feat: honor OCR text and tag scope in search Chapters/Mangas views

The Chapters/Mangas aggregation views accept an OCR ?text= filter and the
page search intersects tags, but the client dropped both: any text query
collapsed every view into a flat page search that discarded the selected tag
and grouping. Add text to AggregateOptions, thread it into
listTaggedChapters/listTaggedMangas, and rework the search loader so a
tag-scoped view runs the aggregation (optionally text-filtered) and the page
search passes the selected tag.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

frontend/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mangalord-frontend",
-  "version": "0.126.0",
+  "version": "0.127.0",
   "private": true,
   "type": "module",
   "scripts": {

frontend/src/lib/api/page_tags.test.ts

@@ -165,6 +165,16 @@ describe('page_tags api client', () => {
         expect(url).toMatch(/\/v1\/me\/page-tags\/mangas\?tag=funny$/);
     });
 
+    it('aggregation calls forward the OCR text filter when provided', async () => {
+        fetchSpy.mockResolvedValueOnce(
+            ok({ items: [], page: { limit: 50, offset: 0, total: 0 } })
+        );
+        await listTaggedChapters({ tag: 'funny', text: 'hello world' });
+        const url = fetchSpy.mock.calls[0][0] as string;
+        expect(url).toContain('tag=funny');
+        expect(url).toContain('text=hello+world');
+    });
+
     it('searchPages serializes tags (CSV), text and content-warning filters', async () => {
         fetchSpy.mockResolvedValueOnce(
             ok({ items: [], page: { limit: 50, offset: 0, total: 0 } })

frontend/src/lib/api/page_tags.ts

@@ -179,6 +179,9 @@ export type AggregateOptions = {
     order?: 'desc' | 'asc';
     limit?: number;
     offset?: number;
+    /** OCR full-text filter: when set, only pages whose analysis search_doc
+     *  matches are aggregated (and `match_count` reflects that). */
+    text?: string;
 };
 
 function aggregateQs(opts: AggregateOptions): string {
@@ -187,6 +190,7 @@ function aggregateQs(opts: AggregateOptions): string {
     if (opts.order != null) params.set('order', opts.order);
     if (opts.limit != null) params.set('limit', String(opts.limit));
     if (opts.offset != null) params.set('offset', String(opts.offset));
+    if (opts.text != null && opts.text !== '') params.set('text', opts.text);
     return params.toString();
 }
 

frontend/src/routes/search/+page.ts

@@ -103,26 +103,44 @@ export const load: PageLoad = async ({ url }) => {
     // every tag / view / sort / text change instead of freezing the old list.
     const results = (async () => {
         try {
-            // Content search takes precedence over tag browsing.
+            // Chapters / Mangas are tag-scoped aggregation views. When a tag is
+            // selected, run the aggregation — optionally filtered by the OCR
+            // `text` — so the selected view AND tag are honoured. Previously any
+            // `text` collapsed every view into a flat page search, discarding
+            // both the tag and the chapters/mangas grouping.
+            if (tag && view === 'chapters') {
+                const r = await listTaggedChapters({
+                    tag,
+                    order,
+                    limit: 100,
+                    text: text || undefined
+                });
+                return { ...emptyResults, chapters: r.items, total: r.page.total ?? 0 };
+            }
+            if (tag && view === 'mangas') {
+                const r = await listTaggedMangas({
+                    tag,
+                    order,
+                    limit: 100,
+                    text: text || undefined
+                });
+                return { ...emptyResults, mangas: r.items, total: r.page.total ?? 0 };
+            }
+            // Pages view (or a content search with no tag): a positive filter
+            // runs the page search — intersecting the selected tag with the OCR
+            // text / content-warning filters instead of ignoring the tag.
             if (contentSearch) {
                 const r = await searchPages({
                     text: text || undefined,
+                    tags: tag ? [tag] : undefined,
                     cwInclude,
                     cwExclude,
                     limit: 100
                 });
                 return { ...emptyResults, results: r.items, total: r.page.total ?? 0 };
             }
-            // No tag selected → just the chip cloud, no results.
+            // No tag selected and no content filter → just the chip cloud.
             if (!tag) return emptyResults;
-            if (view === 'chapters') {
-                const r = await listTaggedChapters({ tag, order, limit: 100 });
-                return { ...emptyResults, chapters: r.items, total: r.page.total ?? 0 };
-            }
-            if (view === 'mangas') {
-                const r = await listTaggedMangas({ tag, order, limit: 100 });
-                return { ...emptyResults, mangas: r.items, total: r.page.total ?? 0 };
-            }
             const r = await listMyPageTags({ tag, limit: 100 });
             return { ...emptyResults, pages: r.items, total: r.page.total ?? 0 };
         } catch (e) {