@supersuit/artifacts 0.44.1 → 0.44.2

package/CHANGELOG.md

@@ -3,6 +3,20 @@
 `@supersuit/artifacts`. One entry per version, newest first. Each entry says what changed, how a
 host can tell whether it is affected (DETECTOR), what a host does about it (REMEDY), and the tests.
 
+## 0.44.2
+
+**An embed's caption no longer throws the read-along out of step.** The caption under a
+` ```simulator ` or ` ```messages ` block was wrapped as read-along words, but the narration skips
+every fenced block, so each caption added words the voice never says and every word after it lit
+early. A page carrying four embeds ran about sixty words ahead by its end. The embed's figure is now
+`data-nospeak`: the caption still shows, and the highlighter skips it as the voice does.
+
+- DETECTOR: on a narrated page with a simulator or messages block, play past the block. Before, the
+  highlight ran ahead of the voice; now it stays on the spoken word.
+- REMEDY: update the dependency. Nothing to change on a page.
+- Tests: "an embed block is not read aloud" in render.test.tsx, both kinds, with and without a
+  player, the wrapped words equal the narrated words.
+
 ## 0.44.1
 
 **Highlighting text works like any page again.** A press in the gap between paragraphs, in a line's

package/lib/artifacts/render.js

@@ -134,7 +134,10 @@ function EmbedFence({ kind, source, ctx }) {
         return null;
     const src = ctx ? embedSrc(kind, ctx.embed, r.block.value) : null;
     const wide = kind === 'simulator' && src != null;
-    return (_jsxs("figure", { "data-artifact-embed": kind, ...(kind === 'simulator' ? { 'data-artifact-simulator': '' } : {}), ...(wide ? { 'data-wide': '' } : {}), className: "my-8", children: [wide ? _jsx("style", { children: WIDE_EMBED_CSS }) : null, src ? _jsx(EmbedFrame, { src: src, title: r.block.caption, height: r.block.height }) : null, _jsx("figcaption", { className: "mt-2 text-sm opacity-80", children: r.block.caption })] }));
+    return (
+    // data-nospeak: the narration skips every fenced block (narration.ts), so the caption must not
+    // be a read-along word either, or every word after it lights out of step with the voice.
+    _jsxs("figure", { "data-artifact-embed": kind, "data-nospeak": true, ...(kind === 'simulator' ? { 'data-artifact-simulator': '' } : {}), ...(wide ? { 'data-wide': '' } : {}), className: "my-8", children: [wide ? _jsx("style", { children: WIDE_EMBED_CSS }) : null, src ? _jsx(EmbedFrame, { src: src, title: r.block.caption, height: r.block.height }) : null, _jsx("figcaption", { className: "mt-2 text-sm opacity-80", children: r.block.caption })] }));
 }
 /** The one image a paragraph holds and nothing else (whitespace aside), or null. */
 function soleImage(node) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@supersuit/artifacts",
-  "version": "0.44.1",
+  "version": "0.44.2",
   "description": "Publish markdown as branded, read-aloud pages from a Next.js app: the artifacts store, the route factory, the read-along reader, brand packs, password and confidential pages, and share cards.",
   "license": "MIT",
   "type": "module",