Added special code to deal with Pandas DataFrames and NumPy RecordArrays

Also:
- Added data_types.py example

TODO:
- Updated README.md
- Add unit tests

Author: tleonhardt

examples/data_types.py

@@ -0,0 +1,78 @@
+#!/usr/bin/env python
+# coding=utf-8
+"""
+tableformatter supports and has been tested with the following tabular data types:
+- lists of lists or other iterables of iterables
+- two-dimensional NumPy array
+- list or another iterable of dicts (dict keys iterated through as column values)
+- dict of iterables (dict keys iterated through as rows where each key must be a hashable iterable)
+- NumPy record arrays (names as columns)
+- pandas.DataFrame
+- list or another iterable of arbitrary non-iterable objects (column specifier required)
+
+This example demonstrates tableformatter working with these data types in the simplest possible manner.
+"""
+import numpy as np
+import pandas as pd
+import tableformatter as tf
+
+iteralbe_of_iterables = [[1, 2, 3, 4],
+                         [5, 6, 7, 8]]
+print("Data type: iterable of iterables")
+print(iteralbe_of_iterables)
+print(tf.generate_table(iteralbe_of_iterables))
+
+np_2d_array = np.array([[1, 2, 3, 4],
+                        [5, 6, 7, 8]])
+print("Data type: NumPy 2D array")
+print(np_2d_array)
+print(tf.generate_table(np_2d_array))
+
+np_rec_array = np.rec.array([(1, 2., 'Hello'),
+                             (2, 3., "World")],
+                            dtype=[('foo', 'i4'), ('bar', 'f4'), ('baz', 'U10')])
+print("Data type: Numpy record array")
+print(np_rec_array)
+print(tf.generate_table(np_rec_array))
+
+d = {'col1': [1, 5], 'col2': [2, 6], 'col3': [3, 7], 'col4': [4, 8]}
+pandas_dataframe = pd.DataFrame(data=d)
+print("Data type: Pandas DataFrame")
+print(pandas_dataframe)
+print(tf.generate_table(pandas_dataframe))
+
+iterable_of_dicts = [ {1: 'a', 2: 'b', 3: 'c', 4: 'd'},
+                      {5: 'e', 6: 'f', 7: 'g', 8: 'h'}]
+print("Data type: iterable of dicts (dict keys iterated through as column values)")
+print(iterable_of_dicts)
+print(tf.generate_table(iterable_of_dicts))
+
+dict_of_iterables = {(1, 2, 3, 4): 'a',
+                     (5, 6, 7, 8): 'b'}
+print("Data type: dict of iterables (dict keys iterated through as rows where each key must be a hashable iterable)")
+print(dict_of_iterables)
+print(tf.generate_table(dict_of_iterables))
+
+
+class MyRowObject(object):
+    """Simple object to demonstrate using a list of non-iterable objects with TableFormatter"""
+    def __init__(self, field1: int, field2: int, field3: int, field4: int):
+        self.field1 = field1
+        self.field2 = field2
+        self._field3 = field3
+        self.field4 = field4
+
+    def get_field3(self):
+        """Demonstrates accessing object functions"""
+        return self._field3
+
+
+rows = [MyRowObject(1, 2, 3, 4),
+        MyRowObject(5, 6, 7, 8)]
+columns = (tf.Column('Col1', attrib='field1'),
+           tf.Column('Col2', attrib='field2'),
+           tf.Column('Col3', attrib='get_field3'),
+           tf.Column('Col4', attrib='field4'))
+print("Data type: iterable of arbitrary non-iterable objects")
+print(rows)
+print(tf.generate_table(rows, columns))

tableformatter.py

@@ -615,6 +615,27 @@ def generate_table(rows: Iterable[Union[Iterable, object]],
     :param row_tagger: decorator function to apply per-row options
     :return: formatted string containing the table
     """
+    # Extract column headers if this is a NumPy record array and columns weren't specified
+    if not columns:
+        try:
+            import numpy as np
+        except ImportError:
+            pass
+        else:
+            if isinstance(rows, np.recarray):
+                columns = rows.dtype.names
+
+    # Deal with Pandas DataFrames not being iterable in a sane way
+    try:
+        import pandas as pd
+    except ImportError:
+        pass
+    else:
+        if isinstance(rows, pd.DataFrame):
+            if not columns:
+                columns = rows.columns
+            rows = rows.values
+
     show_headers = True
     use_attrib = False
     if isinstance(columns, Collection) and len(columns) > 0: