You are not logged in Log in Join
You are here: Home » Members » Martijn Faassen's Page » Parsed XML » README.txt » View Document

Log in
Name

Password

 

README.txt

README for Parsed XML.

What is it?

Parsed XML allows you to use XML objects in the Zope environment. You can create XML documents in Zope and leverage Zope to format, query, and manipulate XML.

Parsed XML consists of a DOM storage, a builder that uses PyExpat to parse XML into the DOM, and a management proxy product that provides Zope management features for a DOM tree. It also includes a system to create paths to nodes in Zope URLs (NodePath).

Requirements, Installation

See INSTALL.txt.

Feedback, discussion, more information

The latest released version can be found at the product site.

For more information, see the Parsed XML wiki.

Bug reports and status is kept by the the Parsed XML tracker.

There is a mailing list with archives.

The latest version is available through CVS.

Features

The Parsed XML product parses XML into a Zopish DOM tree. The elements of this tree support persistence, acquisition, etc.. The document and subnodes are editable and manageable through management proxy objects, and the underlying DOM tree can be directly manipulated via DTML, Python, etc..

DOM and ManageableDOM

We're implementing a lean, mean DOM tree for pure DOM access, and a tree of proxy shells to handle management and take care of the conveniences like publishing and security. The ManageableNodes are the proxy objects. These are what you see in the management interface, and the top object that gets put in the ZODB. Note that only the top proxy object is persistent, the others are transient. The Nodes are pure DOM objects. From a ManageableNode, the DOM Node is retrieved with the getDOMObj() call.

See README.DOMProxy for information about the proxy objects.

DOM API support

The DOM tree created by Zope aims to comply with the DOM level 2 standard. This allows you to access your XML in DTML or External Methods using a standard and powerful API.

We are currently supporting the DOM level 2 Core and Traversal specifications.

The DOM tree is not built with the XML-SIG's DOM package, because it requires significantly different node classes.

DOM attributes are made available according to the Python language mapping for the IDL interfaces described by the DOM recommendation; see the mapping.

URL traversal

Parsed XML implements a NodePath system to create references to XML nodes (most commonly elements).

FIXME include examples here

Currently, traversal uses an element's index within its parent as an URL key. For example,

http://server/myDoc/0/2/mymethod

(Should add comment about NodePath, which is more powerful)

This URL traverses from an XML Document object with id myDoc to it's first sub-element, to that element's second sub-element to an acquired method with id myMethod

DOM methods can also be used in URLs, for example,

http://server/myDoc/firstChild/nextSibling/mymethod

(Not sure this still works..)

Editing XML with the management interface

XML Documents and subnodes are editable via the management interface. Documents and subtrees can be replaced by uploading XML files.

Security

Security is handled at the document level. DOM attributes and methods are protected by the "Access contents information" permission. Subnodes will acquire security settings from the document.

Developing with Parsed XML

We like to think that Parsed XML provides a flexible platform for using a DOM storage and extending that storage to do interesting things. See README.DOMProxy for an explanation of how we're using this for Parsed XML.

We've included a comprehensive unit test suite to make testing for DOM compliance easier. See tests/README for details.

If you want to submit changes to Parsed XML, please use the test suite to make sure that your changes don't break anything.

Bugs

There are bugs in how multiple node references reflect the hierarchy above the node:

  • A reference to a subnode of a DOM document won't reflect some hierarchy changes made on other references to the same node.

    If two references to a node are created, and one is then reparented, the other reference won't reflect the new parent. The parentNode attribute will be incorrect, for example, as well as the ownerDocument and ownerElement attributes.

  • A reference to a subnode of a DOM document can't be properly stored as a persistent attribute of a ZODB object; it will lose hierarchy information about its parent as well.

Entity reference handling is not complete:

  • Entity references do not have child nodes that mirror the child nodes of the referenced entity; they do not have child nodes at all.
  • TreeWalker.expandEntityReferences has no effect, because of the above bug.

Unicode support is still incomplete. It appears to work with Python 2.1 and Zope 2.4, but there are still some issues with the parser.

Traversal support for visibility and roots is not complete.

Credits

see CREDITS.txt and LICENSE.Fourthought.