Save speaking asynchronously

2023-12-10 15:38:26 +00:00
parent 50c39e5f9c
commit 9f4aed52ae
3 changed files with 91 additions and 63 deletions
--- a/app.py
+++ b/app.py
@@ -1,14 +1,13 @@
-import random
+import threading

 from flask import Flask, request
 from flask_jwt_extended import JWTManager, jwt_required
 from functools import reduce
 from helper.api_messages import *
-from helper.constants import *
 from helper.exercises import *
 from helper.file_helper import delete_files_older_than_one_day
 from helper.firebase_helper import *
-from helper.heygen_api import create_video
+from helper.heygen_api import create_videos_and_save_to_db
 from helper.speech_to_text_helper import *
 from helper.token_counter import count_tokens
 from helper.openai_interface import make_openai_call, make_openai_instruct_call
@@ -31,6 +30,8 @@ jwt = JWTManager(app)
 cred = credentials.Certificate(os.getenv("GOOGLE_APPLICATION_CREDENTIALS"))
 firebase_admin.initialize_app(cred)

+thread_event = threading.Event()
+

@app.route('/healthcheck', methods=['GET'])
 def healthcheck():
@@ -158,6 +159,7 @@ def save_listening():
        data = request.get_json()
        parts = data.get('parts')
        template = getListeningTemplate()
+        id = str(uuid.uuid4())
        for i, part in enumerate(parts, start=0):
            file_name = str(uuid.uuid4()) + ".mp3"
            sound_file_path = AUDIO_FILES_PATH + file_name
@@ -170,7 +172,7 @@ def save_listening():
            template["parts"][i]["audio"]["source"] = file_url
            template["parts"][i]["exercises"] = part["exercises"]

-        (result, id) = save_to_db("listening", template)
+        (result, id) = save_to_db_with_id("listening", template, id)
        if result:
            return {**template, "id": id}
        else:
@@ -292,10 +294,11 @@ def save_writing_task():
        data = request.get_json()
        exercises = data.get('exercises')
        template = getWritingTemplate()
+        id = str(uuid.uuid4())
        for i, exercise in enumerate(exercises, start=0):
            template["exercises"][i]["prompt"] = exercise

-        (result, id) = save_to_db("writing", template)
+        (result, id) = save_to_db_with_id("writing", template, id)
        if result:
            return {**template, "id": id}
        else:
@@ -524,63 +527,18 @@ def save_speaking():
        data = request.get_json()
        exercises = data.get('exercises')
        template = getSpeakingTemplate()
+        id = str(uuid.uuid4())

-        # Speaking 1
-        sp1_result = create_video(exercises[0]["question"], random.choice(list(AvatarEnum)))
-        if sp1_result is not None:
-            sound_file_path = VIDEO_FILES_PATH + sp1_result
-            firebase_file_path = FIREBASE_SPEAKING_VIDEO_FILES_PATH + sp1_result
-            url = upload_file_firebase_get_url(FIREBASE_BUCKET, firebase_file_path, sound_file_path)
-            sp1_video_path = firebase_file_path
-            sp1_video_url = url
-            template["exercises"][0]["text"] = exercises[0]["question"]
-            template["exercises"][0]["title"] = exercises[0]["topic"]
-            template["exercises"][0]["video_url"] = sp1_video_url
-            template["exercises"][0]["video_path"] = sp1_video_path
-        else:
-            print("Failed to create video for part 1 question: " + exercises[0]["question"])
+        thread_event.set()
+        thread = threading.Thread(
+            target=create_videos_and_save_to_db,
+            args=(exercises, template, id),
+            name=("thread-save-speaking-" + id)
+        )
+        thread.start()

-        # Speaking 2
-        sp2_result = create_video(exercises[1]["question"], random.choice(list(AvatarEnum)))
-        if sp2_result is not None:
-            sound_file_path = VIDEO_FILES_PATH + sp2_result
-            firebase_file_path = FIREBASE_SPEAKING_VIDEO_FILES_PATH + sp2_result
-            url = upload_file_firebase_get_url(FIREBASE_BUCKET, firebase_file_path, sound_file_path)
-            sp2_video_path = firebase_file_path
-            sp2_video_url = url
-            template["exercises"][1]["prompts"] = exercises[1]["prompts"]
-            template["exercises"][1]["text"] = exercises[1]["question"]
-            template["exercises"][1]["title"] = exercises[1]["topic"]
-            template["exercises"][1]["video_url"] = sp2_video_url
-            template["exercises"][1]["video_path"] = sp2_video_path
-        else:
-            print("Failed to create video for part 2 question: " + exercises[1]["question"])
-
-        # Speaking 3
-        sp3_questions = []
-        avatar = random.choice(list(AvatarEnum))
-        for question in exercises[2]["questions"]:
-            result = create_video(question, avatar)
-            if result is not None:
-                sound_file_path = VIDEO_FILES_PATH + result
-                firebase_file_path = FIREBASE_SPEAKING_VIDEO_FILES_PATH + result
-                url = upload_file_firebase_get_url(FIREBASE_BUCKET, firebase_file_path, sound_file_path)
-                video = {
-                    "text": question,
-                    "video_path": firebase_file_path,
-                    "video_url": url
-                }
-                sp3_questions.append(video)
-            else:
-                print("Failed to create video for part 3 question: " + question)
-        template["exercises"][2]["prompts"] = sp3_questions
-        template["exercises"][2]["title"] = exercises[2]["topic"]
-
-        (result, id) = save_to_db("speaking", template)
-        if result:
+        # Return response without waiting for create_videos_and_save_to_db to finish
        return {**template, "id": id}
-        else:
-            raise Exception("Failed to save speaking: " + template)
    except Exception as e:
        return str(e)

@@ -677,8 +635,8 @@ def save_reading_passage():
        parts = data.get('parts')
        template = getReadingTemplate()
        template["parts"] = parts
-
-        (result, id) = save_to_db("reading", template)
+        id = str(uuid.uuid4())
+        (result, id) = save_to_db_with_id("reading", template, id)
        if result:
            return {**template, "id": id}
        else:
--- a/helper/firebase_helper.py
+++ b/helper/firebase_helper.py
@@ -59,4 +59,18 @@ def save_to_db(collection: str, item):
        return (False, None)


+def save_to_db_with_id(collection: str, item, id: str):
+    db = firestore.client()
+    collection_ref = db.collection(collection)
+    # Reference to the specific document with the desired ID
+    document_ref = collection_ref.document(id)
+    # Set the data to the document
+    document_ref.set(item)
+    if document_ref:
+        print(f"Document added with ID: {document_ref.id}")
+        return (True, document_ref.id)
+    else:
+        return (False, None)
+
+

--- a/helper/heygen_api.py
+++ b/helper/heygen_api.py
@@ -1,9 +1,12 @@
 import os
+import random
+
 import requests
 import time

 from dotenv import load_dotenv
-
+from helper.constants import *
+from helper.firebase_helper import upload_file_firebase_get_url, save_to_db_with_id
 from heygen.AvatarEnum import AvatarEnum

 load_dotenv()
@@ -30,6 +33,59 @@ KAYLA_ABBI = "d688099f8db9472cb4890b0561e81793"
 JEROME_RYAN = "ad41feb2a5c4483085525e3d8907f512"
 TYLER_CHRISTOPHER = "03c796f8ed274bb38f19e893bcbc6121"

+def create_videos_and_save_to_db(exercises, template, id):
+    # Speaking 1
+    sp1_result = create_video(exercises[0]["question"], random.choice(list(AvatarEnum)))
+    if sp1_result is not None:
+        sound_file_path = VIDEO_FILES_PATH + sp1_result
+        firebase_file_path = FIREBASE_SPEAKING_VIDEO_FILES_PATH + sp1_result
+        url = upload_file_firebase_get_url(FIREBASE_BUCKET, firebase_file_path, sound_file_path)
+        sp1_video_path = firebase_file_path
+        sp1_video_url = url
+        template["exercises"][0]["text"] = exercises[0]["question"]
+        template["exercises"][0]["title"] = exercises[0]["topic"]
+        template["exercises"][0]["video_url"] = sp1_video_url
+        template["exercises"][0]["video_path"] = sp1_video_path
+    else:
+        print("Failed to create video for part 1 question: " + exercises[0]["question"])
+
+    # Speaking 2
+    sp2_result = create_video(exercises[1]["question"], random.choice(list(AvatarEnum)))
+    if sp2_result is not None:
+        sound_file_path = VIDEO_FILES_PATH + sp2_result
+        firebase_file_path = FIREBASE_SPEAKING_VIDEO_FILES_PATH + sp2_result
+        url = upload_file_firebase_get_url(FIREBASE_BUCKET, firebase_file_path, sound_file_path)
+        sp2_video_path = firebase_file_path
+        sp2_video_url = url
+        template["exercises"][1]["prompts"] = exercises[1]["prompts"]
+        template["exercises"][1]["text"] = exercises[1]["question"]
+        template["exercises"][1]["title"] = exercises[1]["topic"]
+        template["exercises"][1]["video_url"] = sp2_video_url
+        template["exercises"][1]["video_path"] = sp2_video_path
+    else:
+        print("Failed to create video for part 2 question: " + exercises[1]["question"])
+
+    # Speaking 3
+    sp3_questions = []
+    avatar = random.choice(list(AvatarEnum))
+    for question in exercises[2]["questions"]:
+        result = create_video(question, avatar)
+        if result is not None:
+            sound_file_path = VIDEO_FILES_PATH + result
+            firebase_file_path = FIREBASE_SPEAKING_VIDEO_FILES_PATH + result
+            url = upload_file_firebase_get_url(FIREBASE_BUCKET, firebase_file_path, sound_file_path)
+            video = {
+                "text": question,
+                "video_path": firebase_file_path,
+                "video_url": url
+            }
+            sp3_questions.append(video)
+        else:
+            print("Failed to create video for part 3 question: " + question)
+    template["exercises"][2]["prompts"] = sp3_questions
+    template["exercises"][2]["title"] = exercises[2]["topic"]
+
+    save_to_db_with_id("speaking", template, id)

 def create_video(text, avatar: AvatarEnum):
    # POST TO CREATE VIDEO