teamsync-pdf-viewer 1.1.0 → 1.2.1

package/CHANGELOG.md

@@ -2,6 +2,53 @@
 
 All notable changes to `teamsync-pdf-viewer` are documented here.
 
+## [1.2.1] — 2026-09-01
+
+### Fixed
+- **The stylesheet no longer restyles the host application.** `dist/style.css` bundles
+  `pdfjs-dist/web/pdf_viewer.css`, ~900 selectors written for pdf.js's own standalone viewer —
+  including bare generic class names such as `.sidebar`, `.dialog`, `.toggle-button`,
+  `.closeButton` and `.messageBar`. Only this package's own rules were scoped under `.tspdf-root`;
+  the pdf.js ones shipped globally. Because the stylesheet is injected at runtime it won every
+  cascade tie, so a host with its own `.sidebar` had it silently restyled (wrong width, white
+  background, `position: relative` + `inset-block-start`) as soon as the viewer module loaded —
+  which happens on preference selection, before any PDF is opened.
+
+  Every selector is now scoped at build time, after nesting is flattened. `:root`/`html`/`body`
+  are re-targeted to `.tspdf-root` rather than prefixed, so the custom properties the viewer's
+  descendants read still resolve. `npm run verify:package` fails the build if any selector is
+  unscoped, or prefixed more than once (which would match nothing and break the viewer's own
+  styling).
+
+## [1.2.0] — 2026-08-31
+
+### Added
+- **`AnnotationManager`** (`instance.annotationManager`, also `instance.Core.annotationManager`):
+  the canonical annotation list with undo/redo, granular `annotationChanged` events
+  (`add` | `modify` | `delete`, with an `imported` flag), authorship stamping
+  (`currentUser`), edit permissions (read-only viewer, read-only annotation, other authors —
+  `permissions.canEditOthers`) and programmatic `addAnnotations` / `updateAnnotation` /
+  `deleteAnnotations` / `getAnnotationById`.
+- **Native XFDF**: `annotationManager.importAnnotations(xfdf)` and
+  `exportAnnotations({ annotList?, useDisplayAuthor? })` (bare fragments per annotation, or a full
+  document). Reads Acrobat/Apryse XFDF (square, circle, line/arrow, ink, highlight quads, freetext,
+  callout, sticky note, stamp with image data, link) and keeps everything else as read-only
+  `opaque` annotations that re-export verbatim; unknown attributes/children (e.g. Apryse
+  `trn-custom-data`) are preserved across a round-trip. Stable UUID ids double as the XFDF `name`.
+- **Page geometry module** (`createPageGeometry`, `rectToPdf`, `quadPointsToRects`, …): the single
+  authority for base-page-space ↔ PDF-user-space conversion, mirroring pdf.js's viewport transform.
+- **Print** (menu item enabled, `instance.print()`): the exported document is printed through the
+  browser's PDF viewer in a hidden frame.
+- `instance.on('annotationChanged', …)` typed event; `Annotation.author/authorId/createdAt/
+  modifiedAt/readOnly/strokes/rects` fields; multi-stroke ink and multi-rect highlights render.
+
+### Fixed
+- Export geometry: annotations are baked exactly where they were drawn (any page size, CropBox
+  offset or intrinsic /Rotate) — replaces the hard-coded `scale = 1.5` conversion.
+- Mixed page sizes: layout, scrolling, hit-testing and fit-to-width/page now use per-page
+  dimensions instead of page 1's.
+- Annotation ids were timestamps (collision-prone); now `crypto.randomUUID()`.
+
 ## [1.1.0] — 2026-08-31
 
 ### Breaking

package/README.md

@@ -1,234 +1,254 @@
 # TeamSync PDF Viewer SDK
 
-Client-side PDF viewing, annotation, search, redaction and watermarking for React — built on
-[pdf.js](https://mozilla.github.io/pdf.js/) and [pdf-lib](https://pdf-lib.js.org/).
-
 <div align="center">
 
 ![TeamSync PDF Viewer UI](https://raw.githubusercontent.com/angelbot-ai/TeamSync-PDF-Viewer/main/docs/images/viewer_ui.png)
 
+### **100% Open Source. 100% Client-Side. Enterprise PDF Engine.**
+
 [![License: CPAL-1.0](https://img.shields.io/badge/License-CPAL--1.0-blue.svg)](LICENSE)
 [![Live Demo](https://img.shields.io/badge/Live%20Demo-pdfviewer.teamsync.com-000000.svg)](https://pdfviewer.teamsync.com/)
-[![npm](https://img.shields.io/npm/v/teamsync-pdf-viewer.svg)](https://www.npmjs.com/package/teamsync-pdf-viewer)
-[![React 18/19](https://img.shields.io/badge/React-18%20%7C%2019-61dafb.svg)](https://react.dev/)
-[![TypeScript](https://img.shields.io/badge/TypeScript-types%20included-blue.svg)](https://www.typescriptlang.org/)
+[![NPM Version](https://img.shields.io/badge/npm-teamsync--pdf--viewer-blue.svg)](https://www.npmjs.com/package/teamsync-pdf-viewer)
+[![React 19](https://img.shields.io/badge/React-19-61dafb.svg)](https://reactjs.org/)
+[![Vite 8](https://img.shields.io/badge/Vite-8.2-646cff.svg)](https://vitejs.dev/)
+[![TypeScript](https://img.shields.io/badge/TypeScript-6.0-blue.svg)](https://www.typescriptlang.org/)
+
+*No paywalls. No vendor lock-in. No server dependencies.*  
+Try the **[Interactive Live Demo](https://pdfviewer.teamsync.com/)** or integrate the high-performance PDF Viewer SDK directly into your React applications.
 
 </div>
 
 ---
 
-## Features
-
-- **Viewing** — continuous or page-by-page scrolling, single / double / cover-facing layouts,
-  thumbnails panel, zoom 10 % – 6400 % with fit-to-width / fit-to-page, pinch and wheel zoom,
-  rotation, pan tool, high-DPI canvas rendering, progressive loading of linearized PDFs (HTTP Range).
-- **Text** — real selectable text layer, client-side search with highlighting and a results panel.
-- **Annotations** — highlighter, freehand ink, shapes (rectangle, ellipse, line, arrow), text,
-  sticky notes, callouts, hyperlinks (page-jump or URL), signature images, undo/redo, keyboard shortcuts.
-- **Redaction** — manual and regex/PII-driven redactions, burned in on export by rasterizing the
-  affected pages so the underlying content is destroyed; redacted text is also stripped from the
-  text layer and search index.
-- **Watermarks** — single or tiled forensic watermarks, rendered live and baked on export.
-- **Export** — download the annotated / redacted / watermarked document; no server round-trip.
-- **Embeddable** — `<TeamSyncViewer>` sizes to its container, several viewers can share a page,
-  keyboard shortcuts apply to the focused viewer only, and importing the package is side-effect free
-  (safe under SSR).
-
-Not yet supported: printing (planned for 1.2), AcroForm filling, rendering of annotations already
-embedded in the PDF, document outline/bookmarks, password prompts (password-protected files raise
-`onPasswordRequired`).
+## 🚀 Why TeamSync PDF Viewer?
+
+Traditional commercial PDF viewers force developers into expensive per-domain licenses, opaque sales calls, and heavy server-side processing dependencies. **TeamSync PDF Viewer** is built from the ground up to solve these pain points.
+
+- 💰 **Zero-Cost, Transparent Licensing**: Uncapped usage, no per-server fees, and no commercial paywalls. 100% free and open-source under CPAL 1.0.
+- ⚡ **100% Client-Side WebAssembly**: Documents never leave the browser sandbox. Process, redact, and render PDFs entirely on the client—guaranteeing instant **HIPAA**, **GDPR**, and **FINRA** compliance.
+- 🔌 **Universal WebViewer API**: Clean, intuitive `WebViewer({...})` API interface that makes integrating into your web applications effortless.
+- 🎨 **Headless & Customizable**: Built for modern React 19 SPA lifecycles. No bloated iFrames to fight—customize toolbars, sidebars, context menus, and controls natively.
 
 ---
 
-## Installation
+## 📦 Installation via NPM
+
+Install the package into your React / TypeScript project:
 
 ```bash
-npm install teamsync-pdf-viewer pdfjs-dist
+# Using npm
+npm install teamsync-pdf-viewer
+
+# Using yarn
+yarn add teamsync-pdf-viewer
+
+# Using pnpm
+pnpm add teamsync-pdf-viewer
 ```
 
-`react`, `react-dom` (18 or 19) and `pdfjs-dist` (6.x) are **peer dependencies**. `pdf-lib` and
-`lucide-react` are installed automatically.
+### Installing from GitHub Packages (Alternative)
 
-### Serve the pdf.js assets
+If installing from GitHub Packages registry:
 
-The library never bundles the pdf.js worker (bundling it produced a 5 MB package with a `data:`
-URL worker that strict Content-Security-Policies reject). Serve the worker — and, for full font /
-CJK / JPEG 2000 support, the CMaps, standard fonts and wasm decoders — from your own origin and tell
-the viewer where they are.
+1. Create or update your `.npmrc` file:
+   ```ini
+   @angelbot-ai:registry=https://npm.pkg.github.com
+   ```
+2. Install via npm:
+   ```bash
+   npm install teamsync-pdf-viewer
+   ```
 
-**Vite / Rollup**
+---
 
-```ts
-import workerSrc from 'pdfjs-dist/build/pdf.worker.min.mjs?url';
-import { configurePdfAssets } from 'teamsync-pdf-viewer';
+## 📖 Official Documentation & Wiki
 
-configurePdfAssets({ workerSrc });
-```
+Explore detailed SDK guides, parameter options, code snippets, and architecture deep dives:
 
-**Next.js (or any static host)** — copy the files at build time and reference them by URL:
+- 🏠 **[Wiki Home](docs/wiki/Home.md)**: Main documentation index & overview.
+- 🚀 **[Getting Started Guide](docs/wiki/Getting-Started.md)**: Installation, CDN, and framework setups.
+- ⚙️ **[API Reference & Options](docs/wiki/API-Reference-&-Options.md)**: Complete parameter list for `WebViewerOptions` & `SDKPermissions`.
+- 🎨 **[Annotations & Markup Guide](docs/wiki/Annotations-&-Markup-Guide.md)**: Freehand drawing, vector shapes, callouts & links.
+- 🛡️ **[Redactions & PII Sanitization Guide](docs/wiki/Redactions-&-PII-Sanitization.md)**: Binary redactions, regex PII scanning & discard options.
+- 💧 **[Forensic Watermarking Guide](docs/wiki/Forensic-Watermarking.md)**: Dynamic single and tiled watermark setup.
+- 🔌 **[Plugin Architecture Guide](docs/wiki/Plugin-Architecture.md)**: Writing custom extension plugins.
 
-```bash
-# e.g. in a prebuild script
-cp node_modules/pdfjs-dist/build/pdf.worker.min.mjs public/pdfjs/
-cp -r node_modules/pdfjs-dist/cmaps node_modules/pdfjs-dist/standard_fonts node_modules/pdfjs-dist/wasm public/pdfjs/
-```
+---
 
-```ts
-configurePdfAssets({
-  workerSrc: '/pdfjs/pdf.worker.min.mjs',
-  cMapUrl: '/pdfjs/cmaps/',
-  standardFontDataUrl: '/pdfjs/standard_fonts/',
-  wasmUrl: '/pdfjs/wasm/',
-});
-```
+## ✨ Uncompromised Feature Set
 
-The worker **must** come from the same `pdfjs-dist` version the viewer runs against
-(`import { pdfjsVersion } from 'teamsync-pdf-viewer'` tells you which one is loaded).
+Everything you need to build collaborative, secure document workflows.
+
+### 🎨 Smart Markup & Annotations
+- **Full Drawing Toolkit**: Freehand ink (`brush`), highlighters, geometric shapes (rectangles, ellipses), arrows, and lines.
+- **Notes & Callouts**: Sticky notes, callout text boxes with directional arrows, and text annotations.
+- **Interactive Hyperlinks**: Create internal page-jump links (`#page=N`) or external web URL links directly on document selections.
+
+### 🛡️ Military-Grade Secure Redaction
+- **Binary-Level Data Obliteration**: We don't just place black boxes over text—redactions are permanently rasterized and burned into the underlying PDF vector structure.
+- **Text Layer Sanitization**: Redacted text is automatically stripped from the DOM `textLayer` and PDF content streams, preventing copy/paste extraction and search indexing.
+- **Automatic Regex Redactions**: Programmatically locate and redact sensitive PII (Aadhaar numbers, SSNs, credit cards, dates of birth) in a single click.
+- **Discard Unapplied Redactions**: Easily discard individual pending redactions or bulk-discard all unapplied redaction marks before committing.
+
+### 📐 Dynamic Layout & Viewport Controls
+- **Fit to Width & Fit to Page**: Container-aware dynamic scaling calculations for responsive reading on any screen size.
+- **Synchronized Page Rotation**: 360° page rotation with real-time vector coordinate transforms so annotations rotate seamlessly with document content.
+
+### 🔍 High-DPI Canvas Rendering & Search
+- **Retina 60 FPS Zoom & Pan**: High-DPI `devicePixelRatio` scaling eliminates blurred text. Micro-debounced rendering enables buttery-smooth 60 FPS trackpad pinching and 360° mouse drag panning.
+- **Contextual Document Search**: Instant client-side text search with real-time match highlighting, result jumping, and match counting.
+
+### 💧 Dynamic Forensic Watermarking
+- Programmatic, non-destructive watermarks rendered on the fly (single centered or full-page tiled) with customizable text, color, opacity, font size, and rotation.
 
 ---
 
-## Usage
+## 💻 Developer SDK Usage
 
-### React component
+### Standard WebViewer Initialization
 
-```tsx
-'use client'; // Next.js App Router: the viewer is client-only
+Drop the SDK into any container element with a single function call:
+
+```typescript
+import { WebViewer } from 'teamsync-pdf-viewer';
+import 'teamsync-pdf-viewer/style.css';
+
+WebViewer({
+  initialDoc: '/sample.pdf',
+  initialScale: 1.0, // 100% natural size
+  enableAnnotations: true,
+  permissions: {
+    canRedact: true,
+    canAddAnnotations: true
+  },
+  watermark: {
+    text: 'CONFIDENTIAL',
+    mode: 'single',
+    size: 48,
+    opacity: 0.1,
+    color: '#dc2626'
+  }
+}, document.getElementById('viewer-container')).then((instance) => {
+  console.log('TeamSync PDF Viewer is ready!', instance);
+  
+  // Programmatic Viewport Controls
+  instance.UI.fitWidth();
+  instance.UI.fitPage();
+
+  // Export annotated PDF buffer
+  instance.Core.documentViewer.getDocument().getFileData().then((pdfBytes) => {
+    console.log('Exported PDF byte length:', pdfBytes.length);
+  });
+});
+```
 
-import { TeamSyncViewer, type WebViewerInstance } from 'teamsync-pdf-viewer';
+### Native React Component Usage
+
+For native React integration, use the `<DocumentViewer />` component:
+
+```tsx
+import { DocumentViewer } from 'teamsync-pdf-viewer';
 import 'teamsync-pdf-viewer/style.css';
 
-export function ContractViewer({ url }: { url: string }) {
+function App() {
   return (
-    <div style={{ height: '80vh' }}>
-      <TeamSyncViewer
-        fileUrl={url}
-        fileName="contract.pdf"
-        assets={{ workerSrc: '/pdfjs/pdf.worker.min.mjs', cMapUrl: '/pdfjs/cmaps/' }}
-        currentUser={{ id: 'u-42', name: 'Jane Doe' }}
-        initialScale="fit-width"
-        watermark={{ text: 'CONFIDENTIAL', mode: 'tiled', opacity: 0.08 }}
-        onReady={(instance: WebViewerInstance) => console.log('ready', instance)}
-        onDocumentLoadError={(error, retry) => {
-          // e.g. mint a fresh presigned URL and retry
-          retry(/* newUrl */);
-        }}
+    <div style={{ width: '100vw', height: '100vh' }}>
+      <DocumentViewer
+        initialDoc="/contract.pdf"
+        scale={1.0}
+        enableAnnotations={true}
+        permissions={{ canRedact: true, canAddAnnotations: true }}
       />
     </div>
   );
 }
 ```
 
-In Next.js, load it with `next/dynamic(() => import('./ContractViewer'), { ssr: false })` or keep
-it inside a client component — the package itself never touches `window` at import time.
-
-### Imperative API
+---
 
-```ts
-import { createWebViewer } from 'teamsync-pdf-viewer';
-import 'teamsync-pdf-viewer/style.css';
+## 🛠️ Local Development Quick Start
 
-const instance = await createWebViewer(
-  { initialDoc: '/sample.pdf', fileName: 'sample.pdf', permissions: { canRedact: true } },
-  document.getElementById('viewer-container')!
-);
+### 1. Clone & Install
 
-instance.on('documentLoaded', ({ numPages }) => console.log(numPages, 'pages'));
-const pdfBytes = await instance.getFileData(); // annotated / redacted / watermarked Uint8Array
-instance.destroy(); // unmounts and releases every listener
+```bash
+git clone https://github.com/angelbot-ai/TeamSync-PDF-Viewer.git
+cd TeamSync-PDF-Viewer
+npm install
 ```
 
-`WebViewer()` remains available as a deprecated alias of `createWebViewer()`.
+### 2. Launch Dev Server
 
----
+```bash
+npm run dev
+```
 
-## API overview
-
-### `<TeamSyncViewer>` props / `WebViewerOptions`
-
-| Prop | Type | Default | Description |
-| :--- | :--- | :--- | :--- |
-| `fileUrl` (`initialDoc`) | `string` | — | URL of the PDF. Fetched by pdf.js (CORS + Range supported). |
-| `fileName` | `string` | `annotated_document.pdf` | Name used for downloads. |
-| `assets` | `PdfAssetPaths` | — | Worker / CMap / font / wasm locations (or call `configurePdfAssets`). |
-| `withCredentials` | `boolean` | `false` | Send cookies with the document request. |
-| `initialScale` | `number \| 'fit-width' \| 'fit-page'` | `1` | Initial zoom. |
-| `initialPage` | `number` | `1` | 1-based page to open. |
-| `readOnly` | `boolean` | `false` | Disables all annotation / redaction editing. |
-| `permissions` | `SDKPermissions` | all `true` | `canAddAnnotations`, `canEditAnnotations`, `canDeleteAnnotations`, `canRedact`. |
-| `currentUser` | `{ id, name }` | — | Author of annotations created in this viewer. |
-| `watermark` | `WatermarkOptions` | — | `{ text, mode: 'single' \| 'tiled', size, opacity, color }`. |
-| `redactions`, `regexRedactions` | `Redaction[]`, `RegExp[]` | `[]` | Pre-set and pattern-driven redactions. |
-| `toolbar`, `sidebars`, `leftPanelOpen` | `boolean` | `true` | Chrome toggles. |
-| `autoFocus` | `boolean` | `false` | Focus the viewer on mount (keyboard shortcuts). |
-| `plugins` | `ViewerPlugin[]` | `[]` | Extension plugins. |
-
-Callbacks: `onReady(instance)`, `onDocumentLoaded({ url, numPages })`,
-`onDocumentLoadError(error, retry(newUrl?), { url, passwordRequired })`, `onFirstPageRendered`,
-`onPasswordRequired`, `onPageChange(page, numPages)`, `onAnnotationsChange(annotations)`.
-
-### `WebViewerInstance`
-
-- `on(type, listener)` / `off` — typed events: `documentLoaded`, `documentLoadError`,
-  `firstPageRendered`, `pageChanged`, `annotationsChanged`, `toolChanged`, `destroy`.
-- `loadDocument(url)` — load or reload (resolves on load, rejects on error).
-- `getFileData({ explicitAnnotations? })` — export the document as `Uint8Array`.
-- `getAnnotations()`, `getCurrentPage()`, `getPageCount()`, `getPdfDocument()`.
-- `destroy()` — unmount (imperative API) and release every listener.
-- `UI.*` / `Core.*` — legacy facades (`UI.openElements`, `UI.fitWidth`,
-  `Core.annotationManager.exportAnnotations()`, `Core.documentViewer.getDocument().getFileData()`).
-
-Full reference: [docs/wiki/API-Reference-&-Options.md](docs/wiki/API-Reference-&-Options.md).
+Open `http://localhost:5173` in your browser to launch the live SDK viewer.
 
 ---
 
-## Documentation
+## 🔌 Extensible Plugin Architecture
 
-- [Getting Started](docs/wiki/Getting-Started.md) · [API Reference & Options](docs/wiki/API-Reference-&-Options.md)
-- [Annotations & Markup](docs/wiki/Annotations-&-Markup-Guide.md) · [Redactions & PII Sanitization](docs/wiki/Redactions-&-PII-Sanitization.md)
-- [Forensic Watermarking](docs/wiki/Forensic-Watermarking.md) · [Plugin Architecture](docs/wiki/Plugin-Architecture.md)
-- [Changelog](CHANGELOG.md)
+TeamSync PDF Viewer features a decoupled **Plugin Architecture**. Core PDF rendering, smart annotations, redaction engines, and search remain lightweight and modular in the main engine, while enterprise extensions plug in as independent modules.
 
----
+```
+┌─────────────────────────────────────────────────────────┐
+│              TeamSync PDF Viewer Core                   │
+│   (Page Rendering, Annotations, Redactions, Search)     │
+└───────────────────────────┬─────────────────────────────┘
+                            │ Plugins Registry API
+┌───────────────────────────▼─────────────────────────────┐
+│                   Custom Extension                      │
+│      (Custom Modals, Action Buttons, Export Hooks)      │
+└─────────────────────────────────────────────────────────┘
+```
 
-## Development
+```typescript
+import { WebViewer } from 'teamsync-pdf-viewer';
+import { myCustomPlugin } from './myCustomPlugin';
 
-```bash
-git clone https://github.com/angelbot-ai/TeamSync-PDF-Viewer.git
-cd TeamSync-PDF-Viewer
-npm install
-npm run dev          # demo application at http://localhost:5173
-npm test             # unit tests (vitest)
-npm run build        # library -> dist/
-npm run verify:package   # publint + arethetypeswrong + artifact guards
-npm run build:app    # demo / iframe package build (Vercel)
+WebViewer({
+  initialDoc: '/sample.pdf',
+  plugins: [
+    myCustomPlugin({ enableFeature: true })
+  ]
+}, document.getElementById('viewer-container'));
 ```
 
-Layout:
+---
 
-```
-src/
-├── index.ts              # public entry (side-effect free)
-├── core/                 # createWebViewer, WebViewerInstance, event bus, pdf.js assets, export
-├── components/           # TeamSyncViewer + internal UI (DocumentViewer, Header, sidebars, modals)
-├── annotations/          # annotation data model
-├── hooks/, utils/        # search, shortcuts, rotation helpers, regex redaction
-├── plugins/              # plugin interfaces
-└── demo/                 # demo & iframe-package bootstrap (not published)
-```
+## 🏛️ Architecture & Tech Stack
 
-### Iframe package (script-tag hosts)
+- **Core Engine**: HTML5, TypeScript, WebAssembly (PDF.js + pdf-lib)
+- **UI & State**: React 19, Lucide Icons, Pure CSS Modules
+- **Build System**: Vite 8 & Oxlint (Sub-500ms tree-shakable builds)
 
-`npm run build:app` produces a self-contained viewer; `public/webviewer.js` wraps it in an iframe
-with a `postMessage` bridge and exposes a `WebViewer(options, element)` function for non-React hosts.
+```
+TeamSync-PDF-Viewer/
+├── src/
+│   ├── components/       # Native React UI (Header, DocumentViewer, Sidebars, Modals)
+│   ├── hooks/            # Search & Keyboard shortcut hooks
+│   ├── utils/            # Redaction algorithms, rotation transforms & vector helpers
+│   ├── plugins/          # Plugin API interfaces & registry
+│   ├── index.ts          # Main SDK package exports
+│   ├── main.tsx          # WebViewer SDK entry point & public API bridge
+│   └── App.tsx           # Demo Application Shell
+├── public/               # Static PDF assets & PDF.js workers
+└── docs/images/          # High-res UI documentation screenshots
+```
 
 ---
 
-## TypeScript notes
+## 📄 License
 
-Declarations ship in `dist/src/`. They resolve under `moduleResolution: "bundler"` (Next.js, Vite,
-webpack) and `node10`; `node16`/`nodenext` resolution is not supported yet because the emitted
-declarations use extension-less relative imports.
+Distributed under the **CPAL 1.0 (Common Public Attribution License)**.  
+100% Free and Open Source for personal and commercial usage.
 
-## License
+---
 
-[CPAL-1.0](LICENSE) — Copyright © 2026 AngelBot Ai Pvt Ltd.
+<div align="center">
+  <b>AngelBot AI • TeamSync • </b> <a href="https://www.teamsync.com" target="_blank" rel="noopener noreferrer">https://www.teamsync.com</a>
+</div>
+
+<div align="center">
+  <b>TeamSync PDF Viewer Engine</b> • Engineered with ❤️ for the Open Web.
+</div>

package/dist/src/annotations/AnnotationManager.d.ts

@@ -0,0 +1,90 @@
+import { PDFDocumentProxy } from 'pdfjs-dist';
+import { Annotation, AnnotationInput } from './types';
+import { XfdfExportOptions } from './xfdf';
+export type AnnotationAction = 'add' | 'modify' | 'delete';
+export interface AnnotationChangedEvent {
+    annotations: Annotation[];
+    action: AnnotationAction;
+    /** True when the change came from importAnnotations()/programmatic APIs flagged as imported. */
+    imported: boolean;
+}
+export interface CommitOptions {
+    /** Mark the change as an import round-trip (not persisted by host bridges, not undoable). */
+    imported?: boolean;
+    /** Bypass read-only / authorship checks (programmatic host operations). */
+    force?: boolean;
+    /** Record in the undo history (default: true unless imported). */
+    undoable?: boolean;
+}
+export interface ViewerUserInfo {
+    id?: string;
+    name?: string;
+}
+type ChangedListener = (event: AnnotationChangedEvent) => void;
+export declare class AnnotationManager {
+    private list;
+    private past;
+    private future;
+    private changedListeners;
+    private listListeners;
+    private user;
+    private readOnly;
+    private canEditOthers;
+    private resolver;
+    setDocument(doc: PDFDocumentProxy | null): void;
+    hasDocument(): boolean;
+    setCurrentUser(name: string, id?: string): void;
+    setCurrentUserInfo(user: ViewerUserInfo | undefined): void;
+    getCurrentUser(): string;
+    getCurrentUserInfo(): ViewerUserInfo;
+    setReadOnly(readOnly: boolean): void;
+    isReadOnly(): boolean;
+    /** Allow editing/deleting annotations authored by other users (default false). */
+    setCanEditOthers(value: boolean): void;
+    /** Whether the current user may modify or delete this annotation through the UI. */
+    canEdit(a: Annotation): boolean;
+    canAdd(): boolean;
+    /** The current list. The array reference is stable until the next change. */
+    getAnnotationsList(): Annotation[];
+    getAnnotationById(id: string): Annotation | undefined;
+    /** React `useSyncExternalStore` pair. */
+    subscribe: (listener: () => void) => (() => void);
+    getSnapshot: () => Annotation[];
+    addEventListener(type: 'annotationChanged', listener: ChangedListener): () => void;
+    removeEventListener(type: 'annotationChanged', listener: ChangedListener): void;
+    /** Apryse-compat no-op: React re-renders from the list. */
+    drawAnnotationsFromList(_annotations: Annotation[]): void;
+    get canUndo(): boolean;
+    get canRedo(): boolean;
+    undo(): boolean;
+    redo(): boolean;
+    /**
+     * Replace the whole list (the UI's edit path). Diffs against the current list, enforces
+     * permissions (rejected edits keep the previous version), records history and emits one
+     * `annotationChanged` event per action.
+     */
+    commit(next: Annotation[], opts?: CommitOptions): Annotation[];
+    addAnnotations(inputs: AnnotationInput[], opts?: CommitOptions): Annotation[];
+    updateAnnotation(id: string, patch: Partial<AnnotationInput>, opts?: CommitOptions): Annotation | undefined;
+    deleteAnnotations(targets: Array<Annotation | string>, opts?: CommitOptions): void;
+    /** Remove every annotation (e.g. on document switch). Not undoable, emits deletes as imported. */
+    clear(): void;
+    /**
+     * Import an XFDF document or bare fragments. Annotations whose `name` already exists are
+     * replaced (modify), new ones added; the change is flagged `imported` and is not undoable.
+     * Returns the imported annotations as they now exist in the list.
+     */
+    importAnnotations(xfdf: string): Promise<Annotation[]>;
+    /**
+     * Export as XFDF. With `annotList`, returns the bare fragment(s) for those annotations (the
+     * form hosts persist per annotation); without it, a complete XFDF document.
+     */
+    exportAnnotations(opts?: XfdfExportOptions): Promise<string>;
+    /** Legacy bespoke JSON export (pre-1.2 `exportAnnotations()` shape). */
+    exportAnnotationsLegacyJson(): string;
+    private requireResolver;
+    private apply;
+    private stampNew;
+    private emit;
+}
+export {};

package/dist/src/annotations/geometry.d.ts

@@ -0,0 +1,51 @@
+import { PDFDocumentProxy, PDFPageProxy } from 'pdfjs-dist';
+export type Matrix = [number, number, number, number, number, number];
+export interface PageGeometry {
+    /** 1-based page number */
+    pageNumber: number;
+    /** PDF user-space bounding box [x0, y0, x1, y1] (the CropBox pdf.js renders). */
+    viewBox: [number, number, number, number];
+    /** Intrinsic page rotation (/Rotate), normalized to 0/90/180/270. */
+    rotate: number;
+    /** Base page size in points (already swapped for 90/270). */
+    width: number;
+    height: number;
+    /** Base-space (viewer) -> PDF user-space transform is the inverse of this viewport transform. */
+    transform: Matrix;
+}
+export interface Rect {
+    x: number;
+    y: number;
+    width: number;
+    height: number;
+}
+export interface Point {
+    x: number;
+    y: number;
+}
+/** PDF rectangle as XFDF writes it: [llx, lly, urx, ury]. */
+export type PdfRect = [number, number, number, number];
+/** Build the geometry for a page from its viewBox and intrinsic rotation (mirrors pdf.js PageViewport at scale 1). */
+export declare function createPageGeometry(pageNumber: number, viewBox: [number, number, number, number], rotate: number): PageGeometry;
+export declare function geometryFromPage(page: PDFPageProxy): PageGeometry;
+/** Small per-document cache so callers can resolve geometry for many annotations cheaply. */
+export declare function createGeometryResolver(pdfDoc: PDFDocumentProxy): (pageNumber: number) => Promise<PageGeometry>;
+export type GeometryResolver = (pageNumber: number) => Promise<PageGeometry>;
+export declare function toPdfPoint(g: PageGeometry, p: Point): Point;
+export declare function fromPdfPoint(g: PageGeometry, p: Point): Point;
+/** Viewer rect (top-left origin) -> PDF rect [llx, lly, urx, ury]. */
+export declare function rectToPdf(g: PageGeometry, r: Rect): PdfRect;
+/** PDF rect [llx, lly, urx, ury] -> viewer rect (top-left origin). */
+export declare function rectFromPdf(g: PageGeometry, r: PdfRect): Rect;
+export declare function pathToPdf(g: PageGeometry, points: Point[]): Point[];
+export declare function pathFromPdf(g: PageGeometry, points: Point[]): Point[];
+/**
+ * XFDF QuadPoints for a viewer rect: x1,y1 (upper-left) x2,y2 (upper-right) x3,y3 (lower-left)
+ * x4,y4 (lower-right) in PDF space — the order Acrobat and Apryse write for text markup.
+ */
+export declare function rectToQuadPoints(g: PageGeometry, r: Rect): number[];
+/** Bounding viewer rect of one or more XFDF quads (8 numbers each). */
+export declare function quadPointsToRects(g: PageGeometry, quads: number[]): Rect[];
+export declare function unionRects(rects: Rect[]): Rect;
+/** Round to 3 decimals for XFDF output (Apryse writes 3, Acrobat up to 6). */
+export declare const fmt: (n: number) => string;

package/dist/src/annotations/ids.d.ts

@@ -0,0 +1,6 @@
+/**
+ * © 2026 AngelBot Ai Pvt Ltd. All rights reserved.
+ * Stable, collision-free annotation ids. They double as the XFDF `name` attribute, which hosts use
+ * to match persisted rows to live annotations, so they must never be timestamps.
+ */
+export declare function newAnnotationId(): string;