Is there a way to iterate through all the sub-elements?

[amry-daud]

Hi, I'm currently giving BigXML a go, and I'm loving it so far. However I'm currently stumped on the subject of iterating through all sub-elements.

For example, here's a test XML:

<animals>
  <cats>
    <name1>Baron Cat</name1>
    <name2>Fireball</name2>
  </cats>
</animals>

Let's say I want the list of all the cats I have, as well as the name of the tags (name1 and name2). Rather than explicitly set the subnode names in the handler (which would not be practical if there's a lot of subnodes), is there a way to iterate the subnodes at the level?

[Rogdham]

Hello, currently this was not a targeted use case, so the way to do it is not optimal nor documented.

I'm wondering if this use case is really prevalent, as usually XML documents tends to follow a specific schema.

If you have public XML items that fall into this use case, please share them so that I have a better understanding of real-world examples.

---

The usual way would be as you said to explicitly set the subnode names when defining the handler, by chaining the decorators:

@xml_handle_element("animals", "cats", "name1")
@xml_handle_element("animals", "cats", "name2")
def handle_cats_names(node):
    yield (node.name, node.text)

>>> with open('animals.xml', 'rb') as f:
...     for item in Parser(f).iter_from(handle_cats_names):
...         print(item)
...
('name1', 'Baron Cat')
('name2', 'Fireball')

---

However the following alternative is possible.

First, define a handler with no decorator (we filter on XMLElement as we don't want to handle others):

def generic_handle(node):
    if isinstance(node, XMLElement):
        yield (node.name, node.text)

Then use it at the animals > cats level:

@xml_handle_element("animals", "cats")
def handle_cats(node):
    yield from node.iter_from(generic_handle)

At this point the following should work:

>>> with open('animals.xml', 'rb') as f:
...     for item in Parser(f).iter_from(handle_cats):
...         print(item)
...
('name1', 'Baron Cat')
('name2', 'Fireball')

[amry-daud]

Thank you very much for your swift response! I cannot explicitly name the subnodes because a) the names might be randomly-generated, and b) they number in the thousands, so the alternative your presented looks very promising.

My shift for the day have ended so I'll give it a try tomorrow and report back my findings.