Why is Elementtree iterating over each element even though it isn't a child?

Question

I have created a model that is used to create an object with data gleaned from an xml file using ElementTree to parse the xml file. My project is a couple thousand lines of code but I was able to quickly recreate my problem using the following example.

Sample XML data:

    <data>
       <country name="Liechtenstein">
          <rank>1</rank>
          <year>2008</year>
          <gdppc>141100</gdppc>
          <neighbor name="Austria" direction="E"/>
          <neighbor name="Switzerland" direction="W"/>
       </country>
       <country name="Singapore">
          <rank>4</rank>
          <year>2011</year>
          <gdppc>59900</gdppc>
          <neighbor name="Malaysia" direction="N"/>
       </country>
       <country name="Panama">
          <rank>68</rank>
          <year>2011</year>
          <gdppc>13600</gdppc>
          <neighbor name="Costa Rica" direction="W"/>
          <neighbor name="Colombia" direction="E"/>
       </country>
   </data>

Model:

class neighbor(object):
   name = ""
   direction = ""

class neighborList(object):
   neighbor = []

class country(object):
   name = ""
   rank = ""
   year = ""
   gdppc = ""
   neighborList = neighborList()

class countryList(object):
   country = []

class data(object):
   countryList = countryList()

Parser:

    from xml.etree import ElementTree as ET
    import countries_model as ctry

    def CountriesCrusher(filename):

        xmldoc = ET.parse(filename)
        element = xmldoc.getroot()

        _data = ctry
        _countryList = ctry.countryList()  

        for firstLevel in element.findall('country'):
            b = ctry.country()
            b.rank = firstLevel.find('rank').text
            b.year = firstLevel.find('year').text
            b.gdppc = firstLevel.find('gdppc').text
            b.neighborList = ctry.neighborList()

            for secondLevel in firstLevel.findall('neighbor'):
                c = ctry.neighbor
                c.direction = secondLevel.attrib.get('direction')
                c.name = secondLevel.attrib.get('name')
                b.neighborList.neighbor.append(c)

            _countryList.country.append(b)

        a = ctry.data()
        a.countryList = _countryList
        _data = a
        return _data

    ictry = CountriesCrusher('countries.xml')

Before I run this I would expect that if I look at ictry.countryList.country I would see three entries and that if I look at ictry.countryList.country[0].neighborList.neighbor I would see two neighbor entries for that country but instead I am seeing all five neighbor elements that are in the entire xml file. Why is this happening ??

It's because you are using class attributes, rather than instance attributes. — ekhumoro, Jan 20 '16 at 18:09

score 1 · Accepted Answer · edited May 23 '17 at 12:07

You arent using instance attributes of the class country.

Write your country class (and all others) like so:

class country:
    def __init__(self):
        self.name = ""
        self.rank = ""
        self.year = ""
        self.gdppc = ""
        self.neighborList = neighborList()

Now b = ctry.country() will give you an instance whose attributes will be decoupled/separate from a second call to b = ctry.country(). Your current approach all instances of ctry.country shared the same attributes since they were class attributes, not instance attributes.

Read more about class vs instance attributes here.

Thank you for your quick reply. This solved my issue. – btathalon Jan 20 '16 at 18:19 — btathalon, Jan 20 '16 at 18:19

Why is Elementtree iterating over each element even though it isn't a child?

1 Answers1