Re: svn commit: rev 4821 - trunk

Philip Martin <philip@codematters.co.uk> writes:

> dlr@tigris.org writes:
>
> > Author: dlr
> > Date: 2003-02-10 13:22:30 -0600 (Mon, 10 Feb 2003)
> > New Revision: 4821
> >
> > Modified:
> > trunk/Makefile.in
> > Log:
> > * Makefile.in
> > (swig-java-ext-clean): Worked around a problem in the interaction of
> > this Makefile with versioned directories in the working copy. If
> >
> > Modified: trunk/Makefile.in
> > ==============================================================================
> > --- trunk/Makefile.in (original)
> > +++ trunk/Makefile.in Mon Feb 10 13:22:33 2003
> > @@ -374,7 +374,8 @@
> > done
> >
> > swig-java-ext-clean:
> > - rm -rf $(SWIG_SRC_DIR)/java
> > + rm -rf $(SWIG_SRC_DIR)/java/build/
> > + rm -rf $(SWIG_SRC_DIR)/java/org/
>
> Deleting from SWIG_SRC_DIR doesn't look right. Why do the Java
> bindings build there? Why don't they build in SWIG_BUILD_DIR?

That's where they've always built to, but not for any particular
reason. Is this more along the lines of how it should be (untested)?:

Index: Makefile.in
===================================================================
--- Makefile.in (revision 4821)
+++ Makefile.in (working copy)
@@ -370,12 +370,12 @@
                 -package 'org.tigris.subversion.swig' \
                 -I$(SWIG_SRC_DIR) \
                 -I$(abs_srcdir)/subversion/include $(SVN_APR_INCLUDES) \
- -o $(SWIG_SRC_DIR)/java/build/$$target $$source; \
+ -o $(SWIG_BUILD_DIR)/java/build/$$target $$source; \
         done
 
 swig-java-ext-clean:
- rm -rf $(SWIG_SRC_DIR)/java/build/
- rm -rf $(SWIG_SRC_DIR)/java/org/
+ rm -rf $(SWIG_BUILD_DIR)/java/build/
+ rm -rf $(SWIG_BUILD_DIR)/java/org/
 
 java:
         @$(MAKE) -C $(abs_srcdir)/subversion/bindings/java "JAVAC=$(JAVAC)"

-- 
Daniel Rall <dlr@finemaltcoding.com>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org