BUG: read_csv not guessing delimiter

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

import pandas as pd
file = "pandas_test.dat"

# Fails
df = pd.read_csv(file, delimiter=None)

# Works
from astropy.io import ascii
df = ascii.read(file)

Issue Description

pandas fails when trying to guess the delimiter (space) in the test file. astropy on the other hand is able to guess it correctly

The test file is here pandas_test.dat.zip

Expected Behavior

The delimiter should be correctly guessed

Installed Versions

pd.__version__
'1.4.2'

The line pd.show_versions() throws the following error:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/pandas/util/_print_versions.py", line 109, in show_versions
    deps = _get_dependency_info()
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/pandas/util/_print_versions.py", line 88, in _get_dependency_info
    mod = import_optional_dependency(modname, errors="ignore")
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/pandas/compat/_optional.py", line 138, in import_optional_dependency
    module = importlib.import_module(name)
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/importlib/__init__.py", line 127, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1030, in _gcd_import
  File "<frozen importlib._bootstrap>", line 1007, in _find_and_load
  File "<frozen importlib._bootstrap>", line 986, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 680, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 850, in exec_module
  File "<frozen importlib._bootstrap>", line 228, in _call_with_frames_removed
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/setuptools/__init__.py", line 8, in <module>
    import _distutils_hack.override  # noqa: F401
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/_distutils_hack/override.py", line 1, in <module>
    __import__('_distutils_hack').do_override()
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/_distutils_hack/__init__.py", line 72, in do_override
    ensure_local_distutils()
  File "/home/gabriel/miniconda3/envs/asteca/lib/python3.9/site-packages/_distutils_hack/__init__.py", line 59, in ensure_local_distutils
    assert '_distutils' in core.__file__, core.__file__
AssertionError: /home/gabriel/miniconda3/envs/asteca/lib/python3.9/distutils/core.py

[simonjayhawkins]

Thanks @Gabriel-p for the report.

it appears that pd.read_csv(io.StringIO(src), sep=None) raises Error: Could not determine delimiter whereas pd.read_csv(io.StringIO(src), delimiter=None) produces a DataFrame with one column.

According to the docs, delimiter is an alias for sep. The default for delimiter is documented as None (which is strange), whereas the default for sep is ','.

I suspect that explicitly specifying delimiter=None is being ignored and treated as not passed and the default for sep is being used (not checked this though)

Expected Behavior

The delimiter should be correctly guessed

In any case, if the seperator/delimitor cannot be determined with sep=None, the same should be true for delimiter=None, so the expected behavior is to raise Error: Could not determine delimiter

contributions and PRs welcome.

[dimitra-karadima]

Right now, tests are failing and the behavior is not as I described above, so I guess I have to wait until it is fixed in the main branch. My question is that it is not very clear to me which test file is responsible for this change, so I do not know where to add the test cases for this bugfix.