BUG: Fixed issue #5156: segfault on read_csv

[guyrt]

closes #5156

Our c parser was not checking array bounds.

Example:

In [1]: from StringIO import StringIO; import pandas as pd

In [2]: data = """\     
1,2,3                      
4,,6
7,8,9                      
10,11,12\n"""

In [3]: df = pd.read_csv(StringIO(data), header=0, names=['a', 'b', 'c', 'd'], engine='c')
CParserError: Too many columns specified: expected 4 and found 3

Previously, this would segfault.

[guyrt]

This was requested here: #5211

[jreback]

thanks! helpful as always