Skip to content
/ vorlauf Public

Minimal python library to create shell-like process pipelines (using subprocess.Popen)

Notifications You must be signed in to change notification settings

alex2/vorlauf

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Vorlauf

What?

Vorlauf is a very minimal tool that helps you to create process pipelines (in the shell sense). It also helps separate out the definition of a Process from the running of a that process with a given stdin, stdout, and stderr.

Why?

Because the subprocess api for chaining processes together is cumbersome and not very well documented.

How?

This library does basically nothing - it's implemented in fewer than 100 lines. There are two classes available, Process and Pipeline.

Process

The Process class is passed args, cwd, and env, and is executed by calling Process.run with optional stdin, stdout, and stderr parameters.

By removing the stdout, stderr, and stdin from the creation of the Process class, we can create reusable Process definitions:

critical_grepper = Process('grep', 'CRITICAL')

syslog = open('/var/log/syslog', 'r')
apachelog = open('/var/log/httpd/error.log', 'r')

filtered = open('critical.log', 'w')

for logfile in (syslog, apachelog):
  critical_grepper.run(stdin=logfile, stdout=filtered)

Pipeline

The Pipeline class stores a list of Process classes which, when run with Pipeline.run with optional stdin and stdout, pipes the processes together. If present, stdin is passed to the first process, and if present, stdout is passed to the last process.

Example

from vorlauf import Pipeline, Process

pipeline = Pipeline()
pipeline.add(Process('cat', 'foo.txt'))
pipeline.add(Process('grep', 'something'))
pipeline.add(Process('uniq'))

with open('new.txt', 'wb') as fd:
    pipeline.run(stdout=fd)

And because of operator overloading built into the Process and Pipeline classes, this can be simplified as:

from vorlauf import Process as P

pipeline = P('cat', 'foo.txt') | P('grep', 'something') | P('uniq')
with open('new.txt', 'wb') as fd:
    pipeline.run(stdout=fd)

Finally, you could use the Process class to create reusable components:

from vorlauf import Process

class GPG(Process):

    def __init__(self, passphrase):
        super(GPG, self).__init__('gpg', '-c', '--passphrase', passphrase, '-')


class MySQLDump(Process):

    def __init__(self, password, dbname, **kwargs):
        super(MySQLDump, self).__init__(
            'mysqldump', '-u', 'root', '-p{}'.format(password), dbname
        )


with open('mysql.dump', 'wb') as fd:
    pipeline = MySQLDump('loldongs', 'foo') | GPG('supersekrit')
    pipeline.run(stdout=fd)

Tests

Run:

virtualenv venv
venv/bin/pip install -e .
venv/bin/python test.py

About

Minimal python library to create shell-like process pipelines (using subprocess.Popen)

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages