How to write to csv file from a text file

Question

I have text file with data like below:

"id":0
"value1":"234w-76-54"
"id":1
"value1":"2354w44-7-54"

I want to have these data in csv file. I tried with below code, but this is writing each ids and value1 as list in csv file.

with open("log.txt", "r") as file:
    file2 = csv.writer (open("file.csv", "w", newline=""), delimiter=",")
    file2.writerow(["id", "value1"])
        for lines in file:
            if "id" in lines:
                ids = re.findall(r'(\d+)', lines)
            if "value1" in lines:
                value1 = re.findall(r'2[\w\.:-]+', lines)
                file2.writerow([ids, value1])

getting output-

 id         value1
['0']   ['234w-76-54']
['1']   ['2354w44-7-54']

Desired output-

  id         value1
  0        234w-76-54
  1        2354w44-7-54

Ehat about using file2.writerow([ids[0], value1[0]])?

Barbaros Özhan
– Barbaros Özhan

2021-04-23 12:45:03 +00:00
Commented Apr 23, 2021 at 12:45 — Barbaros Özhan
– Barbaros Özhan, Commented Apr 23, 2021 at 12:45
use re.search instead of re.findall

mousetail
– mousetail

2021-04-23 12:45:52 +00:00
Commented Apr 23, 2021 at 12:45 — mousetail
– mousetail, Commented Apr 23, 2021 at 12:45
This worked, Thanks @BarbarosÖzhan

Priyx
– Priyx

2021-04-23 12:56:53 +00:00
Commented Apr 23, 2021 at 12:56 — Priyx
– Priyx, Commented Apr 23, 2021 at 12:56

patrickjlong1 · Accepted Answer · 2021-05-11 11:00:52Z

1

The simplest way to do it, in my opionion, is to read in the .txt file using pandas's read_csv() method and write out using the Dataframe.to_csv() method.

Below I've created a fully reproducible example recreating the OP's .txt file, reading it in and then writing out a new .csv file.

import pandas as pd

#Step 0: create the .txt file
file = open("file.txt","w")
input = '''"id":0
"value1":"234w-76-54"
"id":1
"value1":"2354w44-7-54"'''

file.writelines(input)
file.close() 

#Step 1: read in the .txt file and create a dataframe with some manipulation to 
#        get the desired shape
df = pd.read_csv('file.txt', delimiter=':', header=None)
df_out = pd.DataFrame({'id': df.loc[df.iloc[:, 0] == 'id'][1].tolist(),
'value1': df.loc[df.iloc[:, 0] == 'value1'][1].tolist()})
print(df_out)

#Step 2: Create a .csv file
df_out.to_csv('out.csv', index=False)

Expected Outputted .csv file:

edited May 11, 2021 at 11:00

answered Apr 23, 2021 at 16:37

patrickjlong1

3,8231 gold badge21 silver badges34 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Tushar Sadhwani · Accepted Answer · 2021-04-23 12:50:40Z

0

findall returns a list. you probably want to use re.search or re.match depending on your usecase.

answered Apr 23, 2021 at 12:50

Tushar Sadhwani

5654 silver badges9 bronze badges

Comments

AnPla · Accepted Answer · 2021-04-25 08:18:58Z

If your log.txt really has that simple structure you can work with split():

import csv

with open("text.txt", "r") as file:
    file2 = csv.writer(open("file.csv", "w", newline=""), delimiter=",")
    file2.writerow(["id", "value1"])
    for line in file:
        if "id" in line:
            ids = int(line.split(":")[-1])
        else:
            value = line.split(":")[-1].split('"')[1]
            file2.writerow([ids, value])

The resulting csv will contain the following:

id,value1
0,234w-76-54
1,2354w44-7-54

The comma as separator is set by the delimiter argument in the csv.writer call.

First look for a line with "id" in it. This line can easily be splitted at the :. This results in a list with two elements. Take the last part and cast it into an integer.

If no "id" is in the line, it is an "value1" line. First split the line at the :. Again take the last part of the resulting list and split it at ". This results again in a list with three elements, We need the second one.

Collectives™ on Stack Overflow

How to write to csv file from a text file

3 Answers 3

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related