Merge pull request #37 from heymanpreet/main

Issue#4: Updating slicer.py for customize start and end time input params

interviewkit/slicer.py

@@ -1,17 +1,30 @@
 """
-This takes two args, a filepath to an audio file and an integer.
+This takes three args, a filepath to an audio file, audio start and audio end times.
 
-It will sample the file specified for that many minutes.
+It will sample the file specified for that many minutes and seconds. 
+
+The audio start and end times params currently support time in 2 ways either minutes and seconds or minutes only.
 
 `ffmpeg` is a requirement for this to work.
 
 Example usage:
-
-    python interviewkit/sampler.py data/Martine+Barrat_FINAL.mp3 2
+    Format:  python interviewkit/slicer.py path_to_audio_file audio_slice_start_time audio_slice_end_time
+    Example 1: python interviewkit/slicer.py data/MartineBarrat_FINAL.mp3 2 3 
+             (Audio will be sliced from 2mins to 4mins)
+    Example 2: python interviewkit/slicer.py data/Martine+Barrat_FINAL.mp3 70:40 80:40 
+             (Audio will be sliced from 105mins 40secs to 107mins and 40secs)
 
 This generates:
+    data/sampled-70m40s-80m40s-MartineBarrat_FINAL.mp3
+    data/sampled-2m-4m-MartineBarrat_FINAL.mp3
+
+Note: 
+The Script currently doesn't support input in hours format, so hours will need to converted to mins:secs format.
+Example: If we want audio from 1hr 20mins 30 secs to 1hr 30mins 40secs, it can be done as shown below:
+1hr 20mins 30 secs = 80mins 30secs
+1hr 30mins 40sec = 90mins 40secs
 
-    data/sampled-5-Martine+Barrat_FINAL.mp3
+python interviewkit/slicer.py data/Martine+Barrat_FINAL.mp3 80:30 90:40
 
 """
 import sys
@@ -29,22 +42,78 @@
     print("  On mac you can: brew install ffmpeg")
     exit(1)
 
+def convert_audio_time_to_msec(audio_time_split_list):
+    """ Converting mins and secs to msecs for pydub computation """
 
-def main():
-    if len(sys.argv) != 3:
-        print("Usage: python3 sampler.py <filepath> <number of minutes>")
-        return
+    if(audio_time_split_list):
+        if(len(audio_time_split_list) == 1):
+            return int(audio_time_split_list[0]) * 60 * 1000
+        elif(len(audio_time_split_list) == 2):
+            return int(audio_time_split_list[0]) * 60 * 1000 + int(audio_time_split_list[1]) * 1000
+        else:
+            print("Error! Audio slice input params invalid. Audio slice supports start/end time in mins or mins:secs format. Please try again with correct input times.")
+            print("Error inside convert_audio_time_to_msec(audio_time_split_list) funtion.")
+            exit(1)
+    else:
+        print("Error! Audio slice input params invalid. Please try again with correct parameters.")
+        print("Error inside convert_audio_time_to_msec(audio_time_split_list) funtion.")
+        exit(1)
+
+def export_filename(audio_time_list):
+    """ Filename for exported file """
+
+    if audio_time_list and len(audio_time_list) == 2:
+        return f"{audio_time_list[0]}m{audio_time_list[1]}s"
+    elif audio_time_list and len(audio_time_list) == 1:
+        return f"{audio_time_list[0]}m"
+    else:
+        print("Error! Audio slice input params invalid. Please try again with correct parameters.")
+        print("Error inside export_filename(audio_time_list) funtion.")
+        exit(1)
 
-    path = Path(sys.argv[1])
-    minutes = int(sys.argv[2])
 
-    print("Sampling {} for {} minutes".format(path, minutes))
+def audio_slicing(path, audio_slice_start_time, audio_slice_end_time):
+    """ It reads the original audio and uses start and end input time params to generate sliced audio. """
+
+    print("Sampling {} from {} to {}".format(path, audio_slice_start_time, audio_slice_end_time))
 
+    # Reading original audio file
     audio = pydub.AudioSegment.from_file(path)
-    audio = audio[:minutes * 60 * 1000]
-    new_filename = f"{path.parent}/sampled-{minutes}-{path.name}"
+    original_audio_size_ms = audio.duration_seconds * 1000
+
+    # Fetching mins and secs from audio input
+    audio_start_time_list = audio_slice_start_time.split(":")
+    audio_end_time_list = audio_slice_end_time.split(":")
+
+    # Converting audio start and end times in msecs
+    audio_start_time = convert_audio_time_to_msec(audio_start_time_list)
+    audio_end_time = convert_audio_time_to_msec(audio_end_time_list)
+
+    # Check if audio start and end times are within original audio size limits
+    if(audio_start_time > original_audio_size_ms or audio_end_time > original_audio_size_ms):
+        print("Error! Audio slice input params cannot be greater than original audio size. Please try again with correct parameters.")
+        exit(1)
+
+    # Audio slicing process
+    audio = audio[audio_start_time:audio_end_time]
+
+    # Filename for exported file
+    audio_start_time_name = export_filename(audio_start_time_list)
+    audio_end_time_name = export_filename(audio_end_time_list)
+    new_filename =  f"{path.parent}/sampled-{audio_start_time_name}-{audio_end_time_name}-{path.name}"
     audio.export(new_filename, format="mp3")
     print("Created new file: ", new_filename)
 
+def main():
+    if len(sys.argv) != 4:
+        print("Usage: python3 slicer.py <filepath> <audio start time in minutes> <audio end time in minutes>")
+        return
+
+    path = Path(sys.argv[1])
+    audio_slice_start_time = (sys.argv[2])
+    audio_slice_end_time = (sys.argv[3])
+
+    audio_slicing(path, audio_slice_start_time, audio_slice_end_time)
+
 if __name__ == '__main__':
     main()