From rlpowell at digitalkingdom.org  Thu Oct  2 20:27:53 2008
From: rlpowell at digitalkingdom.org (Robin Lee Powell)
Date: Thu, 2 Oct 2008 13:27:53 -0700
Subject: [cl-selenium-devel] Better error handling, please.
Message-ID: <20081002202753.GH16572@digitalkingdom.org>


cl-selenium's response to me having forgotten to start
selenium-server is:

debugger invoked on a SIMPLE-ERROR in thread #<THREAD "initial thread" RUNNING {100266AE01}>:
  Error during processing of --eval option (LOAD #P"sert-spec.cl"):

  end of file on #<FLEXI-STREAMS:FLEXI-IO-STREAM {1004DE1851}>

Could that maybe be a little less opaque?  :)

-Robin

-- 
Lojban Reason #17: http://en.wikipedia.org/wiki/Buffalo_buffalo
Proud Supporter of the Singularity Institute - http://singinst.org/
http://www.digitalkingdom.org/~rlpowell/ *** http://www.lojban.org/


From rlpowell at digitalkingdom.org  Wed Oct 15 00:59:11 2008
From: rlpowell at digitalkingdom.org (Robin Lee Powell)
Date: Tue, 14 Oct 2008 17:59:11 -0700
Subject: [cl-selenium-devel] Better error handling, please.
In-Reply-To: <20081002202753.GH16572@digitalkingdom.org>
References: <20081002202753.GH16572@digitalkingdom.org>
Message-ID: <20081015005911.GK10637@digitalkingdom.org>

On Thu, Oct 02, 2008 at 01:27:53PM -0700, Robin Lee Powell wrote:
> 
> cl-selenium's response to me having forgotten to start
> selenium-server is:
> 
> debugger invoked on a SIMPLE-ERROR in thread #<THREAD "initial thread" RUNNING {100266AE01}>:
>   Error during processing of --eval option (LOAD #P"sert-spec.cl"):
> 
>   end of file on #<FLEXI-STREAMS:FLEXI-IO-STREAM {1004DE1851}>
> 
> Could that maybe be a little less opaque?  :)

Did it myself:

*** cl-selenium-0.4-orig/iedoc.lisp     2008-10-14 17:27:36.000000000 -0700
--- cl-selenium-0.4/iedoc.lisp  2008-10-14 17:58:25.000000000 -0700
***************
*** 150,155 ****
--- 150,156 ----
                       (description c)))))

  (defun execute (url parameters &optional return-type)
+   (handler-case
      (multiple-value-bind (reply status-code headers reply-from stream some-bool reason)
        (drakma:http-request url :method :get :parameters parameters)
        (declare (ignore headers reply-from stream some-bool))
***************
*** 160,166 ****
         (convert-result (subseq reply (min 3 (length reply))) return-type))
        ((starts-with reply "ERROR:")
         (let ((err (second (split-sequence:split-sequence #\: reply))))
!          (error 'execution-error :description err))))))

  (defun convert-function (iedoc-function)
    (let ((function-name (convert-function-name (iedoc-function-name iedoc-function)))
--- 161,169 ----
         (convert-result (subseq reply (min 3 (length reply))) return-type))
        ((starts-with reply "ERROR:")
         (let ((err (second (split-sequence:split-sequence #\: reply))))
!          (error 'execution-error :description err)))))
!     (error (cond-val)
!          (error "Selenium RC can't talk to the Selenium server; is it running?~%~%  Specific error: ~A~%~%" cond-val))))

  (defun convert-function (iedoc-function)
    (let ((function-name (convert-function-name (iedoc-function-name iedoc-function)))

-Robin

-- 
They say:  "The first AIs will be built by the military as weapons."
And I'm thinking:  "Does it even occur to you to try for something
other than the default outcome?" -- http://shorl.com/tydruhedufogre
http://www.digitalkingdom.org/~rlpowell/ *** http://www.lojban.org/



From rlpowell at digitalkingdom.org  Wed Oct 15 02:10:52 2008
From: rlpowell at digitalkingdom.org (Robin Lee Powell)
Date: Tue, 14 Oct 2008 19:10:52 -0700
Subject: [cl-selenium-devel] Optional execution errors.
Message-ID: <20081015021052.GA12759@digitalkingdom.org>


I find it vastly annoying for execution errors to, well, throw
errors, because then my test framework doesn't get to tell me
*where* the error occured.

Arguably a bug in my test framework, but I fixed it in Selenium
anyways, by making an overrideable error routine:

diff -C3 -w -r cl-selenium-0.4-orig/iedoc.lisp cl-selenium-0.4/iedoc.lisp
*** cl-selenium-0.4-orig/iedoc.lisp     2008-10-14 17:27:36.000000000 -0700
--- cl-selenium-0.4/iedoc.lisp  2008-10-14 19:02:47.000000000 -0700
***************
*** 160,166 ****
         (convert-result (subseq reply (min 3 (length reply))) return-type))
        ((starts-with reply "ERROR:")
         (let ((err (second (split-sequence:split-sequence #\: reply))))
!          (error 'execution-error :description err))))))

  (defun convert-function (iedoc-function)
    (let ((function-name (convert-function-name (iedoc-function-name iedoc-function)))
--- 160,167 ----
         (convert-result (subseq reply (min 3 (length reply))) return-type))
        ((starts-with reply "ERROR:")
         (let ((err (second (split-sequence:split-sequence #\: reply))))
!          (funcall *selenium-execution-error-func* :description err)
!        )))))

  (defun convert-function (iedoc-function)
    (let ((function-name (convert-function-name (iedoc-function-name iedoc-function)))
***************
*** 184,186 ****
--- 185,194 ----
    defaults to this special variable.  You can bind a selenium session
    to this variable and use those functions without providing an
    explicit session each time.")
+
+ (defvar *selenium-execution-error-func*
+   (lambda (desc err)
+     (error 'execution-error :description err))
+   "This is the function that is run to create Selenium execution errors.
+   Replace it to get different error handling.")
+

-Robin

-- 
They say:  "The first AIs will be built by the military as weapons."
And I'm thinking:  "Does it even occur to you to try for something
other than the default outcome?" -- http://shorl.com/tydruhedufogre
http://www.digitalkingdom.org/~rlpowell/ *** http://www.lojban.org/