How to write in .csv file from a generator in python

Question

My Generator looks like the following:

def mygen(reader):
    for row in reader:
        yield row[0],row[1],row[2],row[3],row[4]

I'm trying to insert these generator produced values in the following manner:

file1=open(f2,"w")
writes=csv.writer(file1,delimiter=' ',quoting=csv.QUOTE_ALL)
g=mygen(reader)
for x in g:
    writes.writerow([x])

It enters blank rows to the file without producing any output. It works perfectly when I use the same generator to insert code to a table.

Answer 1

Your x is already a sequence, you don't need to wrap it in a list:

for x in g:
    writes.writerow(x)

You could just pass your generator directly to writes.writerows():

with open(f2, "w") as file1:
    writes = csv.writer(file1, delimiter=' ', quoting=csv.QUOTE_ALL)
    writes.writerows(mygen(reader))

Demo:

>>> import sys
>>> import csv
>>> def mygen(reader):
...     for row in reader:
...         yield row[0],row[1],row[2],row[3],row[4]
...
>>> reader = csv.reader('''\
... foo,bar,baz,spam,ham,eggs
... 42,81,13,97,73,100
... '''.splitlines())
>>> writes = csv.writer(sys.stdout, delimiter=' ', quoting=csv.QUOTE_ALL)
>>> writes.writerows(mygen(reader))
"foo" "bar" "baz" "spam" "ham"
"42" "81" "13" "97" "73"

You do need to watch out that your reader is not a generator you already exhausted. If it is a csv.reader() object, and you read all rows before, you'll need to re-wind the underlying file object back to the start with fileobj.seek(0). It'd be better to avoid reading the file twice, however.

If your reader object produces sequences (such as lists or tuples), you could just use slicing to get the first 5 elements, no need to type out every single index:

def mygen(reader):
    for row in reader:
        yield row[:5]