From 57fd081dc29263db8b70bb50eb91b4b7bec9e082 Mon Sep 17 00:00:00 2001
From: kevin9327 <5299031+kevin9327@users.noreply.github.com>
Date: Tue, 29 Sep 2026 06:43:57 +0900
Subject: [PATCH] Keep the space a whitespace-only span puts between two words

A text fragment remembers whether space touched its edges, so mergeText
can glue the pieces a parser split out of one word and space the rest.
A text node that is only whitespace was skipped without a trace, so the
fragments on either side of it looked adjacent and were glued.

Wikipedia's HTML writes every &nbsp; as <span typeof="mw:Entity">&nbsp;
</span>, and the text around it shares a parent, so the Python article
read "Firstappeared", "20February 1991", "Python3.0", and "December1989",
43 blocks in all.

A whitespace-only node now marks the text block before it as ending in
space, which is what it was on the page.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---
 src/distill.js        |  6 ++++++
 tests/distill.test.js | 14 ++++++++++++++
 2 files changed, 20 insertions(+)

diff --git a/src/distill.js b/src/distill.js
index d9274c1..71b93c3 100644
--- a/src/distill.js
+++ b/src/distill.js
@@ -347,6 +347,12 @@ export function distill(html, url = '', { view = null, type = '' } = {}) {
       // space that was never in the page.
       if (text) {
         blocks.push({ type: 'text', text, host: node.parentNode, pre: /^\s/.test(raw), post: /\s$/.test(raw) });
+      } else if (raw && blocks.at(-1)?.type === 'text') {
+        // A node that is only a space still separates the words either side
+        // of it. Wikipedia writes every &nbsp; as a span holding nothing
+        // else, and the fragments around it share a parent, so without this
+        // an infobox read "Firstappeared" and "20February 1991".
+        blocks.at(-1).post = true;
       }
       return;
     }
diff --git a/tests/distill.test.js b/tests/distill.test.js
index d365fda..d73f7fb 100644
--- a/tests/distill.test.js
+++ b/tests/distill.test.js
@@ -347,6 +347,20 @@ test('an entity does not put spaces inside a word', () => {
   assert.ok(text.includes('interface & its restraint'), 'a real space was swallowed');
 });
 
+test('a space in an element of its own still separates two words', () => {
+  // Wikipedia's HTML writes every &nbsp; as <span typeof="mw:Entity">&nbsp;</span>.
+  // That span has no text once whitespace is collapsed, and the words either
+  // side of it were glued: "Firstappeared", "20February 1991".
+  const p = distill(`<html><body><table>
+    <tr><th>First<span typeof="mw:Entity">&nbsp;</span>appeared</th>
+    <td>20<span typeof="mw:Entity">&nbsp;</span>February 1991</td></tr>
+  </table><p>ten<b> </b>items</p></body></html>`, 'https://example.test/');
+  const text = p.blocks.map((b) => b.text).join('\n');
+  assert.ok(text.includes('First appeared'), `words glued:\n${text}`);
+  assert.ok(text.includes('20 February 1991'), `words glued:\n${text}`);
+  assert.ok(text.includes('ten items'), `words glued:\n${text}`);
+});
+
 test('an icon button is named by its aria-label, a nameless one is dropped', () => {
   const blocks = social().blocks;
   assert.ok(blocks.some((b) => b.type === 'button' && b.text === 'Follow'), 'a labelled button went missing');