Hello Jason, and Kim,
That sounds, that looks, and that tastes like a bug.
Could you please raise a JIRA issue for that?
We've got some missing nasty test case ;-)
--
Guillaume Laforge
http://glaforge.free.fr/weblog
Jason Braucht wrote:
I think I'm seeing something similar using Groovy Version: 1.0-beta-7
JVM: 1.4.2_05-b04
'cmd /c dir /b /s \\myProjectName\\trunk\\*.java'.execute()
behaves differently from:
"cmd /c dir /b /s \\myProjectName\\trunk\\*.java"'.execute()
Based on my limited understanding of groovy, I thought those two
statements would work the same, but they don't. In the process of
trying to figure out what I was doing wrong I compiled both versions
and ran the .class through Jad. I noticed an extra slash being
inserted before the "trunk" in second example. The output from Jad is
as follows:
InvokerHelper.invokeNoArgumentsMethod("cmd /c dir /b /s
\\myProjectName\\trunk\\*.java", "execute");
and
InvokerHelper.invokeNoArgumentsMethod("cmd /c dir /b /s
\\myProjectName\\\trunk\\*.java", "execute"); // notice the extra
slash before "trunk"
When I looked at the .class with a hex editor, I noticed there was a
tab character in place of the 't' in trunk in the second example, so I
think Jad is decompiling properly.
Interestingly, when I tried:
"cmd /c dir /b /s \\myProjectName\\\trunk\\*.java"'.execute()
the output is the same in the second example above:
InvokerHelper.invokeNoArgumentsMethod("cmd /c dir /b /s
\\myProjectName\\\trunk\\*.java", "execute");
Is this behavior expected?
thanks,
Jason
| 