Update Design and Decision log for Multipart download for pre-signed URL

Author: jencymaryjoseph

docs/design/core/presignedURL-Get/DecisionLog.md

@@ -53,3 +53,10 @@ The team has decided to implement functionality only for S3 async client and def
 2. Remove the consumer builder pattern from Get Request Model.
 
 3. throw UnsupportedOperationException for Multipart S3 Client and S3 CRT Client for now.
+
+## Design Review: 07/31/2025
+**Source**: Design : Multipart Download Support for Pre-signed URLs  
+**Attendees**: John Viegas, Zoe Wang, Dongie Agnir, Bole Yi, Ran Vaknin, Saranya Somepalli, David Ho, Olivier Lepage Applin
+
+### Decision Addressed
+Initially considered separate discovery request (bytes=0-0) followed by download, but decided to follow AWS Transfer Manager SEP specification using Range: bytes=0-{partSizeInBytes-1} to download first part AND discover total object size simultaneously from Content-Range header response.

docs/design/core/presignedURL-Get/Design.md

@@ -16,7 +16,7 @@ The Java SDK team has decided to implement a separate `AsyncPresignedUrlExtensio
 
 The design introduces a new helper API `AsyncPresignedUrlExtension` which can be instantiated via the existing `S3AsyncClient`. This extension provides a clean abstraction layer that preserves SDK benefits while handling the unique requirements of pre-signed URL requests.
 
-This design will implement only the GET /download function for presigned URLs for the S3AsyncClient. The synchronous S3Client implementation is deferred to future work.
+This design implements GET/download functionality for presigned URLs for the S3AsyncClient, including multipart download support for large objects. The synchronous S3Client implementation is deferred to future work.
 
 
 
@@ -48,6 +48,32 @@ AsyncPresignedUrlExtension presignedUrlExtension = s3Client.presignedUrlExtensio
 CompletableFuture<GetObjectResponse> response = presignedUrlExtension.getObject(request, AsyncResponseTransformer.toBytes());
 ```
 
+### Multipart Download Support
+
+For objects with size greater than minimumPartSizeInBytes, the SDK automatically uses multipart downloads with HTTP Range headers when multipart is enabled:
+
+```java
+// Enable multipart downloads
+S3AsyncClient s3Client = S3AsyncClient.builder()
+    .multipartEnabled(true)
+    .build();
+
+// Or with custom configuration
+MultipartConfiguration config = MultipartConfiguration.builder()
+    .minimumPartSizeInBytes(8 * 1024 * 1024)   // 8MB parts
+    .build();
+
+S3AsyncClient multipartClient = S3AsyncClient.builder()
+    .multipartConfiguration(config)
+    .build();
+
+// Download automatically uses multipart for objects larger than the configured part size
+CompletableFuture<ResponseBytes<GetObjectResponse>> response = 
+    multipartClient.presignedUrlExtension().getObject(request, AsyncResponseTransformer.toBytes());
+```
+
+The multipart implementation uses Range headers (e.g., `bytes=0-8388607`) instead of partNumber parameters to preserve presigned URL signatures. The first request downloads the initial part while discovering total object size from the Content-Range header.
+
 ### AsyncPresignedUrlExtension Interface
 
 ```java