pd.lib 0.0.6

Library of small but useful modules

Short module description
=========================

Module pd.lib content some simple modules
and fucntions.

Description of modules
----------------------

pd.lib.heapsort
...............

The module is simple lazy sort implementation realised therewith heapq
module. Items will be sorted only when accepted and using of computation
resource is minimazed.

Two classes are provided by module:

HeapSort
On initialization accepted item list and comparing
function;

HeapSortByIndex
On initialization accepted item list and index. List
items mapped on index values and can be compared by them. In this
case third parameter - revert - are used. Revert can be True or
False. If revert is True then array will be sorted in reverse order.
Parameter revert is False by default.

HeapSortByIndexSafe
It class liked on HeapSortByIndex above, but use index in extremely safe
manner: if item omited from index it mapped on infinity and any item are
comparable;

Classes, mentioned above, to provide methods:

chunk(n)
Return first n items from begin sorted list.

pd.lib.utility
..............

The module contents followed useful functions:

name2klass
Return klass by name

klass2name
Return name by klass


pd.lib.topsort
..............

The module contents classes developed to provide different kind of topologic
sort. All classes are use as:

TopSort([<Parameters>])(<Items>), где :

Parameters
Sort execution parameters;

Items
Each item is a pair of object reference and object reference list which one depend;

Followed classes are provided by module:

TopSort
Simple topologic sort;

TopSortIndex
Simple topologic sort for any kind of object;

TopSortFuzzy
Fuzzy topologic sort;

TopSortFuzzyIndex
Fuzzy topologic sort for any kind of object;


Fuzzy topologic sort can transform to linear ordered array graph with
cyclomatic value more then zero. Best results achieved on litle values: the
sort is for tree with a few cycles.

pd.lib.lazy
...........
The module contents **Lazy** decorator class used to cache return value on
methods depend on argumetnts. Be carryfully, method arguments are to be
hashable.


Sample of use:

>>> from pd.lib.lazy import Lazy
>>>
>>> class A(object) :
... @Lazy
... def mul(self,a,b) :
... print a,b,"compute"
... return a*b
...
>>> a=A() a.mul(1,2)
1 2 compute 2
>>> a.mul(2,3) 2 3 compute 6
>>> a.mul(1,2) 2 a.mul(2,3) 6
>>>

Using such version of **lazy** decorator get processes incredible speed, but
more memory will be used.


pd.lib.fuzzyobject
..................

The module provide FuzzyObject class. Inheritance from FuzzyObject to allow
forget about typografic errors in class names::

>>> from pd.lib.fuzzyobject import FuzzyObject class A(FuzzyObject) :
... def commmon_function(self) :
... print 'common'
... def other_method(self) :
... print 'method'
...
>>> A().coon_functi() common A().oth_m() method
>>

This base class is toy, but it's worked. More complicated sample published
in pd/lib/fuzzyobject/fuzzyobject_sample.py, some pro and contra of such
programming are viewed from.

Conclusion
-----------

Main target of this product to carry all small, frequently used libraries.
This library must be installed on use other our products by depend. But I
think product have not got independent significance.


* Mon May 12 2008 Andrey Orlov <dbdtdreambot.ru> 0.0.6
- Module Lazy added
- A few documentation wrote

* Sun Jan 13 2008 Andrey Orlov <[e-mail]> 0.0.5
- Module fuzzyobject added

* Sat Dec 22 2007 Andrey Orlov <[e-mail]> 0.0.4
- Module topsort added

* Mon Dec 10 2007 Andrey Orlov <[e-mail]> 0.0.3
- Module utility added

* Tue Nov 13 2007 Andrey Orlov <[e-mail]> 0.0.2
- HeapSortIndexSafe class added

* Mon Nov 12 2007 Andrey Orlov <[e-mail]> 0.0.1
- first official release