read_fwf engine issue

[pmart123]

I am trying to use read_fwf to read in a fixed with text data file. I am hoping to use the dtype argument, which only works with the 'c' engine parser. I tried something like the following:

df = pd.read_fwf(dpath,widths=widths,header=0,skiprows=[0,2,3],thousands=',',engine='c',dtype={'cusip':str})

I received an error that the dtype option is not supported with the 'python-fwf' engine. As I understand, in the API docs, it says the c engine is necessary for the dtype argument. Is there no 'c' engine for read_fwf at the current moment, is is something else going on? I am using Pandas 14.1, python 3.4 as part of the Ancacondas distribution.

[TomAugspurger]

Correct, there is not a dtype argument for read_fwf.

[jreback]

fyi, this is a dupe of #7141

[jreback]

master issue is #5232

essentially someone needs to implement dtype=... for the python parser (its all python code, so pretty straightforward)