Help me with the script? How to find items in csv file A and not in file B and vice versa

[Peter Otten]

Alan Newbie wrote:

> Hello,
> Let's say I want to compare two csv files: file A and file B. They are
> both similarly built - the first column has product IDs (one product per
> row) and the columns provide some stats about the products such as sales
> in # and $.
> 
> I want to compare these files - see which product IDs appear in the first
> column of file A and not in B, and which in B and not A. Finally, it would
> be very great if the result could be written into two new CSV files - one
> product ID per row in the first column. (no other data in the other
> columns needed)
> 
> This is the script I tried:
> ==========================
> 
> import csv
> 
> #open CSV's and read first column with product IDs into variables pointing
> #to lists
> A = [line.split(',')[0] for line in open('Afile.csv')]
> B = [line.split(',')[0] for line in open('Bfile.csv')]
> 
> #create variables pointing to lists with unique product IDs in A and B
> #respectively
> inAnotB = list(set(A)-set(B))
> inBnotA = list(set(B)-set(A))
> 
> print inAnotB
> print inBnotA
> 
> c = csv.writer(open("inAnotB.csv", "wb"))
> c.writerow([inAnotB])
> 
> 
> d = csv.writer(open("inBnotA.csv", "wb"))
> d.writerow([inBnotA])
> 
> print "done!"
> 
> =====================================================
> 
> But it doesn't produce the required results.
> It prints IDs in this format:
> 247158132\n

Python reads lines from a file with the trailing newline included, and 
line.split(",") with only one column (i. e. no comma) keeps the whole line. 
As you already know about the csv module you should use it to read your 
data, e. g. instead of

> A = [line.split(',')[0] for line in open('Afile.csv')]

try

with open("Afile.csv", "rb") as f:
    a = {row[0] for row in csv.reader(f)}
...

I used {...} instead of [...], so a is already a set and you can proceed:


in_a_not_b = a - b

Finally as a shortcut for

for item in in_a_not_b:
    writer.writerow([item])

use the writerows() method to write your data:

with open("inAnotB.csv", "wb") as f:
    writer = csv.writer(f)
    writer.writerows([item] for item in_a_not_b)

Note that I'm wrapping every item in the set rather than the complete set as 
a whole. If you wanted to be clever you could spell that even more succinct 
as

    writer.writerows(zip(in_a_not_b))

> and nothing to the csv files.
> 
> You could probably tell I'm a newbie.
> Could you help me out?
> 
> here's some dummy data:
> 
https://docs.google.com/file/d/0BwziqsHUZOWRYU15aEFuWm9fajA/edit?usp=sharing
> 
> 
https://docs.google.com/file/d/0BwziqsHUZOWRQVlTelVveEhsMm8/edit?usp=sharing
> 
> Thanks a bunch in advance! :)