#!/usr/bin/env bun
/**
* Example usage of the PDF Splitter MCP functions
* This demonstrates how to use all available functions with a real PDF
*/
import { PDFProcessor } from "../src/pdf-processor";
const EXAMPLE_PDF_URL = "https://opensource.adobe.com/dc-acrobat-sdk-docs/pdfstandards/PDF32000_2008.pdf";
async function main() {
const processor = new PDFProcessor();
console.log("=== PDF Splitter MCP Example Usage ===\n");
try {
// 1. Load PDF from URL
console.log("1. Loading PDF from URL...");
const loadResult = await processor.loadPDF(EXAMPLE_PDF_URL);
const pdfId = loadResult.id;
console.log(`✓ PDF loaded successfully!`);
console.log(` - ID: ${pdfId}`);
console.log(` - Pages: ${loadResult.pageCount}\n`);
// 2. Get PDF info
console.log("2. Getting PDF information...");
const info = await processor.getPDFInfo(pdfId);
console.log("✓ PDF Info:");
console.log(` - Path: ${info.path}`);
console.log(` - Page Count: ${info.pageCount}`);
if (info.metadata) {
console.log(" - Metadata:");
Object.entries(info.metadata).forEach(([key, value]) => {
if (value) console.log(` - ${key}: ${value}`);
});
}
console.log();
// 3. Extract document outline/TOC
console.log("3. Extracting document outline...");
const outline = await processor.getFormattedOutline(pdfId);
if (!outline.includes("No outline")) {
console.log("✓ Document Outline (first 10 items):");
const outlineLines = outline.split('\n').filter(line => line.trim());
outlineLines.slice(0, 10).forEach(line => console.log(` ${line}`));
if (outlineLines.length > 10) {
console.log(` ... and ${outlineLines.length - 10} more items`);
}
} else {
console.log("✗ No outline found in this PDF");
}
console.log();
// 4. Extract a specific page
console.log("4. Extracting page 1...");
const page1Content = await processor.extractPage(pdfId, 1);
console.log("✓ Page 1 content (first 200 characters):");
console.log(` "${page1Content.substring(0, 200).replace(/\n/g, ' ').trim()}..."\n`);
// 5. Extract a range of pages
console.log("5. Extracting pages 1-3...");
const rangeContent = await processor.extractRange(pdfId, 1, 3);
const pageCount = (rangeContent.match(/--- Page \d+ ---/g) || []).length;
console.log(`✓ Extracted ${pageCount} pages`);
console.log(` Total content length: ${rangeContent.length} characters\n`);
// 6. Search for text
console.log("6. Searching for 'PDF' in the document...");
const searchResults = await processor.searchPDF(pdfId, "PDF");
console.log(`✓ Found "${searchResults.length}" pages containing "PDF"`);
if (searchResults.length > 0) {
console.log(" First 3 occurrences:");
searchResults.slice(0, 3).forEach(result => {
console.log(` - Page ${result.page}: ${result.matches.length} matches`);
if (result.matches.length > 0) {
console.log(` Context: "...${result.matches[0].context.substring(0, 80)}..."`);
}
});
}
console.log();
// 7. Case-sensitive search
console.log("7. Case-sensitive search for 'adobe'...");
const caseSensitiveResults = await processor.searchPDF(pdfId, "adobe", true);
console.log(`✓ Found ${caseSensitiveResults.length} pages with exact match "adobe"\n`);
// 8. List all loaded PDFs
console.log("8. Listing all loaded PDFs...");
const loadedPDFs = await processor.listLoadedPDFs();
console.log(`✓ Currently loaded PDFs: ${loadedPDFs.length}`);
loadedPDFs.forEach(pdf => {
console.log(` - ${pdf.path}`);
console.log(` ID: ${pdf.id}, Pages: ${pdf.pageCount}`);
});
console.log();
// 9. Render a page as an image
console.log("9. Rendering page 1 as an image...");
const renderedPage = await processor.renderPage(pdfId, 1, 150, 'png');
console.log(`✓ Page rendered successfully!`);
console.log(` - Format: ${renderedPage.format}`);
console.log(` - Dimensions: ${renderedPage.width}x${renderedPage.height}`);
console.log(` - DPI: ${renderedPage.dpi}`);
console.log(` - Size: ${Math.round(renderedPage.base64.length / 1024)}KB (base64)\n`);
// 10. Render multiple pages (thumbnails)
console.log("10. Creating thumbnails of first 3 pages...");
const thumbnails = await processor.renderPages(pdfId, [1, 2, 3], 72, 'jpeg');
console.log(`✓ Created ${thumbnails.length} thumbnails`);
thumbnails.forEach(thumb => {
console.log(` - Page ${thumb.page}: ${thumb.width}x${thumb.height} ${thumb.format}`);
});
console.log();
// 11. Regex search
console.log("11. Searching with regex pattern /\\d+\\.\\d+/ (version numbers)...");
const regexResults = await processor.searchPDF(pdfId, "\\d+\\.\\d+", false, true);
console.log(`✓ Found ${regexResults.length} pages with version number patterns`);
if (regexResults.length > 0) {
const allMatches = regexResults.flatMap(r => r.matches.map(m => m.text));
const uniqueMatches = [...new Set(allMatches)].slice(0, 5);
console.log(` Sample matches: ${uniqueMatches.join(', ')}`);
}
console.log("\n=== Example completed successfully! ===");
} catch (error) {
console.error("Error:", error);
}
}
// Run the example
main().catch(console.error);
