Beta Release

Author: pavanpaik

.DS_Store

@@ -0,0 +1,98 @@
+This is the detailed report of the finalized **Sai Leela Rahasya** web application, encompassing the structure, features, and design decisions made for all components.
+
+## 📝 Detailed Application Report: Sai Leela Rahasya
+
+### I. Application Mission & Core Philosophy
+
+The application's fundamental mission is to move beyond the simple narrative of Sai Baba's miracles (_Leela_) to uncover the hidden, instructional story (_Bodhakatha_). It aims to deliver "crystal clear knowledge" and "True information related to Sai Baba's life fact's" and convey the "anguish of Sai's heart to every heart".
+
+### II. Interface & Navigation Strategy
+
+The application employs a responsive design ensuring accessibility across all device types.
+
+### I. Interface Design & Navigation Strategy
+
+The app utilizes responsive design principles for optimal user experience:
+
+| **Element** | **Mobile (Portrait)** | **Tablet/Desktop (Landscape)** | **Function** |
+| --- | --- | --- | --- |
+| **Primary Navigation** | **Footer Tab Bar** | **Left-Hand Sidebar** | Persistent, one-click access to the five core features. |
+| --- | --- | --- | --- |
+| **Utility Menu** | **Top Right Corner Avatar Icon** (Slide-Out Panel) | **Top Right Corner Avatar Icon** (Slide-Out Panel) | Houses user account actions (Sign In/Profile) and secondary links (FAQ, Donate, Glossary, Contact Us). |
+| --- | --- | --- | --- |
+| **Entry Point** | **Landing Page** | **Landing Page** | Acts as the spiritual welcome and drives users to explore the content. |
+| --- | --- | --- | --- |
+
+#### Entry Point: Landing Page
+
+The app opens with a focused Landing Page that acts as a spiritual welcome mat:
+
+- **Hero Message:** "Sai Leela Rahasya: The Secret of Sai's Plays Revealed."
+- **Motto Inclusion:** Features the driving force: "We have believed Sai for years, now let's learn something. It's not our application, Sai, it's your choice."
+- **Primary CTA:** Directs users to the core content.
+- **Social Integration:** Contains subtle links to the source YouTube Channel, Facebook, and Instagram for audience growth.
+
+#### Primary Navigation (5 Core Features)
+
+The five core features are accessed via persistent navigation:
+
+- **Mobile View:** **Footer Tab Bar** for easy, one-handed access.
+- **Tablet/Desktop View:** **Left-Hand Sidebar** for utilizing horizontal space.
+
+| **Feature Page** | **Icon** | **Focus** |
+| --- | --- | --- |
+| **1\. Leela** | Divine Footprints | Narrative Articles (The "What") |
+| --- | --- | --- |
+| **2\. Bodhakatha** | Radiant Figure (Instruction) | Instructional Articles (The "Why") |
+| --- | --- | --- |
+| **3\. Live Stream** | Live Stream Indicator | Real-Time Video Darshan & Library |
+| --- | --- | --- |
+| **4\. Bhajan/Audio** | Music/Speaker Icon | Auditory Devotion & Library |
+| --- | --- | --- |
+| **5\. Ask Krishnaji** | Guidance/Question Icon | Personalized Ticket System |
+| --- | --- | --- |
+
+## Utility & User Profile Menu (Revised)
+
+#### The Utility Menu is triggered by clicking the User/Avatar icon in the Top Right Corner and slides open from the right
+
+### Menu Content Breakdown
+
+<div class="joplin-table-wrapper"><table><thead><tr><th><h4>User State</h4></th><th><h4>Primary Action (Top of Panel)</h4></th><th><h4>Secondary Links (Below Action)</h4></th></tr><tr><th><h4>Guest User</h4></th><th><h4>Sign In / Register button.</h4></th><th><h4>Glossary | FAQ | Donate | Logout</h4></th></tr></thead></table></div>
+
+#### III. Detailed Feature Implementations
+
+#### 1\. Leela & Bodhakatha Pages (Article Lists)
+
+Both pages display articles in a list view, linking directly to the source YouTube video for content transparency.
+
+- **Leela Page Ordering:** Articles are ordered **Chronologically** by Chapter/Sequence number.
+- **Bodhakatha Page Ordering:** Articles are ordered **Semantically** by the Virtue/Theme (e.g., Faith, Saburi) to prioritize instructional wisdom.
+
+#### 2\. Live Streams & Video Darshan Library
+
+This page acts as the central video hub:
+
+- **Priority Placement:** Features a prominently displayed card for the **LIVE NOW** feed, if active.
+- **Library:** Lists individual video cards for recorded content (past Aartis, Darshans).
+- **Functionality:** Includes an **Upcoming Schedule** section to inform users of the next scheduled stream times.
+
+#### 3\. Bhajan/Audio Page
+
+- **Organization:** Content is structured by function: **Daily Aartis** (Kakad, Madhyan, Dhoop, Shej), **Bhajans**, **Mantras/Chants**, and related **Spiritual Discourses**.
+- **User Experience:** Optimized for background listening with a simple interface.
+
+#### 4\. Ask Krishnaji (Personalized Guidance)
+
+This feature is implemented as a professional **Ticket Submission System**.
+
+- **Guest Conversion:** Guests are prompted to **Sign In** to receive personalized answers via email and track their inquiries.
+- **Signed-In User Workflow:**
+  - **New Ticket Submission:** Users submit a detailed question (the "ticket").
+  - **Ticket History:** Users view a list of all their past requests, filterable by Status (**Open, Answered, Closed**).
+  - **Ticket Detail:** Displays a chronological log of the original question and the guidance provided by the backend team.
+  - **Follow-Up:** Users can submit follow-up questions directly within the ticket detail, maintaining a continuous, organized record of their spiritual correspondence.
+
+### IV. Glossary Page
+
+A dedicated utility page for defining key spiritual terms central to the application's mission, including: **Leela**, **Bodhakatha**, **Sai Leela Rahasya**, **Shraddha**, **Saburi**, and **Darshan**.

.github/workflows/deploy.yml

@@ -0,0 +1,84 @@
+# Krishnasagar
+
+Krishnasagar is a serverless video localization pipeline that automates transcription, translation, and dubbing of content using Google Cloud AI services.
+
+> **Status:** The **Web Project** is currently in **Beta Release**. All other projects (Cloud Functions, etc.) are in **Dev Locked Mode**.
+
+## Setup & Deployment
+
+gcloud services enable storage.googleapis.com \
+  cloudfunctions.googleapis.com \
+  run.googleapis.com \
+  speech.googleapis.com \
+  translate.googleapis.com \
+  texttospeech.googleapis.com \
+  videointelligence.googleapis.com \
+  youtube.googleapis.com \
+  logging.googleapis.com \
+  monitoring.googleapis.com
+
+
+gsutil mb -l asia-south1 gs://krishnasagar-videos-raw
+gsutil mb -l asia-south1 gs://krishnasagar-videos-processed
+gsutil mb -l asia-south1 gs://krishnasagar-videos-final
+
+### extract_audio
+gcloud functions deploy krishnasagar-extract-audio \
+  --runtime python39 \
+  --trigger-resource krishnasagar-videos-raw \
+  --trigger-event google.storage.object.finalize \
+  --entry-point extract_audio \
+  --memory 512MB \
+  --region asia-south1 \
+  --source cloud_functions/extract_audio
+
+### transcribe_audio
+  gcloud functions deploy krishnasagar-transcribe-audio \
+  --runtime python39 \
+  --trigger-resource krishnasagar-videos-processed \
+  --trigger-event google.storage.object.finalize \
+  --entry-point transcribe_audio \
+  --memory 512MB \
+  --region asia-south1 \
+  --source cloud_functions/transcribe_audio
+
+### translate_text
+  gcloud functions deploy krishnasagar-translate-text \
+  --runtime python39 \
+  --trigger-resource krishnasagar-videos-processed \
+  --trigger-event google.storage.object.finalize \
+  --entry-point translate_text \
+  --memory 512MB \
+  --region asia-south1 \
+  --source cloud_functions/translate_text
+
+### generate_audio
+  gcloud functions deploy krishnasagar-generate-voice \
+  --runtime python39 \
+  --trigger-resource krishnasagar-videos-processed \
+  --trigger-event google.storage.object.finalize \
+  --entry-point generate_audio \
+  --memory 512MB \
+  --region asia-south1 \
+  --source cloud_functions/generate_audio
+
+### generate_subtitles
+  gcloud functions deploy krishnasagar-generate-subtitles \
+  --runtime python39 \
+  --trigger-resource krishnasagar-videos-processed \
+  --trigger-event google.storage.object.finalize \
+  --entry-point generate_subtitles \
+  --memory 512MB \
+  --region asia-south1 \
+  --source cloud_functions/generate_subtitles
+
+### Build the Docker Image
+  gcloud builds submit --tag gcr.io/potent-airfoil-454402-e4/krishnasagar-video-merger cloud_run/merge_audio_video
+
+### Deploy the Service
+
+gcloud run deploy krishnasagar-video-merger \
+  --image gcr.io/potent-airfoil-454402-e4/krishnasagar-video-merger \
+  --region asia-south1 \
+  --platform managed \
+  --allow-unauthenticated

Full E2E Plan.md

@@ -0,0 +1,55 @@
+import logging
+import uuid
+import subprocess
+from google.cloud import storage
+
+# Initialize the Cloud Storage client
+storage_client = storage.Client()
+
+def extract_audio(event, context):
+    """
+    Triggered by a change to the 'krishnasagar-videos-raw' bucket.
+    This function extracts audio from an MP4 video file and saves the extracted
+    audio as an MP3 file in the 'krishnasagar-videos-processed' bucket.
+    
+    It uses a unique TRACE_ID for logging to enable end-to-end traceability.
+    """
+    bucket_name = event['bucket']
+    video_file = event['name']
+
+    # Process only .mp4 files
+    if not video_file.endswith(".mp4"):
+        logging.info("File is not an MP4. Skipping...")
+        return
+
+    # Generate a unique trace ID for this operation
+    trace_id = str(uuid.uuid4())
+    logging.info(f"[TRACE_ID: {trace_id}] Starting audio extraction for file: {video_file}")
+
+    # Define local file paths
+    input_path = f"/tmp/{video_file}"
+    output_audio = f"/tmp/{video_file.replace('.mp4', '.mp3')}"
+
+    try:
+        # Download the video file from the source bucket
+        bucket = storage_client.bucket(bucket_name)
+        blob = bucket.blob(video_file)
+        blob.download_to_filename(input_path)
+        logging.info(f"[TRACE_ID: {trace_id}] Downloaded {video_file} to {input_path}")
+
+        # Use FFmpeg to extract audio
+        subprocess.run(
+            ["ffmpeg", "-i", input_path, "-q:a", "0", "-map", "a", output_audio],
+            check=True
+        )
+        logging.info(f"[TRACE_ID: {trace_id}] Audio extraction completed, saved to {output_audio}")
+
+        # Upload the extracted audio to the destination bucket
+        processed_bucket = storage_client.bucket("krishnasagar-videos-processed")
+        # Processing logic (extract audio)
+        output_blob = processed_bucket.blob(video_file.replace(".mp4", ".mp3"))
+        output_blob.upload_from_filename(output_audio)
+        logging.info(f"[TRACE_ID: {trace_id}] Uploaded extracted audio to krishnasagar-videos-processed bucket")
+
+    except Exception as e:
+        logging.error(f"[TRACE_ID: {trace_id}] Error during audio extraction: {e}")

README.md

@@ -0,0 +1,57 @@
+import logging
+from google.cloud import texttospeech, storage
+
+# Initialize clients for Storage and Text-to-Speech
+storage_client = storage.Client()
+tts_client = texttospeech.TextToSpeechClient()
+
+def generate_audio(event, context):
+    """
+    Triggered when a new translated text file (_en.txt) is uploaded.
+    Converts the English text into speech using WaveNet and saves the output as an .mp3 file.
+    """
+    bucket_name = event['bucket']
+    text_file = event['name']
+
+    # Process only _en.txt files
+    if not text_file.endswith("_en.txt"):
+        logging.info("File is not a translated transcript (_en.txt). Skipping...")
+        return
+
+    logging.info(f"Processing translated transcript: {text_file} in bucket: {bucket_name}")
+
+    # Read the translated English text from the file
+    bucket = storage_client.bucket(bucket_name)
+    blob = bucket.blob(text_file)
+    english_text = blob.download_as_text()
+
+    # Configure the TTS request using a WaveNet voice
+    synthesis_input = texttospeech.SynthesisInput(text=english_text)
+    voice = texttospeech.VoiceSelectionParams(
+        language_code="en-US",
+        name="en-US-Wavenet-D",  # Change the voice if desired
+        ssml_gender=texttospeech.SsmlVoiceGender.NEUTRAL
+    )
+    audio_config = texttospeech.AudioConfig(
+        audio_encoding=texttospeech.AudioEncoding.MP3,
+        pitch=0.0,
+        speaking_rate=1.0
+    )
+
+    try:
+        response = tts_client.synthesize_speech(
+            input=synthesis_input,
+            voice=voice,
+            audio_config=audio_config
+        )
+        logging.info("Text-to-Speech synthesis completed successfully.")
+    except Exception as e:
+        logging.error(f"Error during TTS synthesis: {str(e)}")
+        return
+
+    # Save the generated audio as an .mp3 file
+    output_filename = text_file.replace("_en.txt", ".mp3")
+    output_blob = bucket.blob(output_filename)
+    output_blob.upload_from_string(response.audio_content)
+
+    logging.info(f"Generated audio saved as {output_filename}")

cloud_functions/extract_audio/main.py

@@ -0,0 +1,2 @@
+google-cloud-storage
+google-cloud-texttospeech

cloud_functions/extract_audio/requirements.txt

@@ -0,0 +1,67 @@
+import logging
+from google.cloud import videointelligence, storage
+
+# Initialize clients for Cloud Storage and Video Intelligence
+storage_client = storage.Client()
+video_client = videointelligence.VideoIntelligenceServiceClient()
+
+def generate_subtitles(event, context):
+    """
+    Triggered when a new .mp3 file is uploaded to the krishnasagar-videos-processed bucket.
+    This function uses the Video Intelligence API to generate subtitles (SRT format)
+    and saves the output in the krishnasagar-videos-final bucket.
+    """
+    bucket_name = event['bucket']
+    audio_file = event['name']
+
+    # Process only .mp3 files
+    if not audio_file.endswith(".mp3"):
+        logging.info("File is not an mp3. Skipping...")
+        return
+
+    logging.info(f"Processing audio file for subtitles: {audio_file}")
+
+    # Construct GCS URI for the audio file
+    gcs_uri = f"gs://{bucket_name}/{audio_file}"
+
+    # Configure the transcription request for subtitles
+    features = [videointelligence.Feature.SPEECH_TRANSCRIPTION]
+    speech_config = videointelligence.SpeechTranscriptionConfig(
+        language_code="en-US",
+        enable_automatic_punctuation=True
+    )
+    video_context = videointelligence.VideoContext(speech_transcription_config=speech_config)
+
+    # Initiate the asynchronous video annotation request
+    operation = video_client.annotate_video(
+        request={"features": features, "input_uri": gcs_uri, "video_context": video_context}
+    )
+    logging.info("Waiting for subtitle generation operation to complete...")
+    result = operation.result(timeout=600)
+
+    # Process the results and create SRT content
+    srt_lines = []
+    index = 1
+    for transcription in result.annotation_results[0].speech_transcriptions:
+        # Use the first alternative only
+        for alternative in transcription.alternatives:
+            for word_info in alternative.words:
+                start_time = word_info.start_time
+                end_time = word_info.end_time
+                text = word_info.word
+
+                # Format timestamps as hh:mm:ss,ms (SRT format)
+                start = f"{int(start_time.seconds // 3600):02}:{int((start_time.seconds % 3600) // 60):02}:{int(start_time.seconds % 60):02},{int(start_time.nanos/1000000):03}"
+                end = f"{int(end_time.seconds // 3600):02}:{int((end_time.seconds % 3600) // 60):02}:{int(end_time.seconds % 60):02},{int(end_time.nanos/1000000):03}"
+
+                srt_lines.append(f"{index}\n{start} --> {end}\n{text}\n")
+                index += 1
+
+    srt_content = "\n".join(srt_lines)
+
+    # Save the SRT file to the krishnasagar-videos-final bucket
+    output_bucket = storage_client.bucket("krishnasagar-videos-final")
+    srt_filename = audio_file.replace(".mp3", ".srt")
+    output_blob = output_bucket.blob(srt_filename)
+    output_blob.upload_from_string(srt_content)
+    logging.info(f"Subtitles saved as {srt_filename}")

cloud_functions/generate_audio/main.py

@@ -0,0 +1,2 @@
+google-cloud-storage
+google-cloud-videointelligence