Well I am writing tests for my html5 parser. So the script I have is generating
Smalltalk unit test code of all possible combinations a set of html elements. I
am using basic data mining to change the expected output.
To make it clear. The tests I have the elements html head body and p opening
and closing tags. Any output with no p in the input would generate this tree:
Html
Head
Body
I added a little bit of "data mining" to check for any p elements in the input
and if that is true the output would be
Html
Head
Body
P
The reason I have many failing tests is because many tests have more than 1 p
element and I want to file those out and change the expected output to have 2 p
elements ...
Keep in mind I am still a student and have no experience in development so this
might be the worst possible way to do things. I am also using Java code to
generate the Smalltalk code in which I then file in.
I hope i did not make it so complicated by explaining...
-----Original Message-----
From: Pharo-users [mailto:pharo-users-boun...@lists.pharo.org] On Behalf Of
Stephan Eggermont
Sent: Monday, August 26, 2013 4:11 PM
To: pharo-users@lists.pharo.org
Subject: Re: [Pharo-users] Categorizing tests according to test result
Hi Mohammad,
You can take a look at TestResult>>historyFor:
Why are you generating unit tests as code with a script? And why do you have
many failing tests?
Stephan
