On 10/19/06, Pete Robbins <[EMAIL PROTECTED]> wrote:
OK... I've tested this and it works a treat. Is this an issue that you'd like to see fixed in M2 release? We are about to re-spin so it can be included. Cheers. On 19/10/06, Pete Robbins <[EMAIL PROTECTED]> wrote: > > forget that... I fixed up your patch > > On 19/10/06, Pete Robbins <[EMAIL PROTECTED]> wrote: > > > > Simon, what you have done there looks ok to me. However... the patch has > > hard-coded paths to yor c:\... eclipe directories in. Any chance you can > > create a friendly patch for me to apply and test? > > > > Cheers, > > > > > > On 19/10/06, Simon Laws <[EMAIL PROTECTED] > wrote: > > > > > > I have created a patch for this problem ( > > > http://issues.apache.org/jira/browse/TUSCANY-873 ) which solves the > > > immediate > > > problem I have in PHP SDO and passes the PHP SDO tests that exist at > > > the > > > moment. However I'm not too comfortable with it because it is used by > > > numerous parts of data object API function which looks complicated > > > enough > > > already. > > > > > > What I have done is create a internal version of the > > > getPropertyIndex() > > > method called getPropertyIndexInternal() which creates a new property > > > in the > > > data object if the property doesn't already exist and the data object > > > type > > > is open. I use this new method whenever is seems that getPropertyIndex > > > was > > > used during the setting of properties. Can an C++ SDO expert take a > > > look at > > > this and see if this looks OK. Happy to discuss alternative > > > approaches. > > > > > > Regards > > > > > > Simon > > > > > > > > > > > > -- > > Pete > > > > > -- > Pete -- Pete Sorry about the paths in the patch. I can't see how to get rid of them
from the Eclipse SVN patch builder. I'll look in a little more detail. I think we should leave it out of M2. I can pick up a revision from SVN with it in as required. Thanks Simon
