added transcript sorter

Dockerfile

@@ -19,6 +19,9 @@ RUN chmod -R 777 /app/results_vt
 RUN mkdir /app/results_wt
 RUN chmod -R 777 /app/results_wt
 
+RUN mkdir /app/results_st
+RUN chmod -R 777 /app/results_st
+
 COPY . .
 
 CMD ["python", "app.py"]

app.py

@@ -6,7 +6,7 @@ from pathlib import Path
 
 import gradio as gr
 
-from utlis import HHMMSS_to_sec, convert_video_format, molly_xlsx_to_table, xlsx_to_table, table_to_ELAN_tsv, trim_media
+from utlis import HHMMSS_to_sec, convert_video_format, molly_xlsx_to_table, xlsx_to_table, table_to_ELAN_tsv, trim_media, sort_transcript
 
 
 def delete_files(files):
@@ -36,11 +36,11 @@ def classify_input_format(input_string):
     # if the input format is neither seconds nor HH:MM:SS, return None
     gradio.Error("Input format not supported. Please enter a valid time in the format HH:MM:SS or seconds.")
 
-def set_output_file(input_file, output_format, folder):
+def set_output_file(input_file, output_format, folder, insert_string = 'trimmed'):
     # Set output file name and extension
     if not os.path.exists(folder):
         os.makedirs(folder)
-    file_name = f"{Path(input_file.name).stem.partition('.')[0]}_trimmed.{output_format}"
+    file_name = f"{Path(input_file.name).stem.partition('.')[0]}_{insert_string}.{output_format}"
     output_file = os.path.join(folder, file_name)
     print(f"Output folder: {folder}")
     print(f"file_name: {file_name}")
@@ -105,6 +105,21 @@ def trim_video_wt(input_file, input_transcript, output_format, start_time, end_t
     except Exception as e:
         return f"Error: {str(e)}"
 
+def sort_transcript_helper(input_transcript, output_transcript):
+    # sort transcript
+    print("start sorting transcript")
+    print("input_transcript: ", input_transcript)
+    print("output_transcript: ", output_transcript)
+    output_transcript = sort_transcript(input_transcript, output_transcript)
+    print("finish sorting transcript")
+    return output_transcript
+
+def sort_transcript_wrapper(input_file):
+    output_folder = f"{os.getcwd()}/results_st/"
+    output_file_path = set_output_file(input_file, "tsv", output_folder, insert_string = 'sorted')
+    output_file_path = sort_transcript_helper(input_file.name, output_file_path)
+    return output_file_path
+
 def trim_video(input_file, output_format, start_time, end_time):
     try:
         # Set output file
@@ -180,6 +195,12 @@ interface_c = gr.Interface(fn=convert_video, inputs=[input_file_c, output_format
                             description="Convert a video file to a different format. Please wait for the file to upload before clicking the 'Submit' button.")
 
 
+# gr components for transcript sorter
+input_file_s = gr.File(label="Select transcript file")
+output_file_s = gr.File(label="Download sorted transcript")
+interface_s = gr.Interface(fn=sort_transcript_wrapper, inputs=input_file_s, outputs=output_file_s, title="Transcript Sorter", allow_flagging="never",
+                            description="Sort a transcript file by time. Please wait for the file to upload before clicking the 'Submit' button.")
+
 
-demo = gr.TabbedInterface([interface_wt, interface, interface_c], ["Video Trimmer with transcript converted","Video Trimmer", "Video Converter"])
+demo = gr.TabbedInterface([interface_wt, interface, interface_c, interface_s], ["Video Trimmer with transcript converted","Video Trimmer", "Video Converter", "Transcript Sorter"])
 demo.launch(server_name="0.0.0.0", server_port=7860)

utlis.py

@@ -8,6 +8,43 @@ import pandas as pd
 
 import gradio as gr
 
+def sort_transcript(file_path, save_path):
+    """
+    Sorts the rows of a transcript file by start time.
+
+    Parameters:
+    file_path (str): The file path of the transcript file.
+    save_path (str): The file path to save the sorted transcript file.
+
+    Returns:
+    save_path (str): The file path of the sorted transcript file.
+    """
+    file_ext = os.path.splitext(file_path)[1]
+    
+    if file_ext == '.txt' or file_ext == '.tsv':
+        # read the tab-separated plaintext file into a DataFrame
+        table = pd.read_csv(file_path, sep='\t', header=None, \
+                            names=['Speaker', 'Empty', 'Start', 'Start (s)', 'End', 'End (s)', 'Duration', 'Duration (s)', 'Transcript'], 
+                            index_col=False)
+    else:
+        raise ValueError("Unsupported file format. Must be '.txt' or '.tsv'.")
+
+    # make the column names lowercase
+    table.columns = map(str.lower, table.columns)
+
+    # select and reorder the desired columns
+    table = table[['speaker', 'transcript', 'start', 'end']]
+    
+    # extract the start time from the 'start' column
+    table['start_time'] = table['start'].str.split('.', expand=True)[0]
+
+    # sort by start_time
+    sorted_table = table.sort_values('start_time')
+    sorted_table.to_csv(save_path, sep='\t', index=False)
+    print("saved sorted transcript to", save_path)
+    return save_path
+
+
 
 def convert_video_format(media_in, media_out):
     try: