无法在Python中写入.csv文件中的所有行？

Question

I'm trying to write to file2.csv file by values from file1.csv file using a keyfile.csv which contains the mapping between two files as the two files don't have the same column order.

def convert():

    Keyfile = open('keyfile.csv', 'rb')
    file1 = open('file1.csv', 'rb')
    file2 = open('file2.csv', 'w')

    reader_Keyfile = csv.reader(Keyfile, delimiter=",")
    reader_file1 = csv.reader(file1, delimiter=",")
    writer_file2 = csv.writer(file2, delimiter=",")
    for row_file1 in reader_file1:
        for row_Keyfile in reader_Keyfile:
            for index_val in row_Keyfile:
                file2.write(row_file1[int(index_val)-1]+',')
    # Closing all the files     
    file2.close()
    Keyfile.close()
    file1.close()

# keyfile structure: 3,77,65,78,1,10,8...
# so 1st column of file2 is 3rd column of file1 ; 
# col2 of file 2 is col77 of file1 and so on

I'm only able to write only one row in file2.csv . It should have as many rows as there are in file1.csv . How do I move to the next row after one row is finished ? I'm assuming Loop should take care of that but that's not happening.What am I doing wrong ?

Answer 1

You have two problems.

1. You should only read keyfile once and build a dict out of the mapping
2. You need to write a \\n at the end of each line of your output file

Answer 2

I am assuming the KeyFile is just one row, giving the mappings for all rows. Something like the following should work:

def convert():
    with open('keyfile.csv') as Keyfile, open('file1.csv', 'r') as file1, open('file2.csv', 'wb') as file2:
        mappings = next(csv.reader(Keyfile, delimiter=","))
        mappings = [int(x)-1 if x else None for x in mappings]

        reader_file1 = csv.reader(file1, delimiter=",")
        writer_file2 = csv.writer(file2, delimiter=",")

        for row_file1 in reader_file1:
            row = [''] * len(mappings)
            for from_index, to_index in enumerate(mappings):
                if to_index != None:
                    row[to_index] = row_file1[from_index]
            writer_file2.writerow(row)

It assumes column mappings start from 1.

Answer 3

Your nested looping is problematic as others mentioned. Instead, create the mapping outside of the row iteration, then write the rows based on the mapping. I use a dict object for this.

import csv

Keyfile = open('keyfile.csv', 'rb')
file_out = csv.reader(open('file1.csv', 'rb'), delimiter=",")
file_in = csv.writer(open('file2.csv', 'w'), delimiter=",")

mapDict = {}

# the first line in KeyFile convert to dict
reader = csv.reader(Keyfile, delimiter=',')
for i, v in enumerate(reader.next()):
    if v != ' ':
        mapDict[i] = int(v)


# re-index the row in file_in based on mapDict
for row in file_out:
    file_in.writerow([row[c] for c in mapDict.values()])