==============================
Jeep - a General Purpose Class
==============================

$Id$

  >>> from cybertools.util.jeep import Jeep
  >>> jeep = Jeep()

  >>> jeep.first = 'first value'
  >>> jeep.second = 'second value'

  >>> jeep.first
  'first value'

In addition to the usual access via dot notation all attributes can be
accessed via dictionary notation:

The third type of interface provided by Jeep objects is the sequence or
iterator interface. Converting a jeep object to a list iterates over its
values (that is different from the dictionary behaviour, but is what
you want usually; use the ``.keys()`` method to get at the keys, see below):

  >>> list(jeep)
  ['first value', 'second value']

Direct index access to certain entries gives the corresponding value,
not the key:

  >>> jeep[1]
  'second value'

Changing Jeep Objects
---------------------

Assignment by dictionary or attribute access appends the newly assigned
attribute:

  >>> jeep['third'] = 'third value'
  >>> jeep.third
  'third value'

  >>> list(jeep)
  ['first value', 'second value', 'third value']

Assigning a new value to an already existing attribute does not change the
order but only changes the attribute's value

  >>> jeep.second = 'new second value'
  >>> list(jeep)
  ['first value', 'new second value', 'third value']
  >>> jeep[1]
  'new second value'

More Dictionary Methods
-----------------------

  >>> jeep.keys()
  ['first', 'second', 'third']

  >>> jeep.values()
  ['first value', 'new second value', 'third value']

  >>> jeep.items()
  [('first', 'first value'), ('second', 'new second value'), ('third', 'third value')]

  >>> jeep.get('second')
  'new second value'
  >>> jeep.get('fourth', 'default')
  'default'
  >>> jeep.get('fourth') is None
  True
  >>> jeep['fourth']
  Traceback (most recent call last):
  ...
  KeyError: 'fourth'

  >>> dict(jeep)
  {'second': 'new second value', 'third': 'third value', 'first': 'first value'}

  >>> jeep.setdefault('first', 'new first value')
  'first value'
  >>> jeep['first']
  'first value'
  >>> jeep.setdefault('fourth', 'new fourth value')
  'new fourth value'
  >>> jeep['fourth']
  'new fourth value'
  >>> len(jeep)
  4

  >>> jeep.update(dict(first='another first value', fifth='5th value'))
  >>> len(jeep)
  5

  >>> del jeep['fifth']
  >>> len(jeep)
  4
  >>> del jeep['fifth']
  Traceback (most recent call last):
  ...
  ValueError: ...not in list

More Methods and Operators
--------------------------

  >>> 'third' in jeep
  True

  >>> jeep.pop()
  'new fourth value'
  >>> jeep.pop()
  'third value'
  >>> len(jeep)
  2
  >>> jeep.pop('blubb')
  Traceback (most recent call last):
  ...
  KeyError: 'blubb'
  >>> jeep.pop('blubb', None)

  >>> 'third' in jeep
  False

  >>> jeep.index('second')
  1
  >>> jeep.index('third')
  Traceback (most recent call last):
  ...
  ValueError: ...not in list

Sequence Additions with Named Objects
-------------------------------------

Objects that have a ``__name__`` attribute can be appended
to a Jeep object as the dictionary key can be obtained from this attribute.

  >>> class Term(object):
  ...     def __init__(self, token, title=None, value=None):
  ...         self.__name__ = self.token = token
  ...         self.title = title or token
  ...         self.value = value or title or token

  >>> t1 = Term('term1', 'title 1')
  >>> jeep.append(t1)
  >>> jeep.keys()
  ['first', 'second', 'term1']
  >>> jeep.term1.title
  'title 1'

  >>> jeep.insert(1, Term('term2', 'title 2'))
  >>> jeep.keys()
  ['first', 'term2', 'second', 'term1']
  >>> jeep[1].title
  'title 2'

Inserting or appending an object with a name that's already present raises
an exception:

  >>> jeep.append(t1)
  Traceback (most recent call last):
  ...
  ValueError: ...already present

Constructors
------------

A simple way of creating a Jeep object is using a sequence
of key-value pairs.

  >>> jeep2 = Jeep((('1', '1st'), ('2', '2nd'), ('3', '3rd')))
  >>> list(jeep2)
  ['1st', '2nd', '3rd']

Using a Jeep object as the argument of the constructor yields a shallow
copy.

  >>> jeep3 = Jeep(jeep2)
  >>> jeep3['1'] is jeep2['1']
  True
  >>> jeep3['4'] = '4th'
  >>> len(jeep3)
  4
  >>> len(jeep2)
  3

More in-place changes
---------------------

Remove elements in a fault-tolerant way.

  >>> jeep.keys()
  ['first', 'term2', 'second', 'term1']
  >>> jeep.remove('third', 'second')
  >>> len(jeep)
  3
  >>> jeep.keys()
  ['first', 'term2', 'term1']
  >>> 'second' in jeep
  False

Reorder the sequence and remove all elements that are not
in the argument list.

  >>> jeep.select('term1', 'term2', 'nope')
  >>> jeep.keys()
  ['term1', 'term2']
  >>> 'first' in jeep
  False

You may even reposition a certain element.

  >>> jeep3.keys()
  ['1', '2', '3', '4']
  >>> jeep3.reorder(-1, '2', '4')
  >>> jeep3.keys()
  ['2', '1', '4', '3']