Iterate over content / child elements

[hcw70]

In an xsdata representation created from a XSD schema from an XML document like the following

<?xml version="1.0" encoding="UTF-8"?>
<MessageDB xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
                    xsi:schemaLocation="http://deuta.de/config/DeutaMsgDatabase MessageDB.xsd"
                    xmlns="http://deuta.de/config/DeutaMsgDatabase">
                    <module name="root">
                        <datatypes>
                            <packettype name="pckType1">
                                <var name="Var0" type="UNSIGNED32" bitsize="1"/>
                            </packettype>
                        </datatypes>
                    </module>
</MessageDB>

is there a uniform way to iterate over the nested elements, maybe something like

for child in obj.child_elems():
    print (child)

?

In my current TMessageDb class created by xsdata, there is an attribute "modules" of kind list, but
the child elements are stored in different attributes depending on the class, so here TDataTypes would have an list element "packettype".

Regards

[tefra]

No there is no such helper, but it's not very difficult to write one, here something similar that I wrote in order to fetch all the obj children that matched a base class.

xsdata/xsdata/models/mixins.py

Lines 190 to 204 in 026ef55

	def children(self, condition: Condition = None) -> Iterator["ElementBase"]:
	"""Iterate over all the ElementBase children of this element that match
	the given condition if any."""
	for f in fields(self):
	value = getattr(self, f.name)
	if (
	isinstance(value, list)
	and len(value) > 0
	and isinstance(value[0], ElementBase)
	):
	yield from (val for val in value if not condition or condition(val))
	elif isinstance(value, ElementBase):
	if not condition or condition(value):
	yield value

[hcw70]

Agreed, after digging into dataclass deeper, i came to a similair solution, thanks for providing the info!

However, after further investigation, it looks like the order of the elements yielded by that is not in document order
(if the xsd has a sequence of choices of elements) but packet together based on the element type.

Since the order in our XML docs are very relevant in most cases, this is a no-go for us. Should i raise a new issue for this?

[tefra]

Oh I see what you need.

Cool that's part of the serializer logic I could move it to a public helper, in pr #247

xsdata/xsdata/formats/dataclass/serializers/xml.py

Lines 214 to 248 in 623ee6d

	def next_value(cls, meta: XmlMeta, obj: Any) -> Iterator[Tuple[XmlVar, Any]]:
	"""
	Return the fields and their values in the correct order according to
	their definition and sequential metadata.
	Sequential fields need to be rendered together in parallel order
	eg: <a1/><a2/><a1/><a/2></a1>
	"""
	index = 0
	attrs = meta.vars
	stop = len(attrs)
	while index < stop:
	var = attrs[index]
	if not var.sequential:
	yield var, getattr(obj, var.name)
	index += 1
	continue
	end = next(
	(i for i in range(index + 1, stop) if not attrs[i].sequential), stop
	)
	sequence = attrs[index:end]
	index = end
	j = 0
	rolling = True
	while rolling:
	rolling = False
	for var in sequence:
	values = getattr(obj, var.name)
	if j < len(values):
	rolling = True
	yield var, values[j]
	j += 1

[tefra]

Hey man take a look at XmlSerializer.next_value

In [1]: from xsdata.formats.dataclass.serializers import XmlSerializer

In [2]: from xsdata.formats.dataclass.context import XmlContext

In [3]: from tests.fixtures.books.fixtures import books

In [4]: books.book[0]
Out[4]: BookForm(author='Hightower, Kim', title='The First Book', genre='Fiction', price=44.95, pub_date='2000-10-01', review='An amazing story of nothing.', id='bk001', lang='en')

In [7]: context = XmlContext()

In [8]: meta = context.fetch(books.__class__)

In [9]: for var, value in XmlSerializer.next_value(books, meta):
   ...:     print(value)
   ...: 
[BookForm(author='Hightower, Kim', title='The First Book', genre='Fiction', price=44.95, pub_date='2000-10-01', review='An amazing story of nothing.', id='bk001', lang='en'), BookForm(author='Nagata, Suanne', title='Becoming Somebody', genre='Biography', price=33.95, pub_date='2001-01-10', review='A masterpiece of the fine art of gossiping.', id='bk002', lang='en')]

In the generator you can do a check for list instance and create a recursive method, the var holds teh attribute metadata used for binding purposes you can ignore it. The values will be in the order the fields are defined with support for sequential elements as well.

[hcw70]

Thank you. i will have a look!