RE: XSLT line numbers, <code>, etc.

I've made the hack you suggest but there is still a some bytes difference between the size of the document sent by Tomcat and the expected size in the moki document.

I've changed manually the size on the moki document so both sizes have identical value but I get a NullPointerException at org.w3c.mwi.mobileok.basic.OneTestTest.handleKnownHeaderDifference(OneTestTest.java:215). 

I'll look at this error more thoroughly later this evening.


>>-----Mensaje original-----
>>De: public-mobileok-checker-request@w3.org [mailto:public-mobileok-
>>checker-request@w3.org] En nombre de Sean Owen
>>Enviado el: miércoles, 10 de octubre de 2007 18:34
>>Para: Abel Rionda
>>CC: public-mobileok-checker
>>Asunto: Re: XSLT line numbers, <code>, etc.
>>
>>
>>Oh OK I thought I saw some recent changes. It looks like an issue with
>>the code from what I can tell. Well, whoever gets to it first...
>>
>>You get this error after an update? If so, sounds like for some reason
>>the body of the response you are getting is different, which is weird.
>>If this is the case, and you are running on Windows, then I somehow
>>suspect a line separator problem -- maybe the XML serializer is
>>putting in \r\n at the end of lines, only on Windows, and that is
>>making the response bigger.
>>
>>if that is so, then hacking in a change to the line separator might
>>fix it. For example at the very top of Tester, you might include this:
>>
>>static {
>>  System.setProperty("line.separator", "\n");
>>}
>>
>>That is not a long-term fix and shouldn't be checked in but if that
>>fixes the test failure then it confirms the problem.
>>
>>I myself work on OS X so it would not affect me if so.
>>
>>On 10/10/07, Abel Rionda <abel.rionda@fundacionctic.org> wrote:
>>>
>>> Hi Sean,
>>>
>>> >Good catch, I just fixed this and the associated tests.
>>> >There are still some failing tests due to some CSS changes I think?
>>>
>>> We have not made any changes since two weeks ago. On the other hand,
>>> when we launch the test suite, we get the same error in all tests:
>>>
>>> Caused by: org.junit.ComparisonFailure: Line 37
>>> expected:<Content-Length: 4[74]&#xD;> but was:<Content-Length:
>>> 4[97]&#xD;>
>>>
>>> It seems that the Content-Length response header does not match the
>>> specified in the moki test case.
>>> Both Miguel and I have run the test on Windows XP. Can this be a
>>> reason? Anyone else is using Windows?
>>>
>>> Abel.
>>>

Received on Thursday, 11 October 2007 11:54:38 UTC