[Bug gjdoc/22877] NullPointerException in Parser.classClosed with -encoding UTF-8

[gcc-bugzilla at gcc dot gnu dot org]

A gjdoc from today's CVS gets a NullPointerException while processing
the attached file:

$ LANG=C LC_ALL=C gjdoc.kaffe -encoding UTF-8 Bug.java
Loading source file Bug.java ...
java.lang.NullPointerException
   at gnu.classpath.tools.gjdoc.Parser.classClosed (Parser.java:892)
   at gnu.classpath.tools.gjdoc.BracketClose.process (Parser.java:90)
   at gnu.classpath.tools.gjdoc.Parser.parse (Parser.java:783)
   at gnu.classpath.tools.gjdoc.ClassComponent.process (Parser.java:568)
   at gnu.classpath.tools.gjdoc.Parser.parse (Parser.java:783)
   at gnu.classpath.tools.gjdoc.Parser.processSourceFile (Parser.java:743)
   at gnu.classpath.tools.gjdoc.RootDocImpl.build (RootDocImpl.java:244)
   at gnu.classpath.tools.gjdoc.Main.startDoclet (Main.java:564)
   at gnu.classpath.tools.gjdoc.Main.start (Main.java:931)
   at gnu.classpath.tools.gjdoc.Main.main (Main.java:758)

This version of gjdoc was compiled with jikes and run with kaffe.
The same occurs with a version compiled to native code with gcj.

The length of the initial comments in the file is critical.


------- Comment #1 from from-classpath at savannah dot gnu dot org  2005-02-15 
13:50 -------
I can reproduce this, it seems to be a bug in the version of Classpath's UTF-8
CharsetDecoder we're using. There have been a number of fixes to
charset-related code lately, so with a bit of luck it will work with a fresh
gcc checkout. Going to check that now.


------- Comment #2 from from-classpath at savannah dot gnu dot org  2005-02-15 
15:58 -------
Confirmed to be a bug in Classpath's UTF-8 CharsetDecoder implementation. This
bug is fixed by attached patch which is for gcc CVS, but should work for
Classpath and Kaffe too.

I'm going to send a patch similar to this one to cp-patches so that this is
fixed in future runtime versions.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22877