teamsync-pdf-viewer 1.5.1 → 1.6.0

package/CHANGELOG.md

@@ -2,6 +2,18 @@
 
 All notable changes to `teamsync-pdf-viewer` are documented here.
 
+## [1.6.0] — 2026-09-06
+
+### Added
+- **Whitespace-Tolerant Text Search**: Search queries now match across line breaks, multiple whitespace runs, tabs, NBSP (`\u00A0`), and other unicode whitespace variants seamlessly. Matches in normalized space while accurately reporting raw source coordinates and bounds.
+- **Shared `pageText` Builder & Bounding Box Interpolator**: Unified raw text extraction and character coordinate mapping between PDF search and regex redactions via internal `buildPageText` and `boundsForRawRange` helpers.
+- **Page-Scoped Search & Highlighting**:
+  - Added additive 5th parameter `pages?: number[]` to `searchPdfText(doc, query, redactions, onProgress, pages)`.
+  - Added `options.pages?: number[]` to `instance.searchText(query, options)`.
+  - Added `scope?: 'document' | 'page'` to `instance.highlightSnippet(query, options)` allowing strict page restriction when navigating to snippet citations.
+- **Sidebar Highlighting Precision**: Result `matchLength` now reflects the exact consumed raw glyph span, ensuring sidebar bold highlights match rendered page text identically.
+- **Stable Search Results Sorting**: Corrected search result sorting to maintain page index ascending order while preserving intra-page ascending match order.
+
 ## [1.5.1] — 2026-09-06
 
 ### Added

package/dist/src/core/ViewerInstance.d.ts

@@ -70,8 +70,14 @@ export declare class WebViewerInstance {
     setCurrentPage(pageNumber: number): void;
     /**
      * Search the document for text matches, returning snippets and page bounding boxes.
+     *
+     * @param query - Search term.
+     * @param options - Search options:
+     *   - `pages`: Specific 1-based page numbers to search. Defaults to all pages.
      */
-    searchText(query: string): Promise<SearchResult[]>;
+    searchText(query: string, options?: {
+        pages?: number[];
+    }): Promise<SearchResult[]>;
     /** Get the current list of transient highlights. */
     getTransientHighlights(): TransientHighlight[];
     /**
@@ -87,9 +93,21 @@ export declare class WebViewerInstance {
     /**
      * Search for a snippet or citation query, scroll to its page, and highlight it temporarily.
      * Returns the matched SearchResult or null if not found.
+     *
+     * @param query - The text snippet to locate and highlight.
+     * @param options - Navigation and display options:
+     *   - `pageIndex`: Preferred page (1-based). When `scope: 'page'`, restricts search to this page.
+     *   - `scope`: Search scope. `'document'` (default) searches all pages, keeping `pageIndex` as a preference.
+     *     `'page'` restricts the search strictly to `pageIndex` (avoids full-document walk).
+     *     If `scope: 'page'` is provided without `pageIndex`, it gracefully degrades to `'document'`.
+     *   - `pulse`: Animate focus ring (default true).
+     *   - `scrollTo`: Scroll page into view (default true).
+     *   - `color`: Background highlight color.
+     *   - `tooltip`: Tooltip text above the highlight (defaults to matched snippet).
      */
     highlightSnippet(query: string, options?: {
         pageIndex?: number;
+        scope?: 'document' | 'page';
         pulse?: boolean;
         scrollTo?: boolean;
         color?: string;

package/dist/src/hooks/usePdfSearch.d.ts

@@ -10,10 +10,10 @@ export interface SearchResult {
     bounds: SearchBounds[];
 }
 /**
- * Standalone asynchronous text search across all pages of a PDFDocumentProxy.
+ * Standalone asynchronous text search across all pages (or specified pages) of a PDFDocumentProxy.
  * Sanitizes redacted regions and returns matching snippets and bounding boxes.
  */
-export declare function searchPdfText(pdfDoc: pdfjsLib.PDFDocumentProxy, query: string, redactions?: Redaction[], onProgress?: (progress: number) => void): Promise<SearchResult[]>;
+export declare function searchPdfText(pdfDoc: pdfjsLib.PDFDocumentProxy, query: string, redactions?: Redaction[], onProgress?: (progress: number) => void, pages?: number[]): Promise<SearchResult[]>;
 export declare function usePdfSearch(pdfDoc: pdfjsLib.PDFDocumentProxy | null, redactions?: Redaction[]): {
     search: (query: string) => Promise<void>;
     searchResults: SearchResult[];

package/dist/src/utils/pageText.d.ts

@@ -0,0 +1,23 @@
+import { SearchBounds } from '../core/types';
+export interface PageText {
+    fullText: string;
+    itemIndices: number[];
+    charIndices: number[];
+    normText: string;
+    normToRaw: number[];
+}
+/**
+ * Collapses any consecutive whitespace runs to a single space and trims edges.
+ */
+export declare function collapseWhitespace(s: string): string;
+/**
+ * Builds raw page text with item/char index source-mapping, followed by a second pass
+ * creating normalized text (collapsing any maximal whitespace run to a single space)
+ * and a 1-to-1 index map back to raw string indices.
+ */
+export declare function buildPageText(items: any[]): PageText;
+/**
+ * Computes bounding boxes in PDF coordinate points for a matched range [rawStart, rawEndExcl) in fullText.
+ * Separator sentinels (-1) are skipped, and character positions are interpolated across each TextItem.
+ */
+export declare function boundsForRawRange(items: any[], pageText: Pick<PageText, 'itemIndices' | 'charIndices'>, rawStart: number, rawEndExcl: number): SearchBounds[];