Wenn es auf einer Ebene mehrere Kindelemente des gleichen Typs gibt, wird der Name im Pfad mit einem "*" gekennzeichnet. Ist nicht perfekt, war aber hilfreich um Code zu schreiben, der die (für mich relevanten) Daten in eine JSON-Struktur überführt.
Code: Alles auswählen
#!/usr/bin/env python3
"""
Little helper to analyse the result of a webservice.
The service query returns an XML document with wrong XML declaration encoding
and an undocumented, quite large, and nested XML structure.
The elements have no attributes, and only leaf elements contain text other than
the occasional line break (or other whitespace). The `visit()` function tests
for both constraints and raises a `ValueError` on violation.
The output is a sorted list of unique paths to leaf elements found in the
document. Where a node has more than one child of a particular type, a "*" is
added to that name.
Here is a small example which also demonstrates two shortcomings:
>>> root = ET.fromstring('''\\
... <root>
... <item>
... <a>one a</a>
... <a>another a</a>
... <b>a single b</b>
... </item>
... <item>
... <a>in this item there is just one a</a>
... </item>
... </root>
... ''')
>>> count_elements(root)
7
>>> get_paths(root)
['./item*/a', './item*/a*', './item*/b']
The first two paths show not all ``<item>`` elements do contain multiple
``<a>`` elements. And the result doesn't show that not *all* ``<item>``
elements contain a ``<b>`` element, i.e. that ``<b>`` is optional.
"""
from collections import Counter
from pathlib import Path
from xml.etree import ElementTree as ET
def count_elements(node):
return 1 + sum(map(count_elements, node))
def visit(node, path="."):
if node.attrib:
raise ValueError(f"element has attribute(s): {node.attrib!r}")
if node:
if node.text and node.text.strip() or node.tail and node.tail.strip():
sub_node_tags = ", ".join(f"<{sub_node.tag}>" for sub_node in node)
raise ValueError(
f"text content ({node.text.strip()!r}) in an element"
f" (<{node.tag}>) with subelement(s) ({sub_node_tags})"
)
tag_to_count = Counter(sub_node.tag for sub_node in node)
for sub_node in node:
multiple = "*" if tag_to_count[sub_node.tag] > 1 else ""
yield from visit(sub_node, f"{path}/{sub_node.tag}{multiple}")
else:
yield path
def get_paths(root):
return sorted(set(visit(root)))
def main():
root = ET.fromstring(Path("group.xml").read_text(encoding="utf-8"))
print(count_elements(root), "elements.")
for path in get_paths(root):
print(path)
if __name__ == "__main__":
main()Code: Alles auswählen
5046 elements.
./Group/BaseCountry
./Group/FoundMonth
./Group/FoundYear
./Group/Founder/Handle/ID
./Group/Grouptypes/Grouptype*
./Group/ID
./Group/Member*/Handle/AKA
./Group/Member*/Handle/CurrentlyUsedHandle
./Group/Member*/Handle/FreelanceFunctions/Function
./Group/Member*/Handle/FreelanceFunctions/Function*
./Group/Member*/Handle/Handle
./Group/Member*/Handle/HandleStory
./Group/Member*/Handle/ID
./Group/Member*/Handle/Scener/Country
./Group/Member*/Handle/Scener/Handles/Handle*/AKA
./Group/Member*/Handle/Scener/Handles/Handle*/CurrentlyUsedHandle
./Group/Member*/Handle/Scener/Handles/Handle*/FreelanceFunctions/Function
./Group/Member*/Handle/Scener/Handles/Handle*/FreelanceFunctions/Function*
./Group/Member*/Handle/Scener/Handles/Handle*/Handle
./Group/Member*/Handle/Scener/Handles/Handle*/HandleStory
./Group/Member*/Handle/Scener/Handles/Handle*/ID
./Group/Member*/Handle/Scener/Handles/Handle*/Scener/ID
./Group/Member*/Handle/Scener/Handles/Handle/ID
./Group/Member*/Handle/Scener/ID
./Group/Member*/Handle/Scener/Trivia
./Group/Member*/JoinDay
./Group/Member*/JoinMonth
./Group/Member*/JoinYear
./Group/Member*/LeaveDay
./Group/Member*/LeaveMonth
./Group/Member*/LeaveYear
./Group/Member*/Profession
./Group/Member*/Profession*
./Group/Member*/Status
./Group/Name
./Group/Rating
./Group/Release*/Release/AKA
./Group/Release*/Release/Achievement/Compo
./Group/Release*/Release/Achievement/Place
./Group/Release*/Release/GfxType
./Group/Release*/Release/ID
./Group/Release*/Release/Name
./Group/Release*/Release/Rating
./Group/Release*/Release/ReleaseDay
./Group/Release*/Release/ReleaseMonth
./Group/Release*/Release/ReleaseYear
./Group/Release*/Release/ReleasedAt/Event/AKA
./Group/Release*/Release/ReleasedAt/Event/Address
./Group/Release*/Release/ReleasedAt/Event/City
./Group/Release*/Release/ReleasedAt/Event/Country
./Group/Release*/Release/ReleasedAt/Event/EndDay
./Group/Release*/Release/ReleasedAt/Event/EndMonth
./Group/Release*/Release/ReleasedAt/Event/EndYear
./Group/Release*/Release/ReleasedAt/Event/EventType
./Group/Release*/Release/ReleasedAt/Event/EventType*
./Group/Release*/Release/ReleasedAt/Event/ID
./Group/Release*/Release/ReleasedAt/Event/Name
./Group/Release*/Release/ReleasedAt/Event/StartDay
./Group/Release*/Release/ReleasedAt/Event/StartMonth
./Group/Release*/Release/ReleasedAt/Event/StartYear
./Group/Release*/Release/ReleasedAt/Event/State
./Group/Release*/Release/ReleasedAt/Event/Tagline
./Group/Release*/Release/ReleasedAt/Event/Website
./Group/Release*/Release/ReleasedAt/Event/Zip
./Group/Release*/Release/ScreenShot
./Group/Release*/Release/Type
./Group/Release*/Release/Website
./Group/Short
./Group/Slogan
./Group/Trivia
./Group/Website
