Reading a UTF8 CSV file with Python

Question

I am trying to read a CSV file with accented characters with Python (only French and/or Spanish characters). Based on the Python 2.5 documentation for the csvreader (http://docs.python.org/library/csv.html), I came up with the following code to read the CSV file since the csvreader supports only ASCII.

def unicode_csv_reader(unicode_csv_data, dialect=csv.excel, **kwargs):
    # csv.py doesn't do Unicode; encode temporarily as UTF-8:
    csv_reader = csv.reader(utf_8_encoder(unicode_csv_data),
                            dialect=dialect, **kwargs)
    for row in csv_reader:
        # decode UTF-8 back to Unicode, cell by cell:
        yield [unicode(cell, 'utf-8') for cell in row]

def utf_8_encoder(unicode_csv_data):
    for line in unicode_csv_data:
        yield line.encode('utf-8')

filename = 'output.csv'
reader = unicode_csv_reader(open(filename))
try:
    products = []
    for field1, field2, field3 in reader:
        ...

Below is an extract of the CSV file I am trying to read:

0665000FS10120684,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Bleu
0665000FS10120689,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Gris
0665000FS10120687,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Vert
...

Even though I try to encode/decode to UTF-8, I am still getting the following exception:

Traceback (most recent call last):
  File ".\Test.py", line 53, in <module>
    for field1, field2, field3 in reader:
  File ".\Test.py", line 40, in unicode_csv_reader
    for row in csv_reader:
  File ".\Test.py", line 46, in utf_8_encoder
    yield line.encode('utf-8', 'ignore')
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 68: ordinal not in range(128)

How do I fix this?

Answer 1

The .encode method gets applied to a Unicode string to make a byte-string; but you're calling it on a byte-string instead... the wrong way 'round! Look at the codecs module in the standard library and codecs.open in particular for better general solutions for reading UTF-8 encoded text files. However, for the csv module in particular, you need to pass in utf-8 data, and that's what you're already getting, so your code can be much simpler:

import csv

def unicode_csv_reader(utf8_data, dialect=csv.excel, **kwargs):
    csv_reader = csv.reader(utf8_data, dialect=dialect, **kwargs)
    for row in csv_reader:
        yield [unicode(cell, 'utf-8') for cell in row]

filename = 'da.csv'
reader = unicode_csv_reader(open(filename))
for field1, field2, field3 in reader:
  print field1, field2, field3 

PS: if it turns out that your input data is NOT in utf-8, but e.g. in ISO-8859-1, then you do need a "transcoding" (if you're keen on using utf-8 at the csv module level), of the form line.decode('whateverweirdcodec').encode('utf-8') -- but probably you can just use the name of your existing encoding in the yield line in my code above, instead of 'utf-8', as csv is actually going to be just fine with ISO-8859-* encoded bytestrings.

Answer 2

The link to the help page is the same for python 2.6 and as far as I know there was no change in the csv module since 2.5 (besides bug fixes). Here is the code that just works without any encoding/decoding (file da.csv contains the same data as the variable data). I assume that your file should be read correctly without any conversions.

test.py:

## -*- coding: utf-8 -*-
#
# NOTE: this first line is important for the version b) read from a string(unicode) variable
#

import csv

data = \
"""0665000FS10120684,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Bleu
0665000FS10120689,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Gris
0665000FS10120687,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Vert"""

# a) read from a file
print 'reading from a file:'
for (f1, f2, f3) in csv.reader(open('da.csv'), dialect=csv.excel):
    print (f1, f2, f3)

# b) read from a string(unicode) variable
print 'reading from a list of strings:'
reader = csv.reader(data.split('\n'), dialect=csv.excel)
for (f1, f2, f3) in reader:
    print (f1, f2, f3)

da.csv:

0665000FS10120684,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Bleu
0665000FS10120689,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Gris
0665000FS10120687,SD1200IS,Appareil photo numérique PowerShot de 10 Mpx de Canon avec trépied (SD1200IS) - Vert

Answer 3

Python 2.X

There is a unicode-csv library which should solve your problems, with added benefit of not naving to write any new csv-related code.

Here is a example from their readme:

>>> import unicodecsv
>>> from cStringIO import StringIO
>>> f = StringIO()
>>> w = unicodecsv.writer(f, encoding='utf-8')
>>> w.writerow((u'é', u'ñ'))
>>> f.seek(0)
>>> r = unicodecsv.reader(f, encoding='utf-8')
>>> row = r.next()
>>> print row[0], row[1]
é ñ

Python 3.X

In python 3 this is supported out of the box by the build-in csv module. See this example:

import csv
with open('some.csv', newline='', encoding='utf-8') as f:
    reader = csv.reader(f)
    for row in reader:
        print(row)

Answer 4

If you want to read a CSV File with encoding utf-8, a minimalistic approach that I recommend you is to use something like this:

        with open(file_name, encoding="utf8") as csv_file:

With that statement, you can use later a CSV reader to work with.

Answer 5

Looking at the Latin-1 unicode table, I see the character code 00E9 "LATIN SMALL LETTER E WITH ACUTE". This is the accented character in your sample data. A simple test in Python shows that UTF-8 encoding for this character is different from the unicode (almost UTF-16) encoding.

>>> u'\u00e9'
u'\xe9'
>>> u'\u00e9'.encode('utf-8')
'\xc3\xa9'
>>> 

I suggest you try to encode("UTF-8") the unicode data before calling the special unicode_csv_reader(). Simply reading the data from a file might hide the encoding, so check the actual character values.

Answer 6

Using codecs.open as Alex Martelli suggested proved to be useful to me.

import codecs

delimiter = ';'
reader = codecs.open("your_filename.csv", 'r', encoding='utf-8')
for line in reader:
    row = line.split(delimiter)
    # do something with your row ...