Merge most instances of DownloadSaveInfo into DownloadUrlParams.

Merge most instances of DownloadSaveInfo into DownloadUrlParams.

This simplifies the programmatic interface for downloads.

[email protected]


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=166269

[Randy Smith (Not in Mondays), 2012-10-30 21:00:13]

Ben: PTAL?  I didn't get rid of DownloadSaveInfo, because I think it still has a
place for those locations that want to just specify "where the download goes" as
opposed to "everything about a download".   But this does substantially simplify
the programmatic interface, so it still seems worth doing to me.

[benjhayden, 2012-11-01 21:19:11]

Would you mind pointing me to all or some of the places that still use DSI? I'm
curious why some callers would rather not use DUP.

LGTM

[Randy Smith (Not in Mondays), 2012-11-02 14:53:18]

On 2012/11/01 21:19:11, benjhayden_chromium wrote:
> Would you mind pointing me to all or some of the places that still use DSI?
I'm
> curious why some callers would rather not use DUP.
> 
> LGTM

The list I have (which *should* be exhaustive :-}) is:

* Underneath ResourceDispatchHost::BeginDownload (conceptually, the URLRequest
has the information about the download source bits, so we don't need to have
that stuff)
* Calls to ResourceDispatcherHost::BeginDownload; same issue, as it takes a
URLRequest.  (Though see below.)
* DownloadCreateInfo.  I played around with refactoring DCI -> DUP + (DCI-DUP),
but at a first pass I couldn't make it make sense.  It might make more sense if
pushed harder, though.

The callers to RDH::BeginDownload are:
* DMI::DownloadUrl.
* Plugin installer.  I think this should be refactored to use DM::DownloadUrl(),
and that's on my medium term list.
* RenderMessageFilter::OnDownloadUrl.  Not sure what to do about this, as it's
IO thread based, so retargeted to DM::DownloadUrl() doesn't make obvious sense. 
But maybe there's another pathway here.  

I have no architectural objection if you want to see about pushing this further,
but I may well have objections about whether it's worth it.  Having a structure
that represents "where the download goes" doesn't particularly bother me
conceptually.  I'm thinking about creating a structure that represents "where
the download comes from", if for no other reason than to simplify some DI
accessors (a whole lot would be combined into one that returned a reference to
that structure).  Then possibly DUP would turn into "comes from" + "goes".  But
that's just a vague thought so far.

[commit-bot: I haz the power, 2012-11-02 14:54:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/[email protected]/11338045/5001

[commit-bot: I haz the power, 2012-11-02 14:54:29]

Failed to apply patch for content/browser/download/drag_download_file.cc:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file content/browser/download/drag_download_file.cc
  Hunk #1 FAILED at 61.
  1 out of 1 hunk FAILED -- saving rejects to file
content/browser/download/drag_download_file.cc.rej

Patch:       content/browser/download/drag_download_file.cc
Index: content/browser/download/drag_download_file.cc
diff --git a/content/browser/download/drag_download_file.cc
b/content/browser/download/drag_download_file.cc
index
71d4aaea1de3b0471f3f7a6f86d460bc60352dc3..e6b5f40b676a0e3dd8c5354b6630fa8e40482ec6
100644
--- a/content/browser/download/drag_download_file.cc
+++ b/content/browser/download/drag_download_file.cc
@@ -61,19 +61,15 @@ class DragDownloadFile::DragDownloadFileUI
     content::DownloadManager* download_manager =
         BrowserContext::GetDownloadManager(web_contents_->GetBrowserContext());
 
-    scoped_ptr<content::DownloadSaveInfo> save_info(
-        new content::DownloadSaveInfo());
-    save_info->file_path = file_path;
-    save_info->file_stream.reset(file_stream.release());
-
     RecordDownloadSource(INITIATED_BY_DRAG_N_DROP);
     scoped_ptr<content::DownloadUrlParameters> params(
-        content::DownloadUrlParameters::FromWebContents(
-            web_contents_, url_, save_info.Pass()));
+        content::DownloadUrlParameters::FromWebContents(web_contents_, url_));
     params->set_referrer(referrer_);
     params->set_referrer_encoding(referrer_encoding_);
     params->set_callback(base::Bind(&DragDownloadFileUI::OnDownloadStarted,
                                     weak_ptr_factory_.GetWeakPtr()));
+    params->set_file_path(file_path);
+    params->set_file_stream(file_stream.Pass()); // Nulls file_stream.
     download_manager->DownloadUrl(params.Pass());
   }

[commit-bot: I haz the power, 2012-11-02 22:14:21]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/[email protected]/11338045/12001

[commit-bot: I haz the power, 2012-11-02 22:14:26]

Presubmit check for 11338045-12001 failed and returned exit status 1.


Running presubmit commit checks ...

** Presubmit Messages **
If this change has an associated bug, add BUG=[bug number].

** Presubmit ERRORS **
Missing LGTM from an OWNER for files in these directories:
    content/browser/web_contents
    chrome/browser/extensions
    chrome/browser/tab_contents

Presubmit checks took 2.3s to calculate.

Was the presubmit check useful? Please send feedback & hate mail to
[email protected]!

[Randy Smith (Not in Mondays), 2012-11-02 23:35:45]

Looking for rubberstamps:
sky@: content/browser/web_contents
aa@: chrome/browser/extensions
avi@: chrome/browser/tab_contents

PTAL?

[sky, 2012-11-05 14:44:20]

LGTM

[Randy Smith (Not in Mondays), 2012-11-06 16:17:27]

Avi, Aaron: PIng?

On 2012/11/02 23:35:45, rdsmith wrote:
> Looking for rubberstamps:
> sky@: content/browser/web_contents
> aa@: chrome/browser/extensions
> avi@: chrome/browser/tab_contents
> 
> PTAL?

[Aaron Boodman, 2012-11-06 18:53:40]

lgtm

[Avi (use Gerrit), 2012-11-06 18:56:03]

lgtm stamp

[commit-bot: I haz the power, 2012-11-06 19:00:00]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/[email protected]/11338045/12001

[commit-bot: I haz the power, 2012-11-06 21:18:14]

Change committed as 166269