Incorrect extraction of class names from Atom feeds


When Atom parser enumerates Entry elements, it extracts entity type from the "term" attribute of a "category" element (see AtomParser.m, lines 305 - 306), and assumes that there is a class with the same name (see ObjectContext.m, line 1133). If the "term" attribute contain dots, then the name of the class is extracted incorrectly. For example, when a parser encounters the following element:
<category term="Aaa.Bbb.Ccc.Ddd.Xxx.UserTracker" scheme="http://schemas.microsoft.com/ado/2007/08/dataservices/scheme" />
it extracts the class name as "Bbb.Ccc.Ddd.Xxx.UserTracker", while only "UserTracker" is expected.


abub wrote Dec 17, 2010 at 9:58 PM

Thanks for reporting this issue. We shall consider it as part of the current release aimed at January.

CTapang wrote Mar 12, 2011 at 12:35 AM

This is fixed in Version 1.2. The relevant lines are now 417 through 435 of AtomParser.m. We are still extracting the last component of the fully qualified name from the server, but are now prepending it with the service namespace.

wrote Feb 14, 2013 at 7:26 PM