html = HtmlConfig::from_root(&root);
+    c.html.user_htmls = vec!["keep.html".into()];
+    let html = &c.html.html_path;
+    std::fs::create_dir_all(html).unwrap();
+    // A sourced page (Home.wiki exists) → kept.
+    std::fs::write(root.join("Home.wiki"), "= Home =\n").unwrap();
+    std::fs::write(html.join("Home.html"), "<h1>Home</h1>").unwrap();
+    // An orphan (no source) → pruned.
+    std::fs::write(html.join("Ghost.html"), "<h1>Ghost</h1>").unwrap();
+    // A user_htmls-protected orphan → kept.
+    std::fs::write(html.join("keep.html"), "<p>manual</p>").unwrap();
+    // The CSS file → never pruned.
+    std::fs::write(html.join(&c.html.css_name), "body{}").unwrap();
+
+    let pruned = nuwiki_lsp::commands::export_ops::prune_orphan_html(&c);
+    assert_eq!(pruned.len(), 1, "only the orphan: {pruned:?}");
+    assert!(!html.join("Ghost.html").exists(), "orphan deleted");
+    assert!(html.join("Home.html").exists(), "sourced kept");
+    assert!(html.join("keep.html").exists(), "user_htmls kept");
+    assert!(html.join(&c.html.css_name).exists(), "css kept");
+}
+
 #[test]
 fn write_rss_honours_rss_name_and_max_items() {
     use nuwiki_lsp::diary::DiaryEntry;