Apache Axis 1.2 RC3 ChangeLog (Changes after 1.2 RC2)

Designed for use with Ant.

2005-02-12 21:58 dims 
Update to RC3
2005-02-12 18:04 dims 
Removing xsd:language, xsd:name from here as it breaks TCK tests.
2005-02-12 13:39 dims 
WSDL4J 1.5 + 1 late fix mentioned below

2005-02-12 07:09 wsdl4jadmin

Added base URI to InputSource so relative URIs get resolved when parsing the Document.
Made this change to readWSDL, parseImport, parseSchema. Also corrected related error
messages to use full URI where possible, not just the relative URI.
    * src/com/ibm/wsdl/xml/WSDLReaderImpl.java (1.38)
2005-02-12 04:41 gdaniels 

Enable beans with array fields but no indexed accessors,
and along the way clean up some potentially problem-
causing rough spots.  Hopefully I didn't add more of the
same. :)

* Provide an equivalent mechanism to the one Tom and I
  added to the ArraySerializer for the ArrayDeserializer,
  allowing us to specify a component type to the
  factory so that we can customize it at creation time.

  This gets used when figuring out types using
  TypeMappingImpl.getDeserializer() - if we're looking
  for a deserializer for a Java array type, and
  we're not in maxOccurs mode (in which case you just
  let the BeanDeserializer or RPCHandler do it for you),
  we return an ArrayDeserializerFactory preconfigured
  with the correct component type.  This allows us to
  deserialize array items without xsi:types.  There's
  a bit of tricky code in there which checks to see
  that the desired XML type isn't the component type
  of the desired array type - if so, that indicates
  we're using maxOccurs mode.

* Remove digging down into the component type from
  DeserializationContext.getDeserializerForClass()

* Since we use SOAP_ARRAY as a marker type for arrays
  even in literal mode, make sure we don't end up
  inadvertently serializing it on the wire as an
  xsi:type (in SerializationContext)

* Clean up what's going on when we're registering
  the "default" mappings in Service/WSDDService/
  Emitter.  Use a common function,
  TypeMappingRegistryImpl.doRegisterFromVersion(),
  in order to register the correct SOAPENC default
  (which is really what goes on when selecting the
  typemapping switch [insert flame about how awful
  the "1.X" arguments are for these things]). Avoid
  weird delegation chains by forcing the defaults
  in these situations.

* Comment out indexed setter/getter from ArrayOfBook

* For some reason we were forcing "isResponse" to
  false in BodyBuilder when deserializing in "fast"
  mode.  Respect the actual message type, which lets
  us get the right metadata in RPCHandler.

Passes all-tests, need to re-run TCK.
2005-02-12 04:41 gdaniels 

Enable beans with array fields but no indexed accessors,
and along the way clean up some potentially problem-
causing rough spots.  Hopefully I didn't add more of the
same. :)

* Provide an equivalent mechanism to the one Tom and I
  added to the ArraySerializer for the ArrayDeserializer,
  allowing us to specify a component type to the
  factory so that we can customize it at creation time.

  This gets used when figuring out types using
  TypeMappingImpl.getDeserializer() - if we're looking
  for a deserializer for a Java array type, and
  we're not in maxOccurs mode (in which case you just
  let the BeanDeserializer or RPCHandler do it for you),
  we return an ArrayDeserializerFactory preconfigured
  with the correct component type.  This allows us to
  deserialize array items without xsi:types.  There's
  a bit of tricky code in there which checks to see
  that the desired XML type isn't the component type
  of the desired array type - if so, that indicates
  we're using maxOccurs mode.

* Remove digging down into the component type from
  DeserializationContext.getDeserializerForClass()

* Since we use SOAP_ARRAY as a marker type for arrays
  even in literal mode, make sure we don't end up
  inadvertently serializing it on the wire as an
  xsi:type (in SerializationContext)

* Clean up what's going on when we're registering
  the "default" mappings in Service/WSDDService/
  Emitter.  Use a common function,
  TypeMappingRegistryImpl.doRegisterFromVersion(),
  in order to register the correct SOAPENC default
  (which is really what goes on when selecting the
  typemapping switch [insert flame about how awful
  the "1.X" arguments are for these things]). Avoid
  weird delegation chains by forcing the defaults
  in these situations.

* Comment out indexed setter/getter from ArrayOfBook

* For some reason we were forcing "isResponse" to
  false in BodyBuilder when deserializing in "fast"
  mode.  Respect the actual message type, which lets
  us get the right metadata in RPCHandler.

Passes all-tests, need to re-run TCK.
2005-02-11 21:48 dims 
Fix for AXIS-1811 - Attribute field names are not honored from bean type desc
from Steve Green (steve.green@epok.net)

URL: http://issues.apache.org/jira/browse/AXIS-1811
2005-02-11 04:31 toshi 
[Japanese resources for the Axis webapps]
Changing the link path from '/axis/jp' to '/axis/ja'.
2005-02-11 03:39 dims 
fix for AXIS-1812 - XMLBeanSerialization doesn't generate correct WSDL \n from daryoush mehrtash
2005-02-10 20:22 toshi 
Removed a number of unused or unnecessary keys for Japanese.
2005-02-10 19:06 dims 
fix another bad wsdl thanks to new wsdl4j
2005-02-10 18:14 dims 
Fix for AXIS-1624 : XML Serialization Alters XML Causing Signature Validation Failure
from  David Del Vecchio <ddelvecc@virginia.edu>

Notes: Sigh!! one more switch to document...
2005-02-10 17:42 dims 
fix bad wsdl - take3
2005-02-10 17:23 dims 
fix bad wsdl - take2
2005-02-10 17:22 dims 
fix bad wsdl
2005-02-10 16:53 dims 
fix bad wsdl
2005-02-10 16:45 dims 
Patch for public constructor and faults from Jonas team
2005-02-10 15:36 dims 
Fix for AXIS-1518 - XML Declaration missing
from  Jongjin Choi

URL: http://issues.apache.org/jira/browse/AXIS-1518
2005-02-10 15:36 dims 
Fix for AXIS-1518 - XML Declaration missing
from  Jongjin Choi

URL: http://issues.apache.org/jira/browse/AXIS-1518
2005-02-09 10:47 venkat 
Added additional targets to the build file to make it a complete test case for building and running ejb based web services deployed on JBoss.

Also added taskdefs and other missing pieces to the build file.
2005-02-08 18:44 dims 
enum package is now deprecated. Please use constants package from now on. we still have wrapper classes in enum package that extend the ones in constants. But they are marked deprecated.
2005-02-08 18:44 dims 
enum package is now deprecated. Please use constants package from now on. we still have wrapper classes in enum package that extend the ones in constants. But they are marked deprecated.
2005-02-08 18:44 dims 
enum package is now deprecated. Please use constants package from now on. we still have wrapper classes in enum package that extend the ones in constants. But they are marked deprecated.
2005-02-08 18:44 dims 
enum package is now deprecated. Please use constants package from now on. we still have wrapper classes in enum package that extend the ones in constants. But they are marked deprecated.
2005-02-07 18:37 dims 
better test case gen for enum's [pick the first from the reftype]
2005-02-07 14:48 dims 
remove extra code from DateDeserializer.
2005-02-07 14:45 dims 
Date -> Calendar
2005-02-07 13:15 dims 
fix failure in test16Schema2ServiceEchoDocumentTest
2005-02-07 07:54 gdaniels 
* When the BeanSerializer is serializing null values for
  elements with minOccurs > 0 and nillable == false, check
  to see if the desired class is a Number and if so try
  to turn it into a zero instead (a little friendlier than
  an exception).

* Introduce TypeMapping.getQNameForClass(xmlType, javaType),
  which acts like getQNameForClass(xmlType) but will search
  up the TypeMapping hierarchy first for an EXACT match
  for the given type before falling back to the default
  (whatever the last mapping for xmlType was).

* Use the above when syncing methods for greater accuracy.

* Register LANGUAGE and NAME in the default type mapping
  (why not?)

* Remove "HACK" from TypeMappingRegistry

There are likely still a couple of broken things, but we're better off
than we were this morning.  More tomorrow.
2005-02-07 07:54 gdaniels 
* When the BeanSerializer is serializing null values for
  elements with minOccurs > 0 and nillable == false, check
  to see if the desired class is a Number and if so try
  to turn it into a zero instead (a little friendlier than
  an exception).

* Introduce TypeMapping.getQNameForClass(xmlType, javaType),
  which acts like getQNameForClass(xmlType) but will search
  up the TypeMapping hierarchy first for an EXACT match
  for the given type before falling back to the default
  (whatever the last mapping for xmlType was).

* Use the above when syncing methods for greater accuracy.

* Register LANGUAGE and NAME in the default type mapping
  (why not?)

* Remove "HACK" from TypeMappingRegistry

There are likely still a couple of broken things, but we're better off
than we were this morning.  More tomorrow.
2005-02-07 02:18 gdaniels 
Roll back dims' change - this breaks test/wsdl/omit!  It's important
that we respect the schema - if the <element> declaration doesn't say
nillable="true", it ain't nillable no matter what type it is.

If there's a problem here we need another solution.
2005-02-06 23:57 dims 
prevent attribute lookup for basic types
2005-02-06 19:31 dims 
prevent problems with terra test case.
2005-02-06 17:12 dims 
Prevent errors like this one from the test case:
java.io.IOException: Non nillable element 'myBigDecimal' is null.

Patch from jonas team (" isNillable value depends only on the field type")
2005-02-06 17:12 dims 
Prevent errors like this one from the test case:
java.io.IOException: Non nillable element 'myBigDecimal' is null.

Patch from jonas team (" isNillable value depends only on the field type")
2005-02-06 05:12 dims 
Ability to use a JavaBean (w/o SimpleType interface marker, See test/wsdl/schema2/Document.java for example).
2005-02-06 05:12 dims 
Ability to use a JavaBean (w/o SimpleType interface marker, See test/wsdl/schema2/Document.java for example).
2005-02-05 21:14 dims 
Checking in test case (working). Will break if the indexed methods in ArrayOfBook are removed.

[Basically if we have a bean that contains a field that is an array, but the bean does not provide indexed methods to set/get values, only java bean setter and getter (that take/return the entire array) then we fail.]
2005-02-05 17:50 dims 
Make explicit checks for NPE.
2005-02-05 00:14 dims 
Support and test case for JSR 109 style exceptions:

JSR 109 Section 7.3.2 says:
Faults must map to an exception such that it:
• Directly or indirectly inherits from java.lang.Exception, but must not inherit from RuntimeException nor RemoteException.
• Has at most a single property called “message” of type java.lang.String

Notes:
- I don't like this, but checking it in for Glen/Tom's review and comment.
- Any alternate solution should not break the test(s) added in this check-in.
2005-02-04 19:10 dims 
Prevent NPE's.
2005-02-04 18:11 dims 
move code to HandlerChainImpl from JAXRPCHandler since HandlerInfoChainFactory is the one that instantiates stuff from wsdd entries.
2005-02-04 18:10 dims 
Check the java class type and return accordingly (patch from jonas team)
2005-02-04 17:39 dims 
remove the "$Id" lines. since they are only in a few files.
2005-02-04 17:39 dims 
remove the "$Id" lines. since they are only in a few files.
2005-02-04 17:35 dims 
remove the "@version $Revision" lines. since they are only in a few files.
2005-02-04 13:59 dims 
Bye Bye IME
2005-02-04 03:42 dims 
minor changes to how text is treated (patches from Jonas team)
2005-02-04 03:41 dims 
Adding additional methods (ser/de-ser changes will follow tomorrow)
2005-02-04 00:31 dims 
Implementation and test for GetElementsByTagName
2005-02-03 21:57 dims 
More fixes from the jonas team.

- Remove hard-coded UTF-8 (caused BOM problems)
- Add flag to not optimize the soappart stuff (keep the message as a soap envelope)
- throw error if jaxrpc handler modifies the operation name or its children

Notes:
- the hack to check for duplicate "<?xml" is commented out as it breaks performance tests.
- Need to find another way to check for duplicate "<?xml" (maybe using a FilterOutputStream?)
2005-02-03 17:46 tomj 
AXIS-1799 fix - continued.
Add new test WSDL file.
2005-02-03 17:42 tomj 
AXIS-1799 fix - continued.
Fix fumble fingered typo that happened due to bad focus switching in my IDE.
2005-02-03 17:33 tomj 
Fix for AXIS-1799 - Duplicate file exception throw when 2 elements have the
same name and differ only in the capitolization of the first letter.

Fix JavaGeneratorFactory to notice duplicate type entries from the symbol table
in addition to the anonymous types list.

Add a test case to wsdl/clash.
2005-02-03 14:49 dims 
- fix typo when registering XSD_DATETIME
- Add separate ser/der stuff for XSD_TIME (patch from jonas team)
2005-02-03 14:34 dims 
DefaultJAXRPC11TypeMappingImpl should be extending DefaultSOAPEncodingTypeMappingImpl (code-fix from Jonas Team/Ias)
2005-02-03 14:16 dims 
Substitute RemoteException with Exception (code-fix from Jonas Team/Ias)
2005-02-03 14:00 dims 
WS-I BP R2728 (code-fix from Jonas Team/Ias)
2005-02-03 11:06 dims 
Tighten the condition when the builtin ByteArrayHolder is used.
2005-02-03 11:05 dims 
make sure there is something left in the local part
2005-02-03 03:24 dims 
Avoid  ill formed element-name (fix from jonas team)
2005-02-02 17:58 dims 
Allow not only SOAPBodyElement but also (more general) MessageElement. (code-fix from Jonas Team/Ias)
2005-02-02 17:54 dims 
Treat both HTTP return code 302 and 307 the same. (code-fix from Jonas Team)
2005-02-02 13:39 dims 
Fix for AXIS-1798 - Multiple AxisServers per ServletContext - suggested EngineConfigurationFactoryServlet changes
from Clive Brettingham-Moore (apache@brettingham-moore.net)

URL: http://issues.apache.org/jira/browse/AXIS-1798
2005-01-29 20:50 dims 
switch the order.
2005-01-29 03:00 dims 
allow <sysproperty> for setting for example the new axis.ws-i.bp11.compatibility flag.
2005-01-28 21:47 dims 
Change the order of registering the typemapping.
2005-01-28 20:11 dims 
- Fix for overloading methods for WS-I BP compliance.
- Additional typemapping for jaxrpc11 compliance.
2005-01-28 09:39 venkat 
Ref AXIS-1790:
The writeFileBody method uses just "Object" where it should have used "java.lang.Object". The created compilation issues in the code generated by WSDL2Java. The code generated already has "Object" class (because WSDL has it) in the same package as the stub class is.
2005-01-27 16:00 dims 
set isreturn if INOUT or OUT (prev change broke build)
2005-01-27 07:30 toshi 
Adding a few messages and  a little fix to enhance its reusability.
2005-01-27 03:19 dims 
Fix for AXIS-1785 - DII doesn't allow you to register IN/OUT parameters
from daryoush mehrtash (dmehrtas@bea.com)

URL: http://issues.apache.org/jira/browse/AXIS-1785
2005-01-27 00:35 dims 
Fix for AXIS-1783 - ant task class path error leads to a ClassNotFoundException
2005-01-26 00:52 dims 
Fix for AXIS-1732 - JAXRPC Handler modification not propagated to actual service implmentation

AND

additional SAAJ problem reported by Michael Merz (working on JSR 181 impl in beehive)
2005-01-25 18:47 dims 
Fix for AXIS-1714 - Java2WSDL throws NPE on WSDLs that import schema from a file
from Mark Hansen (m.hansen23@verizon.net)
2005-01-25 18:15 dims 
Fix for AXIS-1776 - Client configuration '.wsdd' files not found when packaged in an ejb-jar file.
2005-01-25 06:52 gawor 
test for qname on the wire with whitespace
2005-01-25 06:51 gawor 
trim the qname value first
2005-01-25 04:06 dims 
Fix for  AXIS-1446 - WSDL2Java ignores <xsd:any> elements in types
from  Thanh-Phong Lam (phong.lam@t-online.de)
2005-01-24 12:57 venkat 
Fix for the issue AXIS-1775 - If the WSDL has same value for name attribute for both <service> and <portType>, WSDL2Java with -p option results in DuplicateFileException.
2005-01-23 23:11 dims 
Bad News: Ugly Hack!!! Need to fix as part of Glen/Tom's review of my code modifications.
Good News: "ant clean all-tests" works now and back to where we were for TCK status.
2005-01-23 07:55 toshi 
A feedback from the i18n task on the Japanese Axis Community.
The original contributor is Shouhei Kodama [kodama@isp.co.jp].
Some Messages don't need to be translated ...
2005-01-22 20:38 dims 
hmmm....forgot the TypeMappingDelegate
2005-01-21 20:09 dims 
remove commented out code.
2005-01-21 19:39 dims 
back out the org.apache.axis.configuration.EngineConfigurationFactoryEJB change...kills "ant clean all-tests"
2005-01-21 18:20 dims 
undo Fix for AXIS-1754....what was i thinking?
2005-01-21 17:54 dims 
Fix for AXIS-1754 - Service ThreadLocal previousCall should not be static
from Tim Kagle (tkapache@idpax.com)

URL: http://issues.apache.org/jira/browse/AXIS-1754
2005-01-21 17:50 dims 
Fix for AXIS-1776 - Client configuration '.wsdd' files not found when packaged in an ejb-jar file.
from Andrei Iltchenko (andrei.iltchenko@nl.compuware.com)

URL: http://issues.apache.org/jira/browse/AXIS-1776
2005-01-21 17:41 dims 
Fix for  AXIS-1764 - Add an XmlBean Serializer and Deserializer to Axis
from Jonathan Colwell (beehive@superflaco.com)
2005-01-21 16:57 dims 
Fix for AXIS-1769 - Null Pointer when creating a SOAPEnvelope with a SOAP message containing a SOAP Fault
from Joshua Hendrick (jhendrick@parasoft.com)

URL: http://issues.apache.org/jira/browse/AXIS-1769
2005-01-21 16:52 dims 
Fix for AXIS-1709 - Client#setOperation(String) needs to take header params into account (Patch)
from Michael Schuerig (michael@schuerig.de)

URL: http://issues.apache.org/jira/browse/AXIS-1709
2005-01-21 15:57 dims 
shuffle the deck
2005-01-21 07:57 toshi 
For JDK 1.3 compliance.
2005-01-21 06:56 gawor 
fixed wrong class
2005-01-21 06:32 gawor 
synchronization fixes
2005-01-21 04:01 dims 
delegate to its own singleton
2005-01-20 21:35 dims 
Better to do this in the stub constructor. since locator may not be used.
2005-01-20 14:59 dims 
Ensure Typemapping is set.
2005-01-20 14:18 dims 
- Eliminate the axis.jaxrpc11Compliance flag
- Support typeMappingVersion setting per service
- At codegen time, add more information in deploy.wsdd and Locator for setting typemappingversion

Notes:
- Glen/Tom, PLEASE review the code.
- Still check/need to fix wrong xsi:types at runtime for JAXRPC11
- Need to run TCK.
2005-01-20 14:18 dims 
- Eliminate the axis.jaxrpc11Compliance flag
- Support typeMappingVersion setting per service
- At codegen time, add more information in deploy.wsdd and Locator for setting typemappingversion

Notes:
- Glen/Tom, PLEASE review the code.
- Still check/need to fix wrong xsi:types at runtime for JAXRPC11
- Need to run TCK.
2005-01-20 02:38 dims 
fix return values
2005-01-19 21:47 dims 
OOPS!
2005-01-19 21:04 dims 
Reduce the # of places Default***TypeMappingImpl's are created.

Note:
- Am not done cleaning up yet :)
2005-01-18 14:49 dims 
Key off of the typemapping impl and not of the system property.
2005-01-18 07:58 toshi 
A feedback from the i18n task on the Japanese Axis Community.
The original contributor is Shouhei Kodama [kodama@isp.co.jp].
2005-01-18 07:51 toshi 
These're for i18n. The charset should be UTF-8.
2005-01-18 07:46 toshi 
Just correcting typo: from 'seFrver' to 'server'.
2005-01-18 03:38 dims 
- Move JAXRPC11 stuff into its own little typemappingimpl ("1.3")
- Move duplicate code for creating TypeMappingImpl in J2W Ant Task and Command line into Emitter
- Remove the old commented out jaxrpc11mappings code
- Eliminated need for the axis.jaxrpc11Compliance when running the J2W/W2J ant/commandline tools.
- Adding a sample (start from WSDL) for testing the new JAXRPC11TypeMappingImpl.

TODO:
- Looking that the wire traces, some of them have weird xsi:types, need to fix them, before updating the docs
- Need be able to switch axis.jaxrpc11Compliance runtime flag on for specific services or atleast from AxisProperties
- Need to seriously fix which Defautl*TypeMappingImpl gets created under what conditions.
- Need to get rid of the axis.jaxrpc11Compliance system property (codegen more stuff?)
2005-01-18 03:38 dims 
- Move JAXRPC11 stuff into its own little typemappingimpl ("1.3")
- Move duplicate code for creating TypeMappingImpl in J2W Ant Task and Command line into Emitter
- Remove the old commented out jaxrpc11mappings code
- Eliminated need for the axis.jaxrpc11Compliance when running the J2W/W2J ant/commandline tools.
- Adding a sample (start from WSDL) for testing the new JAXRPC11TypeMappingImpl.

TODO:
- Looking that the wire traces, some of them have weird xsi:types, need to fix them, before updating the docs
- Need be able to switch axis.jaxrpc11Compliance runtime flag on for specific services or atleast from AxisProperties
- Need to seriously fix which Defautl*TypeMappingImpl gets created under what conditions.
- Need to get rid of the axis.jaxrpc11Compliance system property (codegen more stuff?)
2005-01-16 23:38 dims 
Fix for AXIS-1767 - Style and other Enums are not serializable enumerated types
2005-01-15 13:33 dims 
- trying to standardize which typemappingimpl is used and when
- making sure that existing test harness and tck harness don't break
- roundtrip has it's own typemapping version "1.0"
- sync'ed all the ant tasks and command line tools
- looks like we need more cleanup :(
2005-01-15 13:29 dims 
cleanup comments a bit.
2005-01-14 15:34 dims 
Try to resolve files w.r.t the base directory
2005-01-14 13:39 dims 
enable fork in adminclient
2005-01-14 13:06 dims 
hmmm...we were not setting the target during "fork"
2005-01-14 04:14 dims 
wrong class
2005-01-14 03:04 dims 
Enhance AdminClient Task to deploy multiple wsdd's (usage similar to FixCRLF task)
2005-01-13 14:59 dims 
ignore axis-1_2RC2
2005-01-11 23:35 gawor 
just some minor changes
2005-01-11 23:31 gawor 
added choice tests
2005-01-11 23:31 gawor 
added choice tests
2005-01-11 23:27 gawor 
applied 1753, enables proper serialization of choice elements
2005-01-11 16:44 gdaniels 
When writing WSDL, a) use maxOccurs="unbounded" for literal arrays, and
b) fix a bug where we were writing encoded-style arrays when in doc/lit
mode.
2005-01-11 03:56 dims 
log error if class specified in HandlerInfoChain is not found
2005-01-10 20:24 gawor 
added a call to deassociate the last call object from the current thread so that jvm can collect this object sooner
2005-01-10 19:02 gdaniels 
* Move jaxrpc 1.1 compatibility mode switch to a publically-accessible static variable on AxisEngine.   Might eventually want to move this to AxisProperties.

* When generating type names for arrays, use "MyArrayOf" instead of "ArrayOf" when in JAXRPC mode to avoid stupid WS-I rule
2005-01-07 11:05 toshi 
For i18n of Axis webapps.
Added some methods to enhance the reusability.
2005-01-06 18:49 gawor 
added deserialization tests for wrapped primitive types
2005-01-06 18:47 gawor 
fixed simple deserializer issues with wrapped types
2005-01-06 07:29 toshi 
For i18n of Axis webapps.
A feedback from the i18n task on the Japanese Axis Community.
Japanese resouces for;
 - [index.jsp] was contributed by
                             Shouhei Kodama [kodama@isp.co.jp]
 - [happyaxis.jsp] was contributed by
                             Manabu Yoneyama [yoneyama@kronos-jp.net]
2005-01-06 07:28 toshi 
For i18n of Axis webapps.
2005-01-06 07:27 toshi 
For i18n of Axis webapps.
Changed the priority [html->jsp->jws] to [jsp->html->jws].
2005-01-06 07:27 toshi 
For i18n of Axis webapps.
2005-01-06 07:27 toshi 
For i18n of Axis webapps.
2005-01-05 06:55 gawor 
minor fix to generated javadoc comments (so that javadoc doesn't complain anymore about invalid arguments)
2005-01-05 06:25 gawor 
just cleanup: untabify + added copyright statment
2005-01-05 06:17 gawor 
simple tests for cheching fault metadata
2005-01-05 06:15 gawor 
don't lookup super classses when generating fault metadata for service operations. super class lookup might lead to missing fault metadata
2004-12-31 23:57 ias 
Initial drop of Koreanized tcpmon.properties.
2004-12-30 23:38 dims 
some input streams (coyote/tomcat) are a bit flaky, so buffer the stream
2004-12-30 15:59 dims 
TCK expects IllegalStateException
2004-12-30 02:03 toshi 
Added language choice - en(English) and ja(Japanese).
2004-12-30 02:01 toshi 
Fixing the wrong or deprecated links.
 - xml.apache.org to ws.apache.org
 - new Wiki page(http://wiki.apache.org/ws/FrontPage/Axis)
2004-12-30 02:01 toshi 
Fixing the wrong or deprecated links.
 - xml.apache.org to ws.apache.org
 - new Wiki page(http://wiki.apache.org/ws/FrontPage/Axis)
 - new discovery page(http://jakarta.apache.org/commons/discovery/)
 - new logging page(http://jakarta.apache.org/commons/logging/)
2004-12-29 17:53 dims 
Sort the names to make generated code predictable
2004-12-29 07:03 toshi 
For i18n of tcpmon.
2004-12-28 05:31 toshi 
A few messages are updated.
2004-12-28 05:23 toshi 
A feedback from the i18n task on the Japanese Axis Community.
The original contributor is Takashi Okamoto [tora@gloria.or.tp]
2004-12-24 05:56 dominik 
PR: AXIS-1397
Submitted by: AJ Banck (ajbanck@informatica.com)

Added servlet-config.wsdd parameter axis.servicesPath to replace hardcoded "/services/" used in AxisServlet.
Fix for AXIS-1397 from AJ Banck (ajbanck@informatica.com).

- Dominik
2004-12-23 21:27 dominik 
Deleted unused files. Neither one of these jar files is referenced from Axis code or used in the build process. I checked revision history and there is no indication how they were supposed to be used.
2004-12-23 19:59 dominik 
Fix for AXIS-843.
Changed how the request path is determined to accommodate .jws deployed in /services/. While at it, I reformatted the code - the six character indentations and some of the curly braces violated our coding standards.

- Dominik
2004-12-23 04:22 dims 
minor nit.
2004-12-23 00:18 dims 
JDK13IO is optional too...
2004-12-23 00:06 dims 
Get build to work under Kaffe
2004-12-21 20:17 dominik 
Added Locator constructor to be called to avoid ServiceException in getCalls(QName).
Fix for AXIS-1711 from Guillaume Sauthier (guillaume.sauthier@objectweb.org).
2004-12-21 01:56 dominik 
As a follow up for AXIS-1674, I updated happyaxis page to detect which version of SAAJ API is available in the classpath. If SAAJ API is available but does not pass the test, happyaxis reports an error.
2004-12-17 19:17 tomj 
Fix logic error when checking to see if an explicit type mapping has been
set in the service description that resulted in the default type mapping
being used instead of the one set explicitly!

Clean up some unnecessary assignments.
2004-12-17 17:34 tomj 
Update the reference page with the correct dotNetSoapEncFix switch case
and mention the static variable that can be set via Java.

Curses to editors that don't automatically save!
2004-12-17 17:00 tomj 
Add a switch to turn on a preference for XML Schema types: "dotNetSoapEncFix"

This addresses Axis-1467:  http://nagoya.apache.org/jira/browse/AXIS-1467

This is a global switch which sets a static flag in the TypeMappingImpl class
and causes it to skip over SOAP encoded types when looking for the QName
of a Java class.

Updated the reference guide with the new switch.

Fix a problem in the JavaServiceDesc class where the type mapping
used by default (for .JWS files for instance) was not the SOAP encoding
mapping when use=encoded.  For deployed services, we did use the soapenc
types if they were encoded.  If someone explicitly sets a type mapping, we
don't mess around with that.
2004-12-16 18:54 tomj 
Revert the patch for AXIS-1720 - http://nagoya.apache.org/jira/browse/AXIS-1720

This breaks the WSDL for doc/lit Object arrays and throws warnings when
Object is encountered by Java2WSDL.

Commented out test cases and added one for Object and Object[].
2004-12-16 18:54 tomj 
Revert the patch for AXIS-1720 - http://nagoya.apache.org/jira/browse/AXIS-1720

This breaks the WSDL for doc/lit Object arrays and throws warnings when
Object is encountered by Java2WSDL.

Commented out test cases and added one for Object and Object[].
2004-12-15 22:45 tomj 
Just a little code cleanup of unused imports and unnecessary assignments.
2004-12-15 22:44 tomj 
Back out my change from Nov 18: do NOT add a trailing slash to generated
namespaces as users (like mine) may depend on the Axis 1.1 namespaces
and get unhappy when they change for no good reason.

Motivation for the original change was WS-I validation,
but this was a false alarm.

Fix namespaces in the TestAutoTypes test to match defaults.
2004-12-15 12:15 dims 
Fix for  AXIS-1664 - Allow doAutoTypes to be configured through AxisProperties
Reworked patch and test cases from Patrick Martin (patrick.martin@aquilauk.co.uk)
2004-12-14 21:27 dims 
Fix for AXIS-1722 - MethodCache does not consider the classloader in the cache key
from Eric Norman (enorman@webmethods.com)

JIRA URL: http://nagoya.apache.org/jira/browse/AXIS-1722
2004-12-14 20:22 tomj 
Remove the jaxrpc 1.1 switch from the WSDL2Java Emitter for Axis 1.2.
This switch did some of the right things, but no where did we switch the engine
to correctly interpret the XML on the wire, nor did it handle the WSDL generation
of the right Schema types.  Basically it did more harm than good.  This is going
to have to be done correctly when someone has the time.

Also, fix up the usages of the DefaultSOAPEncodingTypeMapping so that
we use createWithDelegate when we are supposed to and when we have
"use" of encoded, we get *that* type mapping instead of the default Schema
type mapping.

These changes are to prep for trying to fix Axis 1.2 so that it prefers
XML Schema types again (per 1.1) instead of soap encoded types now that
the encoding type mapping is "in front" of the Schema types in the encoding
type mapping.
2004-12-14 20:22 tomj 
Remove the jaxrpc 1.1 switch from the WSDL2Java Emitter for Axis 1.2.
This switch did some of the right things, but no where did we switch the engine
to correctly interpret the XML on the wire, nor did it handle the WSDL generation
of the right Schema types.  Basically it did more harm than good.  This is going
to have to be done correctly when someone has the time.

Also, fix up the usages of the DefaultSOAPEncodingTypeMapping so that
we use createWithDelegate when we are supposed to and when we have
"use" of encoded, we get *that* type mapping instead of the default Schema
type mapping.

These changes are to prep for trying to fix Axis 1.2 so that it prefers
XML Schema types again (per 1.1) instead of soap encoded types now that
the encoding type mapping is "in front" of the Schema types in the encoding
type mapping.
2004-12-14 20:22 tomj 
Remove the jaxrpc 1.1 switch from the WSDL2Java Emitter for Axis 1.2.
This switch did some of the right things, but no where did we switch the engine
to correctly interpret the XML on the wire, nor did it handle the WSDL generation
of the right Schema types.  Basically it did more harm than good.  This is going
to have to be done correctly when someone has the time.

Also, fix up the usages of the DefaultSOAPEncodingTypeMapping so that
we use createWithDelegate when we are supposed to and when we have
"use" of encoded, we get *that* type mapping instead of the default Schema
type mapping.

These changes are to prep for trying to fix Axis 1.2 so that it prefers
XML Schema types again (per 1.1) instead of soap encoded types now that
the encoding type mapping is "in front" of the Schema types in the encoding
type mapping.
2004-12-14 17:11 dims 
Additional patch from jongjin for all-tests failure.
2004-12-14 15:14 dims 
Fix for AXIS-1720 - Java2WSDL, repeated types auto-mapping to xsd:anyType causes bad definition in schema
from Brett Randall (brett_s_r@yahoo.com.au)
2004-12-14 15:14 dims 
Fix for AXIS-1720 - Java2WSDL, repeated types auto-mapping to xsd:anyType causes bad definition in schema
from Brett Randall (brett_s_r@yahoo.com.au)
2004-12-14 13:27 dims 
Fix for AXIS-1208 - JAXRPCHandler doesn't evaluate return value of handleRequest/handleReponse
from Jongjin Choi (gunsnroz@hotmail.com)
2004-12-14 12:58 dims 
Fix for (AXIS-1713) SAAJ 1.2 (DOM) node trees are invalid - missing parents
from Mark Hansen

URL: http://nagoya.apache.org/jira/browse/AXIS-1713
2004-12-14 12:58 dims 
Fix for (AXIS-1713) SAAJ 1.2 (DOM) node trees are invalid - missing parents
from Mark Hansen

URL: http://nagoya.apache.org/jira/browse/AXIS-1713
2004-12-14 06:22 gawor 
update test to work with latest code. getEncoder(blah) returns a default encoder now
2004-12-10 15:20 dims 
Fix NPE. Note to self, think before you check-in :)
2004-12-10 14:54 dims 
Adding a default encoder (nusoap sends ISO-8859-1 to us, working with FR^2 on IRC to see if this fixes the problem)
2004-12-10 00:35 dims 
Fix for AXIS-1208 - JAXRPCHandler doesn't evaluate return value of handleRequest/handleReponse
from Shantanu Sen (ssen@pacbell.net)

JIRA: http://nagoya.apache.org/jira/browse/AXIS-1208
2004-12-10 00:35 dims 
Fix for AXIS-1208 - JAXRPCHandler doesn't evaluate return value of handleRequest/handleReponse
from Shantanu Sen (ssen@pacbell.net)

JIRA: http://nagoya.apache.org/jira/browse/AXIS-1208
2004-12-09 22:30 gawor 
symbol table optimizations, including caching of dervied types suggested by someone long time ago
2004-12-09 22:29 gawor 
symbol table optimizations, including caching of dervied types suggested by someone long time ago
2004-12-09 21:19 dims 
Make both test/wsdl/jaxrpchandler and test/jaxrpc happy!
2004-12-09 17:09 dims 
Patch for AXIS-1527 - Using multiple AxisServlet in webapp share the same AxisEngine
from Guillaume Sauthier (guillaume.sauthier@objectweb.org)

JIRA: http://nagoya.apache.org/jira/browse/AXIS-1527
2004-12-09 14:59 tomj 
Applied patch from Dominik Kacprzak [dominik@opentoolbox.com].
Update the installation documentation with WebLogic 8.1 info.
2004-12-09 14:36 dims 
Fix for  AXIS-1696 - JAXRPC Handlers behavior different from server-side to client-side
from Guillaume Sauthier (guillaume.sauthier@objectweb.org)

JIRA: http://nagoya.apache.org/jira/browse/AXIS-1696
2004-12-09 13:56 dims 
Fix for  AXIS-1700 - axis performance in multithreading env
from Sebastien Bocq (sebastien.bocq@alcatel.be)

JIRA: http://nagoya.apache.org/jira/browse/AXIS-1700
2004-12-09 12:54 dims 
Fix for AXIS-1684 - Error in class : org.apache.axis.wsdl.toJava.JavaTypeWriter
from Venkat Reddy (vreddyp@gmail.com)
2004-12-09 12:49 dims 
Fix for AXIS-1527 - Using multiple AxisServlet in webapp share the same AxisEngine
from  Guillaume Sauthier (guillaume.sauthier@objectweb.org)
2004-12-09 03:23 dims 
Prevent jikes from croaking...
2004-12-08 22:55 dims 
Fix for AXIS-1608 - Test case fro comparing Axis 1.2 against Axis 1.1
2004-12-08 16:42 tomj 

Applied patch from Dominik Kacprzak [dominik@opentoolbox.com]:

As Dims noticed earlier, this issue is caused by a weblogic.xml.jaxp.RegistryDocumentBuilder
not allowing for null values to be passed to setEntityResolver. The attached patch adds a few
more catch statements along the lines of what Dims already did.
I also added log.debug to the catch statements.
With my patch, I was able to access ?wsdl URLs,
deploy/undeploy and use a few sample applications.

All tests pass.
2004-12-06 18:24 gawor 
fix for bug 1697
2004-12-06 06:38 gawor 
specialized RPCParam target implementation to avoid reflection
2004-12-06 06:23 gawor 
call getRead/WriteMethod once
2004-12-05 06:17 gawor 
avoid reflection if possible
2004-12-04 03:02 dims 
Don't increment the count twice...found when testing with IKVM.NET
2004-12-03 21:54 gawor 
when talking to http1.1 server, the server may send 100 continue header without client explicitely requesting it
2004-12-02 18:07 gawor 
changed slightly the order of how serializers are discovered. the class lookup for serializer is now done after the type mapping lookup and before class interface lookup. before this change the class lookup was done only at the very end as a last resort if no serializer was found. that led to problems where sometimes a subclass serializer was used to serialize the object even though a specific serializer was associated with the class (in metadata)
2004-12-02 15:28 dims 
Sometimes you need to set the JIKESPATH before running ant to get things working.
2004-12-02 13:28 dims 
Fix for AXIS-1689 - Broken NTLM support in CommonsHTTPSender
from Per Thomas Jahr (perja@nr.no)
2004-12-02 13:16 dims 
Fix for AXIS-1691 - add two new targets to build.xml to build api and src jars
from  Ian Springer (ian.springer@hp.com)
2004-12-02 13:13 dims 
Fix for AXIS-1692 - .cvsignore should contain build.properties
from Ian Springer (ian.springer@hp.com)
2004-12-02 13:04 dims 
Fix for AXIS-1694 - problem with accentued characters : server-config.wsdd is not a valid UTF-8 file
from Cédric Chabanois (cchabanois@natsystem.fr)
2004-12-02 12:53 dims 
Fix for AXIS-1680 - Add an offline/unchecked-url mode to WSDL2Java
from Guillaume Sauthier (guillaume.sauthier@objectweb.org)
2004-12-02 12:37 dims 
Fix for AXIS-1690 - HandlerInfo.getHandlerConfig() MUST return a non-null Map
from Guillaume Sauthier (guillaume.sauthier@objectweb.org)
2004-12-01 23:16 gawor 
use the method cache to get the serializer/deserializer from class
2004-12-01 23:14 gawor 
cache the result of previous getTypeDescForClass() lookup, reduced the number of reflection calls
2004-11-30 21:54 gawor 
don't need to call response handlers when doing one way call on the client
2004-11-30 01:57 dims 
reworked patch for AXIS-1017 - org.apache.axis.types.Duration
from Dominik Kacprzak
2004-11-29 22:29 gawor 
nicer error message when socket factory was not found
2004-11-29 22:29 gawor 
nicer error message when socket factory was not found
2004-11-29 22:28 gawor 
allocated the vector only when needed
2004-11-29 20:54 dims 
Trying to get rid of JDK1.4 dependency.
2004-11-29 03:40 dims 
- adding traverse dom test case
- fixed getChildNodes in SOAPDocumentImpl
2004-11-29 03:09 dims 
Fix for AXIS-1674 - NullPointerException on WebLogic81
2004-11-27 21:08 dims 
Fix for AXIS-1017 - org.apache.axis.types.Duration
from Dominik Kacprzak (dominik@opentoolbox.com)
2004-11-24 17:09 gawor 
added a function to reset the Call object associated with the thread to make axis release the object sooner then later (when a new call is made)
2004-11-23 23:30 dims 
Fix for AXIS-1466 - Error with array of objects with <xsd:extension base="xsd:string"> in definition
Fix for AXIS-1464 - Wrong arrays deserialisation (all elements equal to each other)
Original patch was from Ashutosh Shahi (ashutosh.shahi@gmail.com), but checked in a slightly modified version (make local copy of original attributes within the loop)
2004-11-23 23:25 dims 
Fix for AXIS-1344 - UnsignedLong corrupts data
from  Tommy Alander (tommy@alander.se)
2004-11-23 23:22 dims 
Fix for AXIS-1663 - LocalSender does not deserialize response message
from  Steve Green (steve.green@epok.net)
2004-11-23 01:37 dims 
Fix for AXIS-1669 - some samples require Junit but dependency is not in build.xml
from Jayachandra Sekhara Rao Sunkara (jayachandra@gmail.com)
2004-11-23 01:37 dims 
Fix for AXIS-1669 - some samples require Junit but dependency is not in build.xml
from Jayachandra Sekhara Rao Sunkara (jayachandra@gmail.com)
2004-11-23 01:29 dims 
Fix for AXIS-1679 - Error serializing an array of elements with attributes
from David R Robison (drrobison@openroadsconsulting.com)
2004-11-23 00:39 ias 
Fix - if an action URI doesn't have a scheme, it causes an exception (and hence failure) during undeploying a service with WS-FX Addressing.
2004-11-23 00:08 dims 
Fix for AXIS-1676 - SAXParseException for message containing more than 2 bytes UTF-8 chars and DIME Attachment
from Damien (apache-jira@ootoogo.com)
2004-11-22 23:59 dims 
Fix for AXIS-1668 - SerializationContext does not force xsi:type when the WSDLJava generated metadata, and shouldSendType=false
from Steve Green <steveg@epok.net>
2004-11-22 23:56 dims 
Fix for AXIS-1678 - Reports invalid qname part
from Tom Gordon (tmg@nc.rr.com)
2004-11-22 21:39 gawor 
doesn't need a reference to surrounding class
2004-11-22 18:11 gawor 
more small optimizations
2004-11-22 17:57 gawor 
just some minor optimizations
2004-11-22 17:57 gawor 
just some minor optimizations
2004-11-21 21:05 gdaniels 
Improve handling of arrays for (mostly) doc/lit cases.

* Make it OK to convert to Object arrays from individual Objects

* Move SOAP-Array mappings into the SOAP encoding
  TypeMapping - except for one, which is there just to allow
  the default Java-only case to kick in and get the
  ArraySerializer.
2004-11-21 21:05 gdaniels 
Improve handling of arrays for (mostly) doc/lit cases.

* Make it OK to convert to Object arrays from individual Objects

* Move SOAP-Array mappings into the SOAP encoding
  TypeMapping - except for one, which is there just to allow
  the default Java-only case to kick in and get the
  ArraySerializer.
2004-11-21 20:00 tomj 
Add a missing break in the look of field descriptors - save some effort.
2004-11-21 19:47 gdaniels 
Back out change from Jonathan Colwell on 10/27.

I'm not exactly sure what was being done here, but it caused invalid WSDL
to be generated.  In particular, we saw stuff like :

<element name="tns:elementName"...>

...which is just wrong.  If we want to emit elements in different
namespaces, we need to do it right, which means emitting the
schemas for those namespaces separately and using
<element ref="qname"> instead of <element name="localName">.
2004-11-21 15:41 tomj 
Patch from Dominik Kacprzak [dominik@opentoolbox.com]:
 I extended test.wsdl.omit.OmitTestCase to test a fix for AXIS-1670.
2004-11-18 23:39 gawor 
just some optimizations to avoid reflection for selected builtin types
2004-11-18 23:34 gawor 
just minor optimizations
2004-11-18 21:37 tomj 
Fix .NET 2D array interop problem caused by previous changes.

Update the array serializer to not add the subdimentions to the array if
the component type is set as we will already have them there.

Back out my change to the Deserializer that compensates for the incorrect
serialization.
2004-11-18 17:00 tomj 

Patch from Dominik Kacprzak <dominik@sonic.net>, with additional
changes to the encoding unit tests to pass.

This patch provides a fix for various bugs related to Axis treatment of element's
nillable attribute. Although the default value of nillable attribute is "false". Axis
assumes that a nillable attribute is always true and it does not keep track of its
real value.  It leads to scenarios in which Axis sets a non-nillable, omittable
element to be null. For Axis there is no difference between the following
two definitions:
<element name="areaCode" type="string" minOccurs="0" maxOccurs="1"/>
and
<element name="areaCode" type="string" minOccurs="0" maxOccurs="1" nillable="true"/>

The fix adds an additional field to org.apache.axis.description.ElementDesc.  The nillable property defaults to false and is populated based on web services deployment definition or element's type.  As a result, Axis enforces stricter rules when generating web services.  An exception is thrown when a null value is passed as a value of a non-nillable, required element.
This patch fixes the following JIRA issues:
 - AXIS-530
 - AXIS-1321
 - AXIS-1332
 - AXIS-1357
 - AXIS-1670

Details:
- Java/Types.java -
  NOTE: This is most likely the most controversial change.
  1) I changed the method signature to static since it does not access
      any class properties.  This is harmless.
  2) I changed the logic which checks if type can be nullable.
      I couldn't find any justification for why a byte array cannot be null.
     To the contrary, I found an example in JAXRPC patterns that shows a nillable byte array: http://java.sun.com/developer/technicalArticles/xml/jaxrpcpatterns/index3.html

- I had to fix quite a few schemas and wsdls in functional tests to allow for null elements.
 The test were passing null values for non-nillable elements.

- Extended wsdl.omit test case to cover non-nillable and nillable scenarios.
2004-11-18 17:00 tomj 

Patch from Dominik Kacprzak <dominik@sonic.net>, with additional
changes to the encoding unit tests to pass.

This patch provides a fix for various bugs related to Axis treatment of element's
nillable attribute. Although the default value of nillable attribute is "false". Axis
assumes that a nillable attribute is always true and it does not keep track of its
real value.  It leads to scenarios in which Axis sets a non-nillable, omittable
element to be null. For Axis there is no difference between the following
two definitions:
<element name="areaCode" type="string" minOccurs="0" maxOccurs="1"/>
and
<element name="areaCode" type="string" minOccurs="0" maxOccurs="1" nillable="true"/>

The fix adds an additional field to org.apache.axis.description.ElementDesc.  The nillable property defaults to false and is populated based on web services deployment definition or element's type.  As a result, Axis enforces stricter rules when generating web services.  An exception is thrown when a null value is passed as a value of a non-nillable, required element.
This patch fixes the following JIRA issues:
 - AXIS-530
 - AXIS-1321
 - AXIS-1332
 - AXIS-1357
 - AXIS-1670

Details:
- Java/Types.java -
  NOTE: This is most likely the most controversial change.
  1) I changed the method signature to static since it does not access
      any class properties.  This is harmless.
  2) I changed the logic which checks if type can be nullable.
      I couldn't find any justification for why a byte array cannot be null.
     To the contrary, I found an example in JAXRPC patterns that shows a nillable byte array: http://java.sun.com/developer/technicalArticles/xml/jaxrpcpatterns/index3.html

- I had to fix quite a few schemas and wsdls in functional tests to allow for null elements.
 The test were passing null values for non-nillable elements.

- Extended wsdl.omit test case to cover non-nillable and nillable scenarios.
2004-11-18 15:31 tomj 
Add a trailing slash to namespaces we create from package names.

The WS-I validation of our WSDL complains about our targetNamespace
while this doesn't fix it, it does make them just a bit more well formed.
2004-11-18 15:26 tomj 
Fix interop regression with .NET rpc/encoded arrays.

Change the way we do array serialization to keep track of the array type
specified in WSDL and use it when serializing an array.

Example Schema from a .NET rpc/encoded array:
<s:complexType name="ArrayOfString">
 <s:complexContent mixed="false">
  <s:restriction base="soapenc:Array">
   <s:attribute d7p1:arrayType="xsd:string[]" ref="soapenc:arrayType"/>
  </s:restriction>
 </s:complexContent>
</s:complexType>

We will now remember the xsd:string type and use it instead of defaulting
to a soapenc:string, which .NET doesn't like and doesn't follow the WSDL.

Details:
- Fix a problem unconvered in the ArrayDeserailizer with array dimensions:
  If the class of the component specifies dimensions, don't add them back
  in again.
- ArraySerializer now has a special constructor that takes a component type
  argument. The component type from this constructor is checked first before
  the context, the type mapping and the super class type mappings.
- ArraySerializerFactory can now be created to produce ArraySerializers with
  the component type set.
- JavaStubWriter will emit code to register a factory for arrays with the
   component type as specified in the WSDL.
- Add componentType to the TypeEntry class with getter and setter.
- Set the component type for an array type that we create from WSD so we can
  use it in JavaStubWriter.
2004-11-18 15:26 tomj 
Fix interop regression with .NET rpc/encoded arrays.

Change the way we do array serialization to keep track of the array type
specified in WSDL and use it when serializing an array.

Example Schema from a .NET rpc/encoded array:
<s:complexType name="ArrayOfString">
 <s:complexContent mixed="false">
  <s:restriction base="soapenc:Array">
   <s:attribute d7p1:arrayType="xsd:string[]" ref="soapenc:arrayType"/>
  </s:restriction>
 </s:complexContent>
</s:complexType>

We will now remember the xsd:string type and use it instead of defaulting
to a soapenc:string, which .NET doesn't like and doesn't follow the WSDL.

Details:
- Fix a problem unconvered in the ArrayDeserailizer with array dimensions:
  If the class of the component specifies dimensions, don't add them back
  in again.
- ArraySerializer now has a special constructor that takes a component type
  argument. The component type from this constructor is checked first before
  the context, the type mapping and the super class type mappings.
- ArraySerializerFactory can now be created to produce ArraySerializers with
  the component type set.
- JavaStubWriter will emit code to register a factory for arrays with the
   component type as specified in the WSDL.
- Add componentType to the TypeEntry class with getter and setter.
- Set the component type for an array type that we create from WSD so we can
  use it in JavaStubWriter.
2004-11-16 17:29 dims 
Updating cvslog
2004-11-16 16:36 dims 
Fix for Perf problems from Archit Shah (http://marc.theaimsgroup.com/?l=axis-dev&m=110061913205350&w=2)
2004-11-16 16:36 dims 
Fix for Perf problems from Archit Shah (http://marc.theaimsgroup.com/?l=axis-dev&m=110061913205350&w=2)
2004-11-16 05:21 gawor 
a fix for qname serialization when a default namespace is set
2004-11-16 05:18 gawor 
qname serialization tests