Normalizing Terms

Main function is dinsort.normalize. It generates a normalized form of any string term given:

>>> from dinsort import normalize
>>> normalize("Löblich")
'loblich'

Variants are defined as constants:

>>> from dinsort import VARIANT1, VARIANT2
>>> normalize("Müßig", variant=VARIANT1)
'mussig'

>>> normalize("Müßig", variant=VARIANT2)
'muessig'

Terms are normalized to lower-case by default. You can request case-sensitiveness:

>>> normalize("Maße", case_sensitive=True)
'Masse'

Sorting

Normalized terms can easily be used for sorting lists of terms:

>>> sorted(["fas", "fair", "fär"], key=lambda x: normalize(x))
['fair', 'fär', 'fas']

We provide a shortcut to avoid (sometimes not easy to read) lambda statements with normalize in your code. Use sort_func for that:

>>> from dinsort import sort_func
>>> sorted(["fas", "fair", "fär"], key=sort_func())
['fair', 'fär', 'fas']

The sort_func accepts the keywords of normalize:

>>> sorted(["Muße", "muß"], key=sort_func())
['muß', 'Muße']

>>> sorted(["Muße", "muß"], key=sort_func(case_sensitive=True))
['Muße', 'muß']

>>> sorted(["far", "fähre"], key=sort_func())
['fähre', 'far']

>>> sorted(["far", "färe", "fast"], key=sort_func())
['far', 'färe', 'fast']

>>> sorted(["far", "färe", "fast"], key=sort_func(variant=VARIANT2))
['färe', 'far', 'fast']

User Install

The latest release of dinsort can be installed via pip:

$ pip install dinsort

The exact way depends on your operating system.

Developer Install

Developers clone the sources from github:

$ git clone https://github.com/ulif/dinsort.git

Create a virtual env (Python versions >= 2.6 supported):

$ cd dinsort
$ virtualenv py27
$ source py27/bin/activate

and install dependencies:

(py27) python setup py dev

You can run tests with py.test:

(py27) py.test

or with tox:

(py27) pip install tox
(py27) tox

0.2 (2015-10-14)

• Add sort_func. It is as shortcut for lambda expressions with normalize.

• Support pypy officially.

0.1 (2015-10-13)

• Initial version.