Hi Mikhail,
Thanks very much for your correction, I'm awfully sorry for the typo, it
would absolutely lead to potential error, I corrected it on my machine
and run the test again. Fortunately the rest part of the test seems OK.
As for the test on the creation of storage from readonly stream, do you
think we add that part to the test now? If you think it should be
integrated in this CWS I will start working on it.
Best Regards,
Felix.
Mikhail Voitenko 写道:
Hi Felix,
The test looks good. There seems to be a small typo, the random
generator fills the wrong array.
If the storage is constructed with XInputStream the storage can be
changed but the commit() call might fail. From other side the stream
object is provided without type information. So in case the object
implements XStream, the storage will detect it and will use the
XStream interface, in this case the commit might work well.
In other words the storage can be always changed. During the commit it
will try to write to the provided stream, and if it is not possible it
will throw an exception.
By the way, the test uses input stream from the TempFile object. It is
sent to the storage as object, so that the type information is lost.
Since the TempFile object implements XStream interface, the storage
will find this interface on the provided object and will use it. So to
really test creation of the storage from readonly stream it makes
sense to use SimpleFileAccess and to open the readonly stream using
the URL of the TempFile.
Best Regards,
Mikhail.
Zhang Xiaofei wrote:
Hi Mikhail,
Thanks very much for your advice, I kind of confused the data source
and data sink, now I can get the right result from the test, please
review it and give me any suggestions on it. :-)
By the way, I still do not quite understand the service. Is a
storage, constructed with a XInputStream, any different from one
constructed with a XStream? Is the first one read only? If that is
the case, maybe it's better I modify the in-idl documentations a
little bit to make that clear.
Best Regards,
Felix.
Mikhail Voitenko 写道:
Hi Felix,
I have recognized that readBytes() and writeBytes() seem to be used
wrongly.
In general readBytes() reads from the stream to the provided array,
and writeBytes() writes to the stream from the provided array.The
streams are empty because they were not changed after opening,
readBytes() is used at the beginning instead of writeBytes(). So to
work as expected the test should first write the data to the stream,
and read it at the end of the test.
Best Regards,
Mikhail.
Zhang Xiaofei wrote:
Hi Mikhail,
Thanks for your help in our last IRC meeting, the exception is
eliminated, now I got another
com.sun.star.io.NotConnectedException in line 96, Test.java, when
the statement below is executed:
xTempStream[i].getOutputStream ().writeBytes (pBytesOut[i]);
I debugged the test and found the sub-streams had always been
zero-lengthed in the entire test, I could not find out why I got
this result, could you please give me some help on this?
Thanks and Best Regards,
Felix.
Zhang Xiaofei 写道:
Hi Mikhail,
The attachment is the unit test for OLESimpleStorage, right now I
have a problem running it: a java.lang.ClassCastException in
Test01.java, line 89, could you please help me on this? Thanks!
Best Regards,
Felix.
Mikhail Voitenko 写道:
Hi Felix,
Sorry, something seems to go wrong with the mail system. I did
not get this mail although I was in CC. Good that it was also
sent to dev@framework.openoffice.org
There seems to be no attachments to review, but we have discussed
the problems already. So since you have adjusted the tests and
they are executed as expected, it is ok to attach them to the
related issue I think. I will integrate them along with the
changes for TempFile service after the vacation.
The only documentation for the OleSimpleStorage service is the
idl-specification written by you :)
If you would like to see an example of service usage, please take
a look to the embeddedobj project, the service is used in the
following files: embeddedobj/source/msole/olepersist.cxx
embeddedobj/source/msole/ownview.cxx
Best Regards,
Mikhail.
Zhang Xiaofei wrote:
Hi Mikhail,
Sorry for the delay, I was tracking a bug all day long yesterday
and did not sent out the TempFile tests. I guess this version is
functionally OK now. Please check the attachment to review them.
And now that I am to write the test for OleSimpleStorage, could
you recommend to me some references on it please? Any example
tests or documents on the service/interface could be helpful to me.
Again, thanks very much for your help during the task. :-)
Best Regards,
Felix.
Mikhail Voitenko 写道:
Hi Felix,
Sorry for the delay with the answer.
I think that it will be enough to introduce only one test for
OleSimpleStorage based on the scenario that I have already
mentioned. Further tests will need additional .doc test files
integrated in the test, that is relative complex. So if you
want to introduce them, let do it after my vacation, so that I
can answer the questions.
Please prepare the test for OleSimpleStorage so we can
integrate it together with the service reimplementation. The
test suite can be extended in future.
By the way, how is it going with the TempFile tests? You was
going to send me the final version. If there are any questions
please do not hesitate to ask, so that I can answer before the
Friday ( my vacation starts on Friday ).
Thanks,
Mikhail.
------------------------------------------------------------------------
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
------------------------------------------------------------------------
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
