Change log Fixed bug in ExpressionParser to handle case when a function is missing the closing ')', instead of crashing with a NullPointerException (reported by Warner Onstine) Fixed a some error reporting issues (reported by Pascal Belloncle) Add a few minor performance improvements Fixed thread safety problem with sharing the DOMParser (reported by David Blondeau) Other miscellaneous changes have been commited over the last serveral months. Fixed whitespace bug in StylesheetHandler (reported by Peter Armstrong). Fixed a new bug in PathExpr which was a result of doing some peformance improvements, A null NodeSet was being returned under certain circumstances. (reported by Assaf Arkin) Fixed bug in FormatterAdapter (Jacques-Alexandre Gerber) In the constructor the passed in DocumentHandler was not getting set. Some more performance tuning...more to come. did some minor performance tuning changed one of the values of the level attribute for xsl:number from "multi" to "multiple" upgraded xsl:use-attrubute-sets element to the xsl:use-attribute-sets attribute added the current() function call added some more namespace support fixed some bugs with the expression parser and lexer updated xsl:pi to xsl:processing-instruction Fixed bug with lexer, wild card was not properly handled when it appeared directly after '/' (reported by Jason Mathews) Upgraded the Axis Indentifier syntax to the latest spec (sorry for the delay) Fixed top-level parameter support Changed the XSLReader to add SAX support Added internal SAX support for efficiency Updated variable, param, with-param to use the "select" attribute instead of the "expr" attribute Changed param element to with-param, changed param-variable element to param separated out XPath into it's own package look for a new XPath package that supports the proposed recommendation soon. fixed bug with DefaultFormatter that was causing Triplicate output submitted by tons of you! ;-) sorry for the delay in getting this fixed added support for xsl:output from the latest draft supported attributes are: doctype-public doctype-system indent method version fixed bug with xsl:entity-ref introduced in a recent version Added fix for XML4J 2.0.15 (Jason Mathews) Added fix for XMLPrinter with respect to " inside attribute values (Donald Ball) fixed source code location of SimpleMessageObserver for proper compilation added MessageObserver interface to handle xsl:message added SimpleMessageObserver to print xsl:messages to a writer created a default MessageObserver, which will just dump messages to the console. The default MessageObserver can be removed passing XSLProcessor.DEFAULT_MESSAGE_OBSERVER, to the XSLProcessor.removeMessageObserver() method. fixed bug pertaining to Entity and DocumentType Nodes When using entities, XSL:P did not filter Entity and DocumentType nodes, from the DOM tree fixed bug with xsl:number with level set to any, to many counts were being generated, instead of a list of length one, containing one plus the number of nodes, a list of length NodeSet#size was being created with each node count. (reported by Jason Crickmer) fixed whitespace bug, when whitespace appears before the document element, under certain circumstances an <xslp:result> wrapper element would be insterted.(reported by Eric van der Vlist) added support for global variables that reference other global variables not yet declared. fixed a bug with processing global variables. I am storing the global variable declarations in a hashtable. This could cause the variables to be processed out of order, since hashtables are not ordered. So if one variable referenced another one, before it was defined, the result was "undefined". I fixed this, by adding in the required support for out-of-order processing, which I had not yet implemented. thanks a bunch to Roger for enabling me to find this bug. Fixed bug with operator precedence (submitted by Donald Ball) Fixed bug with SelectExpr.java (submitted by Roger Costello) patched bug caused by XML for Java 1.1.x. Attr nodes with the same name, hash to the same value. This appears to be fixed with IBM's 2.x version This causes the document order of attributes with the same name to be the same, and therefor causes improper document order sorting, and the same id was being generated for these nodes. Added Michel Casabianca's additions to HTMLPrinter for more HTML elements that shouldn't print end elements fixed bug with XML4J_1xParser and using a custom EntityResolver Unfortunately their SAXDriver class wasn't working, so I had to implement StreamProducer myself. fixed bug with '.' in CNames the result of a bug report submitted by Dan Potter fixed bug when the result of a predicate is a Number, sometimes this wouldn't use the default comparison of that Number with the current Node's position within the current NodeSet added catalog loading to some of the XML parsers code sent by Michel Casabianca Only instantiate XML parsers once inside the DOM wrapper, if they can be reused Changes effect XML4J 2.0.x parsers, and Sun's ProjectX parser code from Michel Casabianca Made some changes to the XSLReader class for performace improvements, and simpler design Removed AVTComponent class, now that we have different types of expression values, this class is no longer needed Removed StringTokenizer from AttributeValueTemplate parsing changed to use parsing code from MITRE's TransforMiiX (TM) processor, a C++ XSL processor which I wrote for MITRE and will be released soon Fixed a bug I found with the ExprLexer added wrapper element for stylesheets that add text nodes before adding a document element fixed bug introduced in 19990729 build with ProcessorState#findParent (submitted by Matt Hoskins) added generate-id() fixed bug with VariableReference names longest token wasn't being returned by ExprLexer if a '-' was encountered in the name submitted by Edas Pazera added default-space functionality...yikes, sorry I forgot about this thanks for pointing this out, Roger (Costello) added the xsl:cdata proprietary element back into XSLReader updated XSLProcessor to make appropriate calls to XSLReader#setErrorStream (submitted by Michel CASABIANCA and Mohan Embar) fixed bug with NodeSet#copy did some minor performance adjustments added format-number function uses default locale, does not yet support the optional locale argument added substring() function for use with XSLT WD 19990709 (XPath) added string-length function (XPath) currently doesn't handle XPath Unicode surrogate pairs added doc() and docref() function calls ( fragment identifiers not yet handled also mapped these to document() function call for use with XSLT WD 19990709 second argument (used for BaseURI) not yet supported changed the behavior of the HTMLPrinter to not use the empty element shorthand (suggested by Donald Ball) fixed null pointer bug with RuleProcessor#isStripSpaceAllowed (submitted by Ray Powell) fixed bug with NodeSet.java (submitted by Frank Peters) prevented NodeSet from growing beyond default size fixed bug with ExpressionParser.java (submitted by Rolande Kendal) prevented absolute path expressions from being evaluated properly added lang(), qname() and function-available() functions added ns attribute support to xsl:functions fixed bug with comment() (submitted by Roger Costello) fixed bug with parent axis (submitted by Xavier Faz) changed default format for xsl:number to always be 1 (Stefan Mintert) added xsl:copy-of fixed package name for NameSpaceUtil (submitted by Xavier Faz) added ceiling(), round(), floor(), and sum() functions added system-property() function finished Axis Identifier support changed main package from "com.kav.*" to "com.kvisco.*" updated xsl:number to support the expr attribute added xsl:param and xsl:param-variable support fixed a bug with parsing union expressions added local-part function call started adding namespace support Added complete scoping for xsl:variable Did a major overhaul of the ExprLexer (was called ExprTokenizer) Did some clean up on the ExpressionParser Added concat(), contains(), starts-with(), substring-after(), substring-before(), and translate() functions Added top-level xsl:variable Added additional functions Added Mulitplicative and relational Expressions Added simple xsl:functions mapping to xsl:script Added position() function Removed Counter related code Removed Macro related code Integrated Peter Ciuffetti's Named Template code STARTED SUPPORT FOR XSLT Fixed bug with xsl:include submitted by Kelly Holford Fixed bug with ExpressionParser discovered by Shane Hathaway Added fix from Shane Hathaway for XSLImport#getDocumentBase fixes a problem with mixing "/" and "\" file separators Added support to the xslp.properties file for declaring Formatters based on code from Franck Mangin Added META to the list of elements not to print ending tags for when HTML is specified. based on feedback from Ian Davis Added support for addition URLs other than http and local files suggested by Calle Wilund Added Attribute Value Fix based on code submitted by Shane Hathaway Fixed bug with xsl:copy where attributes were automatically getting copied when the context node was an element This bug effected releases 19990326 and 19990406(a and b) discovered by Stefano Bovone Fixed bug when using xsl:counter within xsl:element This bug effected releases 19990326 and 19990406(a and b) discovered by Aaron Metzger Fixed bug with priority This bug affected releases 19990326 and 19990406(a and b) When the result namespace is HTML, special characters within CDATA sections are no longer mapped to their respective xml entities suggested by a few people fixed xsl:counter bug discovered by Aaron Metzger fixed xsl:constant bug that effected releases 19990326 and 19990406a discovered by Tim McCune fixed XSLReader() default constructor to initialize the PrintWriter it used for error reporting discovered by Tim McCune fixed the xslp.jar file to include the updated xslp.properties file discovered by too many people fixed a bug with SelectExpr A node could be selected twice discovered by Stefano Bovone Added XSLProcessor#setValidation to turn on/off xml validation Validation is turned off by default Added -val switch to command line to turn on xml validation suggested by Olivier Brand Added better whitespace handling Added indent-size to the xslp.properties file suggested by Aaron Metzger changed the package for DOMPackage implementations to xslp.xml.parsers Added some basic pretty printing for indent-result="yes" Core classes no longer extend DOM Elements I made this decision so that it would be easy to support any DOM package They also run faster now Added the xslp.properties file to the jar archive this file is used to specify XSL:P properties currently only the DOM package information is specified here Added support for OpenXML Added support for XML4J 2.0.3 and 2.0.4 Added support for Oracle XML Parser XSLProcessor#process methods that used to return NodeList now return a DOM Document if you need the NodeList use the getChildNodes method from the returned Document Formatter now takes a Document instead of a NodeList as input sorry it took me so long for this change, I should have done this sooner Added XSLProcessor#process methods that take a Stylesheet as a DOM document. Many changes to the internals of XSL:P (I hope I didn't cause many bugs! ;-) Added indent-result attribute to xsl:stylesheet doesn't really indent yet but add's line breaks if set to "yes" removed default behavior of adding line breaks Added fix from Peter Marks and David King Lassman for "Not" printing line breaks when the first and last child of an element is Text Fixed bug in Absolute MatchExpr pointed out by Wayne Grunes Added fix from Majkel Kretschmar in Whitespace.class Added fix from Majkel Kretschmar for proper UTF-8 encodings Added Warren Hedley's fix for no stylesheet being specified or possible error when opening the file Added feature for dumping to StdOut if no output file is specifed Warren Hedley Fixed SelectExpr for ancestor expressions class archive is now a jar Fixed bug with error printing Fixed bug in BooleanTokenizer keywords (and, or) were not being ignored inside of quotes Added changes from Domagoj Cosic for compatibility issues between XML4J 1.1.9 and XML4J 1.1.14 includes performance enhancements I've finally done some performace tuning, you should notice a significant speed increase for larger files. More optimizations to come. Added support for namespace quoting Updated ScriptHandler to allow scripting access for inserting nodes into the result tree besides text nodes: Three API methods were added to the scripting environment XSLP.addToResultTree(Node),XSLP.createElement(name),XSLP.createText(data) Updated DefaultFormatter this Formatter is used by Default when no Formatter is specified and the result-ns is not equal to "html" or begin with "http://www.w3.org/TR/REC-html" all empty elements are printed using the empty element convention Added BasicHTMLFormatter this Formatter is used when the result-ns attribute equals "html" or begins with "http://www.w3.org/TR/REC-html" this Formatter does not print closing tags for <BR><IMG> and <HR>. If a <P> element has no children it will not print the closing tag Added Numbering in the result tree xsl:counter xsl:counters xsl:counter-increment xsl:counter-reset xsl:counter-scope Note: for Number formatting, only the format attribute is currently implemented Fixed a bug when using pi() and text() Match expressions at the same time Added some more fixes fixed a problem with context for literals fixed xsl:attribute problem with whitespace Added some whitespace handling features xsl:strip-space, xsl:preserve-space Made some enhancements Fixed some of the document ordering when using UnionExpr select patters still need to fix out of order attributes fixed a bug with IdentityExpr with FilterExpr (thanks Aaron) added xsl:sort!!!! uses jdk 1.1 Collator so only languages supported by jdk1.1 are supported fixed a bug with "importance" fixed a bug when selecting on the document root Added xsl:include support Made some minor modifications to the XSLReader and XSLStylesheet classes Added macro-arg support Changed DefaultFormatter to print <?xml version="1.0"?> Added support for <?xsl:result-dtd ?> processing instruction This is proprietary pi that allows the DOCTYPE to be added to the result document added xsl:script support as children of xsl:attribute added support for xsl:entity-ref which allows the creation of an entity reference in the result changed the CDATA section serialization to replace the entities & < > incorporated new XslNumberFormat from Tim Added support for multi level and any level numbering Incorporated support for number formatting uses Tim Taylor's XslNumberFormat Fixed a bug I ran across in MatchExpr The // operator wasn't always match correctly when used in a MatchExpr. SelectExpr's were unaffected. I may have broken this during my change from the 19980818 draft. Added boolean expressions AndExpr, OrExpr and BooleanGroupExpr Updated scripting support Did a little fine tuning, no feature changes Added more of WD19981216 changes added pi,comment, and text functions added xsl:copy I have incorporated many of the WD19981216 changes changed xsl:process and xsl:process-children to xsl:apply-templates added creation of XML nodes (comment,pi,element,attribute) changed Select patterns and Match patterns to support new syntax changes changed attribute to @ added processing modes changed expr attribute to select for xsl:value-of added ancestor-or-self changed priorities to be a real number xsl:text has been added Added xsl:choose Added documentBase fix provided by Tim Taylor Added Macro support Added proprietary support for xsl:script Scripting language is ECMAScript (Uses FESI ECMAScript Interpreter) Added Attribute Value Templates Added xsl:if Added Anchors to xsl:value-of Added xsl:value-of