server.ts

import "dotenv/config";
import express from "express";
import { createServer as createViteServer } from "vite";
import multer from "multer";
import * as pdfParseModule from "pdf-parse";
const pdfParse = (pdfParseModule as any).default || pdfParseModule;
import path from "path";

const upload = multer({ storage: multer.memoryStorage() });

async function startServer() {
  const app = express();
  const PORT = 3000;

  app.use(express.json({ limit: '50mb' }));
  app.use(express.urlencoded({ limit: '50mb', extended: true }));

  // API Routes
  app.get("/api/health", (req, res) => {
    res.json({ status: "ok" });
  });

  // Upload PDF and extract text chunks
  app.post("/api/upload-pdf", upload.single("file"), async (req, res) => {
    try {
      if (!req.file) {
        return res.status(400).json({ error: "No file uploaded" });
      }

      const data = await pdfParse(req.file.buffer);
      const text = data.text;

      // Better chunking logic that preserves paragraphs
      // First, normalize multiple newlines to double newlines to retain structural boundaries.
      // Then chunk by paragraphs, roughly 2000-3000 characters per chunk.
      const normalizedText = text.replace(/\r\n/g, '\n').replace(/\n{3,}/g, '\n\n');
      const paragraphs = normalizedText.split(/\n\n/);
      
      const chunks = [];
      let currentChunk = "";
      const MAX_CHUNK_LENGTH = 3000;
      
      for (const p of paragraphs) {
        // Clean up internal breaks inside the paragraph occasionally caused by PDF artifacts
        const cleanParagraph = p.replace(/(?<!\n)\n(?!\n)/g, ' ').trim();
        if (!cleanParagraph) continue;
        
        if (currentChunk.length + cleanParagraph.length > MAX_CHUNK_LENGTH && currentChunk.length > 0) {
          chunks.push(currentChunk);
          currentChunk = cleanParagraph;
        } else {
          currentChunk += (currentChunk ? "\n\n" : "") + cleanParagraph;
        }
      }
      if (currentChunk) {
        chunks.push(currentChunk);
      }

      res.json({ chunks, numPages: data.numpages, info: data.info });
    } catch (error) {
      console.error("Error parsing PDF:", error);
      res.status(500).json({ error: "Failed to parse PDF" });
    }
  });

  // Vite middleware for development
  if (process.env.NODE_ENV !== "production") {
    const vite = await createViteServer({
      server: { middlewareMode: true },
      appType: "spa",
    });
    app.use(vite.middlewares);
  } else {
    const distPath = path.join(process.cwd(), "dist");
    app.use(express.static(distPath));
    app.get("*", (req, res) => {
      res.sendFile(path.join(distPath, "index.html"));
    });
  }

  app.listen(PORT, "0.0.0.0", () => {
    console.log(`Server running on http://localhost:${PORT}`);
  });
}

startServer();