#4 Collection meta information


1. Creation of collection data type
1.2 Look for perl modules to implement them
1.1 could use the DOM implementation
1.2 Could use an array paired with an hash

1.3 Ordered collection is important
1.4 be able to parse multiple types of rdf collections

1.5 be able to support reverse relationships
we want to see what objects are contained in what

2. Creation of members types,
Some nodes are to be contained directly by other nodes.
For example, the string_cst and the name of node should
be contained
and not referenced.
Such items things can be identified by nodes

3. Creation of a Pair Class
a pair joins two items into one.
for example identifier_node and integer_cst are joined
into a pair in
the list of values of an enum-

4. Creation of an array and a hash interface to members of
a struct
this declares a named and ordered collection of items
from => "name of the container",
to => "name of the contained",
name => "name of the field in the container",
field => "name of the rdf field",
chain => "name of the chained field"
reverse => "name of the reverse pointer"
reverse_field => "name of the reverse field"

from => "from what clas",
via-field => "the name of the rdf field",
to => "the type of the list"
reverse => "the name of the reverse pointer".
reverse_field => "name of the reverse field"

5. Create a tool to declared perl implementations for RDF
layout of graphs onto data structures
This tool will allow the marking up of RDF so that it can be
read into perl
and instanciated.

6. When a property shows up, we check if it it is in a
check if the element has been added already

If the property is a reverse relationship, the contained
object, look if the container is there,
we might have to create the container as a placeholder.


  • Logged In: YES