bug: WebDAV writer rejects streaming writes larger than one buffer

[KiSetsuFu-PuLiN]

OpenDAL version or commit

v0.58.1 (the same one-shot writer is still present on main)

Affected service or component

services/webdav against a dav-server WebDAV endpoint

Describe the bug

The WebDAV writer is a one-shot writer (oio::OneShotWriter). One writer only accepts a single write call, so any streaming write that exceeds one internal buffer (256 KiB) fails with:

Unsupported (permanent) at write => OneShotWriter doesn't support multiple write
Context:
  service: webdav
  path: ...
  size: 262144
  written: 262144

FuturesAsyncWriter (returned by Operator::writer(...).into_futures_async_write()) flushes its 256 KiB buffer by calling write again, which the one-shot writer rejects. As a result, a single writer cannot upload files larger than 256 KiB.

Steps to Reproduce

use futures::AsyncWriteExt;

let mut writer = op.writer("big.bin").await?.into_futures_async_write();
writer.write_all(&vec![0u8; 600 * 1024]).await?; // fails with the error above
writer.close().await?;

Reproduced against a real dav-server WebDAV endpoint (dav_server::localfs::LocalFs) through opendal::services::Webdav.

Actual Behavior

The second write on the same writer fails with Unsupported (permanent) at write => OneShotWriter doesn't support multiple write.

Expected Behavior

A single writer accepts multiple writes and uploads the data in ordered chunks (for example, the first chunk with a regular PUT and the following chunks with partial PUTs carrying Content-Range), and the service reports write_can_multi = true.

Additional Context

write_can_multi is currently false for webdav even though Operator::writer hands users a writer they can write to multiple times. write_total_max_size / server support for partial PUTs may need to be documented as part of the fix.

Reproduction confirmation

• I reproduced this issue against the affected component. For a service-specific report, I used the actual service rather than only a mock, emulator, or source-code analysis.

Are you willing to submit a PR to fix this bug?

• Yes, I would like to submit a PR.