adding initial cli using click, tests todo

addressing #11

setup.py

@@ -7,7 +7,7 @@ def readme():
 setup(
     name='topic_model_to_Shiny_app',
     url='https://github.com/Sparrow0hawk/topic_model_to_Shiny_app',
-    version="1.3.0",
+    version="1.4.0-dev",
     author='Alex Coleman',
     author_email='a.coleman1@leeds.ac.uk',
     description='An implementation of Gensim topic modelling that identifies the number of topics using coherence scores.',

topic_model_to_Shiny_app/tests/cli_test.py

@@ -0,0 +1,27 @@
+import os
+import unittest
+from unittest.mock import patch
+import pkg_resources
+from click.testing import CliRunner
+from topic2shiny import main
+
+resource_package = 'topic_model_to_Shiny_app'
+
+test_dir = os.path.dirname(os.path.abspath(__file__))
+
+class CLI_test(unittest.TestCase):
+
+    def test_CLI(self):
+        # these are examples and will fail
+        # TODO: update these to actually test CLI
+        runner = CliRunner()
+        result = runner.invoke(cli, ['--debug', 'sync'])
+        assert result.exit_code == 0
+        assert 'Debug mode is on' in result.output
+        assert 'Syncing' in result.output
+
+
+
+if __name__ == "__main__":
+
+    unittest.main(verbosity=2)

topic_model_to_Shiny_app/tests/complete_test.py

@@ -29,15 +29,18 @@ def setUp(self):
 
         self.model = self.class_(corpus, id2word=dictionary, num_topics=3)
 
-    @patch('builtins.input',return_value=pkg_resources.resource_filename(resource_package, 'tests/test_data/test_data.csv'))
-    def test_initial_data_import(self, input):
+    def test_initial_data_import(self):
 
-        self.data = text_preprocessing.initial_data_import()
-
-        print(self.data)
+        self.data = text_preprocessing.initial_data_import(data_path=pkg_resources.resource_filename(resource_package, 'tests/test_data/test_data.csv'))
 
         self.assertTrue(isinstance(self.data, pd.DataFrame))
 
+        with self.assertRaises(SystemExit) as cm:
+
+            self.data = text_preprocessing.initial_data_import(data_path=pkg_resources.resource_filename(resource_package, 'tests/test_data/test_data2.csv'))
+
+        self.assertEqual(cm.exception.code, 0)
+
     def test_validate_input_data(self):
 
         self.data = text_preprocessing.validate_input_data(
@@ -59,10 +62,9 @@ def test_remove_stopwords(self):
 
         self.assertEqual(text_preprocessing.remove_stopwords([['the','and','jackal']]), [['jackal']])
 
-    @patch('builtins.input',return_value=pkg_resources.resource_filename(resource_package,'tests/test_data/test_data.csv'))
-    def test_full_preprocessing(self, input):
+    def test_full_preprocessing(self):
 
-        self.data = text_preprocessing.preprocessing()
+        self.data = text_preprocessing.preprocessing(data_path=pkg_resources.resource_filename(resource_package,'tests/test_data/test_data.csv'))
 
         self.assertTrue(isinstance(self.data, pd.DataFrame))
 
@@ -168,12 +170,11 @@ def test_get_top3(self):
 
 
 
-    @patch('builtins.input', side_effect=[pkg_resources.resource_filename(resource_package, 'tests/test_data/test_data.csv')])
-    def test_int(self, input):
+    def test_int(self):
         # run preprocessing function
         # should save data into right places for next function
 
-        self.data = text_preprocessing.preprocessing()
+        self.data = text_preprocessing.preprocessing(data_path=pkg_resources.resource_filename(resource_package, 'tests/test_data/test_data.csv'))
 
         # run topic selector function using output from above preprocessing
         self.data = topic_number_selex.topic_number_selector(self.data,

topic_model_to_Shiny_app/text_preprocessing.py

@@ -1,6 +1,7 @@
 # script for preprocessing text data for LDA
 # import libraries
 import sys
+import os
 import numpy as np
 import pandas as pd
 import gensim
@@ -17,10 +18,10 @@
 
 
 # main function that performs preprocessing
-def preprocessing():
+def preprocessing(data_path : str):
     """Complete preprocessing script that outputs dataframe with new tokens column."""
 
-    working_dataframe = initial_data_import()
+    working_dataframe = initial_data_import(data_path)
 
     validated_dataframe = validate_input_data(working_dataframe)
 
@@ -51,18 +52,27 @@ def preprocessing():
     return validated_dataframe
 
 
-def initial_data_import():
-    """Imports initial .csv dataframe converts into pandas object."""
+def initial_data_import(data_path : str):
+    """
+    Imports initial .csv dataframe converts into pandas object.
+
+    :param data_path: this is a string of the full path to data file
+
+    :returns: pd.DataFrame object 
+    """
 
     print('Beginning text preprocessing.')
     # specifying data directory
 
-    data_path = str(input('Specify the full path to the input :'))
-
     # validate that the file format is .csv
+    if os.path.isfile(data_path):
 
-    if str(data_path[-4:]) != '.csv':
-        print('Please confirm you are passing a .csv file type.')
+        original_frame = pd.read_csv(data_path.strip(), encoding='latin1')
+
+    else:
+        print('File does not exist.')
+        print('Please check the directory you passed: \n',data_path)
+        sys.exit(0)
 
     # reading data
     original_frame = pd.read_csv(data_path.strip(), encoding='latin1')

topic_model_to_Shiny_app/topic2shiny.py

@@ -0,0 +1,74 @@
+import time
+# consider security imp of subprocess
+import subprocess
+import click
+from topic_model_to_Shiny_app.text_preprocessing import preprocessing
+from topic_model_to_Shiny_app.topic_number_selex import topic_number_selector
+from topic_model_to_Shiny_app.dominant_topic_processing import topic_processing
+import pkg_resources
+import pathlib
+
+# specify top level package folder
+resource_package = 'topic_model_to_Shiny_app'
+
+@click.command()
+@click.option('-i', '--input-path', prompt='Specify input data path', type=str,
+              help='The absolute path to the input data file.')
+@click.option('-o', '--output-directory', prompt='Specify directory for outputs', type=str,
+              help='The absolute path to an directory for outputs, will create a new directory if required.')
+@click.option('-n', '--num-runs', prompt='Please specify the number of LDA repeats for topic number selection', type=int,
+              help='The number of LDA repeat runs desired to establish working topic number.')
+@click.option('-p', '--pretrained', default=False, type=bool,
+              help='A boolean to check if user wishes to use a pretrained model.')
+@click.option('-m', '--pretrained-model', type=str,
+              help='The absolute path to the pretrained model file.')
+@click.option('-s', '--shiny-start', default=True, type=bool,
+              help='A boolean to confirm if user wishes Shiny app instance to start.')
+
+def main(input_path : str, 
+         output_directory : str,
+         num_runs : int, 
+         pretrained : bool, 
+         pretrained_model : str,
+         shiny_start : bool):
+    """
+    The main function handling CLI
+
+    \b
+    :param input_path: this is a first param 
+    :param output_directory: this is a second param 
+    :param n_runs: an integer of the number of repeat runs of LDA to perform
+    :param pretrained: a boolean of whether user wants to use a pretrained model
+    :param pretrained_model: a string of the path to the pretrained model
+    :param shiny_start: a boolean of whether to start shiny app at the end of processing
+    :returns: None (loads shiny app)
+    """
+    # take the output directory and create it if it doesn't exist
+    pathlib.Path(output_directory).mkdir(parents=True, exist_ok=True)
+
+    # check if you need a new model
+    if not pretrained:
+
+        transformed_data = preprocessing(input_path)
+
+        topic_number_selector(processed_data = transformed_data, output_path=output_directory,
+                              narrow_iter = num_runs, wide_iter = 100)
+
+    else:
+        # TODO: write in how to load a pretrained model
+        pass
+
+    topic_processing(output_path=output_directory)
+
+    if shiny_start:
+
+        print('Initialising Shiny dashboard.')
+
+        shiny_dir = pkg_resources.resource_filename(resource_package, "Shiny/combined.R")
+
+        # process to call shiny app
+        subprocess.call([shiny_dir])
+
+if __name__ == '__main__':
+
+    main()