CODEWITHSUNDEEP
pythonweb-scrapinglxml
Zain Khan
Zain Khan

Reputation: 3793

How to handle UnicodeDecodeError without losing any data?

I am using Python & lxml and am stuck with an error

My code

>>>import urllib
>>>from lxml import html

>>>response = urllib.urlopen('http://www.edmunds.com/dealerships/Texas/Grapevine/GrapevineFordLincoln_1/fullservice-505318162.html').read()
>>>dom = html.fromstring(response)

>>>dom.xpath("//div[@class='description item vcard']")[0].xpath(".//p[@class='service-review-paragraph loose-spacing']")[0].text_content()

Traceback

Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/local/lib/python2.7/dist-packages/lxml/html/__init__.py", line 249, in text_content
return _collect_string_content(self)
File "xpath.pxi", line 466, in lxml.etree.XPath.__call__ (src/lxml/lxml.etree.c:119105)
File "xpath.pxi", line 242, in lxml.etree._XPathEvaluatorBase._handle_result (src/lxml/lxml.etree.c:116936)
File "extensions.pxi", line 552, in lxml.etree._unwrapXPathObject (src/lxml/lxml.etree.c:112473)
File "apihelpers.pxi", line 1344, in lxml.etree.funicode (src/lxml/lxml.etree.c:21864)
UnicodeDecodeError: 'utf8' codec can't decode byte 0x93 in position 477: invalid start byte

The problem is the special character which is present in the div I am fetching. How can I encode/decode the text without losing any data?

Upvotes: 2

Views: 3216

Answers (2)

Not_a_Golfer
Not_a_Golfer

Reputation: 49205

The parser assumes this is a utf-8 file, but it's not. the simplest thing to do would be to convert it to unicode first, by knowing the encoding of the page

>>> url =  urllib.urlopen('http://www.edmunds.com/dealerships/Texas/Grapevine/GrapevineFordLincoln_1/fullservice-505318162.html')
>>> url.headers.get('content-type')
'text/html; charset=ISO-8859-1'

>>> response = url.read()
#let's convert to unicode first
>>> response_unicode = codecs.decode(response, 'ISO-8859-1')
>>> dom = html.fromstring(response_unicode)
#and now...
>>> dom.xpath("//div[@class='description item vcard']")[0].xpath(".//p[@class='service-review-paragraph loose-spacing']")[0].text_content()
u'\n                  On December 5th, my vehicle completely shut down.\nI had it towed to Grapevine Ford where they told me that the intak.....

tada!

Upvotes: 6

Oleh Prypin
Oleh Prypin

Reputation: 34116

So it looks like the page is corrupted. It has UTF-8 encoding specified, but is not valid in that encoding.

urlopen(...).read() returns us a byte string (str). When you feed it to lxml, it tries to decode it with UTF-8 and fails.

This might not be the best way, but we can specify a different encoding manually, such as Latin-1:

response = urllib.urlopen(...).read().decode('latin-1')

Now response is a text string (unicode), and that's what LXML wants to work with.

Upvotes: 0

Related Questions