I hope I have the correct forum and this question is not too trivial for the senior people of this forum.
What I am hoping is someone could point me in the right direction.
My dilemma is I am unable to generate
Java Beans from a transformed XML file or find the error within the new XML file. Let me explain. I've taken in an XML file and transformed it into another XML file, which closely resembles a business domain model, yet it has only a portion of the objects from the domain. I have validated this new XML document ensuring it is well formed using the XML Scheme for the entire domain. Using the same XML Scheme, I've generated the domain classes needed using the e-tool in Websphere Studio Application Developer, version 5.1 (XMLParserAPI.jar version 4.0.13, XercesImpl.jar version 4.0.13, and Xsdbeans.har, version 5.1). However, I get different behaviors depending on how I handle the new XML file. If I pass the
String directly to the Factory, it fails to generate the beans as required. However, when I save the "exact" same output from the XSLT transformation into a file and read it into the program, no line breaks added, I can generate the object model as designed. When I do a diff on the two XML documents, there is no differences!!!
This problematic behavior has me stumped. It appears the new XML file has a non-printable character within the stream somewhere and it is removed from the stored file when read back into the program. Herein is my problem. How would one go about capturing this non-printable value? Is there a way to peek or poke to see how these e-tools are pushing the information onto a stack? Do you have a better suggestion on how I can debug this issue?
Thank you for taking the time to read this long input! Any suggestions are better than writing my own parser!!!