From: Robert Klemme <shortcutter@...>
Date: 2008-01-09T02:35:07+09:00
Subject: Re: Using benchmark for wait times

On 08.01.2008 18:02, Max Russell wrote:
> I’m trying to use Benchmark, to return the wait time for a search in an
> application, and thought I could fairly neatly wrap up the wait time
> like this:
> 
> $logger.log(Benchmark.measure($ie.wait()))
> 
> However, when I run this test from my harness, I get the following in
> the console:
> 
> LocalJumpError: no block given
>     C:/ruby/lib/ruby/1.8/benchmark.rb:293:in `measure'
>     ./Tests/test_audittrail.rb:119
>     ./Tests/test_audittrail.rb:61:in `each'
>     ./Tests/test_audittrail.rb:61
>     C:/workspace/CV_Auto/harness.rb:41:in `load'
>     C:/workspace/CV_Auto/harness.rb:41:in `test_audittrail'
> 
> 
> What confuses me further is that line 61 reads:
> message_type.each do |link|
>     $logger.log("")
> 
> which when I diff it against my last version, I can see has not changed.
> 
> What I’d like to know is:
> 
> Am I off the mark wrapping the wait in a measure and then logging it?
> 
> Why does my each statement now fail?

You are not "wrapping the wait in a measure".  Instead you invoke #wait 
and stuff the result into #measure.  Consider this:

robert@fussel ~
$ ruby -r benchmark -e 'puts Benchmark.measure( sleep(1) )'
/usr/lib/ruby/1.8/benchmark.rb:293:in `measure': no block given 
(LocalJumpError)
         from -e:1

robert@fussel ~
$ ruby -r benchmark -e 'puts Benchmark.measure{ sleep(1) }'
   0.000000   0.000000   0.000000 (  1.000000)

Cheers

	robert