WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
49291
Add capability for displaying warnings to autofill popup
https://bugs.webkit.org/show_bug.cgi?id=49291
Summary
Add capability for displaying warnings to autofill popup
Ilya Sherman
Reported
2010-11-09 17:32:02 PST
Add capability for displaying warnings to autofill popup
Attachments
Patch
(10.57 KB, patch)
2010-11-09 17:39 PST
,
Ilya Sherman
no flags
Details
Formatted Diff
Diff
Patch
(17.92 KB, patch)
2010-11-10 03:07 PST
,
Ilya Sherman
no flags
Details
Formatted Diff
Diff
Patch
(11.53 KB, patch)
2010-11-12 03:20 PST
,
Ilya Sherman
no flags
Details
Formatted Diff
Diff
Patch
(11.96 KB, patch)
2010-11-12 15:48 PST
,
Ilya Sherman
no flags
Details
Formatted Diff
Diff
Patch
(14.68 KB, patch)
2010-11-14 19:38 PST
,
Ilya Sherman
no flags
Details
Formatted Diff
Diff
Show Obsolete
(4)
View All
Add attachment
proposed patch, testcase, etc.
Ilya Sherman
Comment 1
2010-11-09 17:39:44 PST
Created
attachment 73445
[details]
Patch
Darin Adler
Comment 2
2010-11-09 18:03:37 PST
Comment on
attachment 73445
[details]
Patch It seems strange that we are adding this feature to PopupMenuClient. Can this be restricted to the Chromium platform code by using some multiple inheritance? Why do we need this concept in the cross-platform sources at all?
Ilya Sherman
Comment 3
2010-11-10 03:07:27 PST
Created
attachment 73478
[details]
Patch
Ilya Sherman
Comment 4
2010-11-10 03:12:03 PST
(In reply to
comment #2
)
> (From update of
attachment 73445
[details]
) > It seems strange that we are adding this feature to PopupMenuClient. Can this be restricted to the Chromium platform code by using some multiple inheritance? Why do we need this concept in the cross-platform sources at all?
You're right, this should probably only touch Chromium code. Does the update version look more along the right lines? (Some parts of it were slapped together in the manner of "sleepily appease the compiler gods", so it'll almost certainly need to be cleaned up some; but the general trajectory should hopefully be better.)
Jay Civelli
Comment 5
2010-11-11 15:10:11 PST
I am not reviewer, but here are my comments:
> WebCore/platform/chromium/PopupMenuChromium.cpp:1089 > + clearSelection();
4 space indent.
> WebCore/platform/chromium/PopupMenuClientChromium.h:31 > + virtual bool itemIsWarning(unsigned listIndex) const = 0;
I think it would have been better to add this to PopupMenuClient (like we did previously for item labels).
> WebKit/chromium/src/WebViewImpl.cpp:-1946 > - m_autoFillPopupClient->setSuggestions(
Is that call not required anymore?
Ilya Sherman
Comment 6
2010-11-11 15:34:54 PST
(In reply to
comment #5
)
> I am not reviewer, but here are my comments:
Thanks for taking a look =)
> > WebCore/platform/chromium/PopupMenuClientChromium.h:31 > > + virtual bool itemIsWarning(unsigned listIndex) const = 0; > > I think it would have been better to add this to PopupMenuClient (like we did previously for item labels).
For the sake of consistency, we should probably either add this to PopupMenuClient or move the labels out of there. I don't know the codebase nearly well enough, so I'll defer to others' wisdom here.
> > WebKit/chromium/src/WebViewImpl.cpp:-1946 > > - m_autoFillPopupClient->setSuggestions( > > Is that call not required anymore?
It's called from within m_autoFillPopupClient->initialize(), so I think this call was redundant.
Ilya Sherman
Comment 7
2010-11-12 03:20:36 PST
Created
attachment 73727
[details]
Patch
Ilya Sherman
Comment 8
2010-11-12 03:24:32 PST
(In reply to
comment #6
)
> (In reply to
comment #5
) > > > WebCore/platform/chromium/PopupMenuClientChromium.h:31 > > > + virtual bool itemIsWarning(unsigned listIndex) const = 0; > > > > I think it would have been better to add this to PopupMenuClient (like we did previously for item labels). > > For the sake of consistency, we should probably either add this to PopupMenuClient or move the labels out of there. I don't know the codebase nearly well enough, so I'll defer to others' wisdom here.
The latest patch sidesteps this issue by moving all the code in question down into AutoFillPopupMenuClient.cpp. The original question is likely to become relevant again in the near future, though, as we might want to distinguish warnings by slightly more than just tweaking the font.
Jay Civelli
Comment 9
2010-11-12 09:53:42 PST
LGTM
Ilya Sherman
Comment 10
2010-11-12 15:48:35 PST
Created
attachment 73784
[details]
Patch Actually make the warning display italicized...
Kent Tamura
Comment 11
2010-11-14 18:03:42 PST
Comment on
attachment 73784
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=73784&action=review
> WebCore/ChangeLog:3 > 2010-11-12 Ilya Sherman <
isherman@chromium.org
> > > + Reviewed by NOBODY (OOPS!).
The ChangeLog diff should start at the top of ChangeLog.
> WebCore/ChangeLog:12 > + * platform/chromium/PopupMenuChromium.cpp: > + (WebCore::PopupListBox::selectIndex): > + * platform/chromium/PopupMenuChromium.h:
The list lacks PopupListBox::getRowFont(). We had better write what we change for each of files/functions as possible.
> WebKit/chromium/ChangeLog:3 > 2010-11-12 Ilya Sherman <
isherman@chromium.org
> > > + Reviewed by NOBODY (OOPS!).
The ChangeLog diff doesn't start at the top of ChangeLog.
> WebKit/chromium/src/AutoFillPopupMenuClient.cpp:254 > + return m_uniqueIDs[index] == -1;
The rule "uniqueID ==-1 -> warning" looks implicit. Can we introduce a symbol for this -1? I also found WebViewClient.h has no explanation of uniqueID.
Ilya Sherman
Comment 12
2010-11-14 19:38:02 PST
Comment on
attachment 73784
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=73784&action=review
Thanks as always for the detailed feedback =)
>> WebCore/ChangeLog:3 >> + Reviewed by NOBODY (OOPS!). > > The ChangeLog diff should start at the top of ChangeLog.
It's the diff tool's fault! =) But, it should be fixed now, with some other changes interspersed between my two.
>> WebKit/chromium/src/AutoFillPopupMenuClient.cpp:254 >> + return m_uniqueIDs[index] == -1; > > The rule "uniqueID ==-1 -> warning" looks implicit. Can we introduce a symbol for this -1? > > I also found WebViewClient.h has no explanation of uniqueID.
Oops, this should actually be uniqueID < 0 -- updated the code to match. Also added comments to WebView.h and WebViewClient.h to explain the role of the uniqueID.
Ilya Sherman
Comment 13
2010-11-14 19:38:53 PST
Created
attachment 73871
[details]
Patch
Kent Tamura
Comment 14
2010-11-14 19:44:46 PST
Comment on
attachment 73871
[details]
Patch ok. good.
WebKit Commit Bot
Comment 15
2010-11-15 03:37:38 PST
Comment on
attachment 73871
[details]
Patch Clearing flags on attachment: 73871 Committed
r72001
: <
http://trac.webkit.org/changeset/72001
>
WebKit Commit Bot
Comment 16
2010-11-15 03:37:44 PST
All reviewed patches have been landed. Closing bug.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug