PyPerler

Quick introduction

PyPerler gives you the power of CPAN, and your perl packages, in Python. Using Perl stuff is as easy as:

>>> import pyperler; i = pyperler.Interpreter()
>>> # use a CPAN module
>>> Table = i.use('Text::Table')
>>> t = Table("Planet", "Radius\nkm", "Density\ng/cm^3")
>>> _ = t.load(
...    [ "Mercury", 2360, 3.7 ],
...    [ "Venus", 6110, 5.1 ],
...    [ "Earth", 6378, 5.52 ],
...    [ "Jupiter", 71030, 1.3 ],
... )
>>> print( t.table() )
Planet  Radius Density
        km     g/cm^3 
Mercury  2360  3.7    
Venus    6110  5.1    
Earth    6378  5.52   
Jupiter 71030  1.3    
<BLANKLINE>

If you install the Class::Inspector CPAN package, then PyPerler will even get you introspection for use in IPython.

Quick install

You'll need cython, perl and the perl headers. On Ubuntu, these are just

$ sudo apt-get install cython perl-base perl

Then compilation and installation is

$ python setup.py build && sudo python setup.py install

This works both for python2 and python3.

Overview

PyPerler allows you to seemlessly interact with Perl code from Python. Python dicts and Perl hashes map into eachother transparently, and so do lists and arrays.

The obvious complications for bridging the differences between Python and Perl are

Weak typing; there's no difference between 1 and "1" in Perl
Void context vs. scalar context vs. list context

For choosing a strategy to deal with this, we follow the Philosophy of Python:

Explicit is better than implicit.
(...)
In the face of ambiguity, refuse the temptation to guess.

On the other hand, we are not shying away from Perl's There Is More Than One Way To Do It. For example, these both give the same results:

>>> [int(x) for x in i["1 .. 10"]]
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
>>> i["1 .. 10"].ints()
(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)

Weak typing

Whenever we return a scalar value from Perl, this remains boxed as an object of type pyperler.ScalarValue. This can either be explicitly cast to bytes, str, int, or float, or that can happen implicitly through type inference in binary operations. We 'refuse the tempation to guess' when applying a binary operator to TWO pyperler.ScalarValue instances: this raises a TypeError.

In case the scalar value is a hashref resp. an arrayref, it supports all the operations that Python's built-in dict object resp. list object support.

In case the scalar value is a blessed reference, it additionally supports methods calls. When there's a naming clash between the Python built-in method and the blessed reference's methods, the Python ones take precedence. The "hidden" methods are still available through an indexable attribute 'F', like so:

>>> arrayref = i('package a; sub append { 42 }; bless [], "a"')
>>> arrayref.F['append'](23)
42

Void context vs. scalar context vs. list context

Any pyperler object that represents a Perl callable has methods void_context, list_context, and scalar_context. In addition, __call__ (ie the function call operator) is a shorthand for scalar_context.

Similarly, the Interpreter object has methods void_context, list_context, and scalar_context. They take a single string as an argument, which is evaluated as Perl code. __call__ is a shorthand for scalar_context, and also __getitem__ is a shorthand for list_context. (This last thing wouldn't be useful for callables, because the case of zero arguments is a syntax error in Python.)

void_context returns None scalar_context returns a pyperler.ScalarValue list_context returns a pyperler.ListValue. This is just a python tuple of pyperler.ScalarValue objects, with a few convenience methods for casting all of them.

License

The file PythonObject.pm contains code from from PyPerl, which is Copyright 2000-2001 ActiveState. This code can be distributed under the same terms as Perl, which includes distribution under the GPL version 1 or later.

For your convenience, the General Public License version 3 is included as the file COPYING.

tkluck / pyperler

readme