How to read csv file in Python if the file can but does not have to contain headers in first line?

I am trying to read a csv file in Python using csv.DictReader. I need to handle both cases:

1. when headers are present in the first line:

col1,col2
foo,bar

2. when they are omitted:

foo,bar

I can assume that headers are always col1,col2 if they are provided.
I tried to use fieldnames parameter, but then headers are treated as values when they are present:

reader = csv.DictReader(csv_file, fieldnames=['col1','col2'])
print(list(reader))

Output:

[{'col1': 'col1', 'col2': 'col2'}, {'col1': 'foo', 'col2': 'bar'}]

instead of:

[{'col1': 'foo', 'col2': 'bar'}]

Using csv.DictReader without fieldnames parameter works when headers are present but returns empty list when there are no headers.

You might look at reading the first row and if it is not the same as your expected headers, go back to the start of the file via seek()

Given "in1.csv"

col1,col2
foo,bar

Given "in2.csv"

foo,bar

Then

import csv

fieldnames=['col1','col2']
for file_name in ["in1.csv", "in2.csv"]:
    with open(file_name, "r") as file_in:
        if file_in.readline().strip() != ",".join(fieldnames):
            file_in.seek(0)
        results = list(csv.DictReader(file_in, fieldnames=fieldnames))
    print(results)

Should give you:

[{'col1': 'foo', 'col2': 'bar'}]
[{'col1': 'foo', 'col2': 'bar'}]

You can use the DictReader and supply it the fieldnames you ultimately want (to handle the implicit no-header case), then handle the explicit case where there was a header and now you need to skip it:

import csv

header_vals = ["col1", "col2"]
header_row = {x: x for x in header_vals}

for fname in ["input1.csv", "input2.csv"]:
    with open(fname, newline="") as f:
        reader = csv.DictReader(f, fieldnames=header_vals)

        print(f"{fname}:")
        for row in reader:
            if row == header_row:
                print("  skipped explicit header")
                continue
            print(f"  {row}")

For these two files:

input1.csv
==========
col1,col2
r1c1,r1c2
r2c1,r2c2

input2.csv
==========
r1c1,r1c2
r2c1,r2c2

that prints:

input1.csv:
  skipped explicit header
  {'col1': 'r1c1', 'col2': 'r1c2'}
  {'col1': 'r2c1', 'col2': 'r2c2'}
input2.csv:
  {'col1': 'r1c1', 'col2': 'r1c2'}
  {'col1': 'r2c1', 'col2': 'r2c2'}