/ 
Gestinanna-0.01
/ parser.PL

Data::DPath::Parser - Parses the DPath subset used by Data::DPath

SYNOPSIS

use Data::DPath::Parser;

my $result = Data::DPath::Parser->parse( $xpath );

$result -> evaluate( $root, $context );

$sub = $result -> compile;
$sub -> ( $root, $context );

DESCRIPTION

Some notes on the parsing and evaluation:

  • Result Objects

    TODO: change to return sets of references into the data tree

    The result expressions alway return true or false. For DPath expressions that would normally return a node-set, the result is true if the current SAX event would build a node that would be in the node set. No floating point or string return objects are supported (this may change).

  • Context

    The DPath context node is the document root.

    Not sure what to do about the context position, but the context size is of necessity undefined.

DPath Expressions

DPath is based on XPath, but has some subtle (or not-so-subtle) differences.

Axis

DPath makes use of the following axis:

ancestor
ancestor-or-self
attribute

Eventually (with Perl 6), we will be able to support arbitrary attributes tied to data objects. But until then, only the following attributes are supported:

can
Example: //*[@can="print" and @can="read"]

This will find all the nodes that have print and read methods.

defined
Example: /foo[@defined]

Returns the object foo in the root of the context if it is defined.

isa
Example: //*[@isa="Apache::Upload"]

This will find all the nodes in the context that are Apache::Upload objects.

size
version
child

This forces the identifier to be a child of a node. A child may be a hash key or an array index, depending on the parent node type. It will not be a method.

descendant
descendant-or-self
following
following-sibling
method

This forces the identifier to be a method of an object.

namespace
parent
preceding
preceding-sibling
self