Improve the CLI API of the target programming language

readme.md

@@ -239,26 +239,24 @@ joblib.dump(clf, 'estimator.pkl')
 After that the model can be transpiled by using the following command:
 
 ```bash
-python -m sklearn_porter --input <pickle_file> [--output <destination_dir>] [--language {c,go,java,js,php,ruby}]
-python -m sklearn_porter -i <pickle_file> [-o <destination_dir>] [-l {c,go,java,js,php,ruby}]
+python -m sklearn_porter --input <PICKLE_FILE> [--output <DEST_DIR>] [--language {c,go,java,js,php,ruby}] [--c] [--java] [--js] [--go] [--php] [--ruby]
+python -m sklearn_porter -i <PICKLE_FILE> [-o <DEST_DIR>] [-l {c,go,java,js,php,ruby}] [--c] [--java] [--js] [--go] [--php] [--ruby]
 ```
 
-The following commands have all the same result:
+For instance the following command transpiles the estimator to the target programming language Java:
 
 ```bash
-python -m sklearn_porter --input estimator.pkl --language java
-python -m sklearn_porter -i estimator.pkl -l java
+python -m sklearn_porter -i estimator.pkl --java
 ```
 
-By changing the language parameter you can set the target programming language:
+You can change the target programming language on the fly:
 
 ```bash
-python -m sklearn_porter -i estimator.pkl -l c
-python -m sklearn_porter -i estimator.pkl -l go
-python -m sklearn_porter -i estimator.pkl -l java
-python -m sklearn_porter -i estimator.pkl -l js
-python -m sklearn_porter -i estimator.pkl -l php
-python -m sklearn_porter -i estimator.pkl -l ruby
+python -m sklearn_porter -i estimator.pkl --c
+python -m sklearn_porter -i estimator.pkl --go
+python -m sklearn_porter -i estimator.pkl --js
+python -m sklearn_porter -i estimator.pkl --php
+python -m sklearn_porter -i estimator.pkl --ruby
 ```
 
 Further information will be shown by using the `--help` parameter:

sklearn_porter/__main__.py

@@ -1,15 +1,16 @@
 # -*- coding: utf-8 -*-
 
 import os
+import sys
 import argparse
 
 from . import Porter
 
 
-def main():
+def parse_args(args):
  parser = argparse.ArgumentParser(
  description=(
- 'Transpile trained scikit-learn models '
+ 'Transpile trained scikit-learn estimators '
  'to C, Java, JavaScript and others. '),
  epilog=(
  'More details on: '
@@ -25,17 +26,36 @@ def main():
  required=False,
  help=(
  'Set the destination directory, '
- 'where the transpiled model will be '
+ 'where the transpiled estimator will be '
  'stored.'))
+ languages = {
+ 'c': 'C',
+ 'java': 'Java',
+ 'js': 'JavaScript',
+ 'go': 'Go',
+ 'php': 'PHP',
+ 'ruby': 'Ruby'
+ }
  parser.add_argument(
  '--language', '-l',
- choices=['c', 'java', 'js', 'go', 'php', 'ruby'],
+ choices=languages.keys(),
  default='java',
  required=False,
  help=(
  'Set the target programming language '
- '("c", "java", "js", "go", "php", "ruby").'))
- args = vars(parser.parse_args())
+ '({}).'.format(', '.join(['"{}"'.format(key)
+ for key in languages.keys()]))))
+ for key, lang in list(languages.items()):
+ parser.add_argument(
+ '--{}'.format(key),
+ action='store_true',
+ help='Set {} as the target programming language.'.format(lang))
+ args = vars(parser.parse_args(args))
+ return args
+
+
+def main():
+ args = parse_args(sys.argv[1:])
 
  input_path = str(args['input'])
  if input_path.endswith('.pkl') and os.path.isfile(input_path):
@@ -44,8 +64,15 @@ def main():
  from sklearn.externals import joblib
  model = joblib.load(input_path)
 
+ # Determine the target programming language:
+ language = str(args['language']) # with default language
+ languages = ['c', 'java', 'js', 'go', 'php', 'ruby']
+ for key in languages:
+ if args.get(key): # found ecplicit assignment
+ language = key
+ break
+
  # Port model:
- language = str(args['language'])
  porter = Porter(model, language=language)
  details = porter.export(details=True)
  filename = details.get('filename')