@jxsuite/search 0.3.0 → 0.4.0

package/LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2026 Kevin Shenk
+Copyright (c) 2026 Avunu LLC and Jx Suite contributors
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

package/dist/search-index.js

@@ -183,6 +183,16 @@ function entryUrl(basePath, id, trailingSlash) {
 function jxTreeToText(children) {
   return collectText(children).join(" ").replaceAll(/\s+/g, " ").trim();
 }
+var ENTITIES = {
+  "$": "$",
+  "&": "&",
+  ">": ">",
+  "&lt;": "<",
+  "&quot;": '"'
+};
+function innerHtmlText(html) {
+  return html.replaceAll(/<[^>]*>/g, " ").replaceAll(/&#36;|&amp;|&gt;|&lt;|&quot;/g, (entity) => ENTITIES[entity]);
+}
 function collectText(children) {
   const parts = [];
   for (const node of children ?? []) {
@@ -195,6 +205,8 @@ function collectText(children) {
     }
     if (typeof node.textContent === "string") {
       parts.push(node.textContent);
+    } else if (typeof node.innerHTML === "string") {
+      parts.push(innerHtmlText(node.innerHTML));
     }
     if (Array.isArray(node.children)) {
       parts.push(...collectText(node.children));
@@ -286,7 +298,8 @@ var SearchIndex = {
           continue;
         }
         const locale = i18n === null ? undefined : canonicalizeLocale(entry._meta?.locale) ?? undefined;
-        const url = entryUrl(`${localeUrlPrefix(locale, i18n)}${collection.basePath}`, entry.id, trailingSlash);
+        const stamped = entry._meta?.url;
+        const url = typeof stamped === "string" && stamped.startsWith("/") ? stamped : entryUrl(`${localeUrlPrefix(locale, i18n)}${collection.basePath}`, entry.id, trailingSlash);
         const title = typeof data.title === "string" ? data.title : entry.id;
         const description = typeof data.description === "string" ? data.description : "";
         const children = entry.$children;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jxsuite/search",
-  "version": "0.3.0",
+  "version": "0.4.0",
   "description": "Static site search for Jx: build-time index emission from content collections and a headless MiniSearch browser client",
   "license": "MIT",
   "repository": {
@@ -36,7 +36,7 @@
     "upgrade": "bun update --latest && bun install"
   },
   "dependencies": {
-    "@jxsuite/schema": "^2.2.0",
+    "@jxsuite/schema": "^3.0.0",
     "minisearch": "^7.2.0"
   },
   "devDependencies": {

package/src/search-index.ts

@@ -137,11 +137,21 @@ export const SearchIndex = {
          */
         const locale =
           i18n === null ? undefined : (canonicalizeLocale(entry._meta?.locale) ?? undefined);
-        const url = entryUrl(
-          `${localeUrlPrefix(locale, i18n)}${collection.basePath}`,
-          entry.id,
-          trailingSlash,
-        );
+        /*
+         * An entry of a content type with a `route` carries its URL (`_meta.url`), stamped by the
+         * loader from the template its page and its links were generated from, so the index cannot
+         * point at a URL the site does not have. Only an entry with none is placed by the
+         * collection's `basePath` + id, the older, flat rule.
+         */
+        const stamped = entry._meta?.url;
+        const url =
+          typeof stamped === "string" && stamped.startsWith("/")
+            ? stamped
+            : entryUrl(
+                `${localeUrlPrefix(locale, i18n)}${collection.basePath}`,
+                entry.id,
+                trailingSlash,
+              );
         const title = typeof data.title === "string" ? data.title : entry.id;
         const description = typeof data.description === "string" ? data.description : "";
         const children = entry.$children as (JxElement | string)[] | undefined;

package/src/shared.ts

@@ -84,6 +84,28 @@ export function jxTreeToText(children?: (JxElement | string)[]): string {
   return collectText(children).join(" ").replaceAll(/\s+/g, " ").trim();
 }
 
+/** The five entities the parser's escaping and an author's markup commonly leave in `innerHTML`. */
+const ENTITIES: Record<string, string> = {
+  "&#36;": "$",
+  "&amp;": "&",
+  "&gt;": ">",
+  "&lt;": "<",
+  "&quot;": '"',
+};
+
+/**
+ * The text of an `innerHTML` string: tags dropped, the common entities decoded.
+ *
+ * The content loader writes text that contains `${` as escaped `innerHTML` so the build does not
+ * read it as a template (parser.md §3.5), which means the words of a code sample live there and not
+ * in `textContent`. An index that skipped `innerHTML` would silently stop finding them.
+ */
+function innerHtmlText(html: string): string {
+  return html
+    .replaceAll(/<[^>]*>/g, " ")
+    .replaceAll(/&#36;|&amp;|&gt;|&lt;|&quot;/g, (entity) => ENTITIES[entity]!);
+}
+
 function collectText(children: (JxElement | string)[] | undefined): string[] {
   const parts: string[] = [];
   for (const node of children ?? []) {
@@ -96,6 +118,8 @@ function collectText(children: (JxElement | string)[] | undefined): string[] {
     }
     if (typeof node.textContent === "string") {
       parts.push(node.textContent);
+    } else if (typeof node.innerHTML === "string") {
+      parts.push(innerHtmlText(node.innerHTML));
     }
     if (Array.isArray(node.children)) {
       parts.push(...collectText(node.children as (JxElement | string)[]));