Files
bookhoard/web/src/reader/formats/pdf/pdf-search.ts
T
john-okeefe e4c18e51f9 refactor(reader): create modular format-specific architecture
Implement complete modularization of reader code by separating format-specific
functionality into dedicated modules. This replaces the monolithic structure
with a clean, maintainable architecture that separates concerns by format type.

## New Architecture

### Format-Specific Modules
- **formats/reflowable/**: EPUB, FB2, TXT, HTML (page-based pagination)
  - types.ts: Shared type definitions for reflowable formats
  - page-calculator.ts: Word-count based pagination with HTML slicing
  - navigation.ts: Page-based navigation logic
  - progress-tracker.ts: CFI-based progress tracking
  - content-renderer.ts: DOM rendering for page content
  - parser.ts: Unified parser interface for all reflowable formats
  - ebook/**: Migrated ebook-specific features

- **formats/pdf/**: PDF format support
  - Core PDF functionality (navigation, text selection, annotations)
  - Advanced features (bookmarks, search, outlines, dual-page)
  - Page cache and rendering optimizations

- **formats/comic/**: Comic format support
  - Background color, chapter markers, page caching
  - Page ordering, gap adjustments

- **formats/manga/**: Manga format support
  - RTL navigation, vertical scrolling, reading direction

## Key Improvements

1. **Separation of Concerns**: Each format has its own dedicated module
2. **No Circular Dependencies**: Clean import structure
3. **Type Safety**: Comprehensive TypeScript types throughout
4. **Functional Programming**: Pure functions, no OOP complexity
5. **Scalability**: Easy to add new formats without touching core code

## Migration Path

- Old format-specific code in reader/, ebook/, pdf/, comic/, manga/
- New code in formats/[format]/ structure
- Maintains backward compatibility during transition
- Core reader logic remains format-agnostic

This change enables the implementation of page-based pagination for reflowable
formats while keeping PDF, comic, and manga functionality unchanged.
2026-04-09 14:53:12 -04:00

137 lines
3.0 KiB
TypeScript

// Full-text search within PDF documents
import { PDFDocumentProxy } from "pdfjs-dist";
interface SearchResult {
pageNumber: number;
text: string;
index: number;
context: string;
}
// Full-text search within PDF documents
// Procedural implementation (no OOP)
interface SearchResult {
pageNumber: number;
text: string;
index: number;
context: string;
}
interface PDFSearchState {
doc: PDFDocumentProxy | null;
searchResults: SearchResult[];
currentResultIndex: number;
}
async function initializePDFSearch(
doc: PDFDocumentProxy,
): Promise<PDFSearchState> {
return {
doc,
searchResults: [],
currentResultIndex: 0,
};
}
async function searchPDF(
state: PDFSearchState,
query: string,
): Promise<PDFSearchState> {
if (!state.doc) return state;
const searchResults: SearchResult[] = [];
const lowerQuery = query.toLowerCase();
for (let pageNum = 1; pageNum <= state.doc.numPages; pageNum++) {
const page = await state.doc.getPage(pageNum);
const textContent = await page.getTextContent();
let fullText = "";
const textItems = textContent.items.map((item) => {
if (typeof item === "string") return "";
fullText += item.str;
return item.str;
});
const pageText = textItems.join(" ");
const matches = findSearchMatches(pageText, lowerQuery, pageNum);
searchResults.push(...matches);
}
return { ...state, searchResults };
}
function findSearchMatches(
text: string,
query: string,
pageNumber: number,
): SearchResult[] {
const matches: SearchResult[] = [];
const lowerText = text.toLowerCase();
let index = 0;
while ((index = lowerText.indexOf(query, index)) !== -1) {
const start = Math.max(0, index - 50);
const end = Math.min(text.length, index + query.length + 50);
const context = text.slice(start, end);
matches.push({
pageNumber,
text: text.slice(index, index + query.length),
index,
context,
});
index += query.length;
}
return matches;
}
function goToNextSearchResult(
state: PDFSearchState,
): PDFSearchState & { result: SearchResult | null } {
if (state.searchResults.length === 0) {
return { ...state, result: null };
}
const newIndex = (state.currentResultIndex + 1) % state.searchResults.length;
return {
...state,
currentResultIndex: newIndex,
result: state.searchResults[newIndex],
};
}
function goToPreviousSearchResult(
state: PDFSearchState,
): PDFSearchState & { result: SearchResult | null } {
if (state.searchResults.length === 0) {
return { ...state, result: null };
}
const newIndex =
(state.currentResultIndex - 1 + state.searchResults.length) %
state.searchResults.length;
return {
...state,
currentResultIndex: newIndex,
result: state.searchResults[newIndex],
};
}
function getSearchResultCount(state: PDFSearchState): number {
return state.searchResults.length;
}
function clearSearchResults(state: PDFSearchState): PDFSearchState {
return {
...state,
searchResults: [],
currentResultIndex: 0,
};
}