CSV

Do this import and set up a path for all examples below:

import csv


path = "demo.csv"

Resources

• csv module in docs.

Newlines

To avoid issues with newlines, especially when writing out, it is a good ideas as per the documentation to specify an empty newline character before making your reader or writer object:

with open('some.csv', newline='') as f:
    # ...

Read CSV file

Given data:

First,Second,Third
abc,def,123
xyz,,1000

Read as a list of lists

Using csv.reader:

This approach will return each row as a list. You need to reference the columns by index. This is fragile as if the columns are reordered or deleted then you’ll get the wrong data.

with open(path, newline='') as f_in:
    reader = csv.reader(f_in)

Data will be like this:

['First', 'Second', 'Third']
["abc", "def", 123]
['xyz', '', '1000']

Read as a list of dictionaries

Using DictReader.

This approach is robust - the fields will be inferred from the first line of the CSV, or you can override them or set them if they are missing. Then you can reference cell values on each row by field name, rather than by index. This is stable even if the columns are reordered or columns are deleted.

with open(path) as f_in:
    reader = csv.DictReader(f_in)

reader.fieldnames
# ['First', 'Second', 'Third']

Here are two rows as dictionaries:

{'First': 'abc', 'Second': 'def', 'Third': '123'}
{'First': 'xyz', 'Second': '', 'Third': '1000'}

Iterate

A few ways to iterate when reading a CSV.

• Using list to read the entire file. This is not memory-efficient, but useful if you want to read the whole file first and then do operations on the data to say filter, reorder, or transform, in the Functional Programming style.

    with open(path) as f_in:
        reader = csv.reader(f_in)
        csv_rows = list(reader)
  
    len(csv_rows)
    # 3
  

• for loop on the iterator. This is memory-efficient as the row is only read when it is needed. You could replace print with some function call that might request an API or insert into a DB.

    with open(path) as f_in:
        reader = csv.reader(f_in)
  
        for row in reader:
            print(row)
            print(row[0])
  
    with open(path) as f_in:
        reader = csv.DictReader(f_in)
  
        for row in reader:
            print(row)
            print(row["First"])
  

• Use next on the iterator.

    f_in = open(path)
    reader = csv.reader(f_in)
  
    reader.line_num
    # 0
    reader.dialect.delimiter
    # ','
  
    next(reader)
    # ['First', 'Second', 'Third']
    reader.line_num
    # 1
    next(reader)
    # ["abc", "def", 123]
    next(reader)
    ['xyz', '', '1000']
    next(reader)
    # Traceback (most recent call last):
    #   File "<stdin>", line 1, in <module>
    # StopIteration
  
    # Reset to the start of the file.
    f_in.seek(0)
    next(reader)
    # ['First', 'Second', 'Third']
    reader.line_num
    # 3
  

Write CSV file

Write a list of lists

Using csv.writer.

import csv


header = ["Foo", "Bar"]
rows = [
  ["foo", "bar"],
  ["fizz", "buzz"]
]

with open(path, 'w', newline='') as f_out:
    writer = csv.writer(f_out)
    write.writerow(header)
    write.writerows(rows)

print(f"Wrote to {path}, rows: {len(rows)}")

You can also write individual rows:

write.writerow(row)

Write a list of dictionaries

Using csv.DictWriter.

import csv


header = ["Foo", "Bar"]
rows = [
    {"Foo": "a", "Bar" 1},
    {"Foo": "b", "Bar": 2}
}

with open(path, 'w') as f_out:
    writer = csv.DictWriter(f_out, fieldnames=header)
    write.writeheader()
    write.writerows(rows)

Request CSV over HTTP

import csv

import requests 


CSV_URL = '' 

with requests.Session() as s:
    download = s.get(CSV_URL) 
    decoded_content = download.content.decode('utf-8') 
    reader = csv.reader(decoded_content.splitlines(), delimiter=',') 
    csv_data = list(reader)