cElementTree <type 'exceptions.SyntaxError'>: not well-formed (invalid token)

trying to scrape this xml file on the web with urllib and cElementTree. I am using Google App Engine but I don't think the problem is relevant to my platform.

This is my error:

<type 'exceptions.SyntaxError'>: not well-formed (invalid token): line 1, column 25
Traceback (most recent call last):
  File "/base/data/home/apps/metautoit/daily-update.353244196034914877/Start_Update.py", line 25, in main
    ShoppingCar.XMLRipper().getNew()
  File "/base/data/home/apps/metautoit/daily-update.353244196034914877/updatecars/sitecrawlers/ShoppingCar.py", line 24, in getNew
    for carDict in newCars:
  File "/base/data/home/apps/metautoit/daily-update.353244196034914877/updatecars/sitecrawlers/ShoppingCar.py", line 67, in _iter_carDicts_in_xml
    tree = self.get_xml()
  File "/base/data/home/apps/metautoit/daily-update.353244196034914877/updatecars/sitecrawlers/ShoppingCar.py", line 63, in get_xml
    return ET.parse(req, parser=parser)
  File "<string>", line 45, in parse
  File "<string>", line 28, in parse

The xml file is long but here is a sample:

<?xml version="1.0" encoding="windows-1252"?><veicoli>
<veicolo>
    <id><![CDATA[16529]]></id>
    <link><![CDATA[http://www.shop开发者_运维技巧pingcar.it/auto_usate_/Chrysler_PT_Cruiser/16529.asp]]></link>
    <marca><![CDATA[Chrysler]]></marca>
    <modello><![CDATA[PT Cruiser]]></modello>
    <versione><![CDATA[2.4 L]]></versione>
    <provincia><![CDATA[Padova]]></provincia>
    <anno><![CDATA[2006]]></anno>
    <mese><![CDATA[4]]></mese>
    <chilometri><![CDATA[26000]]></chilometri>
    <cilindrata><![CDATA[]]></cilindrata>
    <potenza><![CDATA[143]]></potenza>
    <alimentazione><![CDATA[Benzina]]></alimentazione>
    <cambio><![CDATA[Cambio Automatico]]></cambio>
    <colore><![CDATA[nero]]></colore>
    <prezzo><![CDATA[14900]]></prezzo>
    <immagine><![CDATA[http://www.shoppingcar.it/public/Auto%20Usate/Berline/imagesadv/16529_2.jpg]]>
    </immagine>
</veicolo>
</veicoli>

My (simplified) code looks like this:

xml_url =   "http://www.shoppingcar.it/feed/export_vel.asp?parametro=1"
req =       urllib.urlopen(xml_url)
parser =    ET.XMLParser(encoding="windows-1252")
tree =      ET.parse(req, parser=parser).get_xml()

Here's the kicker: I downloaded and uploaded it as a public dropbox file. Using this url the xml parses just fine. I've tried without declaring the encoding, and tried windows-1252 and utf-8 encodings. It's just really strange because column 25 is nothing but the middle of the word "encoding". Any help is appreciated.

---

I tried your code (added imports, so that others can try):

#!/usr/bin/env python
import xml.etree.cElementTree as ET
import urllib

xml_url =   "http://www.shoppingcar.it/feed/export_vel.asp?parametro=1"
req =       urllib.urlopen(xml_url)
parser =    ET.XMLParser(encoding="windows-1252")
tree =      ET.parse(req, parser=parser).get_xml()

and it runs just fine. If your error only happens on the server then you probably hit a limit at the webpage and are trying to parse an error-message. So make sure that you are actually parsing the document (e.g. data = req.read() and then dump data and parse the string as xml.