[xalan-java] branch DTM_EXP deleted (was 0cb81628)

[jk...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

jkesselm pushed a change to branch DTM_EXP
in repository https://gitbox.apache.org/repos/asf/xalan-java.git


     was 0cb81628 Remove 'final' modifier from variables initialized in some constructors; this is to get around Sun JDK 1.1.8 compile error otherwise Reviewed by: sboag

This change permanently discards the following revisions:

 discard 0cb81628 Remove 'final' modifier from variables initialized in some constructors; this is to get around Sun JDK 1.1.8 compile error otherwise Reviewed by: sboag
 discard 86f10236 Warning for users of thread pooling
 discard 57ea21a7 Make use of the thread pool for transform threads and use a feature to set whether or not to do an incremental transform
 discard 6e2c861b Make use of the thread pool for transform threads and use a feature to set whether or not to do an incremental transform
 discard 3ea80733 Update version number.
 discard 519b0d80 Reworked the new end select trace interface a bit, so that there won't be a problem with backwards compatibility. Added TraceListenerEx, which has a selectEnd event, and EndSelectionEvent.  Added fireSelectedEndEvent event to the trace manager.  This event is thrown only from ElemForEach and ElemApplyTemplates.
 discard 5a93efeb Removed StringKey, and references to such.  Minor item that's been long outstanding on my todo list.
 discard 1e8bee57 Just going back to the older str() function (the one that doesn't use BigDecimal.  I believe personally that the one that uses BigDecimal is correct, but right now it will fail the conformance test, and we have to figure out the validation for this once and for all, including in regards to the datapower metrics.xsl test.
 discard 9f121c55 This check-in is probably too large to rightfully do in a single commit.  Sorry. The variable mechanism is been redone, so that references do direct index lookup, instead of lookup by QName comparison.  The variable stack is completely redesigned so that link/unlink is called on the stack before an apply templates or call templates.  I still don't like how parameters are being pushed on the stack... but I'm not sure there's a viable alternative.  All the static variable  [...]
 discard 7f884434 Revised namespace support
 discard b7b9258b Namespace support changes; requires alterations to DOM2DTM and SAX2DTM
 discard ab38b742 Major changes for both performance and robustness -- new Namespace handling, new DOM/DTM treewalker
 discard 2996d1bb Miscellaneous changes to enable tracing for tooling
 discard bd6308c3 Use DTM interface
 discard 1a604f9f Implement DOM Nodelist interface.
 discard f925436e Enabled detach() in a few more places, and fixed bug where detach() was being used with a variable: when owned by a variable you don't want to have the iterator detached, so I had to add a allowDetachToRelease(boolean allowRelease) method on DTMIterator, so that detach() can be called harmlessly by other callers.  (Perhaps detach() should be renamed to release() or some such()?).
 discard 663d96be Fix apparent copy/paste error where reset was being called twice in setRoot.
 discard e1fd0d58 Throw those exceptions, don't System.out them.
 discard 9122a607 Added support for the normalize space arg on dispatchCharacters.
 discard a1ed37fe Fixed a problem with the URI resolver being called twice.
 discard 4b88d2c5 Various minor optimizations or bug fixes.
 discard e724d868 Fixed bug where the DTM doc ID was not being set in the transformer, which we go away with if the ID was zero, but was caught when I forced the root ID to be 1.
 discard 5a83bb44 Added another constructor, so the growth size can be specified.
 discard b2e50bd7 Minor optimizations.
 discard 1ee3754c Minor optimization for getLocalName, where you can assume that the zero index will return an empty string.
 discard d6069585 Changed the m_dtms list to be an array instead of a vector. Always set the first ID to 1, so that bugs with not masking or unmasking the doc ID bits will be caught. Also, turned back on incremental processing.
 discard 1998f9b3 Fixed bug with the previous iterator, where the doc ID was not being masked off.
 discard 35294f19 Various minor optimizations, including setting the m_mask to be a static final.  Playing with the block sizes some.
 discard e1562f30 Set m_last in setRoot, which fixes a bug that shows up in the smoke tests.
 discard bc87eb78 Various performance tweaks, and work related to my previous check-ins from tonight.
 discard 609fec86 High tuning of the for-each loop, with steps towards a non-recursive walk of the stylesheet tree, or at least within the same xsl:template.
 discard d69dad4d Implemented idea by Michael Gratton <mj...@recalldesign.com> to hide all the raw byte writes in writers.  But the big deal here is that there is now a special UTF-8 writer.  I don't think it deals with surrogates yet, so there's a bit more work to be done. Also, there's some other fine tuning, to eek out a few more milleseconds in the benchmarks.
 discard 5f88395c Fixed really ugly dumb bug where, if the output stream was being copied, it would copy from itself instead of the old serializer, which would result in output going to System.out.  Amazing that I haven't heard of this before.
 discard 6b694d74 Folded in functionality from queued sax events directly into the class, for performance reasons.
 discard e56eddbb Added self pooling for iterators.  Instead of cloning all the time, the asNodeIterator can return an iterator from the pool.  This is a major performance improvement.  For the moment this is just used from the ElemForEach#traverse method, but it will be extended to be used universally.  Must be applied against the threading tests, since the pools themselves are kept in the static structure.
 discard 37f8be2f Made identity masks public, so that can be used for performance purposes.
 discard 1f91b74c Added peekOrTrue method.
 discard 0aaf11a6 setEnvironment deleted and passed to setRoot.  This helps with iterator clone pooling.
 discard 7461f11c Whups. Forgot to correct after debugging.
 discard b9be3442 Was invoking a Java 1.2 method...
 discard ecb54ae2 Minor performance tweaks for getNodeType. Minor tweaks for StepPattern#execute. For the moment don't use MatchIterator, but use WalkerIteratorSorted instead. Minor tweaks for DTMWSFilter to pass DTM. Direct call to get string characters for templates if only child is a text literal.
 discard 7cffc083 Are we officially abandoning VJ++?
 discard e7869a9a Implemented new isNodesetExpr and asNode methods on Expression, and getArg0AsNode and Arg0IsNodesetExpr on FunctionDef1Arg. This allows functions that just take a single node to get that node without creating an XNodeset or iterator if the expression is a simple one-step LocationPath (more complicated LocationPaths still need iterators, at least for the moment).  There will be more to come along these lines, trying to reduce the number of XObject derivatives that are are [...]
 discard f956fccc Fix  bug with findGTE where it wasn't properly checking to see if the low was higher than the end.
 discard fa4489ef Set dtm identity on node passed to m_wsfilter.getShouldStripSpace.
 discard 3261af9e Set the dtm identity bits on the way to getShouldStripSpace.
 discard 2c91a9f4 Wasn't setting the dtm identity for getNextIndexed return, which I fixed.
 discard b5a43c09 Fixed ugly bug that I introduced when breaking up the original array... I inverted a statement in nextNode. Also, increment m_size in addNode. In my d2d test, which already has a fair number of exclusions (which I should probably un-exclude for the new stuff), there were five failures, having to do with copying subtrees.  I'll look at these in a while.
 discard 88d597bb Element indexing in the DTM is now implemented.  It's only used for the moment with DescendantTraversers, and taken advantage of by the xpath DescendantIterator.  The signature for getDTM changed to add a boolean argument that tells if indexing is requested (we of course don't want to do indexing for RTFs, and there may be other cases where we don't want to do indexing).  The index lookup is done with a binary search.  We might be able to optimize this for short lists.   [...]
 discard c465ba18 Classes removed in favor of generalization of AxesWalker.  Yay!
 discard 8c3e96a0 This change seems bizarre, and looks like it could easily lead to an infinate loop, so I back out of it.
 discard 70352e63 Return "" instead of "null" for a null object.  (where did "null" come from?)
 discard e1680ee7 Backed out of change from main branch.  I already fixed the bug, and liked my fix better.
 discard 2ba32fc8 Think this file was overwritten from the merge with Myriam's stuff.
 discard 45fb4d8d Myriam's merges from the main branch.
 discard ef577d64 Added onestep iterators which use DTM traversers or iterators, and are applicable to any one-step path.  Added self iterator for "." (can't believe I didn't do this before).  Moved walker portion of LocationPathIterator to WalkingIterator, and made LocationPathIterator abstract.  Removed the waiting logic from AxisWalker (thank god and good riddence). Added many functions to WalkerFactory to determine the structure of the LocationPath, and added much logic to choose the  [...]
 discard a912d06c Foolish fencepost error.
 discard 07d36262 I finally got inverted select patterns passing for 100% of the test suite, except I have them turned off for namespaces, which simply can't be done because of our broken parantage of namespace nodes, and functions, which are easy enough, I've just had it for the time being.
 discard 15ce8983 Clarifying known-issue comments again.
 discard 398bff10 Fixed some bugs.
 discard ad68bc22 Added some missing files.  Sorry.  And some progress on match iterators, but still ongoing.
 discard 66a9bb75 fix access-level clash
 discard b2cd6a07 Moved DTM implementation classes down a directory, into a "ref" package (for reference implementation).  I wanted to clearly seperate the interfaces from the implementation, and make it clear which are which.  SAX2DTM and DOM2DOM went down a level also.  We might want to think about pulling them directly into the ref directory, since I don't think their packaging serves a real purpose at this point.
 discard a94e754c Javadocs. Some namespace issues need a closer look.
 discard d7f98375 Javadocs.
 discard 3bcf195a Slightly clarify architecture of AxisTraversers produced by DTMDefaultBase.
 discard 53d51c12 Clarify description of DTMAxisTraverser
 discard af1dc0fd Made the pattern execution classes fully general, i.e. not dependent on being the child or descendent axis (in reverse terms of match patterns, the parent or ancestor).  The patterns use the DTM traversers, and may be a tiny bit slower.  Also, non-absolute patterns (i.e. not '/' or functions) have the magic "./" added on front, which is needed to generilize the proximity predicates. Predicates are now executed *after* the relative step is executed, so that proximity may  [...]
 discard ffb3cc2e Documentation tweaks
 discard a534165c DTMNodeIterator -- DOM-level NodeIterator over DTM view of document
 discard 6301e380 Docs
 discard 75b372c0 Docs, flag a possible bug
 discard f2c7fd15 Changed the way you call a traverser, so that processSelf is no longer used, instead you call first(context), then next(context, current) for the next one.
 discard b6f8dada Introduction of stateless AxisTraversers, for polymorphic axis traversal, mainly for match patterns/match iterators.
 discard 9b2dd671 Maintaining these is too painful, and I would rather they just go away.  The less code the better.
 discard 31e10948 Don't really need this test anymore, and I would rather not maintain it.
 discard 736de64b Added getTypedAxisIterator and getAxisIterator for DTM.  Added DTMAxisIterator, DTMAxisIteratorBase, and Axes classes based on XSLTC classes.  Added AncestorIterator, AttributeIterator, ChildrenIterator, DescendantIterator, FollowingIterator, FollowingSiblingIterator, NamespaceAttributeIterator, NamespaceChildrenIterator, NthDescendantIterator, ParentIterator, PrecedingIterator, PrecedingSiblingIterator, SingletonIterator, TypedAncestorIterator, TypedAttributeIterator, T [...]
 discard 0fc7b939 Various tweaks for XMLStrings.
 discard 633265bd Too much trouble to maintain for now.
 discard d76004ee DMManagerDefault@getDTMHandleFromNode now tries to find an existing DOM2DTM which contains that node, rather than immediately generating a new one. The cost of searching for it is fairly high, though I've tried to avoid exhaustive search through every DOM2DTM.
 discard 8e0f618a I have implemented the basic XMLString architecture.  Only getStringValue returns an XMLString.  getNodeValue continues to return a string, especially since this is usually used with attribute values, which are already strings from SAX.
 discard 8066aeee Revised text-coalescence code, improved handling of CDATA Sections.
 discard d13684cb Enable redirect, and extensions for dtm-to-dom support.  NodeIterator support really hasn't been enabled yet.  And there is a significant issue regarding node-to-dtm on the way out from extensions, where we should really get the same DTM handle back.  A job for Joe! (see DTMManagerDefault#getDTMHandleFromNode).
 discard d7dcc17c Fixed ref to incrementalSaxParser.
 discard 5c2fecbb Added dtm-to-dom support
 discard d367d3a6 I don't believe this is any longer needed.
 discard 2192900a Updated CoRoutineSaxParser_Xerces so it no longer uses IncrementalXercesSaxParser. The DTMManagerDefault now uses CoRoutineSaxParser_Xerces if it can.
 discard 271061f4 Scott wants to be able to pass in an existing instance of Xerces. Until we get that one method made public, the best we can do is pass in an existing instance if IncrementalXercesSaxParser.
 discard 89df6e64 The Xerces-specific CoroutineParser now avoids the need for a second thread, by taking full advantage of CoroutineParser's encapsulation and Xerces' incremental parsing mode.
 discard a0447cce TransformerHandler now works perfectly.
 discard 79027767 The Xerces-specific CoroutineParser now avoids the need for a second thread, by taking full advantage of CoroutineParser's encapsulation and Xerces' incremental parsing mode.
 discard 9e912ce5 Filter-mode doMore(false) now runs the event stream to completion rather than throwing an exception. I do _not_ like this solution -- it wastes cycles -- but we've decided it would be impolite to do the (almost standardized) trick of using an exception to stop parsing early when we aren't the ones who could catch it and thus know to discard it.
 discard 7d4efef5 The TransformerHandlerImpl is basically working.  It will hang if you throw an error from the tranform... i.e. error handling TBD. You can test by -flavor th from the command line.
 discard 670c3df8 Added missing param to getDTM.
 discard 6db35681 New TransformerHandlerImpl class aggregates with DTM SAX2 event handlers, and is returned from newTransformerHandler.
 discard ed2beade Removed all dependencies of non-stree classes on stree.  RTF construction is now being done with the DTM.
 discard 87b0f39e SAX filtering test written. Mostly works.
 discard 66f34af2 SAX filtering mode written, confirmed not to break old unit test. Still need new unit test to check the new mode.
 discard 6cad0e09 Just comments
 discard 29d7836c Minor tweaking of calls to nextNode().
 discard b7951125 In getElementById, if the id does not yield an element, call nextNode() until either the id does yield an element, or nextNode() returns false.
 discard 51a2bbe0 Check for return of doParse to see if it's an exception, and, if it is call clearCoRoutine and throw the exception.
 discard 658b96ec Call clearCoRoutine in the release method.
 discard 0ce6041e Call mgr.release, so that the manager can call clearCoRoutine on the SAX2DTM object.
 discard 68e77158 > Inactive means no parse in progress. Boolean.TRUE is the parameter meaning we want more nodes. So the question is why we're telling the parser to send us more when the CoroutineParser thinks it has already sent us everything it has... or the parser thinks that it never got a request to begin parsing in the first place.
 discard c538b176 Added very basic CoroutineSAXParser functionality to DTMManagerDefault, and debugged to the point where basic transforms seem to be working.  This involved once again reworking the character handling a bit in SAX2DTM.  Also, I had to pass the appID to the SAX2DTM class, rather than have it be created in the class.
 discard 4e135edf Changed things so that callers use getNodeNameX instead of getPrefix for namespace nodes, to get the prefix being declared.
 discard 179144d5 Modified the whitespace stripping stuff a bit, so it's simpler, and now it actually works right.
 discard 5c230ffa Lurching toward support for the official startup sequence
 discard 22af03ee Links from SAX2DTM to CoroutineParser, for incremental build
 discard e5261a43 Encapsulate some of the coroutine details -- simpler API, and may allow single-thread implementation for Xerces.
 discard 338394da Implement getUnparsedEntityURI.
 discard 9e21e6a6 Fix for dispatchCharactersEvents and getStringValue, to fix termination of loop when iterating over element children.
 discard e7998b5d Added SAX2DTM class, as an alternative to the DTMDocumentImpl class.  Currently about 30 failures in the conformance suite. Also took non-DOM stuff out of DOM2DTM and put it in DTMDefaultBase, so it can share as an abstract base class between DOM2DTM and SAX2DTM. SAX2DTM doesn't have any CoRoutine support yet.
 discard fc4cdf68 New default DTM base, used as a base class for DOM2DTM and SAX2DTM.  Manages traversal, but not data storage.
 discard 8f9c71ec In the constructor, do stringToIndex("") to force empty strings to be index 0. This helps to quickly determine empty strings, and to make "" vs. null a little easier.
 discard 03560c8c Add setSize method.
 discard 35d56ec9 Fixes, including storing the prefix someplace useful.
 discard a5d8f7e1 Known issues:
 discard a324ba93 Sloppy merge. Mea culpa...
 discard c705c7a4 Remove debugging printouts
 discard 73ff6458 DTMManagerDefault now creates a DTMDocumentImpl, and the transformer is now hooked up.  No threading yet.  It crashes in DTMDocumentImpl pretty quickly.
 discard c32106dd Moved to top of src tree.
 discard 27d06b2d Functionality moved to DTMManagerDefault.
 discard 3e4518dd Moved from the DTM directory, since it is not packaged.
 discard 19ce7939 Clean up SAXException handling in the test drivers
 discard fe156ba2 Passes the conformance suite now (at least on my machine, which hasn't had the suite updated for a month).  Many bug fixes.  Anything that requires DTM2DOM still won't work, such as extensions and the TransformState stuff.  DocumentManagement and launching DTMs correctly is next on the agenda.  Note that right now you have to use -Dorg.apache.xml.dtm.DTMManager="org.apache.xalan.stree.StreeDTMManager". DOMSources that hold generic DOMs won't work yet.
 discard ca92504b Moved most of the DTMBuilder logic into DTM itself, as support for sax.ContentHandler interface. Moved code from incomplete implementation of (deprecated) sax.DocumentHandler into this code. Updated TestDTM to build via ContentHandler API. (ComposeDTM not yet updated.)
 discard 71041d9d Stubs for coroutine API encapsulation; need to be fleshed out.
 discard a75ddacc Merge localname tables
 discard 1b436edf Display some DTM API-level information for each node.
 discard 5842bc83 Rough merge and sanity check done. Now to bring it up to date and integrate it with the builder logic.
 discard 4c9e9164 Merge updated code, in progress. Some errors fixed (ands written where ors were intended), some concerns annotated, empty glue routines dropped in pending reconcilliation with the DTMBuilder stub. I've got a bit more to do in the way of fast-once-through sanity checking before attempting that. (This is _not_ a full code review pass. For now, I'm assuming that most of the routines work, or don't work, as documented.)
 discard 538889a1 Merge with updated code, in progress. Still need to reconcile/unify DTMDocumentImpl work with DTMBuilder work, and tie it into the CoroutineParser/incremental-parser hooks.
 discard 4e1d4bc5 Fixed getLastChild and getPreviousSibling.  The run through the test suite produces 55 fails against 1199 tests.
 discard 6c809e7c Added %TBD% marker that I had forgotten.
 discard 19b788c6 Periodic checkin.  Result tree fragments now work, as well as the document() function.  idkey01 goes in infinite loop due to xsl:number issue.  You must use -Dorg.apache.xml.dtm.DTMManager="org.apache.xalan.stree.StreeDTMManager".
 discard fa204a2f Moving the Stree knowledge for the DTM down to this derived class. Use -Dorg.apache.xml.dtm.DTMManager="org.apache.xalan.stree.StreeDTMManager" to enable.
 discard 757dcb21 New interface for allowing the DTM to ask the caller if it should turn whitespace stripping on or off, or inherit the behavior of the containing element.  This will be called for each element.
 discard eaf8d948 Periodic check-in.  Basic axis tests (1-113, haven't been updated for for a month or so) run correctly with DOM2DTM on top of Stree. Should test axis traversal, some positional tests, strip-space, basic match patterns, and rough tree management.  I don't think result tree fragments, and multiple documents will work yet.  Also, extensions are not yet operational.
 discard 6b987656 DTMBuilder to DTMDocumentImpl connections established.
 discard a76b3f7b Copy ChunkedIntArray into this package (from Xalan1 version of DTM)
 discard d4397313 Copy ChunkedIntArray into this package (from Xalan1 version of DTM)
 discard 1c39887f Correcting bad merge.
 discard 29e2ff90 Whups. Some problems merging two separate sets of changes. MERGED.
 discard 869d2f74 Whups. Some problems merging two separate sets of changes. MERGED.
 discard e3b67b6a Whups. Some problems merging two separate sets of changes. NOT STABLE. IN PROGRESS.
 discard 6f8a7f72 Whups. Some problems merging two separate sets of changes. NOT STABLE. IN PROGRESS.
 discard e09f7d29 Yih-Shin Tan's current draft of a DTM implementation, together with a pair of test harnesses.
 discard 5834f5aa These were supposed to have been moved to ..\dtm\ directory. Their package statements say they were. I'm not sure how CVS wound up putting them back in utils.
 discard 90cc5a58 Many changes related to supporting the DTM Builder. String pool and coroutine parser layers now have built-in unit-test drivers; see their main()s.
 discard 996e0134 Built-in unit test (rough, but sufficient as a sanity check)
 discard 37f6d101 1:1 int-to-string-to-int map, used to hold semantically meaningful (and probably frequently reused) strings in the DTM model.
 discard d806c890 Periodic checkin.  Basic traversal works now.
 discard 9c8e4328 Minor reorg for better optimizability, improved safety nets, doc tweaks
 discard 23d6cf9b UnitTest for DOM2DTM.
 discard 13f6e750 Periodic check-in.  Class is almost functional, but has bugs which I am working on.  Some methods have not yet been implemented.
 discard bef98bd8 Added ExpandedNameTable (the class is currently non-functional) and DTMManagerDefault.
 discard 29a77a71 Fixed getDocumentTypeDeclarationPublicIdentifier to return a string.
 discard f82bf6b4 Added setContentHandler, and changed getcontentHandler to getContentHandler.
 discard 3056ebc2 Added addElements(int value, int numberOfElements), addElements(int numberOfElements), and fixed bug with setElementAt.
 discard 7919f3e9 Added peek(int n).
 discard 5081fd2f Added getDTMIdentity.
 discard e735ea7e Fixed call to getAttributeNode, which changed in the DTM interface.
 discard f276a9bf First draft of wrapping a CoroutineParser around a normal SAX stream. Good illustration of why coroutines are a useful concept -- they allow pausing the parse thread at an arbitrary location deep in the call stack, if that happens to be the best time at which to yield control.
 discard 17b6d617 Abstracting the CoroutineParser to an interface. May wind up with a CoroutineSAXParser layer as well. One initial implementation, demonstrating the Xerces parseSome() approach; Need to do an equivalent for pure SAX; that will be messier.
 discard 327fe1f3 First draft of a coroutine control for Xerces incremental parsing -- thanks to Glenn Marcy
 discard 642e2721 Missed a "public". Oops
 discard 0f6c9c7a In progress check-in, so people can see how this is progressing.
 discard c7a3fe74 Added getDTMIdentity, getDTMIdentityMask, getNodeIdentityMask.
 discard a1c7a9cd Fixed getAttributeNode.
 discard 6f8d57bf Empty shell so far.
 discard 19e97306 Startup is still not solid...
 discard 6e7672e8 In progress. Looks workable, maybe.
 discard 820c6532 Added newInstance() method and support.
 discard 796f9387 Added exception classes modeled after javax.xml.transform's exception classes, but these derive from RuntimeException.
 discard 98dbf4ea Javadoc thoughts misplaced during the move from the utils directory to this one
 discard 06e60b7d Minor javadoc tweak
 discard 3f272a3d Branch now compiles clean.  Unfinished code is marked by %TBD%, %OPT%, or %REVIEW%.  DTM stuff moved to org.apache.xml.dtm.
 discard f758a673 Branch now compiles clean.  Unfinished code is marked by %TBD%, %OPT%, or %REVIEW%.  DTM stuff moved to org.apache.xml.dtm.
 discard 2a7e43ee Javadoc
 discard f8458505 Javadoc
 discard de466d88 Javadoc clarifications/issues
 discard 1f12b440 Javadoc clarifications/issues
 discard 0e75d597 Javadoc clarifications/issues
 discard 8c9779d3 Javadoc tweaks
 discard c2211699 Javadoc tweaks, a few typos, one return type
 discard ab9f08d0 Javadoc tweaks
 discard b1be2d73 Javadoc tweaks, mostly just clarification.
 discard 2d466de5 Javadoc tweak
 discard c8cff3fc Added createDocumentFragment() (for creating RTFs in Xalan), added experimental release method, and added createDTMIterator for creating simple iterators that only represent one node.
 discard c8a59f41 Added getDTM(int node) and getDTMManager().
 discard 0ca6da68 Added construction methods, added getDocument and getOwner document.  Defined getPrefix to return the prefix being mapped for namespace nodes, but am not comfortable that this is the right thing to do.
 discard 8c974039 TreeWalker for walking the DTM.  I'm actually not sure if we'll need this much since there is the direct SAX dispatch from the DTM interface.
 discard 03d49988 Changed definition of node handle to include document identity.  Also, made scheme of expanded name opaque.
 discard dd192e8a Non-functional state.  Initial adaption to DTM.
 discard 0e046218 Added supportsPreStripping, isSupported, getNodeType, and NULL.
 discard fc4c8e47 This commit was manufactured by cvs2svn to create branch 'DTM_EXP'.


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@xalan.apache.org
For additional commands, e-mail: commits-help@xalan.apache.org