Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(298)

Issue 2022483003: ABANDONED CL: LoadURLParams: Add |method| + removing LOAD_TYPE_BROWSER_INITIATED_HTTP_POST. (Closed)

Created:
4 years, 6 months ago by Łukasz Anforowicz
Modified:
4 years, 6 months ago
Reviewers:
clamy, Charlie Reis
CC:
chromium-reviews, mlamouri+watch-content_chromium.org, creis+watch_chromium.org, Peter Beverloo, nasko+codewatch_chromium.org, jam, darin-cc_chromium.org, mkwst+moarreviews-shell_chromium.org, jochen+watch_chromium.org
Base URL:
https://chromium.googlesource.com/chromium/src.git@post-data-my-stuff
Target Ref:
refs/pending/heads/master
Project:
chromium
Visibility:
Public.

Description

ABANDONED CL: This CL has been abandoned in favor of just renaming LOAD_TYPE_BROWSER_INITIATED_HTTP_POST into LOAD_TYPE_HTTP_POST. We don't want to introduce an open-ended |std::string method| above //content layer (because we only support GET and POST for page/frame navigations). ======================================================== LoadURLParams: Add |method| + remove LOAD_TYPE_BROWSER_INITIATED_HTTP_POST. This CL prepares for plumbing arbitrary HTTP method and data (i.e. POST body) through OpenURL code path by adding a |std::string method| field to content::NavigationController::LoadURLParams. This CL also removes LOAD_TYPE_BROWSER_INITIATED_HTTP_POST enum value, which was reduntant and overlapping with other fields, as explained below: - POST part was redundant with 1) the newly added |method| field, and also 2) presence of the already existing |browser_initiated_post_data| field. - BROWSER_INITIATED part was reduntant with already existing |is_renderer_initiated| field. BUG=344348 CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_site_isolation

Patch Set 1 #

Patch Set 2 : Rebasing... #

Patch Set 3 : Rebasing... #

Patch Set 4 : Rebasing... #

Patch Set 5 : Rebasing... #

Patch Set 6 : Rebasing... #

Patch Set 7 : Rebasing... #

Unified diffs Side-by-side diffs Delta from patch set Stats (+38 lines, -39 lines) Patch
M chrome/browser/android/tab_android.cc View 1 chunk +1 line, -2 lines 0 comments Download
M chrome/browser/ui/browser_navigator.cc View 1 chunk +1 line, -2 lines 0 comments Download
M chrome/browser/ui/browser_navigator_params.h View 1 chunk +0 lines, -3 lines 0 comments Download
M components/web_contents_delegate_android/web_contents_delegate_android.cc View 1 chunk +1 line, -2 lines 0 comments Download
M content/browser/frame_host/navigation_controller_impl.cc View 1 2 3 4 4 chunks +15 lines, -12 lines 0 comments Download
M content/browser/frame_host/navigation_controller_impl_unittest.cc View 1 chunk +1 line, -3 lines 0 comments Download
M content/public/android/java/src/org/chromium/content_public/browser/LoadUrlParams.java View 1 2 3 5 chunks +12 lines, -3 lines 0 comments Download
M content/public/browser/navigation_controller.h View 3 chunks +6 lines, -7 lines 0 comments Download
M content/public/browser/page_navigator.h View 1 chunk +0 lines, -3 lines 0 comments Download
M content/shell/browser/shell.cc View 1 chunk +1 line, -2 lines 0 comments Download

Depends on Patchset:

Dependent Patchsets:

Messages

Total messages: 7 (4 generated)
Łukasz Anforowicz
clamy@ + creis@, could you please comment on the overall direction of this CL? Approach ...
4 years, 6 months ago (2016-06-01 00:07:40 UTC) #4
clamy
Having a method parameter makes the LoadURlParams closer to the various *Params we use in ...
4 years, 6 months ago (2016-06-01 13:53:31 UTC) #5
Łukasz Anforowicz
4 years, 6 months ago (2016-06-01 16:28:31 UTC) #6
On 2016/06/01 13:53:31, clamy wrote:
> Having a method parameter makes the LoadURlParams closer to the various
*Params
> we use in content/, which looks better from a code unification standpoint. But
> it also opens up the possibility for embedders (in particular other than
> Chromium) to start doing browser-initiated navigations with methods that are
not
> POST or GET. Since it's far from certain that we actually support that in
> content/ (I bet we don't), I'm not sure we want to make it look like we do. To
> me, just renaming the load type seems like the safest option.

Thanks for the feedback and confirming that my initial approach (introducing an
open ended |std::string method| above //content layer) might not be appropriate
here.  I'll abandon the current CL (and fold the
LOAD_TYPE_BROWSER_INITIATED_HTTP_POST -> LOAD_TYPE_HTTP_POST renaming into
another CL working toward OpenURL/POST support -
https://codereview.chromium.org/2030473002/).

I wonder if we might also want to gravitate toward |bool uses_post| in the
//content layer internals as well.  I also added a comment about //content
public API in
https://codereview.chromium.org/1956383003/diff/420001/content/public/browser....

Powered by Google App Engine
This is Rietveld 408576698