Whitespace validation does not work on columns that only contain numbers

[wolces]

If all entries in a column are numeric, then whitespace validation will not find errors in any entries in that column. If a single entry in a column is non-numeric, then whitespace validation will work on all entries in that column. For example:

import pandas as pd
from io import StringIO
from pandas_schema import Column, Schema
from pandas_schema.validation import LeadingWhitespaceValidation, TrailingWhitespaceValidation

schema = Schema([
    Column('col1', [LeadingWhitespaceValidation(), TrailingWhitespaceValidation()]),
    Column('col2', [LeadingWhitespaceValidation(), TrailingWhitespaceValidation()])
])

test_data = pd.read_csv(StringIO('''col1,col2
1,3
4,p
 2 ,3
3, 9
1 ,3
6,2 
'''))

errors = schema.validate(test_data)

for error in errors:
    print(error)

returns

{row: 3, column: "col2"}: " 9" contains leading whitespace
{row: 5, column: "col2"}: "2 " contains trailing whitespace

[multimeric]

The best way to help me with this would be to add a test for LeadingWhitespaceValidation or TrailingWhitespaceValidation in test/test_validation.py that currently fails for this example. Then I can very quickly write a fix for it.

[multimeric]

I've had a look into this, and it's not exactly a bug in PandasSchema. The problem is, pd.read_csv does some automatic type conversion, and sees that, because series 1 entirely consists of integers, it should be converted into an integer series, and thus it loses the whitespace.

If you make sure that everything is parsed as a string, by setting the dtype manually, the validations will work as expected:

test_data = pd.read_csv(StringIO('''col1
3
3
 9
3
2
'''), dtype=str)

[multimeric]

I'll try to update the documentation to make this clearer