fix(search): show context for matches in italicized text (#4594) (#4631)

Fulltext search showed only the matched word with no surrounding context when the match fell inside inline-styled text (e.g. <i>/<em>). The root cause and fix live in the foliate-js submodule's makeExcerpt (readest/foliate-js#25); bump the submodule pointer to pick it up and add a regression test covering both the simpleSearch and segmenterSearch paths. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-18 00:02:02 +08:00
parent b31699d052
commit 5e5564ef3f
2 changed files with 65 additions and 1 deletions
@@ -0,0 +1,64 @@
+// Regression test for readest/readest issue #4594.
+//
+// In fulltext search, a match that falls inside inline-styled text (e.g. a word
+// wrapped in <i>/<em>/<b>) lands in its own text node. `textWalker` maps each
+// text node to one entry of the `strs` array, so such a word becomes a
+// standalone entry with no surrounding text inside it. The excerpt builder used
+// to read the surrounding context only from *within* the start/end node
+// (`start.slice(0, startOffset)` / `end.slice(endOffset)`), which is empty for
+// these standalone nodes — so the search result showed only the word with no
+// context. Context must be gathered across neighbouring text nodes instead.
+import { describe, expect, it } from 'vitest';
+
+import { search } from 'foliate-js/search.js';
+
+// Forces the deterministic substring path (simpleSearch): default granularity is
+// 'grapheme', and 'variant' sensitivity routes search() to simpleSearch.
+const SUBSTRING_OPTS = { sensitivity: 'variant' as const };
+
+const first = (strs: string[], query: string, opts: Record<string, unknown> = {}) => {
+  const results = [...search(strs, query, opts)] as Array<{
+    excerpt: { pre: string; match: string; post: string };
+  }>;
+  expect(results.length).toBeGreaterThan(0);
+  return results[0]!.excerpt;
+};
+
+describe('foliate-js search excerpt context (#4594)', () => {
+  it('shows surrounding context when the match is in its own (styled) text node', () => {
+    // `<p>The quick <i>brown</i> fox jumps over the lazy dog.</p>`
+    const styled = ['The quick ', 'brown', ' fox jumps over the lazy dog.'];
+
+    const excerpt = first(styled, 'brown', SUBSTRING_OPTS);
+
+    expect(excerpt.match).toBe('brown');
+    expect(excerpt.pre).toContain('The quick');
+    expect(excerpt.post).toContain('fox jumps over the lazy dog');
+  });
+
+  it('produces the same excerpt whether or not the word is wrapped in a tag', () => {
+    const plain = ['The quick brown fox jumps over the lazy dog.'];
+    const styled = ['The quick ', 'brown', ' fox jumps over the lazy dog.'];
+
+    expect(first(styled, 'brown', SUBSTRING_OPTS)).toEqual(first(plain, 'brown', SUBSTRING_OPTS));
+  });
+
+  it('also gathers context on the default (segmenter) search path', () => {
+    const styled = ['The quick ', 'brown', ' fox jumps over the lazy dog.'];
+
+    const excerpt = first(styled, 'brown');
+
+    expect(excerpt.match).toBe('brown');
+    expect(excerpt.pre).toContain('The quick');
+    expect(excerpt.post).toContain('fox jumps');
+  });
+
+  it('keeps the full match text when a phrase spans a styling boundary', () => {
+    // Searching across the <i> boundary: "quick brown fox" spans three nodes.
+    const styled = ['The quick ', 'brown', ' fox jumps over the lazy dog.'];
+
+    const excerpt = first(styled, 'quick brown fox', SUBSTRING_OPTS);
+
+    expect(excerpt.match).toBe('quick brown fox');
+  });
+});