[webkit-reviews] review requested: [Bug 44722] Add support for "downloadToFile" to ResourceRequest and ResourceResponse : [Attachment 70287] downloadAsBlob

[bugzilla-daemon at webkit.org]

Michael Nordman <michaeln at google.com> has asked  for review:
Bug 44722: Add support for "downloadToFile" to ResourceRequest and
ResourceResponse
https://bugs.webkit.org/show_bug.cgi?id=44722

Attachment 70287: downloadAsBlob
https://bugs.webkit.org/attachment.cgi?id=70287&action=review

------- Additional Comments from Michael Nordman <michaeln at google.com>
New patch that allows the additional ResourceRequestBase data member to
survivce the lossy call out to the embedder via willSendRequest(). Seems like a
bug in the WebKit layer that this kind of loss occurs.

I'm not looking to put this here simply as a matter of convenience. This
particular functionality seems to belong here for all the reasons Adam has
articulated.

About the dependency on class Blob, which happens to have script bindings as
well as being a primitive object. I think this is a minor issue. When working
up this patch, I had in mind that Blob was actually defined in the platform
layer (ooops). It is a rather primitive class.
* Perhaps it should be moved.
* In time, the Blob infrastructe could be refactored such that there is a
BlobInternal vs Blob (scriptable) with the 'internal' class being defined in
the platform layer.

About the 'adapter' being simple (ap's comments). Yes I have in mind to build
support for such an adpater into ResourceHandle.cpp directly. It's a simple
ResourceHandleClientWrapper.