support for mocking HTTP::header statements #13
Conversation
mocks for HTTP::headers evaluation if available instead of stub execution
separated reading of expectation from supporting unknown statements
|
Good morning, Your code looked promising, but it turns out most of the tests and examples fail. Could you have a look? |
|
I found there I introduced unneeded dependency on HTTP::header, I was testing several iRules, but always including |
fixed return code propagation from expected to unknown proc
removed dependency on HTTP::header in reset_expectations proc
|
I just commited fix. Now all tests are passed without any problem. I hope now it is finally stable ;-) |
|
I'm afraid not. Try running ./tests.sh and ./examples.sh from master. You'll notice all tests pass. When I run same scripts in your branch, I get lots of errors. |
…ith Tcl 8.4, expansion substitution was introduced in Tcl 8.5)
|
I was testing last change on ActiveTcl 8.6 on Windows, in jTcl there is no support for expansion substitution, so I have to use eval to pass args array to another proc. Last commit passed all tests on jTcl and on ActiveTcl. I have left more elegant expansion substitution implementation in comments to use it when jTcl will be compatible with Tcl 8.5 |
|
Good morning, Thanks for the acceptance of pull request and sorry for problems with code quality... I was thinking the differences between latest versions of jTcl and ActiveTcl are not so important in those small changes I made... Cheers, |
|
Actually, that's the only language level change. The stubbing functionality you contributed actually provides a nice foundation for further stubs like HTTP::uri etc. great stuff. Stefan Den 7. aug. 2013 kl. 08:48 skrev Sebastian Brzuzek [email protected]:
|
|
Thanks, I wasn't programming in Tcl for more than 10 years, so I'm not aware of this level change. Seba |
|
Great! Please also add a test case so this will be exposed in the test suite. Sendt fra min iPhone Den 7. aug. 2013 kl. 10:35 skrev Sebastian Brzuzek [email protected]:
|
|
About test case, it is already validated in existing tests, if it will work for 2 or 3 args it will work for any number of them, so I don't think we need to add something more to test how many args to proc will be accepted by JTcl |
No description provided.