Add ignore matching text for name field.

In Alaska Air regsitration page, there is "title" selection. So I added "title/prefix/suffix" for ignore text in name_field and MATCH_SELECT. 



BUG=69912
TEST=Run 20_register_alaskaair.com.html


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=95390

[honten.org, 2011-08-01 01:12:07]

Please review.

[dhollowa, 2011-08-01 17:02:27]

http://codereview.chromium.org/7545003/diff/1/chrome/browser/autofill/credit_...
File chrome/browser/autofill/credit_card_field.cc (right):

http://codereview.chromium.org/7545003/diff/1/chrome/browser/autofill/credit_...
chrome/browser/autofill/credit_card_field.cc:53:
IDS_AUTOFILL_NAME_ON_CARD_CONTEXTUAL_RE);
This is the source of the problem.  This essentially says that any ".*name.*"
following any ".*number.*" should be considered to be a credit card.

This is overly agressive.  We should just get rid of this contextual name
matching entirely.  That should do it.

[Ilya Sherman, 2011-08-01 22:23:12]

This issue is fixed in a different way in http://crrev.com/7531023, which
tightens the credit card number regex instead.

[takano.naoki_gmail.com, 2011-08-01 22:26:48]

Unfortunately I cannot see http://crrev.com/7531023 because of not
googler yet:-)

On Mon, Aug 1, 2011 at 3:23 PM,  <isherman@chromium.org> wrote:
> This issue is fixed in a different way in http://crrev.com/7531023, which
> tightens the credit card number regex instead.
>
> http://codereview.chromium.org/7545003/
>

[Ilya Sherman, 2011-08-01 22:37:10]

On 2011/08/01 22:26:48, takano.naoki_gmail.com wrote:
> Unfortunately I cannot see http://crrev.com/7531023 because of not
> googler yet:-)

Sorry, wrong link.  I meant to link to http://codereview.chromium.org/7531023/

[takano.naoki_gmail.com, 2011-08-01 22:43:43]

I see, but what if the field labe/id/name is "mileage_card_number"
something like that?
Do we consider the case? Of course, still we don't have such a case
for now though.

Thanks,

On Mon, Aug 1, 2011 at 3:37 PM,  <isherman@chromium.org> wrote:
> On 2011/08/01 22:26:48, takano.naoki_gmail.com wrote:
>>
>> Unfortunately I cannot see http://crrev.com/7531023 because of not
>> googler yet:-)
>
> Sorry, wrong link.  I meant to link to
> http://codereview.chromium.org/7531023/
>
> http://codereview.chromium.org/7545003/
>

[Ilya Sherman, 2011-08-02 00:07:25]

On 2011/08/01 22:43:43, takano.naoki_gmail.com wrote:
> I see, but what if the field labe/id/name is "mileage_card_number"
> something like that?
> Do we consider the case? Of course, still we don't have such a case
> for now though.

My intuition is to not worry about this case until we see it show up in the
wild.  The comments indicate that the existing cases are baked in based on real
websites.  So, I'm hesitant to change away from that logic without confirmation
that the new logic is better, in terms of real-world cases.

[takano.naoki_gmail.com, 2011-08-02 00:15:10]

Ok, so I just leave the logic for now.

On Mon, Aug 1, 2011 at 5:07 PM,  <isherman@chromium.org> wrote:
> On 2011/08/01 22:43:43, takano.naoki_gmail.com wrote:
>>
>> I see, but what if the field labe/id/name is "mileage_card_number"
>> something like that?
>> Do we consider the case? Of course, still we don't have such a case
>> for now though.
>
> My intuition is to not worry about this case until we see it show up in the
> wild.  The comments indicate that the existing cases are baked in based on
> real
> websites.  So, I'm hesitant to change away from that logic without
> confirmation
> that the new logic is better, in terms of real-world cases.
>
> http://codereview.chromium.org/7545003/
>

[honten, 2011-08-03 02:24:07]

Please review again.

On 2011/08/02 00:15:10, takano.naoki_gmail.com wrote:
> Ok, so I just leave the logic for now.
> 
> On Mon, Aug 1, 2011 at 5:07 PM,  <mailto:isherman@chromium.org> wrote:
> > On 2011/08/01 22:43:43, http://takano.naoki_gmail.com wrote:
> >>
> >> I see, but what if the field labe/id/name is "mileage_card_number"
> >> something like that?
> >> Do we consider the case? Of course, still we don't have such a case
> >> for now though.
> >
> > My intuition is to not worry about this case until we see it show up in the
> > wild. &nbsp;The comments indicate that the existing cases are baked in based
on
> > real
> > websites. &nbsp;So, I'm hesitant to change away from that logic without
> > confirmation
> > that the new logic is better, in terms of real-world cases.
> >
> > http://codereview.chromium.org/7545003/
> >

[Ilya Sherman, 2011-08-03 07:27:26]

LGTM, thanks

http://codereview.chromium.org/7545003/diff/6001/chrome/browser/autofill/name...
File chrome/browser/autofill/name_field.cc (right):

http://codereview.chromium.org/7545003/diff/6001/chrome/browser/autofill/name...
chrome/browser/autofill/name_field.cc:155: MATCH_DEFAULT | MATCH_SELECT, NULL))
{
nit: Please either indent to the parenthesis or by 4 spaces.

[dhollowa, 2011-08-03 16:33:19]

In the future, please wait for my LGTM before submitting.

[takano.naoki_gmail.com, 2011-08-03 16:34:16]

Sure,

On Wed, Aug 3, 2011 at 9:33 AM,  <dhollowa@chromium.org> wrote:
> In the future, please wait for my LGTM before submitting.
>
> http://codereview.chromium.org/7545003/
>

[dhollowa, 2011-08-03 16:42:34]

LGTM.  This issue came up "Closed" so I thought it had been submitted, but it
appears not to have been.  Weird.  Sorry for the confusion.

[commit-bot: I haz the power, 2011-08-04 06:30:26]

Change committed as 95390