Question

I need to avoid creating double branches in an xml tree when parsing a text file. Let’s say the textfile is as follows (the order of lines is random):

branch1:branch11:message11
branch1:branch12:message12
branch2:branch21:message21
branch2:branch22:message22

So the resulting xml tree should have a root with two branches. Both of those branches have two subbranches. The Python code I use to parse this textfile is as follows:

import string
fh = open ('xmlbasic.txt', 'r')
allLines = fh.readlines()
fh.close()
import xml.etree.ElementTree as ET
root = ET.Element('root')

for line in allLines:
   tempv = line.split(':')
   branch1 = ET.SubElement(root, tempv[0])
   branch2 = ET.SubElement(branch1, tempv[1])
   branch2.text = tempv[2]

tree = ET.ElementTree(root)
tree.write('xmlbasictree.xml')

The problem with this code is, that a branch in xml tree is created with each line from the textfile.

Any suggestions how to avoid creating another branch in xml tree if a branch with this name exists already?

Answer 1

with open("xmlbasic.txt") as lines_file:
    lines = lines_file.read()

import xml.etree.ElementTree as ET

root = ET.Element('root')

for line in lines:
    head, subhead, tail = line.split(":")

    head_branch = root.find(head)
    if not head_branch:
        head_branch = ET.SubElement(root, head)

    subhead_branch = head_branch.find(subhead)
    if not subhead_branch:
        subhead_branch = ET.SubElement(branch1, subhead)

    subhead_branch.text = tail

tree = ET.ElementTree(root)
ET.dump(tree)

The logic is simple — you already stated it in your question! You merely need to check whether a branch already exists in the tree before creating it.

Note that this is likely inefficient, since you are searching up to the entire tree for each line. This is because ElementTree is not designed for uniqueness.

---

If you require speed (which you may not, especially for smallish trees!), a more efficient way would be to use a defaultdict to store the tree structure before converting it to an ElementTree.

import collections
import xml.etree.ElementTree as ET

with open("xmlbasic.txt") as lines_file:
    lines = lines_file.read()

root_dict = collections.defaultdict( dict )
for line in lines:
    head, subhead, tail = line.split(":")
    root_dict[head][subhead] = tail

root = ET.Element('root')
for head, branch in root_dict.items():
    head_element = ET.SubElement(root, head)
    for subhead, tail in branch.items():
        ET.SubElement(head_element,subhead).text = tail

tree = ET.ElementTree(root)
ET.dump(tree)