Unit Mapping

work in progress

One of the simplest ways of getting an overview of a unit is to generate a simple index or map of the major headings and subheadings.

Example of prior work: Generating Mind Maps from OU/OpenLearn Structured Authoring XML Documents

from sqlite_utils import Database

# Open database connection
dbname = "all_openlean_xml.db"
db = Database(dbname)

Let’s get the OU-XML for an arbitary unit:

from lxml import etree
import pandas as pd

# If there are multiple units associated with H807, pick the first
h807_xml_raw = pd.read_sql("SELECT xml FROM xml WHERE code='H807'", con=db.conn).loc[0, "xml"]

# Parse the XML into an xml object
root = etree.fromstring(h807_xml_raw)

Bring in our simple utility function to help flatten elements, if required:

import unicodedata

def unpack(x):
    return etree.tostring(x)

# via http://stackoverflow.com/questions/5757201/help-or-advice-me-get-started-with-lxml/5899005#5899005
def flatten(el):
    """Utility function for flattening XML tags."""
    def _flatten(el):
        if el is None:
            return ""  # Originally returned None; any side effects of move to ''?
        result = [(el.text or "")]
        for sel in el:
            result.append(_flatten(sel))
            result.append(sel.tail or "")
        return unicodedata.normalize("NFKD", "".join(result)) or " "
    return _flatten(el).strip()

We can now grab all the headings and subheadings and render a simple contents list for the unit. To display the contents, we can use a simple tree widget.

Let’s start by parsing out the title of the unit:

title = root.find("ItemTitle").text
code = root.find("CourseCode").text

title, code

('Accessibility of eLearning', 'H807_1')

We can now build up out tree from session and section headings:

#%pip install ipytree
# ipytree provides access to a jstree wdget
from ipytree import Tree, Node

# Create a tree object
tree = Tree()

# Create a unit title node for our tree
node1 = Node(f"{title} ({code})")

# Add the unit title node to the top of the tree
tree.add_node(node1)

sessions = root.findall('.//Unit/Session')

unit_structure = {"title": {}}
for session in sessions:
    title = session.find('.//Title').text
    subnode = Node(title)
    node1.add_node(subnode)
    
    subsessions=session.findall('.//Section')
    for subsession in subsessions:
        heading = subsession.find('.//Title').text
        subnode.add_node( Node(heading) )

tree

The tree widget doesn’t appear to render when I flow this document as part of a Jupyer Book, so I need to find an alternative tree display for this demo. In the meantime, here’s a screehshot to get a flavour of what you’re missing…

It would be easy enough to generate a table contain session and section headings across all units and then use that as a way of providing a heading level search to retrieve items at that level of granularity.

Generating Tables of Contents Derived From Sections in Different Units

As well as generating tree listings of session and section headings related to a single unit, we can also generate table of content views over sections retrieved from multiple units.

For example, TO DO - search around a term to retrieve items from multiple units and generate a “customised” uniti on a topic, eg ordered by level, etc