Script execution concept Opt. 2: Denoted functions

readme.md

Concept

• The algorithm is a function, exposed to the framework through the script decorator.
• The decorator describes the algorithm's metadata.
• Input parameters are given by keyword arguments (all with defaults).
• Output parameters are not declared.
• The documentation is given as a docstring.
• The algorithm is identified as module_name.func_name.

Pros

• Utilizing the language syntax and features to manage execution.
• Concise declaration of the algorithm.
• The direct call to the denoted function might actually call execute for the execution to be managed by the framework. If so, there's no easy way to hack it.

Cons

• More bootstrapping.

load_clouds.py

from glob import glob
from os import cwd
from os.path import join, split, splitext
from frames import execute, script

@script(author='John Doe', version='0.6', multithreaded=False)        # Declare the algorithm's info.
def load_clouds(cloud_extension='copsxml'):                           # Name and input params are given a a function with kwargs.
  ''' Loads all clouds of given format from the CWD. '''              # Documentation in a docstring.

  cloud_ids = []
  
  for path in glob(join(cwd(), '*.'+cloud_extension)):
    name = split(splitext(path)[0])[-1]
    cloud_id = execute('import_cloud', path, name)
    cloud_ids.append(cloud_id)
  
  return {'cloud_ids': cloud_ids}                                     # Results as a return value. 

minimal.py

from frames import script

@script
def hello():
  print 'Hello, framework.'