Quit using 'pdfjs-dist' because it did not support SSR

app/(playground)/p/[agentId]/actions.ts

@@ -9,7 +9,6 @@ import {
 	withCountMeasurement,
 } from "@/lib/opentelemetry";
 import { type ListBlobResult, del, list, put } from "@vercel/blob";
-import { getDocument } from "pdfjs-dist";
 import { UnstructuredClient } from "unstructured-client";
 import { Strategy } from "unstructured-client/sdk/models/shared";
 import { vercelBlobFileFolder } from "./constants";

app/(playground)/p/[agentId]/prev/beta-proto/files/server-actions.ts

@@ -8,7 +8,7 @@ import {
 	withCountMeasurement,
 } from "@/lib/opentelemetry";
 import { put } from "@vercel/blob";
-import { getDocument } from "pdfjs-dist";
+import { PDFDocument } from "pdf-lib";
 import { UnstructuredClient } from "unstructured-client";
 import { Strategy } from "unstructured-client/sdk/models/shared";
 import { elementsToMarkdown } from "../utils/unstructured";
@@ -99,7 +99,7 @@ export async function parseFile(args: ParseFileInput) {
 		startTime,
 		{
 			strategy,
-			pdf: await getDocument(args.blobUrl).promise,
+			pdf: await PDFDocument.load(await content.arrayBuffer()),
 		},
 	);
 	if (partitionResponse.statusCode !== 200) {

lib/opentelemetry/wrapper.ts

@@ -2,9 +2,7 @@ import { getCurrentMeasurementScope, isRoute06User } from "@/app/(auth)/lib";
 import { waitUntil } from "@vercel/functions";
 import type { LanguageModelUsage } from "ai";
 import type { LanguageModelV1 } from "ai";
-import type { PDFDocumentProxy } from "pdfjs-dist";
-import { getDocument } from "pdfjs-dist";
-import * as pdfWorker from "pdfjs-dist/build/pdf.worker.mjs";
+import { PDFDocument } from "pdf-lib";
 import type { UnstructuredClient } from "unstructured-client";
 import type { PartitionResponse } from "unstructured-client/sdk/models/operations/partition";
 import type { Strategy } from "unstructured-client/sdk/models/shared";
@@ -155,13 +153,13 @@ type VercelBlobOperationType =
 
 interface UnstructuredOptions {
 	strategy: Strategy;
-	pdf: PDFDocumentProxy;
+	pdf: PDFDocument;
 }
 
 type ServiceOptions = UnstructuredOptions | VercelBlobOperationType | undefined;
 
-function getNumPages(pdf: PDFDocumentProxy) {
-	return pdf.numPages;
+function getNumPages(pdf: PDFDocument) {
+	return pdf.getPages().length;
 }
 
 export function withCountMeasurement<T>(
@@ -278,25 +276,22 @@ export async function wrappedPartition(
 	}: PartitionParameters,
 	{ logger, startTime }: MeasureParameters,
 ): Promise<PartitionResponse> {
-	const pdf = await getDocument(blobUrl).promise;
-	const pdfContent = new Blob([await pdf.getData()], {
-		type: "application/pdf",
-	});
+	const content = await fetch(blobUrl).then((response) => response.blob());
 
 	return withCountMeasurement(
 		logger,
 		async () =>
 			client.general.partition({
 				partitionParameters: {
-					files: { fileName, content: pdfContent },
+					files: { fileName, content },
 					strategy,
 					splitPdfPage,
 					splitPdfConcurrencyLevel,
 				},
 			}),
 		ExternalServiceName.Unstructured,
 		startTime,
-		{ strategy, pdf },
+		{ strategy, pdf: await PDFDocument.load(await content.arrayBuffer()) },
 	).finally(() => waitForTelemetryExport());
 }
 

package.json

@@ -80,7 +80,7 @@
 		"next-auth": "^5.0.0-beta.20",
 		"next-themes": "0.3.0",
 		"openai": "4.64.0",
-		"pdfjs-dist": "4.9.155",
+		"pdf-lib": "1.17.1",
 		"pino": "9.5.0",
 		"posthog-js": "1.194.2",
 		"react": "19.0.0-rc-4d577fd2-20241104",
@@ -103,7 +103,6 @@
 		"@tailwindcss/typography": "0.5.15",
 		"@types/bun": "1.1.13",
 		"@types/node": "^22",
-		"@types/pdfjs-dist": "^2.10.378",
 		"@types/react": "npm:types-react@rc",
 		"@types/react-dom": "npm:types-react-dom@rc",
 		"drizzle-kit": "0.23.0",