Refactor BrowsingDataRemover creation for clarity.

Refactor BrowsingDataRemover creation for clarity.

We currently have two constructors for BrowsingDataRemover. This CL drops down to a single, private constructor with a few more specific public Create methods.

At the same time we're making that chance, we should explicitly refer to the end of time when we want to remove all browsing data from a user's machine, rather than passing in base::Time::Now() which is vulnerable to clock skew or other simply issues of asynchronocity. The new CreateForPeriod method makes that decision explicit.

base::Time::Max() has the nice property that it automatically compares as greater than any reasonable time we might encounter. This means we don't have to update every subsystem that BrowsingDataRemover relies on: they can continue to do simple comparisons without the complications of special-casing the null Time.

BUG=144946, 130732
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=154744

[Mike West, 2012-08-28 10:00:06]

Hello Jochen, Bernhard, and Markus!

I've talked with all three of you at different points in the distant past about
refactoring BrowsingDataRemover constructors.

Does this proposal make sense? If so, great.

I'll start updating callsites once we agree how the interface should look. :)

-mike

[markusheintz_, 2012-08-28 10:35:50]

https://chromiumcodereview.appspot.com/10898002/diff/1/chrome/browser/browsin...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

https://chromiumcodereview.appspot.com/10898002/diff/1/chrome/browser/browsin...
chrome/browser/browsing_data/browsing_data_remover.h:119: static
BrowsingDataRemover* create(Profile* profile,
Method names should start with capital letters.
s/create/Create

same below.

https://chromiumcodereview.appspot.com/10898002/diff/1/chrome/browser/browsin...
chrome/browser/browsing_data/browsing_data_remover.h:120: base::Time
delete_begin = base::Time(),
I think default values for parameters are discouraged in general. Let me double
check that with the style guide.

https://chromiumcodereview.appspot.com/10898002/diff/1/chrome/browser/browsin...
chrome/browser/browsing_data/browsing_data_remover.h:128:
BrowsingDataRemover(Profile* profile, base::Time delete_begin,
I think you may want to make this constructor private in order to force people
to use the new create methods.

https://chromiumcodereview.appspot.com/10898002/diff/1/chrome/browser/browsin...
chrome/browser/browsing_data/browsing_data_remover.h:292: static base::Time
CalculateBeginDeleteTime(TimePeriod time_period);
After making this static, please order the methods correctly according to this: 
http://code.google.com/p/chromium/issues/detail?id=68682

[Mike West, 2012-08-28 12:31:32]

Markus, would you take another look before I loop in OWNERS for the other
directories?

[markusheintz_, 2012-08-28 13:10:23]

LGTM from my side after addressing the 3 comments below.

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:137: static bool
is_removing() { return removing_; }
Please add a comment about the meaning of the returned status.

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:176: // The constructor is
private: use the static Create* methods instead.
I think this line is not needed. The pure existence of the "Create" methods
explains that.

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:300: static base::Time
CalculateBeginDeleteTime(TimePeriod time_period);
Now that this method is static, please move the method declaration to line 174.

[Mike West, 2012-08-28 13:28:38]

Alright, thanks. On to OWNERS.

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:137: static bool
is_removing() { return removing_; }
On 2012/08/28 13:10:23, markusheintz_ wrote:
> Please add a comment about the meaning of the returned status.

Done.

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:176: // The constructor is
private: use the static Create* methods instead.
On 2012/08/28 13:10:23, markusheintz_ wrote:
> I think this line is not needed. The pure existence of the "Create" methods
> explains that.

Done.

https://chromiumcodereview.appspot.com/10898002/diff/2001/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:300: static base::Time
CalculateBeginDeleteTime(TimePeriod time_period);
On 2012/08/28 13:10:23, markusheintz_ wrote:
> Now that this method is static, please move the method declaration to line
174.

Done.

[Mike West, 2012-08-28 13:44:37]

Hello, friendly OWNERS! Would you mind carefully evaluating (and LGTMing :) )
this BrowsingDataRemover refactoring?

ben, can you take chrome/browser/ui/*?
jam, mind looking at the single files in chrome/browser/* and
chrome/browser/automation/*?
cbentzel, what do you think about the change to chrome/browser/prerender?

Thanks!

[jam, 2012-08-28 16:13:04]

On 2012/08/28 13:44:37, Mike West (chromium) wrote:
> Hello, friendly OWNERS! Would you mind carefully evaluating (and LGTMing :) )
> this BrowsingDataRemover refactoring?
> 
> ben, can you take chrome/browser/ui/*?
> jam, mind looking at the single files in chrome/browser/* and
> chrome/browser/automation/*?

what do you mean by single files?

> cbentzel, what do you think about the change to chrome/browser/prerender?
> 
> Thanks!

https://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_da...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

https://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_da...
chrome/browser/browsing_data/browsing_data_remover.h:131: TimePeriod
deletionPeriod);
nit: google  style is deletion_period

https://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_da...
chrome/browser/browsing_data/browsing_data_remover.h:138: static bool
is_removing() { return removing_; }
nit: this should be beside the set_removing setter. also, it should be called
removing() per the style guide

https://codereview.chromium.org/10898002/diff/1004/chrome/browser/chrome_cont...
File chrome/browser/chrome_content_browser_client.cc (right):

https://codereview.chromium.org/10898002/diff/1004/chrome/browser/chrome_cont...
chrome/browser/chrome_content_browser_client.cc:1608: BrowsingDataRemover*
remover = BrowsingDataRemover::CreateForPeriod(profile,
I'm confused about this change, the previous two calls had the same parameters,
but now you're calling two different functions?

[Ben Goodger (Google), 2012-08-28 16:31:04]

ui LGTM

[Mike West, 2012-08-28 18:17:49]

On 2012/08/28 16:13:04, John Abd-El-Malek wrote:
> On 2012/08/28 13:44:37, Mike West (chromium) wrote:
> > ben, can you take chrome/browser/ui/*?

Thanks Ben!

> > jam, mind looking at the single files in chrome/browser/* and
> > chrome/browser/automation/*?
> 
> what do you mean by single files?

I meant chrome/browser/chrome_content_browser_client.cc and
chrome/browser/automation/automation_provider.cc, but thank you very much for
your feedback on the browsing_data files as well!

You're entirely correct about the last point: I should have called
CreateForUnboundedRegion instead of CreateForPeriod. Copy/paste error. I'll get
it fixed up tomorrow morning, as well as the style nits. I hope you won't mind
if I get someone local to LGTM those files sometime during CEST. :)

[cbentzel, 2012-08-29 01:53:13]

prerender LGTM, but see note.

https://chromiumcodereview.appspot.com/10898002/diff/1004/chrome/browser/brow...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

https://chromiumcodereview.appspot.com/10898002/diff/1004/chrome/browser/brow...
chrome/browser/browsing_data/browsing_data_remover.h:121: static
BrowsingDataRemover* CreateForUnboundedRange(Profile* profile);
Should these return scoped_ptr<BrowsingDataRemover> so ownership expectations
are clear?

[Mike West, 2012-09-03 08:02:57]

Cool. Thanks, folks. I'll clean this up based on comments, and land it once
engedy lands his other BDR refactorings.

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_dat...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_dat...
chrome/browser/browsing_data/browsing_data_remover.h:121: static
BrowsingDataRemover* CreateForUnboundedRange(Profile* profile);
On 2012/08/29 01:53:13, cbentzel wrote:
> Should these return scoped_ptr<BrowsingDataRemover> so ownership expectations
> are clear?

BrowsingDataRemover has very strange semantics, in that it deletes itself after
removal is done. I don't think scoped_ptr is correct in that case, as the
creator doesn't actually retain ownership of the pointer.

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_dat...
chrome/browser/browsing_data/browsing_data_remover.h:131: TimePeriod
deletionPeriod);
On 2012/08/28 16:13:04, John Abd-El-Malek wrote:
> nit: google  style is deletion_period

Will do.

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_dat...
chrome/browser/browsing_data/browsing_data_remover.h:138: static bool
is_removing() { return removing_; }
On 2012/08/28 16:13:04, John Abd-El-Malek wrote:
> nit: this should be beside the set_removing setter. also, it should be called
> removing() per the style guide

Will do.

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/chrome_conte...
File chrome/browser/chrome_content_browser_client.cc (right):

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/chrome_conte...
chrome/browser/chrome_content_browser_client.cc:1608: BrowsingDataRemover*
remover = BrowsingDataRemover::CreateForPeriod(profile,
On 2012/08/28 16:13:04, John Abd-El-Malek wrote:
> I'm confused about this change, the previous two calls had the same
parameters,
> but now you're calling two different functions?

No, you're right: this should also be calling CreateForUnboundedRange. I'll fix
it up.

[markusheintz_, 2012-09-03 08:58:43]

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_dat...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

http://codereview.chromium.org/10898002/diff/1004/chrome/browser/browsing_dat...
chrome/browser/browsing_data/browsing_data_remover.h:121: static
BrowsingDataRemover* CreateForUnboundedRange(Profile* profile);
On 2012/09/03 08:02:57, Mike West (chromium) wrote:
> On 2012/08/29 01:53:13, cbentzel wrote:
> > Should these return scoped_ptr<BrowsingDataRemover> so ownership
expectations
> > are clear?
> 
> BrowsingDataRemover has very strange semantics, in that it deletes itself
after
> removal is done. I don't think scoped_ptr is correct in that case, as the
> creator doesn't actually retain ownership of the pointer.

Maybe you should mention in the comment, that the BrowsingDataRemover deletes
itself.

I'll also file a bug for changing this as it seems more complicated as it looks.
I guess that make a very nice starter bug.

[Bernhard Bauer, 2012-09-03 09:29:24]

Thanks for the cleanup!

https://chromiumcodereview.appspot.com/10898002/diff/16001/chrome/browser/bro...
File chrome/browser/browsing_data/browsing_data_remover.h (right):

https://chromiumcodereview.appspot.com/10898002/diff/16001/chrome/browser/bro...
chrome/browser/browsing_data/browsing_data_remover.h:138: static bool removing()
{ return removing_; }
Nit: If the style guide says that the accessor should have the same name as the
member, can we instead rename *both* to |is_removing_|? :)

https://chromiumcodereview.appspot.com/10898002/diff/16001/chrome/browser/bro...
chrome/browser/browsing_data/browsing_data_remover.h:178: // Setter for
removing_; DCHECKs that we can only start removing if we're not
Nit: |removing_| in pipes please.

https://chromiumcodereview.appspot.com/10898002/diff/16001/chrome/browser/bro...
chrome/browser/browsing_data/browsing_data_remover.h:184:
BrowsingDataRemover(Profile* profile, base::Time delete_begin,
Nit: If you're breaking parameter lists, put each parameter on its own line.

[commit-bot: I haz the power, 2012-09-04 08:17:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/[email protected]/10898002/18003

[commit-bot: I haz the power, 2012-09-04 09:41:34]

Try job failure for 10898002-18003 (retry) on mac_rel for step "unit_tests".
It's a second try, previously, step "unit_tests" failed.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[commit-bot: I haz the power, 2012-09-04 11:07:00]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/[email protected]/10898002/14002

[commit-bot: I haz the power, 2012-09-04 13:25:49]

Change committed as 154744