On 12/22/2015 05:29 AM, KP wrote:
From my first foray into XML with Python:
I would like to retrieve this list from the XML upon searching for the 'config'
with id attribute = 'B'
config = {id: 1, canvas: (3840, 1024), comment: "a comment",
{id: 4, gate: 3, (0, 0, 1280, 1024)},
{id: 5, gate: 2, (1280, 0, 2560, 1024)},
{id: 6, gate: 1, (2560, 0, 3840, 1024)}}
I have started to use this code - but this is beginning to feel very
non-elegant; not the cool Python code I usually see...
import xml.etree.ElementTree as ET
tree = ET.parse('master.xml')
master = tree.getroot()
for config in master:
if config.attrib['id'] == 'B':
...
It much depends on
1/ the xml parser you use.
2/ the xml structure
1/ I'm happily using beautifulSoup. Using it is really simple and yield
simple code.
2/ Whenever the code gets complicated is because the xml is not properly
structured. For instance in you example, 'id' is an attribute of
'config' nodes, that's fine, but for 'panel' nodes it's a child node.
There's no point using a node when only one 'id' can be specified.
Filtering by attributes is much easier than by child nodes.
Anyway here's an example of using beautifulSoup:
python 2.7 (fix the print statement if you're using python3)
import bs4
xmlp = bs4.BeautifulSoup(open('test.xml', 'r'), 'xml')
# print all canvas
for cfg in xmlp.findAll('config'):
print cfg.canvas.text
# find config B panel 6 coordinates
xmlp.find('config', id='B').find(lambda node: node.name=='panel' and
node.id.text=='6').coordinates.text
# if panel id were attributes:
xmlp.find('config', id='B').find('panel', id='6').coordinates.text
If you can change the layout of the xml file it's better that you do,
put every values as attribute whenever you can:
<config id="A" canvas="3840,1024">
<comment> comments can span
on multiple lines, you probably need a node
</comment>
<panel id="1" gate="6" coordinates="0,0,1280,124"/>
<config>
Properly structured xml will yield proper python code.
cheers,
JM
--
https://mail.python.org/mailman/listinfo/python-list