sequential-executor 1.0.0

A concurrent.futures.Executor implementation that performs sequential execution

sequential-executor

A concurrent.futures.Executor implementation that performs sequential execution.

While this micro-library may seem a little oxymoronic (Executors are, after all, meant to execute calls asynchronously), it does have a fairly well-defined use case.

Use case

A Python package that performs multiple operations via callback(s) defined by implementors using that package. The package allows the implementor to define the concurrent execution environment by passing in a concurrent.futures.Executor object.

This allows the implementor to tailor the execution environment based upon the work that their callback will be performing. For example, if they are making web calls (e.g. - IO bound) they could pass in a ThreadPoolExecutor, but if they are doing numpy analysis of large datasets (e.g. - CPU bound), they could pass in a ProcessPoolExecutor. But what if the implementor must enforce sequential execution, such as for ordered message processing?

The SequentialExecutor provided by this package allows for the implementor using the hypothetical Python package to give that package an Executor, but forces the execution environment into seqential ordering.

Installation

pip install sequential-executor

SequentialExecutor Example

import concurrent.futures
import urllib.request

import sequential_executor

URLS = ['http://www.foxnews.com/',
        'http://www.cnn.com/',
        'http://europe.wsj.com/',
        'http://www.bbc.co.uk/',
        'http://some-made-up-domain.com/']

# Retrieve a single page and report the URL and contents
def load_url(url, timeout):
    print('loading %r' % url)
    with urllib.request.urlopen(url, timeout=timeout) as conn:
        return conn.read()

# We can use in a with statement just like other Executors
with sequential_executor.SequentialExecutor() as executor:
    # Start the load operations and mark each future with its URL
    future_to_url = {executor.submit(load_url, url, 60): url for url in URLS}
    for future in concurrent.futures.as_completed(future_to_url):
        url = future_to_url[future]
        try:
            data = future.result()
        except Exception as exc:
            print('%r generated an exception: %s' % (url, exc))
        else:
            print('%r page is %d bytes' % (url, len(data)))

NOTE - Although this looks like it is submitting multiple jobs and then waiting on each job to complete, the concurrent.futures.Future objects returned by the SequentialExecutor will be fully complete upon the return from the submit call!