From 07083086a0272266fba4c057d69a767c054bfa97 Mon Sep 17 00:00:00 2001
From: Rob Hallam <0504004h@student.gla.ac.uk>
Date: Sun, 18 Aug 2024 00:46:56 +0100
Subject: [PATCH] refactor: pull out _run_get_output() FE method

Runs a command returns stout as string, defined to aid testing
---
 pipeline/feature_extractors.py | 21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)

diff --git a/pipeline/feature_extractors.py b/pipeline/feature_extractors.py
index 513386d..45224fb 100644
--- a/pipeline/feature_extractors.py
+++ b/pipeline/feature_extractors.py
@@ -16,6 +16,13 @@ logger = logging.getLogger(__name__)
 class FeatureExtractor(ABC):
     """Feature extractor interface."""
     # TODO: #API -- decide if .features will be a member variable
+    def _run_get_output(self, cmd: list, cwd:str=".") -> str:
+        """Run a command and return the output as a string
+
+        Defined to be mocked out in tests via unittest.mock.patch
+        """
+        return subprocess.run(cmd, stdout=subprocess.PIPE, cwd=cwd).stdout.decode("utf-8")
+
     def setup(self):
         """Setup the feature extractor -- validate input files & config"""
 
@@ -47,8 +54,11 @@ class LaughterFeatureExtractor(FeatureExtractor):
         self.config = config
         self.features = []
 
-    def _laughdetect(self, audio_file):
-        """Run laughter detection on the audio file"""
+    def _laughdetect(self, audio_file) -> list:
+        """Run laughter detection on the audio file
+
+        Returns a list of 2-tuples, each representing a laugh instance in the audio file
+        """
         laugh_detector_dir = "/home/robert/mounts/980data/code/laughter-detection/"
         laugh_detector_script = "segment_laughter.py"
         # fake output for testing
@@ -57,9 +67,9 @@ class LaughterFeatureExtractor(FeatureExtractor):
                               f"--input_audio_file={audio_file}"]
 
         # run command, capture output, ignore exit status
-        laugh_output = subprocess.run(laugh_detector_cmd,
-                                      stdout=subprocess.PIPE,
-                                      cwd=laugh_detector_dir).stdout.decode("utf-8")
+        # use self._run_get_output to allow mocking in tests
+        laugh_output = self._run_get_output(laugh_detector_cmd, laugh_detector_dir)
+
         # ↑ have to include cwd to keep laughter-detection imports happy
         # also, it isn't happy if no output dir is specified but we get laughs so it's grand
 
@@ -112,6 +122,7 @@ class LaughterFeatureExtractor(FeatureExtractor):
         """Extract laughter features for each input file"""
         if self.input_files:
             for file in self.input_files:
+                # adjust this call for better test mocking
                 laughs = self._laughdetect(file.path)
                 for laugh in laughs:
                     start, end = laugh