On Mon, Sep 19, 2011 at 9:20 PM, Sander Sweers <sander.swe...@gmail.com> wrote: > On 17/09/11 13:08, lists wrote: >> >> I have been trying to learn how to parse XML with Python and learn how >> to use xml.etree. Lots of the tutorials seem to be very long winded. >> >> I'm trying to access a UK postcode API at www.uk-postcodes.com to take >> a UK postcode and return the lat/lng of the postcode. This is what the >> XML looks like: http://www.uk-postcodes.com/postcode/HU11AA.xml >> >> The function below returns a dict with the xml tag as a key and the >> text as a value. Is this a correct way to use xml.etree? > > Define correct, does it give the desired result? Then I would say yes it is > correct. There may be alternative ways to get to the same result though. > >> def ukpostcodesapi(postcode): >> import urllib > > Why do the import here, for speed? You are reading an xml file from the > internet, guess where most of the time is spend in your function ;-). > >> import xml.etree.ElementTree as etree >> >> baseURL='http://www.uk-postcodes.com/' >> geocodeRequest='postcode/'+postcode+'.xml' > > You could use string formatting here. > url = 'http://www.uk-postcodes.com/postcode/%s.xml' % postcode > > Also what would happen if postcode includes a space? > >> >> #grab the xml >> tree=etree.parse(urllib.urlopen(baseURL+geocodeRequest)) > > What happens if you get an error (a 404 error perhaps)? You might want to > add a try/except block around reading the xml from the internet. > >> root=tree.getroot() >> results={} >> for child in root[1]: #here's the geo tag >> results.update({child.tag:child.text}) #build a dict >> containing the >> geocode data >> return results > > As you only get 1 set of long/lat tags in the xml you could use find(). See > below an example. > > from xml.etree import ElementTree as ET > import urllib2 > > url = 'http://www.uk-postcodes.com/postcode/HU11AA.xml' > xml = urllib2.urlopen(url).read() > tree = ET.XML(xml) > > geo = {} > > for leaf in tree.find('geo'): > geo[leaf.tag] = leaf.text > > Greets > Sander
Thank you I've been working on this and ended up with http://pastebin.com/Y9keC9tB Chris _______________________________________________ Tutor maillist - Tutor@python.org To unsubscribe or change subscription options: http://mail.python.org/mailman/listinfo/tutor
