Oxygen is a good XML editor, XSLT debugger, and XML file/dir diff.

Setup Oxygen

• Options->Preferences->Editor->Format -- Uncheck Detect indent on open. Set tab-spacing as desired.
• Options->Preferences->Editor->Format -- Check "Hard line wrap...".
• Options->Preferences->Editor->Save -- Check "Enable automatic save"
• Options->Preferences->Editor->Save -- Check "Check errors on save"
• Options->Preferences->Editor->Spell Check -- Check "Automatic spell check" and "Spell check entire document"
• Options->Preferences->Editor->Spell Check -- Check "Apply spell checking on entire document" (slow for large documents)
• Options->Preferences->Editor->Code templates -- Add code fragments that you use all the time. To insert a fragment, type the template name followed by ctrl-space. Oxygen predefines templates for XSD and XSL. Periodically export your code templates. For example, define a template, 'ul', for a list and 'li' for additional list elements (insert caret on the first blank line).
• Project->New -- create a 'Oxygen' project for frequently accessed files. Add an XML file to the project by opening the file then clicking Add Edited File in the project panel. Save to your config directory. It will reopen whenever Oxygen launches.
• Perspective->Show view->Project -- also Outline -- you can edit via the outline view.
• Periodically, close and reopen Oxygen. It has some memory leaks that degrade performance over days or weeks.
• Be careful of rebooting without saving Oxygen first. It appears to ignore the shut messages.

Saxon is an XSLT and XSLT 2.0 processor

• Copy saxon8.jar from home page, or Oxygen/jre/lib. The directory name can not contain spaces
• java -jar dir/saxon8.jar [options] source-document stylesheet [ params...]

Some notes

• col -- Firefox doesn't appear to fully support the col element. It recognizes width, but not text-align. Neither browser supports colspan="0" to span to the end of a colgroup.
• < -- To output entity references with XSLT, use <xsl:text disable-output-escaping="yes">&gt;</xsl:text>
• XHTML does not include the entity reference   for non-breaking space. Nbsp is useful for compound words and explicit control of spacing. Add the following DTD to your header. Other useful entities include copyright (#169;) and mdash (#8212). <!DOCTYPE root [ <!ENTITY nbsp "&#160;"> ]>
• External entities allow easy inclusion of one file inside another file. <!DOCTYPE root [ <!ENTITY chapter3 SYSTEM "chapters/chapter3.xml"> ]> ... &chapter3;
• To automatically produce character references for multi-byte characters, specify ASCII as the output encoding (<xsl:output ... encoding="US-ASCII" ...>)

• IEBlog on script debugging -- good introduction to debugging IE scripts [wwong]
• IE requires the closing script element (</script>) for Javascript.
• Microsoft Script Debugger -- Standalone debugger if you do not have DevStudio. In any case, you need to uncheck Tools->Internet Options->Advanced->Disable script debugging

IE has a nice integration with Dev Studio for Javascript debugging.

• Script Explorer -- Open Debug->Windows->Script Explorer ("Running Documents" in DevStudio 2003). If it is not available, use Tools->Import and Export Settings to change your settings to Web development.
• IE->View->Script Debugger->Break at next statement -- On break, select your debugger. DevStudio will attach to the html file.
• DevStudio attach debug -- Tools->Debug Processes->select IExplore->Attach -- After attaching, Script Explorer should show your IE page. If not, close IE and start over. You can set breakpoints in the Javascript code, view variables, etc.
• DevStudio edit/debug -- Stop the debugger, edit the Javascript file, refresh the browser, reattach the debugger.
• DevStudio missing step-into-- If DevStudio does not show "Step into", select from Toolbar->Add/Remove buttons->Customize->Commands->Debug.
• [Sept 2006 cbb] -- The interface can get hosed. Reboot?

• CSS formating -- Consider using the following formating style (Style Master -- File->Edit Options->Formating)

    h1 {
                  text-align: left;
                  margin-top: 20px;
    }
        

• stylin' CSS -- Wyke-Smith's book stylin' CSS is an excellent introduction []. His web site contains templates and links. A good starting point.
• Get rid of the default browser settings by starting CSS files with body {font: 1em verdana, arial, sans-serif;} * {margin:0; padding:0;}
• Debugging CSS -- Use the Web Developer plugin from . View style information shows the class hierarchy and active definitions for each element.
• res/html.css -- For a quick introduction, look at the default style sheet for html formating in
• IE6 quirks -- Floating divs not cleared correctly http://www.positioniseverything.net/easyclearing.html

XSLT is an XML standard for transforming XML documents into XML, HTML, and text. ,

To quickly learn XSLT, read followed by skimming . To learn XSLT is a production environment, see

XSLT 2.0 has many useful extensions, see . Each XSLT processor implements various extensions, see EXSLT for a repository.

For .NET programming, see

• XslTransform threading. XslTransform is thread-save only during transform. Load() and other methods must be single-threaded.

XSLT 2.0 is a major upgrade of XSLT. See , , and .

New features in XSLT 2.0

• Substituition group -- Write templates for elements that do not exist yet
• Typed variables and parameters -- Use the as attribute to declare the data type of a variable. The basic types (e.g., xs:string) follow XML Schema.

If the input document defines a default namespace, the XSLT must assign a namespace prefix for its XPATH expressions. For example, an XHTML document requires <xsl:stylesheet ... xmlns:xh="http://www.w3.org/1999/xhtml">. Alternatively, in XSLT 2, define xpath-default-namespace="..."

• Does the element contain text? -- normalize-space(text())!=''
• coalesce -- For single element values, the equivalent of coalesce in XPath is ($valueA, $valueB)[1]

To associate an XML schema file, you need a schema name, a file path, and a namespace name. The schema name and namespace can be the same. For example, this file has the following root element:

<rf:topic xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
xsi:schemaLocation="http://schemas.roadintranet.org/road-faq-1 ../xsl/road-faq.xsd"
xmlns:rf="http://schemas.roadintranet.org/road-faq-1"

Use HTML Tidy to convert raw html (e.g., with unclosed <p> and <li> tags) to xml. tidy file.html >file.xml

• Does the element contain text? -- normalize-space(text())!=''
• coalesce -- For single element values, the equivalent of coalesce in XPath is ($valueA, $valueB)[1]

SQL Server 2000 introduced the 'for xml' output format for SELECT. SQL Server 2005 adds a number of improvements

• To output XML -- Prefix the SELECT command with :XML ON and use sqlcmd to run your query. You must use a script file (sqlcmd assumes an interactive session if stdin contains multiple lines) For example

     :XML ON
     select * from testcase_builds where build_id='2005_11_01-095320' 
     and broke_prev is not null and fixed_prev is null
     order by testcase_id, machine for xml auto, root('rows')
     go
  

XML can represent information as attributes or elements. Attributes are more limited than elements. Attributes can not have children or sub-attributes. Attributes are unique within an element. Attributes are unordered. Attribute values may not contain a new-line.

Consider using attributes for metadata, identifiers, numeric data, and URLs. Use elements for human readable text and data.

XML uses UTF-8 by default. For an explanation of the various character sets, including ASCII, HTML, and UTF, see

Cocoon is an XML web site engine that dynamically transforms XML data into web pages. See for an introduction and sample Cocoon application.

A data island is an <xml> element within an HTML document. Microsoft's InternetExplorer can parse data islands with its MSXML parser.

DocBook is an XML standard for technical books and documentation. It is a "rich source of text markup and structuring wisdom"

DOM is an in-memory, document object model for XML documents.

Using naming to communicate the data's structure. Avoid abbreviations, prefer lower-case, hyphenate names, reuse names such as "head"

FOR XML is a Microsoft SQL Server extension for generating XML results from SQL queries.

OPENXML is a Microsoft SQL Server extension to insert rowsets from an XML document. Use it for small documents up to 50 KB in size.

XML Bulk Load is a COM component for loading database tables from XMLM.

NITF is an XML standard for news stories. It can define "who the story is about, when and where the described event occurred, and even why it is considered newsworthy by the story author

SVG is an XML-specifiction for vector graphics. Batik is Apache's Java implementation of SVG. Also consider Adobe Illustrator, Inkscape, and Imagemagick PfaEdit is a good font editor for SVG fonts and other font formats.

An identifier servers multiple roles. The primary roles are reference and summarization. Consider providing multiple names for the same entity. Use a short name for internal reference, a short title for textual references, a URL for external reference, a descriptive title for headers and table of contents, and a full title for archival purposes. To simplify textual references, convert references to lowercase and strip all whitespace and punctuation. See Kirsanov for further advice and

SQL 2003 included XML support, commonly called "SQL/XML". It appears that the ISO Standard has been withdrawn. Individual vendors have implemented different approaches.

• SQL/XML for Oracle -- introduction to using SQL/XML
• SQL/XML for SQL Server -- SQL Server has a proprietary XML support, also called "SQL/XML" but unrelated to the standard.
• XQuery is used for querying XML in a database.
• OpenXML -- Read xml files for querying as a table
• for xml -- Produce results of a query as XML (e.g., select ... from ... where ... for xml raw('row'), root('builds")
• SQLXML Bulk Load -- XML Bulk Load imports an XML document according to an annotated XSD schema. It is as fast as using bcp. For documentation, see SQLXML->bulk loading.
  • Install SQLXMLBULKLOADLib -- Install sqlxml.msi from Microsoft. It is a download for SQL Server 2000. SQL Server 2005 includes SQLXML4.msi, but it does not appear to work as a independent installer.
  • How bulk load populates the database -- See SQL Server 2000 documentation Record Generation Process and the Interpretation of Mapping Schema
  • Example C# code -- See Bulk Loading XML (Centricity) and SQL Server documentation Using SQLXML Bulk Load in the .NET Environment
  • sql namespace -- Add the following namespace to <xs:schema>, xmlns:sql="urn:schemas-microsoft-com:mapping-schema"
  • Annotated XSD Schemas -- Documentation for annotated schemas in SQL Server ... Books Online. The index page is Using Annotations in XSD Schemas
  • sql:relation="..." -- Defines the table corresponding to an xs:element
  • sql:field="..." -- Defines the column corresponding to an xs:attribute or xs:element
  • sql:key-fields -- Each table (sql:relation) should specify its key-fields. It may be required for tables using sql:relationship. Not clear if this is really needed. See SQL Server documentation on Identifying Key Columns Using sql:key-fiel
  • sql:limit-field, sql:limit-value -- selects rows for sql:relation where field=value.
  • sql:relationship -- Defines joins between tables. See SQL Server documentation on Specifying Relationships Using sql:relationship
  • Error: Invalid character value for cast specification -- The data format is incorrect For example, uploading a string field into an int column. SQL Server does not report the column. For example, "T" formatted datetime may require a type annotation, sql:datatype="dateTime". The same error message occurs for SQL bulk load. In older versions of SQL Server, it may indicate missing columns.
  • If adding or modifying data table columns -- As with bcp, XML bulk load can mis-report data typing errors. To fix, delete and recreate the data table.

XForm is an XML-based processing framework for displaying, submitting, and processing forms.

XLink is an attribute-based syntax for attaching links to XML documents.

XML is a document markup standard, based on SGML.

Use XPath to return node sets from an XML document. ,

For .NET programming, see

XPathDocument is an optimized DOM model for processing XPATH expressions. Use CreateNavigator() to create an XPathNavigator object. . These objects implement the IXPathNavigable interface

XPointer allows XPATH references in URLs.

XSD is the prefered XML schema format. It replaces DTD and XDR.

For programming .NET with XSD (e.g., SOAP) see

• Preload schemas with XMLSchemaCollection.
• XSD types map to .NET types
• Use xsd.exe from XML to XSD to C#
• Use an external reference to the XSD file

XSL-FO is an XML syntax for describing page layout. Toot-o-matic uses XSL-FO and FOP to generate PDF files from XML .

CSS is a non-XML syntax for page layout . It works well with XSLT.

A master document is an XML configuration file that defines a web site, configuration, and site directory. Site documents use the master document to determine their context within a web site and to establish mutual links.

Use the master document to associate short mnemonic identifiers with full titles and URLs . Use the master document to define the base URL for images and links . Use the master document to define the configuration for each deployable environment (e.g., staging, production, development, QA) .

See for a detailed example.

.NET allows embedded scripts inside XSL files. msxsl:script supersedes their xsl:eval extension.

MSXML is an older COM module for processing XML. It includes support for XSD, XML Schema and XPath.

XSLT templates may be invoked directly (push processing, e.g., xsl:for-each) or indirectly (pull processing, e.g., xsl:apply-templates). A similar distinction is trunk templates that define the highest level of document and branch templates that define the leaves. Push processing is suitable for trunk templates while pull processing works well for branch templates .

An XML entity names a string value, e.g., &author;. It is defined within a !DOCTYPE node, e.g., <!DOCTYPE book [<!ENTITY author "Dino Esposito">]> It requires a XmlValidatingReader.

SAX is an event-based API for XML documents . The model appears to be losing favor, e.g., .NET uses instead of SAX

SOAP is an XML standard for invoking web services.

For programming SOAP in .NET see

NameTable stores atomized strings for XMLDocument and XMLTextReader. Much faster than string comparisons.

.Net offers a variety of XML-enabled classes:

• DataSet -- a memory-resident database format with tables, columns, rows, and relations. Useful for mirroring database data. Not useful for business rules.
• DiffGram -- XML update representation for a DataSet.
• Updategram -- another XML update representation for a DataSet
• Web.config, AppSettingsReader -- application-specific configuration settings. Prefered over using the registry.

.NET Remoting is a non-XML alternative to SOAP. It is highly optimized for .NET-to-.NET communications, and replaces DCOM.

XML Document is Microsoft's DOM, memory-based representation of an XML document or fragment. It is a hierarchical structure of elements, attributes, and other nodes. See the .Net class XMLDocument. It is used for:

• Random access
• Repeated access to the same XML
• \nComplex transformations using XSLT and XPATH
• Like a DataSet
• XML DOM is overkill for writing XML.
• Avoid calls to PreviousSibling, Item, and Count. They walk all subnodes of a node. Use NextSibling instead.
• The corresponding XMLReader must be positioned at an element instead of an attribute.
• Use XmlNodeReader to process a subset of nodes.

An XMLReader presents a stream input as a sequence of nodes. Use XMLReader for

• Process XML in sequence.
• Process of lots of XML
• Like a DataReader
• A TextReader is not thread-safe.

An XMLWriter writes stream output as well-formed XML.

• Does not check for invalid names.
• XMLWriter's maintain a stack of unclosed elements with the associated namespace

Meyer, O'Reilly, 2004, ISBN 0-596-00777-9

Flanagan, O'Reilly, 2003, ISBN 0-596-00411-7

Excellent coverage of JavaScript 1.5 by the author of The Definitive Guide

Flanagan, O'Reilly, 2001, ISBN: 0-596-00048-0

JavaScript reference. Well regarded.

Goodman, O'Reilly, 2003, ISBN 0-596-00467-2

Wyke-Smith, New Riders, 2005, ISBN 0-321-30525-6

Thorough introduction to CSS and XHTML. Recommended.

Box, Skonnard, and Lam, Addison Wesley, 2000.

A general look at XML as a unifying technology

Tidwell, O'Reilly 2001

Tidwell developed IBM's Toot-o-matic for IBM Tutorials. It generates HTML, PDF, ZIP, and JPEG files from one XML file. It's a good demonstration of XSLT, XSL:FO, and FOP. [page 212-236]

Appendix A, B, C contains a complete reference for XSLT and XPATH 1.0 with multiple examples illustrating each feature.

Esposito, Microsoft Press 2003

Thorough introduction to XML programming in .NET with best practices and lots of examples.

Fitzgerald, O'Reilly, 2003, ISBN: 0-596-00327-7

Excellent, thorough introduction to XSLT. Lots of examples.

Foggon, Maharry, Ullman, Watson, Microsoft, 2004

Step by step introduction to web services. Includes SOAP, WSDL, DISCO, UDDI, HTTP, XML Streaming, serialization, XSD, web methods. ADO.NET, security, WS-* standards,

Harold, Means, O'Reilly 2004, ISBN: 0-596-00764-7

Thorough introduction to XML including DTD, namespaces, XSLT, XPath, XLinks, XPointers, CSS, XSL-FO, RDDL, XSD, DOM, SAX

For XSLT and XPath, has more complete coverage.

Kay, Wiley, 2004, ISBN: 0-7645-6910-4

Definitive coverage of XPath 2.0 by the editor of the XSLT 2.0 specification.

Kay, Wiley, 2004, ISBN: 0-7645-6909-0

Definitive coverage of XSLT 2.0 by the editor of the XSLT 2.0 specification and the author of the Saxon XSLT engine. The introductory chapters are worth a careful read, as are the detailed examples in Chapters 9-12. Chapter 4 provides a quick introduction to XML Schema. Chapter 6, Patterns, provides a quick introduction to XPath. Chapter 5 provides complete definition and usage for each XSLT element.

Kirsanov, Prentice Hall, 2004, ISBN: 0-13-140635-3

Thorough introduction to developing web sites using XML, XSLT, and Schematron. Clear separation between processing and content. Lots of practical advice. Good use of a master document to configure a web site.

Lenz, O'Reilly, 2005, ISBN: 0-596-10050-7

Excellent reference guide to XSLT 1.0. Written from the XSLT specification.

Includes preliminary documentation on XSLT 2.0 -- a significant enhancement to XSLT.

Mangano, O'Reilly 2003

Advanced XSLT programming with lots of examples, many convoluted. Covers string manipuation, numbers and math, dates and times, XPATH, whitespace, XML transformation, XML queries, HTML generation, SVG generation, code generation, Visio VDX, Excel XML, SOAP WSDL documentation, XSLT extensions, testing, debugging, generic programming, and functional programming

Skonnard, Gudgin, Addison Wesley, 2001.

Annotated review of XML standards.

Stanek, Microsoft Press 2002, ISBN 0-7356-1183-1.

Stanek's book on XML.