Re: [cp-patches] META-INF support
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, AFAIK we support not only delivery as glibj.zip but as glibj.jar and as a plain directory. The patch looks as if it only works for the zip-case. Are there any reasons to make this different for the other two variants? cu Robert Andrew John Hughes wrote: > At present, the XML libraries are not aware of the parsers provided > by the inclusion of GNU JAXP in GNU Classpath. Indeed, further inspection > by myself showed that there is no META-INF directory added to glibj.zip > at all, let alone a services subdirectory. The attached patch adds > this. Please give your comments on this; it is potentially useful > for other things as well such as the imageIO framework. > > It also fixes a minor problem with ecj support. > > Changelog > > 2005-04-10 Andrew John Hughes <[EMAIL PROTECTED]> > > * examples/Makefile.am: > Fix command-line options for ecj. > * lib/Makefile.am: > Added a META-INF directory structure to glibj.zip. > > > > > > Index: examples/Makefile.am > === > RCS file: /cvsroot/classpath/classpath/examples/Makefile.am,v > retrieving revision 1.3 > diff -u -3 -p -u -r1.3 Makefile.am > --- examples/Makefile.am 7 Feb 2005 02:32:37 - 1.3 > +++ examples/Makefile.am 9 Apr 2005 23:48:38 - > @@ -11,7 +11,7 @@ if FOUND_GCJX > JCOMPILER = $(GCJX) -bootclasspath '' -sourcepath '' -classpath > $(top_builddir)/lib:. > else > if FOUND_ECJ > -JCOMPILER = $(ECJ) --bootclasspath '$(top_builddir)/lib' --classpath . > +JCOMPILER = $(ECJ) -bootclasspath '$(top_builddir)/lib' -classpath . > else > error dunno how to setup the JCOMPILER and compile > endif > Index: lib/Makefile.am > === > RCS file: /cvsroot/classpath/classpath/lib/Makefile.am,v > retrieving revision 1.79 > diff -u -3 -p -u -r1.79 Makefile.am > --- lib/Makefile.am 7 Feb 2005 02:32:37 - 1.79 > +++ lib/Makefile.am 9 Apr 2005 23:48:40 - > @@ -66,7 +66,7 @@ endif # INSTALL_CLASS_FILES > .PHONY: genclasses > > glibj.zip: classes compile-classes resources > - if test "$(ZIP)" != ""; then $(ZIP) -r -D glibj.zip gnu java javax org > > /dev/null; fi > + if test "$(ZIP)" != ""; then $(ZIP) -r -D glibj.zip gnu java javax org > META-INF > /dev/null; fi > > resources: > if ! [ -e gnu ]; then mkdir gnu; fi > @@ -78,6 +78,8 @@ resources: > @list='$(propertyfiles)'; for p in $$list; do \ > cp $(top_srcdir)/resource/$$p $$p; \ > done > + if [ -e META-INF ]; then rm -rf META-INF; fi > + cp -r $(top_srcdir)/resource/META-INF . > touch resources > > classes: genclasses > > > > > ___ > Classpath-patches mailing list > Classpath-patches@gnu.org > http://lists.gnu.org/mailman/listinfo/classpath-patches -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFCWISPG9cfwmwwEtoRAlJPAJ9TlULI+MdDsXn2vaun5a0w8bqsYwCeOGxp bsJt5j2OwZegpdWrPkKUkP4= =EVPo -END PGP SIGNATURE- ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] META-INF support
At present, the XML libraries are not aware of the parsers provided by the inclusion of GNU JAXP in GNU Classpath. Indeed, further inspection by myself showed that there is no META-INF directory added to glibj.zip at all, let alone a services subdirectory. The attached patch adds this. Please give your comments on this; it is potentially useful for other things as well such as the imageIO framework. It also fixes a minor problem with ecj support. Changelog 2005-04-10 Andrew John Hughes <[EMAIL PROTECTED]> * examples/Makefile.am: Fix command-line options for ecj. * lib/Makefile.am: Added a META-INF directory structure to glibj.zip. -- Andrew :-) Please avoid sending me Microsoft Office (e.g. Word, PowerPoint) attachments. See http://www.fsf.org/philosophy/no-word-attachments.html No software patents in Europe -- http://nosoftwarepatents.com "Value your freedom, or you will lose it, teaches history. `Don't bother us with politics' respond those who don't want to learn." -- Richard Stallman Escape the Java Trap with GNU Classpath! http://www.gnu.org/philosophy/java-trap.html public class gcj extends Freedom implements Java { ... } Index: examples/Makefile.am === RCS file: /cvsroot/classpath/classpath/examples/Makefile.am,v retrieving revision 1.3 diff -u -3 -p -u -r1.3 Makefile.am --- examples/Makefile.am7 Feb 2005 02:32:37 - 1.3 +++ examples/Makefile.am9 Apr 2005 23:48:38 - @@ -11,7 +11,7 @@ if FOUND_GCJX JCOMPILER = $(GCJX) -bootclasspath '' -sourcepath '' -classpath $(top_builddir)/lib:. else if FOUND_ECJ -JCOMPILER = $(ECJ) --bootclasspath '$(top_builddir)/lib' --classpath . +JCOMPILER = $(ECJ) -bootclasspath '$(top_builddir)/lib' -classpath . else error dunno how to setup the JCOMPILER and compile endif Index: lib/Makefile.am === RCS file: /cvsroot/classpath/classpath/lib/Makefile.am,v retrieving revision 1.79 diff -u -3 -p -u -r1.79 Makefile.am --- lib/Makefile.am 7 Feb 2005 02:32:37 - 1.79 +++ lib/Makefile.am 9 Apr 2005 23:48:40 - @@ -66,7 +66,7 @@ endif # INSTALL_CLASS_FILES .PHONY: genclasses glibj.zip: classes compile-classes resources - if test "$(ZIP)" != ""; then $(ZIP) -r -D glibj.zip gnu java javax org > /dev/null; fi + if test "$(ZIP)" != ""; then $(ZIP) -r -D glibj.zip gnu java javax org META-INF > /dev/null; fi resources: if ! [ -e gnu ]; then mkdir gnu; fi @@ -78,6 +78,8 @@ resources: @list='$(propertyfiles)'; for p in $$list; do \ cp $(top_srcdir)/resource/$$p $$p; \ done + if [ -e META-INF ]; then rm -rf META-INF; fi + cp -r $(top_srcdir)/resource/META-INF . touch resources classes: genclasses signature.asc Description: Digital signature ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
[cp-patches] Re: -ko on libjava/external/sax/org/xml/sax/helpers/NamespaceSupport.java
On Sat, 2005-04-09 at 13:07 -0700, Geoffrey Keating wrote: > I did > > cvs admin -ko > libjava/external/sax/org/xml/sax/helpers/NamespaceSupport.java > > because it contains an $Id$ tag. (I would have just removed the tag, > or used -kk, but I understand this comes from elsewhere and we want to > minimise local changes.) GNU Classpath contains the same series of files, and this has been suggested as an appropriate solution in the past (although I don't think anyone has actually gone through with it...) The full list of files that contain these keywords are as follows: * external/sax/org/xml/sax/AttributeList.java * external/sax/org/xml/sax/Attributes.java * external/sax/org/xml/sax/ContentHandler.java * external/sax/org/xml/sax/DTDHandler.java * external/sax/org/xml/sax/DocumentHandler.java * external/sax/org/xml/sax/EntityResolver.java * external/sax/org/xml/sax/ErrorHandler.java * external/sax/org/xml/sax/HandlerBase.java * external/sax/org/xml/sax/InputSource.java * external/sax/org/xml/sax/Locator.java * external/sax/org/xml/sax/Parser.java * external/sax/org/xml/sax/SAXException.java * external/sax/org/xml/sax/SAXNotRecognizedException.java * external/sax/org/xml/sax/SAXNotSupportedException.java * external/sax/org/xml/sax/SAXParseException.java * external/sax/org/xml/sax/XMLFilter.java * external/sax/org/xml/sax/XMLReader.java * external/sax/org/xml/sax/package.html * external/sax/org/xml/sax/ext/Attributes2.java * external/sax/org/xml/sax/ext/Attributes2Impl.java * external/sax/org/xml/sax/ext/DeclHandler.java * external/sax/org/xml/sax/ext/DefaultHandler2.java * external/sax/org/xml/sax/ext/EntityResolver2java * external/sax/org/xml/sax/ext/LexicalHandler.java * external/sax/org/xml/sax/ext/Locator2.java * external/sax/org/xml/sax/ext/package.html * external/sax/org/xml/sax/helpers/AttributeListImpl.java * external/sax/org/xml/sax/helpers/AttributesImpl.java * external/sax/org/xml/sax/helpers/DefaultHandler.java * external/sax/org/xml/sax/helpers/LocatorImpl.java * external/sax/org/xml/sax/helpers/NamespaceSupport.java * external/sax/org/xml/sax/helpers/NewInstance.java * external/sax/org/xml/sax/helpers/ParserAdapter.java * external/sax/org/xml/sax/helpers/ParserFactory.java * external/sax/org/xml/sax/helpers/XMLFilterImpl.java * external/sax/org/xml/sax/helpers/XMLReaderAdapterjava * external/sax/org/xml/sax/helpers/XMLReaderFactory.java * external/sax/org/xml/sax/helpers/package.html If others agree (and such action has not already taken place), I volunteer to do the same procedure on each of these for the GNU Classpath tree. -- Andrew :-) Please avoid sending me Microsoft Office (e.g. Word, PowerPoint) attachments. See http://www.fsf.org/philosophy/no-word-attachments.html No software patents in Europe -- http://nosoftwarepatents.com "Value your freedom, or you will lose it, teaches history. `Don't bother us with politics' respond those who don't want to learn." -- Richard Stallman Escape the Java Trap with GNU Classpath! http://www.gnu.org/philosophy/java-trap.html public class gcj extends Freedom implements Java { ... } signature.asc Description: This is a digitally signed message part ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
Re: [cp-patches] Reindented native/jni code
On Sat, Apr 09, 2005 at 02:19:03AM +0200, Mark Wielaard wrote: > Hi Guilhem. > > On Fri, 2005-04-08 at 14:34 +0200, Guilhem Lavaux wrote: > > GNU Indent (GNU style) seems to show a lot of differences after having > > been run on the native/jni code. I have run it on classpath, java-* > > directories and here is a patch. > > Thanks. But next time please give people a little while before actually > checking a large automatic change like this in. Someone might be working > on one of the files and then you would create an unnecessary merge > conflict. > > Now that this is in I get a build failure since the JNI check script > cannot find all native methods. JNI methods have to start at the > position zero of a line. And strangely enough it now also finds a couple > of native methods that don't have a corresponding java method > definition. > > The following patch seems to fix the build failure. And all code seem to > still work correctly even though I just removed a large collection of > code. The code was probably dead. And without it things seem to work > fine. But could someone (Michael?) take a quick look? I won't be online > till Sunday night so could someone check it in if it isn't too bogus. A > broken build is really not acceptable. > > 2005-04-08 Mark Wielaard <[EMAIL PROTECTED]> > > * native/jni/java-lang/java_lang_Double.c: Reindent. > * native/jni/java-lang/java_lang_Math.c: Likewise. > * native/jni/java-lang/java_lang_Object.c: Likewise. > * native/jni/java-lang/java_lang_VMDouble.c: Likewise. > * native/jni/java-lang/java_lang_VMFloat.c: Likewise. > * native/jni/java-lang/java_lang_VMSystem.c: Likewise. > * native/jni/java-lang/java_lang_reflect_Array.c: Likewise. > * native/jni/java-nio/java_nio.c > (Java_gnu_java_nio_FileChannelImpl_nio_1mmap_1file): Removed. > (Java_gnu_java_nio_FileChannelImpl_nio_1unmmap_1file): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketCreate): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketConnect): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketBind): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketListen): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketAvailable): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketClose): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketRead): Likewise. > (Java_gnu_java_nio_SocketChannelImpl_SocketWrite): Likewise. > * native/jni/java-nio/java_nio_VMDirectByteBuffer.c: Reindent. I tested and commited this now. I found no bug with it and it makes HEAD buildable again. Michael -- Escape the Java Trap with GNU Classpath! http://www.gnu.org/philosophy/java-trap.html Join the community at http://planet.classpath.org/ ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches
PING: Re: [cp-patches] RFC: Fix for "XXX" use in java.util.Currency
On Mon, 2005-03-28 at 01:41 +0100, Andrew John Hughes wrote: > The attached patch allows examples such as the following: > > NumberFormat.getCurrencyInstance(Locale.ENGLISH).getCurrency() > > to succeed with the result of a Currency instance containing "XXX" > rather than throwing a NullPointerException. This is Sun's behaviour, > which is made possible by allowing "XXX" to be supplied to > Currency.getInstance(String). This is something we have discussed > before, which has lead to the introduction of the String version > of this value in java.text.DecimalFormatSymbols (11 Jan - Fix for Bug #11545). > > Should we allow this horrible design decision to creep into > GNU Classpath? Comments please. > > Changelog: > > 2005-03-28 Andrew John Hughes <[EMAIL PROTECTED]> > > * java/text/DecimalFormatSymbols.java: > Added retrieval of "XXX" instance in place of null. > * java/util/Currency.java, > (Currency(String)): New constructor for the XXX special case. > (getInstance(String)): Allow special case of "XXX". > > ___ > Classpath-patches mailing list > Classpath-patches@gnu.org > http://lists.gnu.org/mailman/listinfo/classpath-patches Does anyone have any comments on this patch, posted March 28th? Should I commit it or not? Thanks, -- Andrew :-) Please avoid sending me Microsoft Office (e.g. Word, PowerPoint) attachments. See http://www.fsf.org/philosophy/no-word-attachments.html No software patents in Europe -- http://nosoftwarepatents.com "Value your freedom, or you will lose it, teaches history. `Don't bother us with politics' respond those who don't want to learn." -- Richard Stallman Escape the Java Trap with GNU Classpath! http://www.gnu.org/philosophy/java-trap.html public class gcj extends Freedom implements Java { ... } signature.asc Description: This is a digitally signed message part ___ Classpath-patches mailing list Classpath-patches@gnu.org http://lists.gnu.org/mailman/listinfo/classpath-patches