为什么 csvwriter.writerow() 在每个字符后加一个逗号？

Question

This code opens the URL and appends the /names at the end and opens the page and prints the string to test1.csv :

import urllib2
import re
import csv

url = ("http://www.example.com")
bios = [u'/name1', u'/name2', u'/name3']
csvwriter = csv.writer(open("/test1.csv", "a"))

for l in bios:
    OpenThisLink = url + l
    response = urllib2.urlopen(OpenThisLink)
    html = response.read()
    item = re.search('(JD)(.*?)(\d+)', html)
    if item:
        JD = item.group()
        csvwriter.writerow(JD)
    else:
        NoJD = "NoJD"
        csvwriter.writerow(NoJD)

But I get this result:

J,D,",", ,C,o,l,u,m,b,i,a, ,L,a,w, ,S,c,h,o,o,l,....

If I change the string to ("JD", "Columbia Law School"....) then I get

JD, Columbia Law School...)

I couldn't find in the documentation how to specify the delimeter.

If I try to use delimeter I get this error:

TypeError: 'delimeter' is an invalid keyword argument for this function

Answer 1

It expects a sequence (eg: a list or tuple) of strings. You're giving it a single string. A string happens to be a sequence of strings too, but it's a sequence of 1 character strings, which isn't what you want.

If you just want one string per row you could do something like this:

csvwriter.writerow([JD])

This wraps JD (a string) with a list.

Answer 2

The csv.writer class takes an iterable as it's argument to writerow; as strings in Python are iterable by character, they are an acceptable argument to writerow, but you get the above output.

To correct this, you could split the value based on whitespace (I'm assuming that's what you want)

csvwriter.writerow(JD.split())

Answer 3

This happens, because when group() method of a MatchObject instance returns only a single value, it returns it as a string. When there are multiple values, they are returned as a tuple of strings.

If you are writing a row, I guess, csv.writer iterates over the object you pass to it. If you pass a single string (which is an iterable), it iterates over its characters, producing the result you are observing. If you pass a tuple of strings, it gets an actual string, not a single character on every iteration.

Answer 4

To put it another way - if you add square brackets around the whole output, it will be treated as one item, so commas won't be added. eg instead of:

spamwriter.writerow(matrix[row]['id'],matrix[row]['value'])

use:

spamwriter.writerow([matrix[row]['id'] + ',' + matrix[row]['value']])