From e2b5ad15b48f039c9b340b192e2088a54a4f68e3 Mon Sep 17 00:00:00 2001
From: ganders <g.anders@iwm-tuebingen.de>
Date: Tue, 10 Oct 2023 14:17:38 +0200
Subject: [PATCH] Feature request: added json config

---
 README.md                   |  8 ++++++--
 data/jobs/example_job.json  |  6 ++++++
 main.py                     | 28 ++++++++++++++++++++++------
 src/transcript_generator.py | 13 +++++++------
 4 files changed, 41 insertions(+), 14 deletions(-)
 create mode 100644 data/jobs/example_job.json

diff --git a/README.md b/README.md
index 5a51fcc..d7ebe41 100644
--- a/README.md
+++ b/README.md
@@ -52,14 +52,18 @@ Jobs specify tasks for transcript creation. A job contains the following informa
 
 - `name`: string: The name of the job
 - `source`: List[string]: The sources of the videos (paths, YouTube links, TibAV links)
-- `modelpath`: Optional[string]: The path to the model if a specific model should be used.
 - `language`: Optional[string]: The language of the video (If not specified employs auto-detection)
 
 ## to-do
 
+### Feature: specific model
+
+add an opitonal model path to the json config
+- `model_path`:Optional[string]: The path to a specific language model to be used (mainly for testing)
+
+### Other
 - seperate data production form code
 - remember to accept the conditions on models if you use huggingface
-- do a config json
 - check warnings
 - downloader for tib-av
 - more optional setting
diff --git a/data/jobs/example_job.json b/data/jobs/example_job.json
new file mode 100644
index 0000000..4a27811
--- /dev/null
+++ b/data/jobs/example_job.json
@@ -0,0 +1,6 @@
+{
+  "name": "myjob1",
+  "source": [
+    "https://www.youtube.com/watch?v=BVVo1BaL6JA"
+  ]
+}
\ No newline at end of file
diff --git a/main.py b/main.py
index 2c5c6f1..4f91315 100644
--- a/main.py
+++ b/main.py
@@ -1,20 +1,36 @@
 import os
+import json
+import argparse
 
 from src.transcript_generator import TranscriptGenerator
 
 
 def main():
-    # input could be a YouTube link or a local file
-    input_sources = ["https://www.youtube.com/watch?v=BVVo1BaL6JA"]
+    # Create a parser
+    parser = argparse.ArgumentParser(description='Transcript Generator for Videos')
 
+    # Add the arguments
+    parser.add_argument('--job', type=str, help='The name of the job', required=True)
+
+    # Parse the arguments
+    args = parser.parse_args()
+
+    # Load the job config
+    with open(f'jobs/{args.job}.json', 'r') as f:
+        job_config = json.load(f)
+
+    # Check the hf_token
     hf_token = os.environ.get("HF_TOKEN")
     if not hf_token:
         hf_token = input("Please enter your HF_TOKEN: ")
     os.environ["HF_TOKEN"] = hf_token
 
-    transcript_generator = TranscriptGenerator()
-    transcript_generator.generate(input_sources)
+    transcript_generator = TranscriptGenerator(
+        name=job_config['name'],
+        language_code=job_config['language_code']
+    )
 
+    transcript_generator.generate(job_config['source'])
 
-if __name__ == "__main__":
-    main()
+    if __name__ == "__main__":
+        main()
diff --git a/src/transcript_generator.py b/src/transcript_generator.py
index 51cd113..03594d3 100644
--- a/src/transcript_generator.py
+++ b/src/transcript_generator.py
@@ -15,9 +15,10 @@ class TranscriptGenerator:
     to generate transcriptions from given input sources.
     """
 
-    def __init__(self):
+    def __init__(self, job_name, language_code="en"):
         self.file_manager = FileManager()
-        self.transcriber = Transcriber()
+        self.transcriber = Transcriber(language_code)
+        self.job_name = job_name
 
     def generate(self, input_sources: List[str]) -> None:
         """
@@ -29,15 +30,15 @@ class TranscriptGenerator:
         for i, source in enumerate(input_sources):
             try:
                 if source.startswith("http"):
-                    mp4_filename = f"audio_{i + 1}.mp4"
-                    wav_filename = f"audio_{i + 1}.wav"
+                    mp4_filename = f"audio_{self.job_name}_{i + 1}.mp4"
+                    wav_filename = f"audio_{self.job_name}_{i + 1}.wav"
                     self.file_manager.download_youtube_audio(source, mp4_filename)
                     self.file_manager.convert_to_wav(mp4_filename, wav_filename)
                 else:
                     wav_filename = self.file_manager.load_local_file(source)
                 self.transcriber.transcribe(wav_filename)
-                json_output_file = f"result_{i + 1}.json"
-                txt_output_file = f"transcripts_{i + 1}.txt"
+                json_output_file = f"result_{self.job_name}_{i + 1}.json"
+                txt_output_file = f"transcripts_{self.job_name}_{i + 1}.txt"
                 self.transcriber.save_transcript(json_output_file, txt_output_file)
             except Exception as e:
                 logger.exception(