fetching

Fetching is a small library that enables easy lazy initialization.

To use Fetching, define your class with an __init__ function that stores all necessary seed data, and a fetch function that does the computationally expensive initializations. Don't call fetch from __init__. If you designed the class correctly, the following sequence should result in a fully initialized class instance:

someClass = SomeClass(args) # can have args
someClass.fetch()           # can't have args

An example:

# not compatible with Fetching
class Example(object):
    def __init__(self,i):
        self.parsed = []
        for e in range(i):
            self.parsed.append(e)
            time.sleep(0.5)

    def getParsed(self):
        return self.parsed

# compatible with Fetching
class Example(object):
    def __init__(self,i):
        self.i = i

    def fetch(self):
        self.parsed = []
        for e in range(self.i):
            self.parsed.append(e)
            time.sleep(0.5)       # the slow process is moved to fetch()

    def getParsed(self):
        return self.parsed

A more practical example:

# not compatible with Fetching
class Webpage(object):
    def __init__(self,url):
        self.page = requests.get(url).text

    def pageText(self):
        return self.page

# compatible with Fetching
class Webpage(object):
    def __init__(self,url):
        self.url = url

    def fetch(self):
        self.page = requests.get(url).text

    def pageText(self):
        return self.page

Once your class is formatted correctly, use simple wrapper classes to give the desired behavior:

from fetching import FetchWhenNeeded

example = Example(5)
example = FetchWhenNeeded(example)

# example.parsed doesn't exist yet

print(example.getParsed()) # runs Example.fetch and returns .getParsed()

from fetching import FetchInBackground

example = Example(5)
example = FetchInBackground(example)

# Example.fetch is being run in a separate thread

# some time later...
print(example.getParsed())