Pyplyn is an MIT Licensed simple flow-based data handling structure for making data handling repetitive tasks, easily without repeating yourself for every different scenario.
It is based on Python's lovely generators, so for every data flow into the pipe is in an iterative fashion. It is currently used in a research project to handle some repetitive daily tasks. (Moving, filtering, altering the data)
Still the pyplyn module that is used in the project has some dirty but useful components like progressbar, ML based classification filter and so on, with this simple library, I think there can be a common simple ground for handling our repetitive tasks.
In order to install pyplyn, just simply:
pip install pyplyn
Or alternatively, download the package from pypi, extract and execute:
python setup.py install
Pyplyn aims to make data handling in a flow based fashion:
import pyplyn as p
pipe = p.Pipe()
pipe.add(p.LineReader("hello.txt"))
pipe.add(p.LambdaFilter(lambda line: len(line) < 50))
pipe.add(p.LineWriter("small_hello.txt"))
pipe.run()
You can even write your own Pyp modules as simple as this:
import pyplyn as p
import pymongo
class MongoCollection(p.InPypElement):
def __init__(self, db, collection):
self.collection = pymongo.MongoClient()[db][collection]
def grasp(self):
for document in self.collection:
yield document
Add this new pipe element to your current flow by:
pipe = p.Pipe()
pipe.add(MongoCollection("data","raw"))
pipe.add(p.LambdaExtension(lambda document: document["text"])
pipe.add(p.LineWriter("data_text.txt"))
Has a simple API:
Simplicity is the ultimate aim, however there is a experimental branch multi-pyplyn in the project currently, I am just experimenting to find an elegant and easy to use API for the library.
Sorry, it is currently not available, but I recommend you to check the source, it is pretty straightforward for now.
Any contribution is welcome.