Refactor and enhance type safety across various modules

- Updated function parameters and return types in `page-preview.ts`, `pdf-decrypt.ts`, and `pymupdf-loader.ts` for improved type safety.
- Introduced type definitions for `CpdfInstance`, `PyMuPDFInstance`, and other related types to ensure better type checking.
- Enhanced error handling in `sanitize.ts` by creating a utility function for error messages.
- Removed unnecessary type assertions and improved type inference in `editor.ts`, `serialization.ts`, and `tools.test.ts`.
- Added type definitions for markdown-it plugins to improve compatibility and type safety.
- Enforced stricter TypeScript settings by enabling `noImplicitAny` in `tsconfig.json`.
- Cleaned up test files by refining type assertions and ensuring consistency in type usage.

src/tests/hocr-transform.test.ts

@@ -9,6 +9,7 @@ import {
   calculateWordTransform,
   calculateSpaceTransform,
 } from '../js/utils/hocr-transform';
+import type { OcrWord } from '@/types';
 
 describe('hocr-transform', () => {
   describe('parseBBox', () => {
@@ -210,7 +211,7 @@ describe('hocr-transform', () => {
       bbox: { x0: 0, y0: 100, x1: 500, y1: 130 },
       baseline: { slope: 0, intercept: 0 },
       textangle: 0,
-      words: [],
+      words: [] as OcrWord[],
       direction: 'ltr' as const,
       injectWordBreaks: true,
     };
@@ -302,7 +303,7 @@ describe('hocr-transform', () => {
       bbox: { x0: 0, y0: 100, x1: 500, y1: 130 },
       baseline: { slope: 0, intercept: 0 },
       textangle: 0,
-      words: [],
+      words: [] as OcrWord[],
       direction: 'ltr' as const,
       injectWordBreaks: true,
     };