Table.print_table errors after homogenize with ints

[nbedi]

Found while investigating #699

Seems related to #411. My guess is homogenizing isn't converting int values from default_row into Decimal.

[jpmckinney]

I'm not sure that this is a bug. agate doesn't really have much functionality for adding explicit values to an existing table. The user must provide a default row with decimals.

This errors:

import agate
data = [
    {'year': 1997, 'female_count': 2, 'male_count': 1},
    {'year': 2000, 'female_count': 3, 'male_count': 3}
]
key = 'year'
expected_values = (1997, 1998, 2000)
default_row = (0, 0)
t = agate.Table.from_object(data)
t.homogenize(key, expected_values, default_row).print_table()

This works:

import agate
from decimal import Decimal
data = [
    {'year': 1997, 'female_count': 2, 'male_count': 1},
    {'year': 2000, 'female_count': 3, 'male_count': 3}
]
key = 'year'
expected_values = [Decimal(num) for num in (1997, 1998, 2000)]
default_row = [Decimal(num) for num in (0, 0)]
t = agate.Table.from_object(data)
t.homogenize(key, expected_values, default_row).print_table()

[jpmckinney]

I got part-way with this patch:

diff --git a/agate/table/homogenize.py b/agate/table/homogenize.py
index 0e04ec4..57ceaf4 100644
--- a/agate/table/homogenize.py
+++ b/agate/table/homogenize.py
@@ -53,8 +53,16 @@ def homogenize(self, key, compare_values, default_row=None):
         if any(not utils.issequence(compare_value) for compare_value in compare_values):
             compare_values = [[compare_value] for compare_value in compare_values]
 
-    column_values = [self._columns.get(name) for name in key]
-    column_indexes = [self._column_names.index(name) for name in key]
+    column_values = []
+    column_indexes = []
+    for i, name in enumerate(key):
+        column = self._columns.get(name)
+        column_values.append(column)
+        column_indexes.append(self._column_names.index(name))
+        for values in compare_values:
+            values = list(values)
+            if i < len(values):
+                values[i] = column.data_type.cast(values[i])
 
     column_values = zip(*column_values)
     differences = list(set(map(tuple, compare_values)) - set(column_values))

However, it causes some tests to fail. It also doesn't deal with the default row, which is much more complicated to cast, since you'd have to map the default row's indices to the table's indices to cast them correctly.

To test, we can just add homogenized.print_table() in test_homogenize.py.