slicing a bsddb table, eg. for rec in bsddb["AArdvark":"zebra"]: print rec

[Skip Montanaro]

    Neville> # I was expecting a slice of an index file to yield a
    Neville> # generator so not all the records need to be read from disk....

Slicing is a feature of sequence types, not mapping types.

    >>> import string
    >>> d = dict(zip(string.lowercase, string.uppercase))
    >>> d
    {'a': 'A', 'c': 'C', 'b': 'B', 'e': 'E', 'd': 'D', ...}
    >>> d["a":"z"]
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    TypeError: unhashable type
    >>> import UserDict
    >>> d2 = UserDict.UserDict(d)
    >>> d2
    {'a': 'A', 'c': 'C', 'b': 'B', 'e': 'E', 'd': 'D', ...}
    >>> d2["a":"z"]
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
      File "/Users/skip/local/lib/python2.5/UserDict.py", line 17, in __getitem__
        def __getitem__(self, key): return self.data[key]
    TypeError: unhashable type

The unhashable type it's referring to is the slice object generated by the
"a":"z" notation:

    >>> hash(slice("a", "z"))
    Traceback (most recent call last):
      File "<stdin>", line 1, in ?
    TypeError: unhashable type

If you look at help on the slice type, you'll see that it takes three args:

    class slice(object)
     |  slice([start,] stop[, step])
     |  
     | Create a slice object.  This is used for extended slicing
     | (e.g. a[0:10:2]).
     |  

Step sizes really only make sense for sequence indices.

Skip