[cp-patches] Re: Bug in getEncoding testcase

[Robert Schuster]

Hi Torben,
sorry for answering so late. I read your first request but had no stable
internet access at that time.

Regarding your request: I think you're right. The test is more sane if the
optional charset's name mapping is only tested if the charset is available.

I fixed the doc as well.

Thanks for reporting!

Committed as:

2005-09-27  Robert Schuster <address@hidden>
        Suggested by: address@hidden
        
        * gnu/testlet/java/io/InputStreamReader/getEncoding.java: Check names
        of extended (= optional) charset only if it is available.

cu
Robert

address@hidden wrote:
> A few weeks ago I posted this patch to mauve-patches. Since it is still
> unchanged in cvs, I would like to ask if I posted to the wrong list, or
> if there is a problem with the patch. The problem with the existing
> testcase
> is that contradictory to the test description in the top of the file, it
> requires the extended charsets to be present, causing numerous false error
> reports.
> 
> Index: gnu/testlet/java/io/InputStreamReader/getEncoding.java
> ===================================================================
> RCS file:
> /cvs/mauve/mauve/gnu/testlet/java/io/InputStreamReader/getEncoding.java,v
> retrieving revision 1.1
> diff -u -F^f -r1.1 getEncoding.java
> --- gnu/testlet/java/io/InputStreamReader/getEncoding.java      17 Jun 2005
> 01:49:48 -0000      1.1
> +++ gnu/testlet/java/io/InputStreamReader/getEncoding.java      7 Sep 2005
> 14:06:06 -0000
> @@ -165,8 +165,8 @@
>                         } catch (UnsupportedCharsetException uce) {
>                                 supported = false;
>                         }
> -                       harness.check(name, extIoNames[i]);
> -                       harness.check(supported, true);
> +                       if ( supported )
> +                               harness.check(name, extIoNames[i]);
>                 }
>         }
> 
> 
>