Bug 111860 – [EFL][GTK][Blackberry] Update validation message related to number range in number-validation-message.html

• Status: RESOLVED
• Resolution: LATER
• Component: WebKit EFL

[Sudarsana Nagineni (babu)]

fast/forms/number/number-validation-message.html started failing after r145025.

--- /home/buildslave-1/webkit-buildslave/efl-linux-64-release-wk2/build/layout-test-results/fast/forms/number/number-validation-message-expected.txt
+++ /home/buildslave-1/webkit-buildslave/efl-linux-64-release-wk2/build/layout-test-results/fast/forms/number/number-validation-message-actual.txt
@@ -10,9 +10,9 @@
 Type mismatch
 PASS testIt("foo", "", "") is ""
 Range overflow
-PASS testIt("200", "", "100") is "range overflow"
+FAIL testIt("200", "", "100") should be range overflow. Was Value must be less than or equal to 100.
 Range underflow
-PASS testIt("10", "50", "100") is "range underflow"
+FAIL testIt("10", "50", "100") should be range underflow. Was Value must be greater than or equal to 50.
 Step mismatch
 PASS testIt("55", "0", "100", "10") is "step mismatch"
 PASS successfullyParsed is true

[Santosh Mahto]

Hi 
I have  found the root cause for this issue.
I will be updating patch very soon.

[Gyuyoung Kim]

Created attachment 201181 [details]
Patch

[Gyuyoung Kim]

In number-validation-message.html, range validation message is different between 'EFL, blackberry' and GTK. So, EFL and blackberry ports have handled this test as failing or wrong expectation file. If GTK port changes the message, EFL, GTK and Blackberry port can pass this test successfully.

[Gyuyoung Kim]

CC'ing Martin and Rob, could you guys take a look this patch ?

[kov's GTK+ EWS bot]

Comment on attachment 201181 [details]
Patch

Attachment 201181 [details] did not pass gtk-ews (gtk):
Output: http://webkit-queues.appspot.com/results/382154

[Gyuyoung Kim]

Created attachment 201182 [details]
Patch

[Chris Dumez]

Comment on attachment 201182 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=201182&action=review

> LayoutTests/platform/blackberry/fast/forms/number/number-validation-message-expected.txt:-9
> -FAIL testIt("", "", "") should be value missing. Was Input value missing.

The blackberry expectation should probably be updated, not removed. As far as I can tell, it still has 2 checks failing after your patch, right?

> LayoutTests/platform/blackberry/fast/forms/number/number-validation-message-expected.txt:-17
> -FAIL testIt("55", "0", "100", "10") should be step mismatch. Was Step mismatch.

and here.

[Build Bot]

Comment on attachment 201182 [details]
Patch

Attachment 201182 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/264032

New failing tests:
fast/forms/number/number-validation-message.html

[Build Bot]

Created attachment 201185 [details]
Archive of layout-test-results from webkit-ews-01 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-01  Port: mac-mountainlion  Platform: Mac OS X 10.8.2

[Gyuyoung Kim]

Created attachment 201191 [details]
Patch

[Gyuyoung Kim]

(In reply to comment #7)
> (From update of attachment 201182 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=201182&action=review
> 
> > LayoutTests/platform/blackberry/fast/forms/number/number-validation-message-expected.txt:-9
> > -FAIL testIt("", "", "") should be value missing. Was Input value missing.
> 
> The blackberry expectation should probably be updated, not removed. As far as I can tell, it still has 2 checks failing after your patch, right?
> 

Yes, right. My bad. Fixed.

[Rob Buis]

(In reply to comment #4)
> CC'ing Martin and Rob, could you guys take a look this patch ?

This makes sense from a BB point of view.

[Benjamin Poulain]

Comment on attachment 201191 [details]
Patch

No! This is not how it works.

All ports must align to LocalizedStrings. GTK/QT/EFL decided to have their own implementation, then it is their job to track the main file. You should never modify localization expectations based on anything but main.

[Gyuyoung Kim]

(In reply to comment #13)
> (From update of attachment 201191 [details])
> No! This is not how it works.
> 
> All ports must align to LocalizedStrings. GTK/QT/EFL decided to have their own implementation, then it is their job to track the main file. You should never modify localization expectations based on anything but main.

Ah, I missed it. Thank you for pointing it out. BTW, as you can see, EFL and BB ports have different string against main string. Is there any method to test it ? 

In BB port case, expectation file includes "FAIL" message as below,

FAIL testIt("200", "", "100") should be range overflow. Was Value must be less than or equal to 100.

Do you think this is fine ?

[Gyuyoung Kim]

(In reply to comment #14)
 
> Do you think this is fine ?

Or, can we change main string with this ? I think this string shows more detailed information.

[Gyuyoung Kim]

Ok, there is an issue to change main string because Mac port checks if string is valid via platform specific APIs. I think it is a little difficult to change the platform specific APIs at the moment. When there is a solution, I will file a bug again. Close resolved later.