BUG: sniffing a csv raises with only a header

[Nodd]

I'm trying to read a bunch of csv files, but some of them have a header line without any data. I expect to get empty data but Pandas raises an exception on these files:

Traceback (most recent call last):
  File "/usr/bin/positionneur_rng", line 9, in <module>
    load_entry_point('pymela==0.1.0', 'console_scripts', 'positionneur_rng')()
  File "/data/libs/python/pymela_scripts/positionneur_rng.py", line 203, in main
    = importer_csv_pandas(fichier)
  File "/data/libs/python/pymela_scripts/positionneur_rng.py", line 179, in importer_csv_pandas
    engine='python')
  File "/usr/lib/python3.4/site-packages/pandas/io/parsers.py", line 443, in parser_f
    return _read(filepath_or_buffer, kwds)
  File "/usr/lib/python3.4/site-packages/pandas/io/parsers.py", line 235, in _read
    return parser.read()
  File "/usr/lib/python3.4/site-packages/pandas/io/parsers.py", line 686, in read
    ret = self._engine.read(nrows)
  File "/usr/lib/python3.4/site-packages/pandas/io/parsers.py", line 1506, in read
    self.index_names)
  File "/usr/lib/python3.4/site-packages/pandas/io/parsers.py", line 2129, in _get_empty_meta
    index = MultiIndex.from_arrays([[]] * len(index_col),
TypeError: object of type 'bool' has no len()

I used this command: pd.read_csv(fichier_data, sep=None, index_col=False, engine='python')
The file content is a single line:

TEMPS;MODE;STATUS;LATITUDE;LONGITUDE;ALTITUDE;VNORD;VEST;VHAUT;ROLL;PITCH;YAW;DELTA YAW;RATEX;RATEY;RATEZ;ACCX;ACCY;ACCZ;GISEMENT;SITE;CONAZ;CONEL;POSAZ;POSEL;LATCIBLE;LONGICIBLE;ALTICIBLE;DISTANCE;

>>> pandas.show_versions()

INSTALLED VERSIONS
------------------
commit: None
python: 3.4.1.final.0
python-bits: 64
OS: Linux
OS-release: 3.10.43-1-MANJARO
machine: x86_64
processor: 
byteorder: little
LC_ALL: None
LANG: fr_FR.UTF-8

pandas: 0.14.0
nose: 1.3.3
Cython: 0.20.2
numpy: 1.8.1
scipy: 0.14.0
statsmodels: None
IPython: 2.1.0
sphinx: 1.2.2
patsy: None
scikits.timeseries: None
dateutil: 2.2
pytz: 2014.4
bottleneck: 0.8.0
tables: 3.1.1
numexpr: 2.4
matplotlib: 1.3.1
openpyxl: None
xlrd: 0.9.3
xlwt: None
xlsxwriter: 0.5.5
lxml: 3.3.5
bs4: 4.3.2
html5lib: 0.999
bq: None
apiclient: None
rpy2: None
sqlalchemy: 0.9.6
pymysql: None
psycopg2: None

[jreback]

you are specifying a separator of None, using sep=';' seems to work, no? (I have an extra column as I have a trailing ;). The engine doesn't matter.

In [4]: pd.read_csv(StringIO(data), sep=';', index_col=False)
Out[4]: 
Empty DataFrame
Columns: [TEMPS, MODE, STATUS, LATITUDE, LONGITUDE, ALTITUDE, VNORD, VEST, VHAUT, ROLL, PITCH, YAW, DELTA YAW, RATEX, RATEY, RATEZ, ACCX, ACCY, ACCZ, GISEMENT, SITE, CONAZ, CONEL, POSAZ, POSEL, LATCIBLE, LONGICIBLE, ALTICIBLE, DISTANCE, Unnamed: 29]
Index: []

[Nodd]

I set the engine to remove a warning.

I'll try to set the separator, but I expected pandas to guess it.

[jreback]

well the warning is telling u that u are not doing the right thing
u need to specify a sep
not trivial to sniff (though it can be done) just not recommended

[jreback]

this might be a sniffing error (as it in theory should work).

[sergey-komissarov]

Hello, we are facing the same bug with pandas 0.15.2, simple way to reproduce:

>pandas.io.parsers.read_csv(StringIO("a,b,c"), index_col=False, engine="python")
TypeError: object of type 'bool' has no len()
>pandas.io.parsers.read_csv(StringIO("a,b,c"), index_col=False, engine="c")
Empty DataFrame
Columns: [a, b, c]
Index: []

now we patched _get_empty_meta function in pandas.io.parsers to treat False and None the same way, but I wonder if this is the right solution.