diff --git a/.gitignore b/.gitignore index ac12668..b67a7dd 100644 --- a/.gitignore +++ b/.gitignore @@ -1 +1,2 @@ -.flashrank_cache* \ No newline at end of file +.flashrank_cache* +podcasts/* diff --git a/surfsense_backend/.env.example b/surfsense_backend/.env.example index 6dfcc99..8e834bf 100644 --- a/surfsense_backend/.env.example +++ b/surfsense_backend/.env.example @@ -15,6 +15,9 @@ FAST_LLM="openai/gpt-4o-mini" STRATEGIC_LLM="openai/gpt-4o" LONG_CONTEXT_LLM="gemini/gemini-2.0-flash" +#LiteLLM TTS Provider: https://docs.litellm.ai/docs/text_to_speech#supported-providers +TTS_SERVICE="openai/tts-1" + # Chosen LiteLLM Providers Keys OPENAI_API_KEY="sk-proj-iA" GEMINI_API_KEY="AIzaSyB6-1641124124124124124124124124124" diff --git a/surfsense_backend/alembic/versions/6_change_podcast_content_to_transcript.py b/surfsense_backend/alembic/versions/6_change_podcast_content_to_transcript.py new file mode 100644 index 0000000..991948f --- /dev/null +++ b/surfsense_backend/alembic/versions/6_change_podcast_content_to_transcript.py @@ -0,0 +1,44 @@ +"""Change podcast_content to podcast_transcript with JSON type + +Revision ID: 6 +Revises: 5 +Create Date: 2023-08-15 00:00:00.000000 + +""" +from typing import Sequence, Union + +from alembic import op +import sqlalchemy as sa +from sqlalchemy.dialects.postgresql import JSON + + +# revision identifiers, used by Alembic. +revision: str = '6' +down_revision: Union[str, None] = '5' +branch_labels: Union[str, Sequence[str], None] = None +depends_on: Union[str, Sequence[str], None] = None + + +def upgrade() -> None: + # Drop the old column and create a new one with the new name and type + # We need to do this because PostgreSQL doesn't support direct column renames with type changes + op.add_column('podcasts', sa.Column('podcast_transcript', JSON, nullable=False, server_default='{}')) + + # Copy data from old column to new column + # Convert text to JSON by storing it as a JSON string value + op.execute("UPDATE podcasts SET podcast_transcript = jsonb_build_object('text', podcast_content) WHERE podcast_content != ''") + + # Drop the old column + op.drop_column('podcasts', 'podcast_content') + + +def downgrade() -> None: + # Add back the original column + op.add_column('podcasts', sa.Column('podcast_content', sa.Text(), nullable=False, server_default='')) + + # Copy data from JSON column back to text column + # Extract the 'text' field if it exists, otherwise use empty string + op.execute("UPDATE podcasts SET podcast_content = COALESCE((podcast_transcript->>'text'), '')") + + # Drop the new column + op.drop_column('podcasts', 'podcast_transcript') \ No newline at end of file diff --git a/surfsense_backend/alembic/versions/7_remove_is_generated_column.py b/surfsense_backend/alembic/versions/7_remove_is_generated_column.py new file mode 100644 index 0000000..c5d25ad --- /dev/null +++ b/surfsense_backend/alembic/versions/7_remove_is_generated_column.py @@ -0,0 +1,28 @@ +"""Remove is_generated column from podcasts table + +Revision ID: 7 +Revises: 6 +Create Date: 2023-08-15 01:00:00.000000 + +""" +from typing import Sequence, Union + +from alembic import op +import sqlalchemy as sa + + +# revision identifiers, used by Alembic. +revision: str = '7' +down_revision: Union[str, None] = '6' +branch_labels: Union[str, Sequence[str], None] = None +depends_on: Union[str, Sequence[str], None] = None + + +def upgrade() -> None: + # Drop the is_generated column + op.drop_column('podcasts', 'is_generated') + + +def downgrade() -> None: + # Add back the is_generated column with its original constraints + op.add_column('podcasts', sa.Column('is_generated', sa.Boolean(), nullable=False, server_default='false')) \ No newline at end of file diff --git a/surfsense_backend/app/agents/podcaster/__init__.py b/surfsense_backend/app/agents/podcaster/__init__.py new file mode 100644 index 0000000..8459b29 --- /dev/null +++ b/surfsense_backend/app/agents/podcaster/__init__.py @@ -0,0 +1,8 @@ +"""New LangGraph Agent. + +This module defines a custom graph. +""" + +from .graph import graph + +__all__ = ["graph"] diff --git a/surfsense_backend/app/agents/podcaster/configuration.py b/surfsense_backend/app/agents/podcaster/configuration.py new file mode 100644 index 0000000..6bbb4ce --- /dev/null +++ b/surfsense_backend/app/agents/podcaster/configuration.py @@ -0,0 +1,28 @@ +"""Define the configurable parameters for the agent.""" + +from __future__ import annotations + +from dataclasses import dataclass, fields +from typing import Optional + +from langchain_core.runnables import RunnableConfig + + +@dataclass(kw_only=True) +class Configuration: + """The configuration for the agent.""" + + # Changeme: Add configurable values here! + # these values can be pre-set when you + # create assistants (https://langchain-ai.github.io/langgraph/cloud/how-tos/configuration_cloud/) + # and when you invoke the graph + podcast_title: str + + @classmethod + def from_runnable_config( + cls, config: Optional[RunnableConfig] = None + ) -> Configuration: + """Create a Configuration instance from a RunnableConfig object.""" + configurable = (config.get("configurable") or {}) if config else {} + _fields = {f.name for f in fields(cls) if f.init} + return cls(**{k: v for k, v in configurable.items() if k in _fields}) diff --git a/surfsense_backend/app/agents/podcaster/graph.py b/surfsense_backend/app/agents/podcaster/graph.py new file mode 100644 index 0000000..d102432 --- /dev/null +++ b/surfsense_backend/app/agents/podcaster/graph.py @@ -0,0 +1,31 @@ +from langgraph.graph import StateGraph + +from .configuration import Configuration +from .state import State + + +from .nodes import create_merged_podcast_audio, create_podcast_transcript + + +def build_graph(): + + # Define a new graph + workflow = StateGraph(State, config_schema=Configuration) + + # Add the node to the graph + workflow.add_node("create_podcast_transcript", create_podcast_transcript) + workflow.add_node("create_merged_podcast_audio", create_merged_podcast_audio) + + # Set the entrypoint as `call_model` + workflow.add_edge("__start__", "create_podcast_transcript") + workflow.add_edge("create_podcast_transcript", "create_merged_podcast_audio") + workflow.add_edge("create_merged_podcast_audio", "__end__") + + # Compile the workflow into an executable graph + graph = workflow.compile() + graph.name = "Surfsense Podcaster" # This defines the custom name in LangSmith + + return graph + +# Compile the graph once when the module is loaded +graph = build_graph() diff --git a/surfsense_backend/app/agents/podcaster/nodes.py b/surfsense_backend/app/agents/podcaster/nodes.py new file mode 100644 index 0000000..19a233a --- /dev/null +++ b/surfsense_backend/app/agents/podcaster/nodes.py @@ -0,0 +1,197 @@ +from typing import Any, Dict +import json +import os +import uuid +from pathlib import Path +import asyncio + +from langchain_core.messages import HumanMessage, SystemMessage +from langchain_core.runnables import RunnableConfig +from litellm import aspeech +from ffmpeg.asyncio import FFmpeg + +from .configuration import Configuration +from .state import PodcastTranscriptEntry, State, PodcastTranscripts +from .prompts import get_podcast_generation_prompt +from app.config import config as app_config + + +async def create_podcast_transcript(state: State, config: RunnableConfig) -> Dict[str, Any]: + """Each node does work.""" + + # Initialize LLM + llm = app_config.long_context_llm_instance + + # Get the prompt + prompt = get_podcast_generation_prompt() + + # Create the messages + messages = [ + SystemMessage(content=prompt), + HumanMessage(content=f"{state.source_content}") + ] + + # Generate the podcast transcript + llm_response = await llm.ainvoke(messages) + + # First try the direct approach + try: + podcast_transcript = PodcastTranscripts.model_validate(json.loads(llm_response.content)) + except (json.JSONDecodeError, ValueError) as e: + print(f"Direct JSON parsing failed, trying fallback approach: {str(e)}") + + # Fallback: Parse the JSON response manually + try: + # Extract JSON content from the response + content = llm_response.content + + # Find the JSON in the content (handle case where LLM might add additional text) + json_start = content.find('{') + json_end = content.rfind('}') + 1 + if json_start >= 0 and json_end > json_start: + json_str = content[json_start:json_end] + + # Parse the JSON string + parsed_data = json.loads(json_str) + + # Convert to Pydantic model + podcast_transcript = PodcastTranscripts.model_validate(parsed_data) + + print(f"Successfully parsed podcast transcript using fallback approach") + else: + # If JSON structure not found, raise a clear error + error_message = f"Could not find valid JSON in LLM response. Raw response: {content}" + print(error_message) + raise ValueError(error_message) + + except (json.JSONDecodeError, ValueError) as e2: + # Log the error and re-raise it + error_message = f"Error parsing LLM response (fallback also failed): {str(e2)}" + print(f"Error parsing LLM response: {str(e2)}") + print(f"Raw response: {llm_response.content}") + raise + + return { + "podcast_transcript": podcast_transcript.podcast_transcripts + } + + +async def create_merged_podcast_audio(state: State, config: RunnableConfig) -> Dict[str, Any]: + """Generate audio for each transcript and merge them into a single podcast file.""" + + configuration = Configuration.from_runnable_config(config) + + starting_transcript = PodcastTranscriptEntry( + speaker_id=1, + dialog=f"Welcome to {configuration.podcast_title} Podcast." + ) + + transcript = state.podcast_transcript + + # Merge the starting transcript with the podcast transcript + # Check if transcript is a PodcastTranscripts object or already a list + if hasattr(transcript, 'podcast_transcripts'): + transcript_entries = transcript.podcast_transcripts + else: + transcript_entries = transcript + + merged_transcript = [starting_transcript] + transcript_entries + + # Create a temporary directory for audio files + temp_dir = Path("temp_audio") + temp_dir.mkdir(exist_ok=True) + + # Generate a unique session ID for this podcast + session_id = str(uuid.uuid4()) + output_path = f"podcasts/{session_id}_podcast.mp3" + os.makedirs("podcasts", exist_ok=True) + + # Map of speaker_id to voice + voice_mapping = { + 0: "alloy", # Default/intro voice + 1: "echo", # First speaker + # 2: "fable", # Second speaker + # 3: "onyx", # Third speaker + # 4: "nova", # Fourth speaker + # 5: "shimmer" # Fifth speaker + } + + # Generate audio for each transcript segment + audio_files = [] + + async def generate_speech_for_segment(segment, index): + # Handle both dictionary and PodcastTranscriptEntry objects + if hasattr(segment, 'speaker_id'): + speaker_id = segment.speaker_id + dialog = segment.dialog + else: + speaker_id = segment.get("speaker_id", 0) + dialog = segment.get("dialog", "") + + # Select voice based on speaker_id + voice = voice_mapping.get(speaker_id, "alloy") + + # Generate a unique filename for this segment + filename = f"{temp_dir}/{session_id}_{index}.mp3" + + try: + # Generate speech using litellm + response = await aspeech( + model=app_config.TTS_SERVICE, + voice=voice, + input=dialog, + max_retries=2, + timeout=600, + ) + + # Save the audio to a file - use proper streaming method + with open(filename, 'wb') as f: + f.write(response.content) + + return filename + except Exception as e: + print(f"Error generating speech for segment {index}: {str(e)}") + raise + + # Generate all audio files concurrently + tasks = [generate_speech_for_segment(segment, i) for i, segment in enumerate(merged_transcript)] + audio_files = await asyncio.gather(*tasks) + + # Merge audio files using ffmpeg + try: + # Create FFmpeg instance with the first input + ffmpeg = FFmpeg().option("y") + + # Add each audio file as input + for audio_file in audio_files: + ffmpeg = ffmpeg.input(audio_file) + + # Configure the concatenation and output + filter_complex = [] + for i in range(len(audio_files)): + filter_complex.append(f"[{i}:0]") + + filter_complex_str = "".join(filter_complex) + f"concat=n={len(audio_files)}:v=0:a=1[outa]" + ffmpeg = ffmpeg.option("filter_complex", filter_complex_str) + ffmpeg = ffmpeg.output(output_path, map="[outa]") + + # Execute FFmpeg + await ffmpeg.execute() + + print(f"Successfully created podcast audio: {output_path}") + + except Exception as e: + print(f"Error merging audio files: {str(e)}") + raise + finally: + # Clean up temporary files + for audio_file in audio_files: + try: + os.remove(audio_file) + except: + pass + + return { + "podcast_transcript": merged_transcript, + "final_podcast_file_path": output_path + } diff --git a/surfsense_backend/app/agents/podcaster/prompts.py b/surfsense_backend/app/agents/podcaster/prompts.py new file mode 100644 index 0000000..c08d38e --- /dev/null +++ b/surfsense_backend/app/agents/podcaster/prompts.py @@ -0,0 +1,111 @@ +import datetime + + +def get_podcast_generation_prompt(): + return f""" +Today's date: {datetime.datetime.now().strftime("%Y-%m-%d")} + +You are a master podcast scriptwriter, adept at transforming diverse input content into a lively, engaging, and natural-sounding conversation between two distinct podcast hosts. Your primary objective is to craft authentic, flowing dialogue that captures the spontaneity and chemistry of a real podcast discussion, completely avoiding any hint of robotic scripting or stiff formality. Think dynamic interplay, not just information delivery. + + +- '': A block of text containing the information to be discussed in the podcast. This could be research findings, an article summary, a detailed outline, user chat history related to the topic, or any other relevant raw information. The content might be unstructured but serves as the factual basis for the podcast dialogue. + + + +A JSON object containing the podcast transcript with alternating speakers: +{{ + "podcast_transcripts": [ + {{ + "speaker_id": 0, + "dialog": "Speaker 0 dialog here" + }}, + {{ + "speaker_id": 1, + "dialog": "Speaker 1 dialog here" + }}, + {{ + "speaker_id": 0, + "dialog": "Speaker 0 dialog here" + }}, + {{ + "speaker_id": 1, + "dialog": "Speaker 1 dialog here" + }} + ] +}} + + + +1. **Establish Distinct & Consistent Host Personas:** + * **Speaker 0 (Lead Host):** Drives the conversation forward, introduces segments, poses key questions derived from the source content, and often summarizes takeaways. Maintain a guiding, clear, and engaging tone. + * **Speaker 1 (Co-Host/Expert):** Offers deeper insights, provides alternative viewpoints or elaborations on the source content, asks clarifying or challenging questions, and shares relevant anecdotes or examples. Adopt a complementary tone (e.g., analytical, enthusiastic, reflective, slightly skeptical). + * **Consistency is Key:** Ensure each speaker maintains their distinct voice, vocabulary choice, sentence structure, and perspective throughout the entire script. Avoid having them sound interchangeable. Their interaction should feel like a genuine partnership. + +2. **Craft Natural & Dynamic Dialogue:** + * **Emulate Real Conversation:** Use contractions (e.g., "don't", "it's"), interjections ("Oh!", "Wow!", "Hmm"), discourse markers ("you know", "right?", "well"), and occasional natural pauses or filler words. Avoid overly formal language or complex sentence structures typical of written text. + * **Foster Interaction & Chemistry:** Write dialogue where speakers genuinely react *to each other*. They should build on points ("Exactly, and that reminds me..."), ask follow-up questions ("Could you expand on that?"), express agreement/disagreement respectfully ("That's a fair point, but have you considered...?"), and show active listening. + * **Vary Rhythm & Pace:** Mix short, punchy lines with longer, more explanatory ones. Vary sentence beginnings. Use questions to break up exposition. The rhythm should feel spontaneous, not monotonous. + * **Inject Personality & Relatability:** Allow for appropriate humor, moments of surprise or curiosity, brief personal reflections ("I actually experienced something similar..."), or relatable asides that fit the hosts' personas and the topic. Lightly reference past discussions if it enhances context ("Remember last week when we touched on...?"). + +3. **Structure for Flow and Listener Engagement:** + * **Natural Beginning:** Start with dialogue that flows naturally after an introduction (which will be added manually). Avoid redundant greetings or podcast name mentions since these will be added separately. + * **Logical Progression & Signposting:** Guide the listener through the information smoothly. Use clear transitions to link different ideas or segments ("So, now that we've covered X, let's dive into Y...", "That actually brings me to another key finding..."). Ensure topics flow logically from one to the next. + * **Meaningful Conclusion:** Summarize the key takeaways or main points discussed, reinforcing the core message derived from the source content. End with a final thought, a lingering question for the audience, or a brief teaser for what's next, providing a sense of closure. Avoid abrupt endings. + +4. **Integrate Source Content Seamlessly & Accurately:** + * **Translate, Don't Recite:** Rephrase information from the `` into conversational language suitable for each host's persona. Avoid directly copying dense sentences or technical jargon without explanation. The goal is discussion, not narration. + * **Explain & Contextualize:** Use analogies, simple examples, storytelling, or have one host ask clarifying questions (acting as a listener surrogate) to break down complex ideas from the source. + * **Weave Information Naturally:** Integrate facts, data, or key points from the source *within* the dialogue, not as standalone, undigested blocks. Attribute information conversationally where appropriate ("The research mentioned...", "Apparently, the key factor is..."). + * **Balance Depth & Accessibility:** Ensure the conversation is informative and factually accurate based on the source content, but prioritize clear communication and engaging delivery over exhaustive technical detail. Make it understandable and interesting for a general audience. + +5. **Length & Pacing:** + * **Six-Minute Duration:** Create a transcript that, when read at a natural speaking pace, would result in approximately 6 minutes of audio. Typically, this means around 1000 words total (based on average speaking rate of 150 words per minute). + * **Concise Speaking Turns:** Keep most speaking turns relatively brief and focused. Aim for a natural back-and-forth rhythm rather than extended monologues. + * **Essential Content Only:** Prioritize the most important information from the source content. Focus on quality over quantity, ensuring every line contributes meaningfully to the topic. + + + +Input: "Quantum computing uses quantum bits or qubits which can exist in multiple states simultaneously due to superposition." + +Output: +{{ + "podcast_transcripts": [ + {{ + "speaker_id": 0, + "dialog": "Today we're diving into the mind-bending world of quantum computing. You know, this is a topic I've been excited to cover for weeks." + }}, + {{ + "speaker_id": 1, + "dialog": "Same here! And I know our listeners have been asking for it. But I have to admit, the concept of quantum computing makes my head spin a little. Can we start with the basics?" + }}, + {{ + "speaker_id": 0, + "dialog": "Absolutely. So regular computers use bits, right? Little on-off switches that are either 1 or 0. But quantum computers use something called qubits, and this is where it gets fascinating." + }}, + {{ + "speaker_id": 1, + "dialog": "Wait, what makes qubits so special compared to regular bits?" + }}, + {{ + "speaker_id": 0, + "dialog": "The magic is in something called superposition. These qubits can exist in multiple states at the same time, not just 1 or 0." + }}, + {{ + "speaker_id": 1, + "dialog": "That sounds impossible! How would you even picture that?" + }}, + {{ + "speaker_id": 0, + "dialog": "Think of it like a coin spinning in the air. Before it lands, is it heads or tails?" + }}, + {{ + "speaker_id": 1, + "dialog": "Well, it's... neither? Or I guess both, until it lands? Oh, I think I see where you're going with this." + }} + ] +}} + + +Transform the source material into a lively and engaging podcast conversation. Craft dialogue that showcases authentic host chemistry and natural interaction (including occasional disagreement, building on points, or asking follow-up questions). Use varied speech patterns reflecting real human conversation, ensuring the final script effectively educates *and* entertains the listener while keeping within a 5-minute audio duration. + +""" \ No newline at end of file diff --git a/surfsense_backend/app/agents/podcaster/state.py b/surfsense_backend/app/agents/podcaster/state.py new file mode 100644 index 0000000..d77270d --- /dev/null +++ b/surfsense_backend/app/agents/podcaster/state.py @@ -0,0 +1,38 @@ +"""Define the state structures for the agent.""" + +from __future__ import annotations + +from dataclasses import dataclass +from typing import List, Optional +from pydantic import BaseModel, Field + + +class PodcastTranscriptEntry(BaseModel): + """ + Represents a single entry in a podcast transcript. + """ + speaker_id: int = Field(..., description="The ID of the speaker (0 or 1)") + dialog: str = Field(..., description="The dialog text spoken by the speaker") + + +class PodcastTranscripts(BaseModel): + """ + Represents the full podcast transcript structure. + """ + podcast_transcripts: List[PodcastTranscriptEntry] = Field( + ..., + description="List of transcript entries with alternating speakers" + ) + +@dataclass +class State: + """Defines the input state for the agent, representing a narrower interface to the outside world. + + This class is used to define the initial state and structure of incoming data. + See: https://langchain-ai.github.io/langgraph/concepts/low_level/#state + for more information. + """ + + source_content: str + podcast_transcript: Optional[List[PodcastTranscriptEntry]] = None + final_podcast_file_path: Optional[str] = None diff --git a/surfsense_backend/app/config/__init__.py b/surfsense_backend/app/config/__init__.py index 8c457e1..bdc370e 100644 --- a/surfsense_backend/app/config/__init__.py +++ b/surfsense_backend/app/config/__init__.py @@ -1,10 +1,12 @@ import os from pathlib import Path +import shutil from chonkie import AutoEmbeddings, CodeChunker, RecursiveChunker from dotenv import load_dotenv from langchain_community.chat_models import ChatLiteLLM from rerankers import Reranker +from litellm import speech # Get the base directory of the project BASE_DIR = Path(__file__).resolve().parent.parent.parent @@ -13,8 +15,27 @@ env_file = BASE_DIR / ".env" load_dotenv(env_file) +def is_ffmpeg_installed(): + """ + Check if ffmpeg is installed on the current system. + + Returns: + bool: True if ffmpeg is installed, False otherwise. + """ + return shutil.which("ffmpeg") is not None + + class Config: + # Check if ffmpeg is installed + if not is_ffmpeg_installed(): + import static_ffmpeg + # ffmpeg installed on first call to add_paths(), threadsafe. + static_ffmpeg.add_paths() + # check if ffmpeg is installed again + if not is_ffmpeg_installed(): + raise ValueError("FFmpeg is not installed on the system. Please install it to use the Surfsense Podcaster.") + # Database DATABASE_URL = os.getenv("DATABASE_URL") @@ -61,6 +82,9 @@ class Config: # Firecrawl API Key FIRECRAWL_API_KEY = os.getenv("FIRECRAWL_API_KEY", None) + # Litellm TTS Configuration + TTS_SERVICE = os.getenv("TTS_SERVICE") + # Validation Checks # Check embedding dimension if hasattr(embedding_model_instance, 'dimension') and embedding_model_instance.dimension > 2000: diff --git a/surfsense_backend/app/db.py b/surfsense_backend/app/db.py index b4ee3e7..7327c3a 100644 --- a/surfsense_backend/app/db.py +++ b/surfsense_backend/app/db.py @@ -110,8 +110,7 @@ class Podcast(BaseModel, TimestampMixin): __tablename__ = "podcasts" title = Column(String, nullable=False, index=True) - is_generated = Column(Boolean, nullable=False, default=False) - podcast_content = Column(Text, nullable=False, default="") + podcast_transcript = Column(JSON, nullable=False, default={}) file_location = Column(String(500), nullable=False, default="") search_space_id = Column(Integer, ForeignKey("searchspaces.id", ondelete='CASCADE'), nullable=False) diff --git a/surfsense_backend/app/routes/podcasts_routes.py b/surfsense_backend/app/routes/podcasts_routes.py index 7ac1da1..bc82e21 100644 --- a/surfsense_backend/app/routes/podcasts_routes.py +++ b/surfsense_backend/app/routes/podcasts_routes.py @@ -1,12 +1,16 @@ -from fastapi import APIRouter, Depends, HTTPException +from fastapi import APIRouter, BackgroundTasks, Depends, HTTPException from sqlalchemy.ext.asyncio import AsyncSession from sqlalchemy.future import select from sqlalchemy.exc import IntegrityError, SQLAlchemyError from typing import List -from app.db import get_async_session, User, SearchSpace, Podcast -from app.schemas import PodcastCreate, PodcastUpdate, PodcastRead +from app.db import get_async_session, User, SearchSpace, Podcast, Chat +from app.schemas import PodcastCreate, PodcastUpdate, PodcastRead, PodcastGenerateRequest from app.users import current_active_user from app.utils.check_ownership import check_ownership +from app.tasks.podcast_tasks import generate_chat_podcast +from fastapi.responses import StreamingResponse +import os +from pathlib import Path router = APIRouter() @@ -119,4 +123,121 @@ async def delete_podcast( raise he except SQLAlchemyError: await session.rollback() - raise HTTPException(status_code=500, detail="Database error occurred while deleting podcast") \ No newline at end of file + raise HTTPException(status_code=500, detail="Database error occurred while deleting podcast") + +async def generate_chat_podcast_with_new_session( + chat_id: int, + search_space_id: int, + podcast_title: str = "SurfSense Podcast" +): + """Create a new session and process chat podcast generation.""" + from app.db import async_session_maker + + async with async_session_maker() as session: + try: + await generate_chat_podcast(session, chat_id, search_space_id, podcast_title) + except Exception as e: + import logging + logging.error(f"Error generating podcast from chat: {str(e)}") + +@router.post("/podcasts/generate/") +async def generate_podcast( + request: PodcastGenerateRequest, + session: AsyncSession = Depends(get_async_session), + user: User = Depends(current_active_user), + fastapi_background_tasks: BackgroundTasks = BackgroundTasks() +): + try: + # Check if the user owns the search space + await check_ownership(session, SearchSpace, request.search_space_id, user) + + if request.type == "CHAT": + # Verify that all chat IDs belong to this user and search space + query = select(Chat).filter( + Chat.id.in_(request.ids), + Chat.search_space_id == request.search_space_id + ).join(SearchSpace).filter(SearchSpace.user_id == user.id) + + result = await session.execute(query) + valid_chats = result.scalars().all() + valid_chat_ids = [chat.id for chat in valid_chats] + + # If any requested ID is not in valid IDs, raise error immediately + if len(valid_chat_ids) != len(request.ids): + raise HTTPException( + status_code=403, + detail="One or more chat IDs do not belong to this user or search space" + ) + + # Only add a single task with the first chat ID + for chat_id in valid_chat_ids: + fastapi_background_tasks.add_task( + generate_chat_podcast_with_new_session, + chat_id, + request.search_space_id, + request.podcast_title + ) + + return { + "message": "Podcast generation started", + } + except HTTPException as he: + raise he + except IntegrityError as e: + await session.rollback() + raise HTTPException(status_code=400, detail="Podcast generation failed due to constraint violation") + except SQLAlchemyError as e: + await session.rollback() + raise HTTPException(status_code=500, detail="Database error occurred while generating podcast") + except Exception as e: + await session.rollback() + raise HTTPException(status_code=500, detail=f"An unexpected error occurred: {str(e)}") + +@router.get("/podcasts/{podcast_id}/stream") +async def stream_podcast( + podcast_id: int, + session: AsyncSession = Depends(get_async_session), + user: User = Depends(current_active_user) +): + """Stream a podcast audio file.""" + try: + # Get the podcast and check if user has access + result = await session.execute( + select(Podcast) + .join(SearchSpace) + .filter(Podcast.id == podcast_id, SearchSpace.user_id == user.id) + ) + podcast = result.scalars().first() + + if not podcast: + raise HTTPException( + status_code=404, + detail="Podcast not found or you don't have permission to access it" + ) + + # Get the file path + file_path = podcast.file_location + + # Check if the file exists + if not os.path.isfile(file_path): + raise HTTPException(status_code=404, detail="Podcast audio file not found") + + # Define a generator function to stream the file + def iterfile(): + with open(file_path, mode="rb") as file_like: + yield from file_like + + # Return a streaming response with appropriate headers + return StreamingResponse( + iterfile(), + media_type="audio/mpeg", + headers={ + "Accept-Ranges": "bytes", + "Content-Disposition": f"inline; filename={Path(file_path).name}" + } + ) + + except HTTPException as he: + raise he + except Exception as e: + raise HTTPException(status_code=500, detail=f"Error streaming podcast: {str(e)}") \ No newline at end of file diff --git a/surfsense_backend/app/schemas/__init__.py b/surfsense_backend/app/schemas/__init__.py index 07adf24..21688df 100644 --- a/surfsense_backend/app/schemas/__init__.py +++ b/surfsense_backend/app/schemas/__init__.py @@ -10,7 +10,7 @@ from .documents import ( DocumentRead, ) from .chunks import ChunkBase, ChunkCreate, ChunkUpdate, ChunkRead -from .podcasts import PodcastBase, PodcastCreate, PodcastUpdate, PodcastRead +from .podcasts import PodcastBase, PodcastCreate, PodcastUpdate, PodcastRead, PodcastGenerateRequest from .chats import ChatBase, ChatCreate, ChatUpdate, ChatRead, AISDKChatRequest from .search_source_connector import SearchSourceConnectorBase, SearchSourceConnectorCreate, SearchSourceConnectorUpdate, SearchSourceConnectorRead @@ -39,6 +39,7 @@ __all__ = [ "PodcastCreate", "PodcastUpdate", "PodcastRead", + "PodcastGenerateRequest", "ChatBase", "ChatCreate", "ChatUpdate", diff --git a/surfsense_backend/app/schemas/chats.py b/surfsense_backend/app/schemas/chats.py index ad7829b..f5eefc5 100644 --- a/surfsense_backend/app/schemas/chats.py +++ b/surfsense_backend/app/schemas/chats.py @@ -1,8 +1,10 @@ from typing import Any, Dict, List, Optional -from pydantic import BaseModel -from sqlalchemy import JSON -from .base import IDModel, TimestampModel + from app.db import ChatType +from pydantic import BaseModel + +from .base import IDModel, TimestampModel + class ChatBase(BaseModel): type: ChatType diff --git a/surfsense_backend/app/schemas/podcasts.py b/surfsense_backend/app/schemas/podcasts.py index fbec548..4132fb2 100644 --- a/surfsense_backend/app/schemas/podcasts.py +++ b/surfsense_backend/app/schemas/podcasts.py @@ -1,10 +1,10 @@ from pydantic import BaseModel +from typing import Any, List, Literal from .base import IDModel, TimestampModel class PodcastBase(BaseModel): title: str - is_generated: bool = False - podcast_content: str = "" + podcast_transcript: List[Any] file_location: str = "" search_space_id: int @@ -16,4 +16,10 @@ class PodcastUpdate(PodcastBase): class PodcastRead(PodcastBase, IDModel, TimestampModel): class Config: - from_attributes = True \ No newline at end of file + from_attributes = True + +class PodcastGenerateRequest(BaseModel): + type: Literal["DOCUMENT", "CHAT"] + ids: List[int] + search_space_id: int + podcast_title: str = "SurfSense Podcast" \ No newline at end of file diff --git a/surfsense_backend/app/tasks/podcast_tasks.py b/surfsense_backend/app/tasks/podcast_tasks.py new file mode 100644 index 0000000..e148f54 --- /dev/null +++ b/surfsense_backend/app/tasks/podcast_tasks.py @@ -0,0 +1,94 @@ +from sqlalchemy.ext.asyncio import AsyncSession +from app.schemas import PodcastGenerateRequest +from typing import List +from sqlalchemy import select +from app.db import Chat, Podcast +from app.agents.podcaster.graph import graph as podcaster_graph +from surfsense_backend.app.agents.podcaster.state import State + + +async def generate_document_podcast( + session: AsyncSession, + document_id: int, + search_space_id: int, + user_id: int +): + # TODO: Need to fetch the document chunks, then concatenate them and pass them to the podcast generation model + pass + + + +async def generate_chat_podcast( + session: AsyncSession, + chat_id: int, + search_space_id: int, + podcast_title: str +): + # Fetch the chat with the specified ID + query = select(Chat).filter( + Chat.id == chat_id, + Chat.search_space_id == search_space_id + ) + + result = await session.execute(query) + chat = result.scalars().first() + + if not chat: + raise ValueError(f"Chat with id {chat_id} not found in search space {search_space_id}") + + # Create chat history structure + chat_history_str = "" + + for message in chat.messages: + if message["role"] == "user": + chat_history_str += f"{message['content']}" + elif message["role"] == "assistant": + # Last annotation type will always be "ANSWER" here + answer_annotation = message["annotations"][-1] + answer_text = "" + if answer_annotation["type"] == "ANSWER": + answer_text = answer_annotation["content"] + # If content is a list, join it into a single string + if isinstance(answer_text, list): + answer_text = "\n".join(answer_text) + chat_history_str += f"{answer_text}" + + chat_history_str += "" + + # Pass it to the SurfSense Podcaster + config = { + "configurable": { + "podcast_title" : "Surfsense", + } + } + # Initialize state with database session and streaming service + initial_state = State( + source_content=chat_history_str, + ) + + # Run the graph directly + result = await podcaster_graph.ainvoke(initial_state, config=config) + + # Convert podcast transcript entries to serializable format + serializable_transcript = [] + for entry in result["podcast_transcript"]: + serializable_transcript.append({ + "speaker_id": entry.speaker_id, + "dialog": entry.dialog + }) + + # Create a new podcast entry + podcast = Podcast( + title=f"{podcast_title}", + podcast_transcript=serializable_transcript, + file_location=result["final_podcast_file_path"], + search_space_id=search_space_id + ) + + # Add to session and commit + session.add(podcast) + await session.commit() + await session.refresh(podcast) + + return podcast + diff --git a/surfsense_backend/pyproject.toml b/surfsense_backend/pyproject.toml index c447a74..cecf709 100644 --- a/surfsense_backend/pyproject.toml +++ b/surfsense_backend/pyproject.toml @@ -21,9 +21,11 @@ dependencies = [ "notion-client>=2.3.0", "pgvector>=0.3.6", "playwright>=1.50.0", + "python-ffmpeg>=2.0.12", "rerankers[flashrank]>=0.7.1", "sentence-transformers>=3.4.1", "slack-sdk>=3.34.0", + "static-ffmpeg>=2.13", "tavily-python>=0.3.2", "unstructured-client>=0.30.0", "unstructured[all-docs]>=0.16.25", diff --git a/surfsense_backend/uv.lock b/surfsense_backend/uv.lock index a5621ab..5f90ed9 100644 --- a/surfsense_backend/uv.lock +++ b/surfsense_backend/uv.lock @@ -590,6 +590,15 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/68/1b/e0a87d256e40e8c888847551b20a017a6b98139178505dc7ffb96f04e954/dnspython-2.7.0-py3-none-any.whl", hash = "sha256:b4c34b7d10b51bcc3a5071e7b8dee77939f1e878477eeecc965e9835f63c6c86", size = 313632 }, ] +[[package]] +name = "docutils" +version = "0.21.2" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/ae/ed/aefcc8cd0ba62a0560c3c18c33925362d46c6075480bfa4df87b28e169a9/docutils-0.21.2.tar.gz", hash = "sha256:3a6b18732edf182daa3cd12775bbb338cf5691468f91eeeb109deff6ebfa986f", size = 2204444 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/8f/d7/9322c609343d929e75e7e5e6255e614fcc67572cfd083959cdef3b7aad79/docutils-0.21.2-py3-none-any.whl", hash = "sha256:dafca5b9e384f0e419294eb4d2ff9fa826435bf15f15b7bd45723e8ad76811b2", size = 587408 }, +] + [[package]] name = "effdet" version = "0.4.1" @@ -1144,6 +1153,18 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/f0/0f/310fb31e39e2d734ccaa2c0fb981ee41f7bd5056ce9bc29b2248bd569169/humanfriendly-10.0-py2.py3-none-any.whl", hash = "sha256:1697e1a8a8f550fd43c2865cd84542fc175a61dcb779b6fee18cf6b6ccba1477", size = 86794 }, ] +[[package]] +name = "id" +version = "1.5.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "requests" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/22/11/102da08f88412d875fa2f1a9a469ff7ad4c874b0ca6fed0048fe385bdb3d/id-1.5.0.tar.gz", hash = "sha256:292cb8a49eacbbdbce97244f47a97b4c62540169c976552e497fd57df0734c1d", size = 15237 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/9f/cb/18326d2d89ad3b0dd143da971e77afd1e6ca6674f1b1c3df4b6bec6279fc/id-1.5.0-py3-none-any.whl", hash = "sha256:f1434e1cef91f2cbb8a4ec64663d5a23b9ed43ef44c4c957d02583d61714c658", size = 13611 }, +] + [[package]] name = "idna" version = "3.10" @@ -1165,6 +1186,48 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/79/9d/0fb148dc4d6fa4a7dd1d8378168d9b4cd8d4560a6fbf6f0121c5fc34eb68/importlib_metadata-8.6.1-py3-none-any.whl", hash = "sha256:02a89390c1e15fdfdc0d7c6b25cb3e62650d0494005c97d6f148bf5b9787525e", size = 26971 }, ] +[[package]] +name = "jaraco-classes" +version = "3.4.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "more-itertools" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/06/c0/ed4a27bc5571b99e3cff68f8a9fa5b56ff7df1c2251cc715a652ddd26402/jaraco.classes-3.4.0.tar.gz", hash = "sha256:47a024b51d0239c0dd8c8540c6c7f484be3b8fcf0b2d85c13825780d3b3f3acd", size = 11780 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/7f/66/b15ce62552d84bbfcec9a4873ab79d993a1dd4edb922cbfccae192bd5b5f/jaraco.classes-3.4.0-py3-none-any.whl", hash = "sha256:f662826b6bed8cace05e7ff873ce0f9283b5c924470fe664fff1c2f00f581790", size = 6777 }, +] + +[[package]] +name = "jaraco-context" +version = "6.0.1" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/df/ad/f3777b81bf0b6e7bc7514a1656d3e637b2e8e15fab2ce3235730b3e7a4e6/jaraco_context-6.0.1.tar.gz", hash = "sha256:9bae4ea555cf0b14938dc0aee7c9f32ed303aa20a3b73e7dc80111628792d1b3", size = 13912 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/ff/db/0c52c4cf5e4bd9f5d7135ec7669a3a767af21b3a308e1ed3674881e52b62/jaraco.context-6.0.1-py3-none-any.whl", hash = "sha256:f797fc481b490edb305122c9181830a3a5b76d84ef6d1aef2fb9b47ab956f9e4", size = 6825 }, +] + +[[package]] +name = "jaraco-functools" +version = "4.1.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "more-itertools" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/ab/23/9894b3df5d0a6eb44611c36aec777823fc2e07740dabbd0b810e19594013/jaraco_functools-4.1.0.tar.gz", hash = "sha256:70f7e0e2ae076498e212562325e805204fc092d7b4c17e0e86c959e249701a9d", size = 19159 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/9f/4f/24b319316142c44283d7540e76c7b5a6dbd5db623abd86bb7b3491c21018/jaraco.functools-4.1.0-py3-none-any.whl", hash = "sha256:ad159f13428bc4acbf5541ad6dec511f91573b90fba04df61dafa2a1231cf649", size = 10187 }, +] + +[[package]] +name = "jeepney" +version = "0.9.0" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/7b/6f/357efd7602486741aa73ffc0617fb310a29b588ed0fd69c2399acbb85b0c/jeepney-0.9.0.tar.gz", hash = "sha256:cf0e9e845622b81e4a28df94c40345400256ec608d0e55bb8a3feaa9163f5732", size = 106758 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/b2/a3/e137168c9c44d18eff0376253da9f1e9234d0239e0ee230d2fee6cea8e55/jeepney-0.9.0-py3-none-any.whl", hash = "sha256:97e5714520c16fc0a45695e5365a2e11b81ea79bba796e26f9f1d178cb182683", size = 49010 }, +] + [[package]] name = "jinja2" version = "3.1.5" @@ -1269,6 +1332,23 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/d1/0f/8910b19ac0670a0f80ce1008e5e751c4a57e14d2c4c13a482aa6079fa9d6/jsonschema_specifications-2024.10.1-py3-none-any.whl", hash = "sha256:a09a0680616357d9a0ecf05c12ad234479f549239d0f5b55f3deea67475da9bf", size = 18459 }, ] +[[package]] +name = "keyring" +version = "25.6.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "jaraco-classes" }, + { name = "jaraco-context" }, + { name = "jaraco-functools" }, + { name = "jeepney", marker = "sys_platform == 'linux'" }, + { name = "pywin32-ctypes", marker = "sys_platform == 'win32'" }, + { name = "secretstorage", marker = "sys_platform == 'linux'" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/70/09/d904a6e96f76ff214be59e7aa6ef7190008f52a0ab6689760a98de0bf37d/keyring-25.6.0.tar.gz", hash = "sha256:0b39998aa941431eb3d9b0d4b2460bc773b9df6fed7621c2dfb291a7e0187a66", size = 62750 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/d3/32/da7f44bcb1105d3e88a0b74ebdca50c59121d2ddf71c9e34ba47df7f3a56/keyring-25.6.0-py3-none-any.whl", hash = "sha256:552a3f7af126ece7ed5c89753650eec89c7eaae8617d0aa4d9ad2b75111266bd", size = 39085 }, +] + [[package]] name = "kiwisolver" version = "1.4.8" @@ -1754,6 +1834,15 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/cd/76/c8575f90f521017597c5e57e3bfef61e3f27d9cb6c741a82a24d72b10a60/model2vec-0.4.1-py3-none-any.whl", hash = "sha256:04a397a17da9b967082b6baa4c494f0be48c89ec4e1a3975b4f290f045238a38", size = 41972 }, ] +[[package]] +name = "more-itertools" +version = "10.7.0" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/ce/a0/834b0cebabbfc7e311f30b46c8188790a37f89fc8d756660346fe5abfd09/more_itertools-10.7.0.tar.gz", hash = "sha256:9fddd5403be01a94b204faadcff459ec3568cf110265d3c54323e1e866ad29d3", size = 127671 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/2b/9f/7ba6f94fc1e9ac3d2b853fdff3035fb2fa5afbed898c4a72b8a020610594/more_itertools-10.7.0-py3-none-any.whl", hash = "sha256:d43980384673cb07d2f7d2d918c616b30c659c089ee23953f601d6609c67510e", size = 65278 }, +] + [[package]] name = "mpmath" version = "1.3.0" @@ -1829,6 +1918,37 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/b9/54/dd730b32ea14ea797530a4479b2ed46a6fb250f682a9cfb997e968bf0261/networkx-3.4.2-py3-none-any.whl", hash = "sha256:df5d4365b724cf81b8c6a7312509d0c22386097011ad1abe274afd5e9d3bbc5f", size = 1723263 }, ] +[[package]] +name = "nh3" +version = "0.2.21" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/37/30/2f81466f250eb7f591d4d193930df661c8c23e9056bdc78e365b646054d8/nh3-0.2.21.tar.gz", hash = "sha256:4990e7ee6a55490dbf00d61a6f476c9a3258e31e711e13713b2ea7d6616f670e", size = 16581 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/7f/81/b83775687fcf00e08ade6d4605f0be9c4584cb44c4973d9f27b7456a31c9/nh3-0.2.21-cp313-cp313t-macosx_10_12_x86_64.macosx_11_0_arm64.macosx_10_12_universal2.whl", hash = "sha256:fcff321bd60c6c5c9cb4ddf2554e22772bb41ebd93ad88171bbbb6f271255286", size = 1297678 }, + { url = "https://files.pythonhosted.org/packages/22/ee/d0ad8fb4b5769f073b2df6807f69a5e57ca9cea504b78809921aef460d20/nh3-0.2.21-cp313-cp313t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:31eedcd7d08b0eae28ba47f43fd33a653b4cdb271d64f1aeda47001618348fde", size = 733774 }, + { url = "https://files.pythonhosted.org/packages/ea/76/b450141e2d384ede43fe53953552f1c6741a499a8c20955ad049555cabc8/nh3-0.2.21-cp313-cp313t-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:d426d7be1a2f3d896950fe263332ed1662f6c78525b4520c8e9861f8d7f0d243", size = 760012 }, + { url = "https://files.pythonhosted.org/packages/97/90/1182275db76cd8fbb1f6bf84c770107fafee0cb7da3e66e416bcb9633da2/nh3-0.2.21-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:9d67709bc0d7d1f5797b21db26e7a8b3d15d21c9c5f58ccfe48b5328483b685b", size = 923619 }, + { url = "https://files.pythonhosted.org/packages/29/c7/269a7cfbec9693fad8d767c34a755c25ccb8d048fc1dfc7a7d86bc99375c/nh3-0.2.21-cp313-cp313t-musllinux_1_2_armv7l.whl", hash = "sha256:55823c5ea1f6b267a4fad5de39bc0524d49a47783e1fe094bcf9c537a37df251", size = 1000384 }, + { url = "https://files.pythonhosted.org/packages/68/a9/48479dbf5f49ad93f0badd73fbb48b3d769189f04c6c69b0df261978b009/nh3-0.2.21-cp313-cp313t-musllinux_1_2_i686.whl", hash = "sha256:818f2b6df3763e058efa9e69677b5a92f9bc0acff3295af5ed013da544250d5b", size = 918908 }, + { url = "https://files.pythonhosted.org/packages/d7/da/0279c118f8be2dc306e56819880b19a1cf2379472e3b79fc8eab44e267e3/nh3-0.2.21-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:b3b5c58161e08549904ac4abd450dacd94ff648916f7c376ae4b2c0652b98ff9", size = 909180 }, + { url = "https://files.pythonhosted.org/packages/26/16/93309693f8abcb1088ae143a9c8dbcece9c8f7fb297d492d3918340c41f1/nh3-0.2.21-cp313-cp313t-win32.whl", hash = "sha256:637d4a10c834e1b7d9548592c7aad760611415fcd5bd346f77fd8a064309ae6d", size = 532747 }, + { url = "https://files.pythonhosted.org/packages/a2/3a/96eb26c56cbb733c0b4a6a907fab8408ddf3ead5d1b065830a8f6a9c3557/nh3-0.2.21-cp313-cp313t-win_amd64.whl", hash = "sha256:713d16686596e556b65e7f8c58328c2df63f1a7abe1277d87625dcbbc012ef82", size = 528908 }, + { url = "https://files.pythonhosted.org/packages/ba/1d/b1ef74121fe325a69601270f276021908392081f4953d50b03cbb38b395f/nh3-0.2.21-cp38-abi3-macosx_10_12_x86_64.macosx_11_0_arm64.macosx_10_12_universal2.whl", hash = "sha256:a772dec5b7b7325780922dd904709f0f5f3a79fbf756de5291c01370f6df0967", size = 1316133 }, + { url = "https://files.pythonhosted.org/packages/b8/f2/2c7f79ce6de55b41e7715f7f59b159fd59f6cdb66223c05b42adaee2b645/nh3-0.2.21-cp38-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d002b648592bf3033adfd875a48f09b8ecc000abd7f6a8769ed86b6ccc70c759", size = 758328 }, + { url = "https://files.pythonhosted.org/packages/6d/ad/07bd706fcf2b7979c51b83d8b8def28f413b090cf0cb0035ee6b425e9de5/nh3-0.2.21-cp38-abi3-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:2a5174551f95f2836f2ad6a8074560f261cf9740a48437d6151fd2d4d7d617ab", size = 747020 }, + { url = "https://files.pythonhosted.org/packages/75/99/06a6ba0b8a0d79c3d35496f19accc58199a1fb2dce5e711a31be7e2c1426/nh3-0.2.21-cp38-abi3-manylinux_2_17_ppc64.manylinux2014_ppc64.whl", hash = "sha256:b8d55ea1fc7ae3633d758a92aafa3505cd3cc5a6e40470c9164d54dff6f96d42", size = 944878 }, + { url = "https://files.pythonhosted.org/packages/79/d4/dc76f5dc50018cdaf161d436449181557373869aacf38a826885192fc587/nh3-0.2.21-cp38-abi3-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:6ae319f17cd8960d0612f0f0ddff5a90700fa71926ca800e9028e7851ce44a6f", size = 903460 }, + { url = "https://files.pythonhosted.org/packages/cd/c3/d4f8037b2ab02ebf5a2e8637bd54736ed3d0e6a2869e10341f8d9085f00e/nh3-0.2.21-cp38-abi3-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:63ca02ac6f27fc80f9894409eb61de2cb20ef0a23740c7e29f9ec827139fa578", size = 839369 }, + { url = "https://files.pythonhosted.org/packages/11/a9/1cd3c6964ec51daed7b01ca4686a5c793581bf4492cbd7274b3f544c9abe/nh3-0.2.21-cp38-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a5f77e62aed5c4acad635239ac1290404c7e940c81abe561fd2af011ff59f585", size = 739036 }, + { url = "https://files.pythonhosted.org/packages/fd/04/bfb3ff08d17a8a96325010ae6c53ba41de6248e63cdb1b88ef6369a6cdfc/nh3-0.2.21-cp38-abi3-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:087ffadfdcd497658c3adc797258ce0f06be8a537786a7217649fc1c0c60c293", size = 768712 }, + { url = "https://files.pythonhosted.org/packages/9e/aa/cfc0bf545d668b97d9adea4f8b4598667d2b21b725d83396c343ad12bba7/nh3-0.2.21-cp38-abi3-musllinux_1_2_aarch64.whl", hash = "sha256:ac7006c3abd097790e611fe4646ecb19a8d7f2184b882f6093293b8d9b887431", size = 930559 }, + { url = "https://files.pythonhosted.org/packages/78/9d/6f5369a801d3a1b02e6a9a097d56bcc2f6ef98cffebf03c4bb3850d8e0f0/nh3-0.2.21-cp38-abi3-musllinux_1_2_armv7l.whl", hash = "sha256:6141caabe00bbddc869665b35fc56a478eb774a8c1dfd6fba9fe1dfdf29e6efa", size = 1008591 }, + { url = "https://files.pythonhosted.org/packages/a6/df/01b05299f68c69e480edff608248313cbb5dbd7595c5e048abe8972a57f9/nh3-0.2.21-cp38-abi3-musllinux_1_2_i686.whl", hash = "sha256:20979783526641c81d2f5bfa6ca5ccca3d1e4472474b162c6256745fbfe31cd1", size = 925670 }, + { url = "https://files.pythonhosted.org/packages/3d/79/bdba276f58d15386a3387fe8d54e980fb47557c915f5448d8c6ac6f7ea9b/nh3-0.2.21-cp38-abi3-musllinux_1_2_x86_64.whl", hash = "sha256:a7ea28cd49293749d67e4fcf326c554c83ec912cd09cd94aa7ec3ab1921c8283", size = 917093 }, + { url = "https://files.pythonhosted.org/packages/e7/d8/c6f977a5cd4011c914fb58f5ae573b071d736187ccab31bfb1d539f4af9f/nh3-0.2.21-cp38-abi3-win32.whl", hash = "sha256:6c9c30b8b0d291a7c5ab0967ab200598ba33208f754f2f4920e9343bdd88f79a", size = 537623 }, + { url = "https://files.pythonhosted.org/packages/23/fc/8ce756c032c70ae3dd1d48a3552577a325475af2a2f629604b44f571165c/nh3-0.2.21-cp38-abi3-win_amd64.whl", hash = "sha256:bb0014948f04d7976aabae43fcd4cb7f551f9f8ce785a4c9ef66e6c2590f8629", size = 535283 }, +] + [[package]] name = "nltk" version = "3.9.1" @@ -2366,6 +2486,12 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/bc/2b/e944e10c9b18e77e43d3bb4d6faa323f6cc27597db37b75bc3fd796adfd5/playwright-1.50.0-py3-none-win_amd64.whl", hash = "sha256:1859423da82de631704d5e3d88602d755462b0906824c1debe140979397d2e8d", size = 34784546 }, ] +[[package]] +name = "progress" +version = "1.6" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/2a/68/d8412d1e0d70edf9791cbac5426dc859f4649afc22f2abbeb0d947cf70fd/progress-1.6.tar.gz", hash = "sha256:c9c86e98b5c03fa1fe11e3b67c1feda4788b8d0fe7336c2ff7d5644ccfba34cd", size = 7842 } + [[package]] name = "propcache" version = "0.2.1" @@ -2705,6 +2831,19 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/6a/3e/b68c118422ec867fa7ab88444e1274aa40681c606d59ac27de5a5588f082/python_dotenv-1.0.1-py3-none-any.whl", hash = "sha256:f7b63ef50f1b690dddf550d03497b66d609393b40b564ed0d674909a68ebf16a", size = 19863 }, ] +[[package]] +name = "python-ffmpeg" +version = "2.0.12" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "pyee" }, + { name = "typing-extensions" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/dd/4d/7ecffb341d646e016be76e36f5a42cb32f409c9ca21a57b68f067fad3fc7/python_ffmpeg-2.0.12.tar.gz", hash = "sha256:19ac80af5a064a2f53c245af1a909b2d7648ea045500d96d3bcd507b88d43dc7", size = 14126292 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/7f/6d/02e817aec661defe148cb9eb0c4eca2444846305f625c2243fb9f92a9045/python_ffmpeg-2.0.12-py3-none-any.whl", hash = "sha256:d86697da8dfb39335183e336d31baf42fb217468adf5ac97fd743898240faae3", size = 14411 }, +] + [[package]] name = "python-iso639" version = "2025.2.18" @@ -2770,6 +2909,15 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/eb/38/ac33370d784287baa1c3d538978b5e2ea064d4c1b93ffbd12826c190dd10/pytz-2025.1-py2.py3-none-any.whl", hash = "sha256:89dd22dca55b46eac6eda23b2d72721bf1bdfef212645d81513ef5d03038de57", size = 507930 }, ] +[[package]] +name = "pywin32-ctypes" +version = "0.2.3" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/85/9f/01a1a99704853cb63f253eea009390c88e7131c67e66a0a02099a8c917cb/pywin32-ctypes-0.2.3.tar.gz", hash = "sha256:d162dc04946d704503b2edc4d55f3dba5c1d539ead017afa00142c38b9885755", size = 29471 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/de/3d/8161f7711c017e01ac9f008dfddd9410dff3674334c233bde66e7ba65bbf/pywin32_ctypes-0.2.3-py3-none-any.whl", hash = "sha256:8a1513379d709975552d202d942d9837758905c8d01eb82b8bcc30918929e7b8", size = 30756 }, +] + [[package]] name = "pyyaml" version = "6.0.2" @@ -2834,6 +2982,20 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/4b/43/ca3d1018b392f49131843648e10b08ace23afe8dad3bee5f136e4346b7cd/rapidfuzz-3.12.2-cp313-cp313-win_arm64.whl", hash = "sha256:69f6ecdf1452139f2b947d0c169a605de578efdb72cbb2373cb0a94edca1fd34", size = 863535 }, ] +[[package]] +name = "readme-renderer" +version = "44.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "docutils" }, + { name = "nh3" }, + { name = "pygments" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/5a/a9/104ec9234c8448c4379768221ea6df01260cd6c2ce13182d4eac531c8342/readme_renderer-44.0.tar.gz", hash = "sha256:8712034eabbfa6805cacf1402b4eeb2a73028f72d1166d6f5cb7f9c047c5d1e1", size = 32056 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/e1/67/921ec3024056483db83953ae8e48079ad62b92db7880013ca77632921dd0/readme_renderer-44.0-py3-none-any.whl", hash = "sha256:2fbca89b81a08526aadf1357a8c2ae889ec05fb03f5da67f9769c9a592166151", size = 13310 }, +] + [[package]] name = "referencing" version = "0.36.2" @@ -2927,6 +3089,15 @@ flashrank = [ { name = "flashrank" }, ] +[[package]] +name = "rfc3986" +version = "2.0.0" +source = { registry = "https://pypi.org/simple" } +sdist = { url = "https://files.pythonhosted.org/packages/85/40/1520d68bfa07ab5a6f065a186815fb6610c86fe957bc065754e47f7b0840/rfc3986-2.0.0.tar.gz", hash = "sha256:97aacf9dbd4bfd829baad6e6309fa6573aaf1be3f6fa735c8ab05e46cecb261c", size = 49026 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/ff/9a/9afaade874b2fa6c752c36f1548f718b5b83af81ed9b76628329dab81c1b/rfc3986-2.0.0-py2.py3-none-any.whl", hash = "sha256:50b1502b60e289cb37883f3dfd34532b8873c7de9f49bb546641ce9cbd256ebd", size = 31326 }, +] + [[package]] name = "rich" version = "14.0.0" @@ -3083,6 +3254,19 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/e4/1f/5d46a8d94e9f6d2c913cbb109e57e7eed914de38ea99e2c4d69a9fc93140/scipy-1.15.1-cp313-cp313t-win_amd64.whl", hash = "sha256:bc7136626261ac1ed988dca56cfc4ab5180f75e0ee52e58f1e6aa74b5f3eacd5", size = 43181730 }, ] +[[package]] +name = "secretstorage" +version = "3.3.3" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "cryptography", marker = "sys_platform != 'darwin'" }, + { name = "jeepney", marker = "sys_platform != 'darwin'" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/53/a4/f48c9d79cb507ed1373477dbceaba7401fd8a23af63b837fa61f1dcd3691/SecretStorage-3.3.3.tar.gz", hash = "sha256:2403533ef369eca6d2ba81718576c5e0f564d5cca1b58f73a8b23e7d4eeebd77", size = 19739 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/54/24/b4293291fa1dd830f353d2cb163295742fa87f179fcc8a20a306a81978b7/SecretStorage-3.3.3-py3-none-any.whl", hash = "sha256:f356e6628222568e3af06f2eba8df495efa13b3b63081dafd4f7d9a7b7bc9f99", size = 15221 }, +] + [[package]] name = "sentence-transformers" version = "3.4.1" @@ -3192,6 +3376,20 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/d9/61/f2b52e107b1fc8944b33ef56bf6ac4ebbe16d91b94d2b87ce013bf63fb84/starlette-0.45.3-py3-none-any.whl", hash = "sha256:dfb6d332576f136ec740296c7e8bb8c8a7125044e7c6da30744718880cdd059d", size = 71507 }, ] +[[package]] +name = "static-ffmpeg" +version = "2.13" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "filelock" }, + { name = "progress" }, + { name = "requests" }, + { name = "twine" }, +] +wheels = [ + { url = "https://files.pythonhosted.org/packages/09/39/1a5d0603280dd681ec52a2a6717c05dab530190dff7887b7603740a1741b/static_ffmpeg-2.13-py3-none-any.whl", hash = "sha256:3bed55a7979f9de9d1eec1126b98774a1d41c2e323811f59973d54b9c94d6dac", size = 7586 }, +] + [[package]] name = "surf-new-backend" version = "0.0.6" @@ -3213,9 +3411,11 @@ dependencies = [ { name = "notion-client" }, { name = "pgvector" }, { name = "playwright" }, + { name = "python-ffmpeg" }, { name = "rerankers", extra = ["flashrank"] }, { name = "sentence-transformers" }, { name = "slack-sdk" }, + { name = "static-ffmpeg" }, { name = "tavily-python" }, { name = "unstructured", extra = ["all-docs"] }, { name = "unstructured-client" }, @@ -3242,9 +3442,11 @@ requires-dist = [ { name = "notion-client", specifier = ">=2.3.0" }, { name = "pgvector", specifier = ">=0.3.6" }, { name = "playwright", specifier = ">=1.50.0" }, + { name = "python-ffmpeg", specifier = ">=2.0.12" }, { name = "rerankers", extras = ["flashrank"], specifier = ">=0.7.1" }, { name = "sentence-transformers", specifier = ">=3.4.1" }, { name = "slack-sdk", specifier = ">=3.34.0" }, + { name = "static-ffmpeg", specifier = ">=2.13" }, { name = "tavily-python", specifier = ">=0.3.2" }, { name = "unstructured", extras = ["all-docs"], specifier = ">=0.16.25" }, { name = "unstructured-client", specifier = ">=0.30.0" }, @@ -3549,6 +3751,26 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/c7/30/37a3384d1e2e9320331baca41e835e90a3767303642c7a80d4510152cbcf/triton-3.2.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e5dfa23ba84541d7c0a531dfce76d8bcd19159d50a4a8b14ad01e91734a5c1b0", size = 253154278 }, ] +[[package]] +name = "twine" +version = "6.1.0" +source = { registry = "https://pypi.org/simple" } +dependencies = [ + { name = "id" }, + { name = "keyring", marker = "platform_machine != 'ppc64le' and platform_machine != 's390x'" }, + { name = "packaging" }, + { name = "readme-renderer" }, + { name = "requests" }, + { name = "requests-toolbelt" }, + { name = "rfc3986" }, + { name = "rich" }, + { name = "urllib3" }, +] +sdist = { url = "https://files.pythonhosted.org/packages/c8/a2/6df94fc5c8e2170d21d7134a565c3a8fb84f9797c1dd65a5976aaf714418/twine-6.1.0.tar.gz", hash = "sha256:be324f6272eff91d07ee93f251edf232fc647935dd585ac003539b42404a8dbd", size = 168404 } +wheels = [ + { url = "https://files.pythonhosted.org/packages/7c/b6/74e927715a285743351233f33ea3c684528a0d374d2e43ff9ce9585b73fe/twine-6.1.0-py3-none-any.whl", hash = "sha256:a47f973caf122930bf0fbbf17f80b83bc1602c9ce393c7845f289a3001dc5384", size = 40791 }, +] + [[package]] name = "types-requests" version = "2.32.0.20250328" diff --git a/surfsense_web/app/dashboard/[search_space_id]/chats/chats-client.tsx b/surfsense_web/app/dashboard/[search_space_id]/chats/chats-client.tsx index c481bd6..7b14c7e 100644 --- a/surfsense_web/app/dashboard/[search_space_id]/chats/chats-client.tsx +++ b/surfsense_web/app/dashboard/[search_space_id]/chats/chats-client.tsx @@ -3,7 +3,7 @@ import { useState, useEffect } from 'react'; import { motion, AnimatePresence } from 'framer-motion'; import { useSearchParams } from 'next/navigation'; -import { MessageCircleMore, Search, Calendar, Tag, Trash2, ExternalLink, MoreHorizontal } from 'lucide-react'; +import { MessageCircleMore, Search, Calendar, Tag, Trash2, ExternalLink, MoreHorizontal, Radio, CheckCircle, Circle, Podcast } from 'lucide-react'; import { format } from 'date-fns'; // UI Components @@ -42,6 +42,9 @@ import { SelectTrigger, SelectValue, } from "@/components/ui/select"; +import { Checkbox } from "@/components/ui/checkbox"; +import { Label } from "@/components/ui/label"; +import { toast } from "sonner"; interface Chat { created_at: string; @@ -92,6 +95,18 @@ export default function ChatsPageClient({ searchSpaceId }: ChatsPageClientProps) const [chatToDelete, setChatToDelete] = useState<{ id: number, title: string } | null>(null); const [isDeleting, setIsDeleting] = useState(false); + // New state for podcast generation + const [selectedChats, setSelectedChats] = useState([]); + const [selectionMode, setSelectionMode] = useState(false); + const [podcastDialogOpen, setPodcastDialogOpen] = useState(false); + const [podcastTitle, setPodcastTitle] = useState(""); + const [isGeneratingPodcast, setIsGeneratingPodcast] = useState(false); + + // New state for individual podcast generation + const [currentChatIndex, setCurrentChatIndex] = useState(0); + const [podcastTitles, setPodcastTitles] = useState<{[key: number]: string}>({}); + const [processingChat, setProcessingChat] = useState(null); + const chatsPerPage = 9; const searchParams = useSearchParams(); @@ -234,6 +249,177 @@ export default function ChatsPageClient({ searchSpaceId }: ChatsPageClientProps) // Get unique chat types for filter dropdown const chatTypes = ['all', ...Array.from(new Set(chats.map(chat => chat.type)))]; + // Generate individual podcasts from selected chats + const handleGeneratePodcast = async () => { + if (selectedChats.length === 0) { + toast.error("Please select at least one chat"); + return; + } + + const currentChatId = selectedChats[currentChatIndex]; + const currentTitle = podcastTitles[currentChatId] || podcastTitle; + + if (!currentTitle.trim()) { + toast.error("Please enter a podcast title"); + return; + } + + setIsGeneratingPodcast(true); + try { + const token = localStorage.getItem('surfsense_bearer_token'); + if (!token) { + toast.error("Authentication error. Please log in again."); + setIsGeneratingPodcast(false); + return; + } + + // Create payload for single chat + const payload = { + type: "CHAT", + ids: [currentChatId], // Single chat ID + search_space_id: parseInt(searchSpaceId), + podcast_title: currentTitle + }; + + const response = await fetch(`${process.env.NEXT_PUBLIC_FASTAPI_BACKEND_URL}/api/v1/podcasts/generate/`, { + method: 'POST', + headers: { + 'Authorization': `Bearer ${token}`, + 'Content-Type': 'application/json', + }, + body: JSON.stringify(payload) + }); + + if (!response.ok) { + const errorData = await response.json().catch(() => ({})); + throw new Error(errorData.detail || "Failed to generate podcast"); + } + + const data = await response.json(); + toast.success(`Podcast "${currentTitle}" generation started!`); + + // Move to the next chat or finish + if (currentChatIndex < selectedChats.length - 1) { + // Set up for next chat + setCurrentChatIndex(currentChatIndex + 1); + + // Find the next chat from the chats array + const nextChatId = selectedChats[currentChatIndex + 1]; + const nextChat = chats.find(chat => chat.id === nextChatId) || null; + setProcessingChat(nextChat); + + // Default title for the next chat + if (!podcastTitles[nextChatId]) { + setPodcastTitle(nextChat?.title || `Podcast from Chat ${nextChatId}`); + } else { + setPodcastTitle(podcastTitles[nextChatId]); + } + + setIsGeneratingPodcast(false); + } else { + // All done + finishPodcastGeneration(); + } + } catch (error) { + console.error('Error generating podcast:', error); + toast.error(error instanceof Error ? error.message : 'Failed to generate podcast'); + setIsGeneratingPodcast(false); + } + }; + + // Helper to finish the podcast generation process + const finishPodcastGeneration = () => { + toast.success("All podcasts are being generated! Check the podcasts tab to see them when ready."); + setPodcastDialogOpen(false); + setSelectedChats([]); + setSelectionMode(false); + setCurrentChatIndex(0); + setPodcastTitles({}); + setProcessingChat(null); + setPodcastTitle(""); + setIsGeneratingPodcast(false); + }; + + // Start podcast generation flow + const startPodcastGeneration = () => { + if (selectedChats.length === 0) { + toast.error("Please select at least one chat"); + return; + } + + // Reset the state for podcast generation + setCurrentChatIndex(0); + setPodcastTitles({}); + + // Set up for the first chat + const firstChatId = selectedChats[0]; + const firstChat = chats.find(chat => chat.id === firstChatId) || null; + setProcessingChat(firstChat); + + // Set default title for the first chat + setPodcastTitle(firstChat?.title || `Podcast from Chat ${firstChatId}`); + setPodcastDialogOpen(true); + }; + + // Update the title for the current chat + const updateCurrentChatTitle = (title: string) => { + const currentChatId = selectedChats[currentChatIndex]; + setPodcastTitle(title); + setPodcastTitles(prev => ({ + ...prev, + [currentChatId]: title + })); + }; + + // Skip generating a podcast for the current chat + const skipCurrentChat = () => { + if (currentChatIndex < selectedChats.length - 1) { + // Move to the next chat + setCurrentChatIndex(currentChatIndex + 1); + + // Find the next chat + const nextChatId = selectedChats[currentChatIndex + 1]; + const nextChat = chats.find(chat => chat.id === nextChatId) || null; + setProcessingChat(nextChat); + + // Set default title for the next chat + if (!podcastTitles[nextChatId]) { + setPodcastTitle(nextChat?.title || `Podcast from Chat ${nextChatId}`); + } else { + setPodcastTitle(podcastTitles[nextChatId]); + } + } else { + // All done (all skipped) + finishPodcastGeneration(); + } + }; + + // Toggle chat selection + const toggleChatSelection = (chatId: number) => { + setSelectedChats(prev => + prev.includes(chatId) + ? prev.filter(id => id !== chatId) + : [...prev, chatId] + ); + }; + + // Select all visible chats + const selectAllVisibleChats = () => { + const visibleChatIds = currentChats.map(chat => chat.id); + setSelectedChats(prev => { + const allSelected = visibleChatIds.every(id => prev.includes(id)); + return allSelected + ? prev.filter(id => !visibleChatIds.includes(id)) // Deselect all visible if all are selected + : [...new Set([...prev, ...visibleChatIds])]; // Add all visible, ensuring no duplicates + }); + }; + + // Cancel selection mode + const cancelSelectionMode = () => { + setSelectionMode(false); + setSelectedChats([]); + }; + return ( -
- +
+ {selectionMode ? ( + <> + + + + + ) : ( + <> + + + + )}
@@ -334,44 +565,79 @@ export default function ChatsPageClient({ searchSpaceId }: ChatsPageClientProps) animate="animate" exit="exit" transition={{ duration: 0.2, delay: index * 0.05 }} - className="overflow-hidden hover:shadow-md transition-shadow" + className={`overflow-hidden hover:shadow-md transition-shadow + ${selectionMode && selectedChats.includes(chat.id) + ? 'ring-2 ring-primary ring-offset-2' : ''}`} + onClick={(e) => { + if (!selectionMode) return; + // Ignore clicks coming from interactive elements + if ((e.target as HTMLElement).closest('button, a, [data-stop-selection]')) return; + toggleChatSelection(chat.id); + }} >
-
- {chat.title || `Chat ${chat.id}`} - - - - {format(new Date(chat.created_at), 'MMM d, yyyy')} - - +
+ {selectionMode && ( +
+ {selectedChats.includes(chat.id) + ? + : } +
+ )} +
+ {chat.title || `Chat ${chat.id}`} + + + + {format(new Date(chat.created_at), 'MMM d, yyyy')} + + +
- - - - - - window.location.href = `/dashboard/${chat.search_space_id}/researcher/${chat.id}`}> - - View Chat - - - { - setChatToDelete({ id: chat.id, title: chat.title || `Chat ${chat.id}` }); - setDeleteDialogOpen(true); - }} - > - - Delete Chat - - - + {!selectionMode && ( + + + + + + window.location.href = `/dashboard/${chat.search_space_id}/researcher/${chat.id}`}> + + View Chat + + { + setSelectedChats([chat.id]); + setPodcastTitle(chat.title || `Chat ${chat.id}`); + setPodcastDialogOpen(true); + }} + > + + Generate Podcast + + + { + e.stopPropagation(); + setChatToDelete({ id: chat.id, title: chat.title || `Chat ${chat.id}` }); + setDeleteDialogOpen(true); + }} + > + + Delete Chat + + + + )}
@@ -505,6 +771,104 @@ export default function ChatsPageClient({ searchSpaceId }: ChatsPageClientProps) + + {/* Podcast Generation Dialog */} + { + if (!isOpen) { + // Cancel the process if dialog is closed + setPodcastDialogOpen(false); + setSelectedChats([]); + setSelectionMode(false); + setCurrentChatIndex(0); + setPodcastTitles({}); + setProcessingChat(null); + setPodcastTitle(""); + } else { + setPodcastDialogOpen(true); + } + }} + > + + + + + Generate Podcast {currentChatIndex + 1} of {selectedChats.length} + + + {selectedChats.length > 1 ? ( + <>Creating individual podcasts for each selected chat. Currently processing: {processingChat?.title || `Chat ${selectedChats[currentChatIndex]}`} + ) : ( + <>Create a podcast from this chat. The podcast will be available in the podcasts section once generated. + )} + + + +
+
+ + updateCurrentChatTitle(e.target.value)} + /> +
+ + {selectedChats.length > 1 && ( +
+
+
+ )} +
+ + + {selectedChats.length > 1 && !isGeneratingPodcast && ( + + )} + + + + + ); } \ No newline at end of file diff --git a/surfsense_web/app/dashboard/[search_space_id]/layout.tsx b/surfsense_web/app/dashboard/[search_space_id]/layout.tsx index 7449e10..a3c344a 100644 --- a/surfsense_web/app/dashboard/[search_space_id]/layout.tsx +++ b/surfsense_web/app/dashboard/[search_space_id]/layout.tsx @@ -73,6 +73,13 @@ export default function DashboardLayout({ }, ], }, + { + title: "Podcasts", + url: `/dashboard/${search_space_id}/podcasts`, + icon: "Podcast", + items: [ + ], + } // TODO: Add research synthesizer's // { // title: "Research Synthesizer's", diff --git a/surfsense_web/app/dashboard/[search_space_id]/podcasts/page.tsx b/surfsense_web/app/dashboard/[search_space_id]/podcasts/page.tsx new file mode 100644 index 0000000..394177c --- /dev/null +++ b/surfsense_web/app/dashboard/[search_space_id]/podcasts/page.tsx @@ -0,0 +1,22 @@ +import { Suspense } from 'react'; +import PodcastsPageClient from './podcasts-client'; + +interface PageProps { + params: { + search_space_id: string; + }; +} + +export default async function PodcastsPage({ params }: PageProps) { + // Access dynamic route parameters + // Need to await params before accessing its properties in an async component + const { search_space_id: searchSpaceId } = await Promise.resolve(params); + + return ( + +
+
}> + + + ); +} diff --git a/surfsense_web/app/dashboard/[search_space_id]/podcasts/podcasts-client.tsx b/surfsense_web/app/dashboard/[search_space_id]/podcasts/podcasts-client.tsx new file mode 100644 index 0000000..72e50ab --- /dev/null +++ b/surfsense_web/app/dashboard/[search_space_id]/podcasts/podcasts-client.tsx @@ -0,0 +1,789 @@ +'use client'; + +import { useState, useEffect, useRef } from 'react'; +import { motion, AnimatePresence } from 'framer-motion'; +import { format } from 'date-fns'; +import { + Search, Calendar, Trash2, MoreHorizontal, Podcast, + Play, Pause, SkipForward, SkipBack, Volume2, VolumeX +} from 'lucide-react'; + +// UI Components +import { Input } from '@/components/ui/input'; +import { Button } from '@/components/ui/button'; +import { Card, CardContent, CardFooter, CardHeader, CardTitle } from '@/components/ui/card'; +import { Slider } from '@/components/ui/slider'; +import { + DropdownMenu, + DropdownMenuContent, + DropdownMenuItem, + DropdownMenuTrigger, + DropdownMenuSeparator +} from '@/components/ui/dropdown-menu'; +import { + Dialog, + DialogContent, + DialogDescription, + DialogFooter, + DialogHeader, + DialogTitle, +} from "@/components/ui/dialog"; +import { + Select, + SelectContent, + SelectGroup, + SelectItem, + SelectTrigger, + SelectValue, +} from "@/components/ui/select"; +import { toast } from "sonner"; + +interface PodcastItem { + id: number; + title: string; + created_at: string; + file_location: string; + podcast_transcript: any[]; + search_space_id: number; +} + +interface PodcastsPageClientProps { + searchSpaceId: string; +} + +const pageVariants = { + initial: { opacity: 0 }, + enter: { opacity: 1, transition: { duration: 0.3, ease: 'easeInOut' } }, + exit: { opacity: 0, transition: { duration: 0.3, ease: 'easeInOut' } } +}; + +const podcastCardVariants = { + initial: { y: 20, opacity: 0 }, + animate: { y: 0, opacity: 1 }, + exit: { y: -20, opacity: 0 } +}; + +const MotionCard = motion(Card); + +export default function PodcastsPageClient({ searchSpaceId }: PodcastsPageClientProps) { + const [podcasts, setPodcasts] = useState([]); + const [filteredPodcasts, setFilteredPodcasts] = useState([]); + const [isLoading, setIsLoading] = useState(true); + const [error, setError] = useState(null); + const [searchQuery, setSearchQuery] = useState(''); + const [sortOrder, setSortOrder] = useState('newest'); + const [deleteDialogOpen, setDeleteDialogOpen] = useState(false); + const [podcastToDelete, setPodcastToDelete] = useState<{ id: number, title: string } | null>(null); + const [isDeleting, setIsDeleting] = useState(false); + + // Audio player state + const [currentPodcast, setCurrentPodcast] = useState(null); + const [audioSrc, setAudioSrc] = useState(undefined); + const [isAudioLoading, setIsAudioLoading] = useState(false); + const [isPlaying, setIsPlaying] = useState(false); + const [currentTime, setCurrentTime] = useState(0); + const [duration, setDuration] = useState(0); + const [volume, setVolume] = useState(0.7); + const [isMuted, setIsMuted] = useState(false); + const audioRef = useRef(null); + const currentObjectUrlRef = useRef(null); + + // Add podcast image URL constant + const PODCAST_IMAGE_URL = "https://static.vecteezy.com/system/resources/thumbnails/002/157/611/small_2x/illustrations-concept-design-podcast-channel-free-vector.jpg"; + + // Fetch podcasts from API + useEffect(() => { + const fetchPodcasts = async () => { + try { + setIsLoading(true); + + // Get token from localStorage + const token = localStorage.getItem('surfsense_bearer_token'); + + if (!token) { + setError('Authentication token not found. Please log in again.'); + setIsLoading(false); + return; + } + + // Fetch all podcasts for this search space + const response = await fetch( + `${process.env.NEXT_PUBLIC_FASTAPI_BACKEND_URL}/api/v1/podcasts/`, + { + headers: { + 'Authorization': `Bearer ${token}`, + 'Content-Type': 'application/json', + }, + cache: 'no-store', + } + ); + + if (!response.ok) { + const errorData = await response.json().catch(() => null); + throw new Error(`Failed to fetch podcasts: ${response.status} ${errorData?.detail || ''}`); + } + + const data: PodcastItem[] = await response.json(); + setPodcasts(data); + setFilteredPodcasts(data); + setError(null); + } catch (error) { + console.error('Error fetching podcasts:', error); + setError(error instanceof Error ? error.message : 'Unknown error occurred'); + setPodcasts([]); + setFilteredPodcasts([]); + } finally { + setIsLoading(false); + } + }; + + fetchPodcasts(); + }, [searchSpaceId]); + + // Filter and sort podcasts based on search query and sort order + useEffect(() => { + let result = [...podcasts]; + + // Filter by search term + if (searchQuery) { + const query = searchQuery.toLowerCase(); + result = result.filter(podcast => + podcast.title.toLowerCase().includes(query) + ); + } + + // Filter by search space + result = result.filter(podcast => + podcast.search_space_id === parseInt(searchSpaceId) + ); + + // Sort podcasts + result.sort((a, b) => { + const dateA = new Date(a.created_at).getTime(); + const dateB = new Date(b.created_at).getTime(); + + return sortOrder === 'newest' ? dateB - dateA : dateA - dateB; + }); + + setFilteredPodcasts(result); + }, [podcasts, searchQuery, sortOrder, searchSpaceId]); + + // Cleanup object URL on unmount or when currentPodcast changes + useEffect(() => { + return () => { + if (currentObjectUrlRef.current) { + URL.revokeObjectURL(currentObjectUrlRef.current); + currentObjectUrlRef.current = null; + } + }; + }, []); + + // Audio player time update handler + const handleTimeUpdate = () => { + if (audioRef.current) { + setCurrentTime(audioRef.current.currentTime); + } + }; + + // Audio player metadata loaded handler + const handleMetadataLoaded = () => { + if (audioRef.current) { + setDuration(audioRef.current.duration); + } + }; + + // Play/pause toggle + const togglePlayPause = () => { + if (audioRef.current) { + if (isPlaying) { + audioRef.current.pause(); + } else { + audioRef.current.play(); + } + setIsPlaying(!isPlaying); + } + }; + + // Seek to position + const handleSeek = (value: number[]) => { + if (audioRef.current) { + audioRef.current.currentTime = value[0]; + setCurrentTime(value[0]); + } + }; + + // Volume change + const handleVolumeChange = (value: number[]) => { + if (audioRef.current) { + const newVolume = value[0]; + audioRef.current.volume = newVolume; + setVolume(newVolume); + + if (newVolume === 0) { + setIsMuted(true); + } else if (isMuted) { + setIsMuted(false); + } + } + }; + + // Toggle mute + const toggleMute = () => { + if (audioRef.current) { + audioRef.current.muted = !isMuted; + setIsMuted(!isMuted); + } + }; + + // Skip forward 10 seconds + const skipForward = () => { + if (audioRef.current) { + audioRef.current.currentTime = Math.min(audioRef.current.duration, audioRef.current.currentTime + 10); + } + }; + + // Skip backward 10 seconds + const skipBackward = () => { + if (audioRef.current) { + audioRef.current.currentTime = Math.max(0, audioRef.current.currentTime - 10); + } + }; + + // Format time in MM:SS + const formatTime = (time: number) => { + const minutes = Math.floor(time / 60); + const seconds = Math.floor(time % 60); + return `${minutes}:${seconds < 10 ? '0' : ''}${seconds}`; + }; + + // Play podcast - Fetch blob and set object URL + const playPodcast = async (podcast: PodcastItem) => { + // If the same podcast is selected, just toggle play/pause + if (currentPodcast && currentPodcast.id === podcast.id) { + togglePlayPause(); + return; + } + + // Revoke previous object URL if exists + if (currentObjectUrlRef.current) { + URL.revokeObjectURL(currentObjectUrlRef.current); + currentObjectUrlRef.current = null; + } + + // Reset player state and show loading + setCurrentPodcast(podcast); + setAudioSrc(undefined); + setCurrentTime(0); + setDuration(0); + setIsPlaying(false); + setIsAudioLoading(true); + + try { + const token = localStorage.getItem('surfsense_bearer_token'); + if (!token) { + toast.error('Authentication token not found.'); + setIsAudioLoading(false); + return; + } + + const response = await fetch( + `${process.env.NEXT_PUBLIC_FASTAPI_BACKEND_URL}/api/v1/podcasts/${podcast.id}/stream`, + { + headers: { + 'Authorization': `Bearer ${token}`, + }, + } + ); + + if (!response.ok) { + throw new Error(`Failed to fetch audio stream: ${response.statusText}`); + } + + const blob = await response.blob(); + const objectUrl = URL.createObjectURL(blob); + currentObjectUrlRef.current = objectUrl; + + // Wait for React to commit the new `src` + setAudioSrc(objectUrl); + + // Use requestAnimationFrame instead of setTimeout for more reliable DOM updates + requestAnimationFrame(() => { + if (audioRef.current) { + // The