test

App/Transcription/Model.py

@@ -10,7 +10,13 @@ class Transcriptions(orm.Model):
     fields = {
         "id": orm.Integer(primary_key=True),
         "task_id": orm.String(max_length=100, index=True, default=""),
-        "tl_file_id": orm.String(max_length=100, index=True, default=""),
+        "file_name": orm.String(max_length=100, index=True, default=""),
+        "language": orm.String(max_length=100, index=True, default="-"),
+        "youtubeLink": orm.String(max_length=100, index=True, allow_null=True),
+        "tl_file_id": orm.String(
+            max_length=100, index=True, default="", allow_null=True
+        ),
+        "duration": orm.Integer(index=True, default=0),
         "user": orm.ForeignKey(User, on_delete=orm.CASCADE),
         "createdAt": orm.DateTime(index=True, default=datetime.datetime.now),
         "content": orm.JSON(default={}),

App/Transcription/Schemas.py

@@ -1,14 +1,27 @@
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
+from datetime import datetime
 
 
 class TranscriptionMetadata(BaseModel):
     duration: int = 0
-    language: str = "0"
-    logs: str = 0
-    percentage: str = "0"
+    language: str = "-"
+    logs: str = ""
+    percentage: str = "-"
     transcription: dict = {}
     state: str = "PENDING"
 
 
+class TranscriptionResult(BaseModel):
+    created_at: datetime = Field(default_factory=datetime.utcnow)
+    duration: int = 0
+    language: str = "-"
+    transcription_state: str = "SUCCESS"
+    transcript: list = []
+
+    @property
+    def content(self):
+        return self.transcript
+
+
 class UserDetails(BaseModel):
     userId: str

App/Transcription/TranscriptionRoutes.py

@@ -1,6 +1,6 @@
 from fastapi import APIRouter, status, Depends, UploadFile, File, Query, BackgroundTasks
 from typing_extensions import Annotated
-from .Schemas import UserDetails
+from .Schemas import UserDetails, TranscriptionMetadata, TranscriptionResult
 from App import bot
 import aiofiles, os, re
 import tempfile
@@ -54,10 +54,12 @@ async def download_audio(
     }
 
     task = downloadfile.delay(url, ydl_opts, model)
+    response = {"task_id": task.id, "file_name": f"{video_title}.mp3"}
     transcription_enrty = await Transcriptions.objects.create(
-        task_id=task.id, user=user
+        user=user, youtubeLink=url, **response
     )
-    return {"task_id": task.id, "file_name": filename}
+
+    return response
 
 
 @transcription_router.post("/uploadfile/")
@@ -101,13 +103,16 @@ async def create_file(
 @transcription_router.get("/tasks/{task_id}")
 async def get_status(task_id):
     task_result = AsyncResult(task_id)
+
     entry: Transcriptions = await Transcriptions.objects.filter(task_id=task_id).first()
     if task_result.status == "SUCCESS":
-        await entry.update(
-            content=task_result.result, transcription_state=task_result.status
-        )
+        trans = TranscriptionResult(**task_result.result)
+        trans
+        await entry.update(**trans.dict(exclude={"transcript"}, content=trans.content))
     else:
-        await entry.update(transcription_state=task_result.status)
+        _trans = TranscriptionMetadata(**task_result.result)
+        await entry.update(**_trans.dict(exclude={"logs", "transcription"}))
+
     result = {
         "task_id": task_id,
         "task_status": task_result.status,

App/Transcription/Utils/audio_transcription.py

@@ -1,7 +1,8 @@
 from faster_whisper import WhisperModel
+from datetime import datetime, timedelta
 from tqdm import tqdm
 import os, time
-from App.Transcription.Schemas import TranscriptionMetadata
+from App.Transcription.Schemas import TranscriptionMetadata, TranscriptionResult
 
 current_time = time.localtime()
 model_size = "tiny"
@@ -11,6 +12,7 @@ def transcribe_file(state, file_path, model_size="tiny"):
     result = []
     metadata = TranscriptionMetadata()
     metadata.logs = "Transcription STARTING"
+    start_time = datetime.now()
     state.update_state(
         state="PROGRESS",
         meta=metadata.dict(),
@@ -43,8 +45,8 @@ def transcribe_file(state, file_path, model_size="tiny"):
                         "text": word.word,
                     }
                     result.append(temp)
-
-                    if current_time.tm_sec % 5 == 0:
+                    time_difference = datetime.now() - start_time
+                    if time_difference >= timedelta(seconds=30):
                         metadata.logs = "Transcribing.."
                         metadata.percentage = f"{((word.end / total_duration)*100)}"
                         state.update_state(state="PROGRESS", meta=metadata.dict())
@@ -61,4 +63,6 @@ def transcribe_file(state, file_path, model_size="tiny"):
     # delete file
     os.remove(file_path)
 
-    return result
+    transcription_result = TranscriptionResult(**metadata.dict())
+    transcription_result.transcript = result
+    return transcription_result.dict()

App/app.py

@@ -25,14 +25,6 @@ logging.basicConfig(
 )
 
 
-async def create_async_model(model):
-    import asyncio
-
-    # until something better comes along
-    url = model._get_database_url()
-    await asyncio.gather(asyncio.create_task(model._create_all(url)))
-
-
 app = FastAPI()
 origins = ["*"]