wiki-search-index 0.1.1 → 0.1.3

package/README.md

@@ -54,6 +54,8 @@ Run locally: `python3 -m http.server` from the repo root, then open
 
 ## Release notes
 
+- 0.1.3 _Fix: HTML entities in headings now decode too, so anchors and section titles match GitHub (0.1.1 only handled body text)._
+- 0.1.2 _Set the published CLI's execute bit (tidy hygiene; npm already chmods bins on install)._
 - 0.1.1 _HTML entities (`—`, `Ӓ`, …) are decoded so they no longer pollute the index._
 - 0.1.0 _Initial release of the `wiki-search-index` builder._
 

package/builder/lib/build-index.mjs

@@ -7,16 +7,18 @@
 
 import {readdir, readFile} from 'node:fs/promises';
 import {join} from 'node:path';
-import {splitSections} from './markdown.mjs';
+import {splitSections, decodeEntities} from './markdown.mjs';
 import {createSlugger} from './slug.mjs';
 
 // GitHub stores page "Foo Bar" as Foo-Bar.md and special pages (_Sidebar,
 // _Footer, …) start with an underscore — those are chrome, not content.
 const isContentPage = name => name.endsWith('.md') && !name.startsWith('_');
 
+// Entity-decoded so a title carrying e.g. & displays the glyph, matching
+// splitSections' headings (see decodeEntities).
 const firstH1 = md => {
   const m = /^#\s+(.+?)\s*#*\s*$/m.exec(md);
-  return m ? m[1].trim() : null;
+  return m ? decodeEntities(m[1]).trim() : null;
 };
 
 export const buildIndex = async ({wikiDir, urlTemplate, siteName, fragments = true}) => {

package/builder/lib/markdown.mjs

@@ -7,14 +7,16 @@ const ATX = /^(#{1,6})\s+(.*?)\s*#*\s*$/;
 
 // Split markdown into sections. Text before the first heading becomes a
 // preamble section with heading=null, level=0. `#` inside fenced code is
-// ignored so code comments don't masquerade as headings.
+// ignored so code comments don't masquerade as headings. Headings are
+// entity-decoded so the display text and the derived slug match GitHub, which
+// renders entities before slugging (see decodeEntities).
 export const splitSections = md => {
   const sections = [{level: 0, heading: null, lines: []}];
   let inFence = false;
   for (const line of md.split(/\r?\n/)) {
     if (FENCE.test(line)) inFence = !inFence;
     const m = inFence ? null : ATX.exec(line);
-    if (m) sections.push({level: m[1].length, heading: m[2].trim(), lines: []});
+    if (m) sections.push({level: m[1].length, heading: decodeEntities(m[2]).trim(), lines: []});
     else sections.at(-1).lines.push(line);
   }
   return sections
@@ -81,6 +83,14 @@ const decodeEntity = (_m, body) => {
   return cp > 0 && cp <= 0x10ffff ? String.fromCodePoint(cp) : ' ';
 };
 
+// Resolve every HTML entity in a string to its character. Shared by toPlainText
+// (the term index) and the heading path (display text + slug). GitHub renders a
+// heading's entities to glyphs before slugging, so "4.2.2 &mdash; 2026-05-29"
+// must decode to "4.2.2 — 2026-05-29" first — then the slugger drops the em dash
+// and the two flanking spaces collapse to "--" (#422--2026-05-29), exactly as
+// GitHub does. Slugging the raw "&mdash;" instead leaks the junk token "mdash".
+export const decodeEntities = s => s.replace(ENTITY_RE, decodeEntity);
+
 // Reduce Markdown to plain, collapsed text. Code *text* is kept (API names are
 // worth searching) — only the fence delimiters are removed.
 export const toPlainText = md =>

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "wiki-search-index",
-  "version": "0.1.1",
+  "version": "0.1.3",
   "description": "Build a self-describing search index from a GitHub wiki (or any Markdown docs) — the indexer for wiki-search.",
   "type": "module",
   "bin": {