[gxml] Created tag GXML_0_0_2

[Richard Hans Schwarting <rschwart src gnome org>]

The unsigned tag 'GXML_0_0_2' was created.

Tagger: Richard Schwarting <aquarichy gmail com>
Date: Thu Sep 29 20:25:56 2011 -0400

    Tagging 0.0.2

Changes since the last tag 'GXML_0_0_1':

Olav Vitters (1):
      fix userid in doap file

Richard Schwarting (81):
      * inc. version to 0.0.2
      * update potential patch for libxml2's vapi
      * advance doctype support, and add Stream support for saving and File support for reading, requiring libxml2's vapi patch
      * start adding support for functionality based on what we can expose in libxml2's bindings
      * stub out tests for Document's properties, now that we'll be able to support them with libxml2 changes
      * add gxml.doap
      * add .pc file
      * correct cflags for .pc
      * add documentation for Document
      * add documentation for Attr
      * add documentation for BackedNode     * throw some additional errors
      * add documentation for CDATASection
      * add documentation for CharacterData
      * add documentation for Comment
      * add documentation for DocumentFragment
      * add documentation for DocumentType
      * document DomError
      * fix some whitespace
      * fix more trailing whitespace
      * document Element
      * document Entity, EntityReference, Implementation
      * document NodeList     * throw some exceptions
      * document NodeType, Notation, ProcessingInstruction, Text and XNode
      * change to a shared library from a static library.
      * comment out a test until we can figure out how libxml2 is handling DTDs
      * add rule to build .pc so I can easily use gxml for other software     * add hacky call to valadoc, finally something that works
      * add bindings for xmlBuffer so we can stringify nodes easily
      * make to_string return an XML representation of a node or document
      * test to_string for Document and Element
      * change content_to_string () to a content property, for user convenience.
      * revolutionise the stringification of NodeLists
      * dispense with unnecessary lines
      * implement NameTagNodeList, a NodeList suitable for get_elements_by_tag_name; still needs live monitoring
      * make Document's get_elements_by_tag_name use NodeList
      * make Element use NodeList for get_elements_by_tag_name
      * make version parametre default to null
      * in libxml2 patch, don't use absolute paths
      * use NodeList now for get_elements_by_tag_name test
      * fill in tests for implementation and document_element     * fill in tests for constructing by string     * start writing generic document contents test to be reused to make sure documents were read in correctly
      * yay, Document loading is tested well now.
      * figured out how to neatly run outside processes synchronously for testing, yay
      * fill out tests for Document for creating other classes
      * fix DocumentTest
      * add test_out_{path,stream}_expected.xml test files
      * settle an upwards notification of element removal and addition to maintain live-requirement for nodelists from get_elements_by_tag_name
      * figured out a way to upwards notify ancestors with tag-name node-lists on the addition and removal of relevant elements.
      * change from NameTagNodeList to TagNameNodeList
      * fixed NodeList for addition of DocumentFragment     * successfully test get_elements_by_tag_name's live nature
      * add 'waf' binary
      * add some TODOs after talking to Philip Withnall
      * make more IO cancellable
      * add name space properties from DOM Level 2 Core.  Note that it's possible to have a list of namespaces, but these only support the first encountered.
      * expose a constructor to create a GXml Document from a libxml2 Xml.Node*.  Even if the internals change, it can still be useful to convert between the two.
      * add tests for Element's namespace support
      * handle DomErrors in tests
      * actually implement syncing of attributes from hashtable to libxml2 structures (needs more testing)
      * build out the Attr namespace tests.  They currently FAIL because while GXml keeps a map between libxml2 Xml.Nodes to GXML XNodes, we don't do that for attrs right now, since before namespace support, we could really get away with creating string-based attrs. Nuts\!
      * trivially easy to map the two correctly.  D'oh
      * since libxml2's attr isn't really a node, we have to override the functions.  Really want to find out why Xml.Attr uses ns and why Xml.Node uses ns_def
      * make sure we handle nulls, rather than just saying we should eventually
      * fix namespace usage: definitions are actually attributes
      * add NamespaceAttr, an Attr that defines an NS
      * improve error and cancellation handling for Document
      * remove superfluous bits
      * add an AttrNodeList, TagNameNodeList, and NamespaceAttrNodeList based on new GListNodeList
      * add dbg_inspect method to print what a node contains     * add namespace_definitions property
      * add NamespaceAttr.vala to wscript_build
      * fix ElementTest's testing for namespaces
      * last minute documentation fixes to unbreak compilation
      * add actual libxml-2.0.vapi that we're using rather than just the patch, because the patch would need to be applied on a specific version, and I don't want to account for which version until I upstream it
      * add useful instructions for compilation and testing
      * add a note about how to run the test correctly
      * add some notes on Attr behaviour: we really need to implement NamedNodeMap and not use GHashTable :D
      * remove proposed function
      * access _attributes directly, rather than ultimately calling ourself (we know _attributes has been initialised because we initialised it here)
      * do not unset all and try to readd the current ones, because we'd be deleting the actual attribute data out from underneath the ones that were there from the start.  Instead, just unset libxml2 ones that we don't have GXml attributes for anymore.
      * test dirtying of an element and its syncing of attributes
      * note possible null return value.
      * add a test for escaped characters
      * silence debugging noise
      * update release note