TypeError: expected a character buffer object

Question

I have the following code:

import os
import csv


listing = os.listdir('/directory/my/files/are/in')
os.chdir('/directory/my/files/are/in')

for file in listing[1:]:
    r = csv.reader(open(file, 'rU')) 

    for row in r:
        stuff = [str.split('\t', 1) for row in r]
        print stuff

which returns TypeError: expected a character buffer object. How do I fix this? Basically I want to open all (csv) files in a directory and then cut out certain junk from each cell (everything after and including /t.)

Answer 1

str.split('\t', 1) is the equivalent of '\t'.split(1), which tries to split the string '\t' by the non-string 1, which fails. And it's not entirely clear what you expect this to do: you're iterating over r, and then in the body of that loop you are iterating over r again (in the list comprehension.) Did you instead mean to do something like this?

for row in r:
    stuff = [item.split('\t', 1) for item in row]
    print stuff

(Which still doesn't make much sense, but at least it does something.)

Answer 2

import os
import csv

listing = os.listdir('/directory/my/files/are/in')
os.chdir('/directory/my/files/are/in')

for file in listing[1:]:
    if os.path.isdir(file):
        continue
    with open(file, 'rU') as csvfile:
        for row in csv.reader(csvfile, delimiter='\t'):
            print row
            #now you can extract data from the row

This fixes several problems.

First, you need to make sure each file to try to open is actually a file, not a directory.

Second, csv.reader can use tabs as the delimiter. This means you don't need to use split.

Third, this version actually closes the files.