Add outer join option for dataframe merge

[stress-tess]

A customer requested we add df.merge(df2, how='outer') like is possible in pandas

An (ugly) workaround for the time being:

>>> df1 = ak.DataFrame(
...:     {
...:         "key": ak.arange(4),
...:         "value1": ak.array(["A", "B", "C", "D"]),
...:         "value3":ak.arange(4,0,-1)
...:     }
...: )
...:
...: df2 = ak.DataFrame(
...:    {
...:         "key": ak.arange(2, 6, 1),
...:         "value1": ak.array(["A", "B", "D", "F"]),
...:         "value2": ak.array(["apple", "banana", "cherry", "date"]),
...:     }
...: )

>>> left_join = df1.merge(df2, how="left", on="key")

>>> right_join = df1.merge(df2, how="right", on="key")

>>> def convert_int_col_to_float(df):
...:     for col in df.columns:
...:         if df[col].dtype is ak.int64:
...:             df[col] = ak.cast(df[col], ak.float64)

>>> convert_int_col_to_float(left_join)

>>> convert_int_col_to_float(right_join)

>>> my_df = ak.DataFrame.append(left_join, right_join).drop_duplicates(subset=["key"]).reset_index()

>>> my_df
   key value1_y  value2 value1_x  value3
0  0.0      nan     nan        A     4.0
1  1.0      nan     nan        B     3.0
2  2.0        A   apple        C     2.0
3  3.0        B  banana        D     1.0
4  4.0        D  cherry      nan     NaN
5  5.0        F    date      nan     NaN (6 rows x 5 columns)

# verify it matches pandas
>>> df1.to_pandas().merge(df2.to_pandas(), how="outer", on="key")
   key value1_x  value3 value1_y  value2
0    0        A     4.0      NaN     NaN
1    1        B     3.0      NaN     NaN
2    2        C     2.0        A   apple
3    3        D     1.0        B  banana
4    4      NaN     NaN        D  cherry
5    5      NaN     NaN        F    date