Odtnhj

I'm using lxml to create a json file based on a xml. The xml file has this kind of structure :

<spots_list>
 <spot id="001" latitude="2011464" longitude="979511">
 <adress>Somewhere</adress>
 <city>BOSTON</city>
 <price category="Intermediate" value="782"/>
 <price category="Expensive" value="2765"/>
 <price category="Cheap" value="12"/>
 </spot>
 <spot id="002" latitude="2101644" longitude="915971">
 <adress>Somewhere else (very very far away)</adress>
 <city>CAMBRIDGE</city>
 <price category="Intermediate" value="472"/>
 <price category="Intermediate (but less expensive)" value="422"/>
 <price category="Expensive" value="20275"/>
 <price category="Cheap" value="12"/>
 </spot>
</spots_list>

The number of price elements in every can change, so I tried to use a while loop in Python. Here's the associate code :

from lxml import etree

tree = etree.parse("my_file.xml")

for node in tree.xpath("//spots_list/spot"):
 for adress in node.xpath("adress"):
 adr = adress.text
 while node.xpath("price"):
 print(adr)

I know it's wrong, because the first adress appears over and over, but I don't figure how formulate this loop to switch for next elements...

Thanks in advance.

The basic problem in the while statement is that node.xpath(...) returns a list, which is considered True if it is not empty. You just have to do the same thing as at the top level, i.e. iterate over the elements you are interested in, e.g.

def parse_spot(el):
 adr = el.find('adress')
 return dict(
 address=adr.text if adr is not None else None, # error handling if not found
 price=[dict(p.attrib) for p in el.findall('price')]
 )

tree = etree.fromstring(xml) # xml is your example as string

[parse_spot(el) for el in tree.findall('./spot')]

You can also use xpath instead of findall like you did.