When using IDEs or tools like xdoclet to generate HBM files, it is easy to get XML like this:
However the SchemaValidator barfs on this because the HbmBinder does not trim the sequence name.
The fix is similar to the fix for I believe, i.e. in HbmBinder.java in method makeIdentifier(...), use getTextTrim() instead of getText().
In 3.1 branch, this is at line 1902. On trunk this is line 1990.
Oracle 10g, JDK 1.5.0_06, linux, hibernate 3.1.3
I just commented on the 3.1.4 release planned question and the wrong assumption about 3.2 and annotations. Not about the quality of the fix.
I was wondering if this issue fell through the gaps? 3.2.1GA still has this issue. The patch is a trivial one-liner, I can't see any reason not to apply it.
Fixed in 3.2 branch / trunk.
I added 3.3 as another fix version, since it was also committed to trunk.
Bulk closing stale resolved issues