Open source video conferencing app powered by LiveKit. Built with Django and React.
✨(summary) extend v2 routes to support visio usecase
Florent Chehabecf8f0fe3f50
diff --git a/src/summary/summary/api/route/tasks_v2.py b/src/summary/summary/api/route/tasks_v2.pyindex e8f8cc78..eebe88e8 100644--- a/src/summary/summary/api/route/tasks_v2.py+++ b/src/summary/summary/api/route/tasks_v2.py@@ -1,15 +1,19 @@"""API routes related to application tasks (V2 / tenant friendly)."""+import logging+from datetime import datetime, timezone+from celery.result import AsyncResult-from fastapi import APIRouter, Depends, HTTPException, Request+from fastapi import APIRouter, Depends, HTTPException, Request, status+from summary.core.analytics import get_analyticsfrom summary.core.celery_worker import (celery,process_audio_transcribe_v2_task,summarize_v2_task,)-from summary.core.config import AuthorizedTenant-from summary.core.models import SummarizeTaskV2Request, TranscribeTaskV2Request+from summary.core.config import AuthorizedTenant, get_settings+from summary.core.models import SummarizeTaskApiRequest, TranscribeTaskApiRequestfrom summary.core.security import verify_tenant_api_key_v2from summary.core.shared_models import (SummarizeWebhookFailurePayload,@@ -20,17 +24,52 @@ from summary.core.shared_models import (TranscribeWebhookSuccessPayload,)+logger = logging.getLogger(__name__)router_tasks_v2 = APIRouter()+analytics = get_analytics()+settings = get_settings()+@router_tasks_v2.post("/async-jobs/transcribe")async def create_transcribe_task_v2(- request: TranscribeTaskV2Request,+ request: TranscribeTaskApiRequest,request_tenant: AuthorizedTenant = Depends(verify_tenant_api_key_v2),):"""Create a transcription task."""+ if (+ request.push_to_docs_config is not None+ and not request_tenant.allowed_push_to_docs+ ):+ logger.warning(+ "Push to docs is not allowed for this tenant (%s).", request_tenant.id+ )+ raise HTTPException(+ status_code=status.HTTP_400_BAD_REQUEST,+ detail="Push to docs is not allowed for this tenant.",+ )+task = process_audio_transcribe_v2_task.apply_async(- args=[{**request.model_dump(), "tenant_id": request_tenant.id}]+ args=[+ {+ **request.model_dump(),+ "tenant_id": request_tenant.id,+ "received_at": datetime.now(timezone.utc),+ }+ ]+ )++ # We track the request, this also properly initializes the user in the+ # analytics system, so that later feature flags work properly+ analytics.capture(+ settings.posthog_event_request,+ request.user_sub,+ properties={+ "kind": "transcribe",+ "$set": {+ "email": request.user_email,+ },+ },)return TranscribeWebhookPendingPayload(job_id=task.id).model_dump()@@ -38,14 +77,32 @@ async def create_transcribe_task_v2(@router_tasks_v2.post("/async-jobs/summarize")async def create_summarize_task_v2(- request: SummarizeTaskV2Request,+ request: SummarizeTaskApiRequest,request_tenant: AuthorizedTenant = Depends(verify_tenant_api_key_v2),):"""Create a summarization task."""task = summarize_v2_task.apply_async(- args=[{**request.model_dump(), "tenant_id": request_tenant.id}]+ args=[+ {+ **request.model_dump(),+ "tenant_id": request_tenant.id,+ "received_at": datetime.now(timezone.utc),+ }+ ])+ # We track the request, this also properly initializes the user in the+ # analytics system, so that later feature flags work properly+ analytics.capture(+ settings.posthog_event_request,+ request.user_sub,+ properties={+ "kind": "summarize",+ "$set": {+ "email": request.user_email,+ },+ },+ )return SummarizeWebhookPendingPayload(job_id=task.id).model_dump()diff --git a/src/summary/summary/core/analytics.py b/src/summary/summary/core/analytics.pyindex cffe1a37..fb52ef8a 100644--- a/src/summary/summary/core/analytics.py+++ b/src/summary/summary/core/analytics.py@@ -4,12 +4,14 @@ import jsonimport timefrom collections import Counterfrom functools import lru_cache+from urllib.parse import urlsplit, urlunsplitimport redisfrom celery.utils.log import get_task_loggerfrom posthog import Posthogfrom summary.core.config import get_settings+from summary.core.models import TranscribeTaskJoblogger = get_task_logger(__name__)settings = get_settings()@@ -107,23 +109,24 @@ class MetadataManager:"""Check if task_id exists in tasks metadata cache."""return self._redis.exists(self._get_redis_key(task_id))- def create(self, task_id, task_args):+ def create(self, task_id: str, task_payload: TranscribeTaskJob):"""Create initial metadata entry for a new task."""if self._is_disabled or self.has_task_id(task_id):return- # Positional args mirror process_audio_transcribe_summarize_v2 signature:- # owner_id, recording_filename, metadata_filename, email, sub, received_at, ...- _, filename, _, email, _, received_at, *_ = task_args-start_time = time.time()+ parts = urlsplit(task_payload.cloud_storage_url)+ clean_url = urlunsplit((parts.scheme, parts.netloc, parts.path, "", ""))initial_metadata = {"start_time": start_time,"asr_model": settings.whisperx_asr_model,"retries": 0,- "filename": filename,- "email": email,- "queuing_time": round(start_time - received_at, 2),+ "filename": clean_url,+ "sub": task_payload.user_sub,+ # avoid None in redis, it shouldn't happen anyway in prod+ "email": task_payload.user_email or "",+ "tenant_id": task_payload.tenant_id,+ "queuing_time": round(start_time - task_payload.received_at.timestamp(), 2),}self._save_metadata(task_id, initial_metadata)@@ -210,6 +213,6 @@ class MetadataManager:self.clear(task_id)try:- self._analytics.capture(event_name, metadata.get("email"), metadata)+ self._analytics.capture(event_name, metadata.get("sub"), metadata)except AnalyticsException:logger.exception("Failed to capture analytics event")diff --git a/src/summary/summary/core/celery_worker.py b/src/summary/summary/core/celery_worker.pyindex ff803cac..091c5fae 100644--- a/src/summary/summary/core/celery_worker.py+++ b/src/summary/summary/core/celery_worker.py@@ -4,7 +4,7 @@import jsonimport time-from datetime import datetime+from datetime import datetime, timezonefrom typing import Anyfrom urllib.parse import urljoin@@ -17,16 +17,15 @@ from requests import exceptionsfrom summary.core.analytics import MetadataManager, get_analyticsfrom summary.core.config import get_settings-from summary.core.file_service import (- FileService,- FileServiceException,- TranscribeError,-)+from summary.core.docs_service import create_document_in_lasuite_docs+from summary.core.file_service import FileService, FileServiceException, TranscribeErrorfrom summary.core.llm_service import LLMException, LLMObservability, LLMServicefrom summary.core.locales import get_localefrom summary.core.models import (- SummarizeTaskV2Payload,- TranscribeTaskV2Payload,+ PushToDocsBaseConfig,+ RecordingMetadata,+ SummarizeTaskJob,+ TranscribeTaskJob,)from summary.core.prompt import (FORMAT_NEXT_STEPS,@@ -85,29 +84,22 @@ file_service = FileService()def transcribe_audio(*,task_id: str,- recording_filename: str | None = None,language: str,- cloud_storage_url=None,+ cloud_storage_url: str,raises: bool = False,):"""Transcribe an audio file using WhisperX.- Downloads the audio from MinIO or a cloud storage URL, sends it to+ Downloads the audio from a cloud storage URL, sends it toWhisperX for transcription, and tracks metadata throughout the process.Returns the transcription object, or None if the file could not be retrieved."""- if bool(recording_filename) == bool(cloud_storage_url):- raise ValueError(- "Either filename or cloud_storage_url must be provided, but not both."- )-logger.info("Initiating WhisperX client")# Transcriptiontry:with file_service.prepare_audio_file(- remote_object_key=recording_filename,cloud_storage_url=cloud_storage_url,) as (audio_file, metadata):metadata_manager.track(task_id, {"audio_length": metadata["duration"]})@@ -196,11 +188,7 @@ def transcribe_audio(cloud_storage_url.split("?", 1)[0] if cloud_storage_url else None)logger.exception(- (- "Unexpected error while preparing file | filename: %s "- "| cloud_storage_url: %s"- ),- recording_filename,+ ("Unexpected error while preparing file %s "),redacted_cloud_storage_url,)return None@@ -210,44 +198,34 @@ def transcribe_audio(def resolve_speaker_identities_and_apply_to(- transcription, recording_start_at, recording_end_at, metadata_filename, task_id-):+ *, transcription: WhisperXResponse, recording_metadata: RecordingMetadata, task_id+) -> WhisperXResponse:"""Assign users to detected speakers and rewrite the transcriptions.Args:transcription: output of meet-whisperx after transcription and diarization- recording_start_at: sourced from LiveKit FileInfo via the egress_ended webhook- recording_end_at: sourced from LiveKit FileInfo via the egress_ended webhook- metadata_filename: name of metadata file containing VAD information in S3+ recording_metadata: Metadata of the recordingtask_id: current task id, for logging purposes"""- recording_start_dt = (- datetime.fromisoformat(recording_start_at) if recording_start_at else None- )- recording_end_dt = (- datetime.fromisoformat(recording_end_at) if recording_end_at else None- )-logger.debug("recording_start_dt: %s ; recording_end_dt: %s",- recording_start_dt,- recording_end_dt,+ recording_metadata.started_at,+ recording_metadata.ended_at,)- if (recording_start_dt is None) or (recording_end_dt is None):- logger.debug("Skipping resolve_speaker_identities")- return transcriptionlogger.debug("Running resolve_speaker_identities")try:- metadata = file_service.read_json(metadata_filename)+ metadata = file_service.read_cloud_storage_json(+ recording_metadata.cloud_storage_url+ )speaker_mapping = resolve_speaker_identities(metadata,- transcription,- recording_start_dt,- recording_end_dt,+ transcription.model_dump(),+ recording_metadata.started_at,+ recording_metadata.ended_at,)new_transcription = speaker_mapping.apply_to(transcription.model_dump())- return new_transcription+ return WhisperXResponse.model_validate(new_transcription)except FileServiceException as exc:logger.error(@@ -272,12 +250,9 @@ def format_transcript(transcription,context_language: str | None,language: str,- room: str | None,- recording_datetime: str | None,- owner_timezone: str | None,download_link: str | None,form_link: str | None,-) -> tuple[str, str]:+) -> str:"""Format a transcription into readable content with a title.Resolves the locale from context_language / language, then uses@@ -290,9 +265,6 @@ def format_transcript(return formatter.format(transcription,- room=room,- recording_datetime=recording_datetime,- owner_timezone=owner_timezone,download_link=download_link,form_link=form_link,)@@ -317,7 +289,7 @@ def format_actions(llm_output: dict) -> str:def summarize_transcription_internals(- *, owner_id: str, transcript: str, session_id: str+ *, distinct_id: str, transcript: str, session_id: str) -> str:"""Generate a summary from the provided transcription text.@@ -328,11 +300,11 @@ def summarize_transcription_internals("""logger.info("Starting summarization task | Owner: %s",- owner_id,+ distinct_id,)user_has_tracing_consent = analytics.is_feature_enabled(- "summary-tracing-consent", distinct_id=owner_id+ "summary-tracing-consent", distinct_id=distinct_id)# NOTE: We must instantiate a new LLMObservability client for each task invocation@@ -343,7 +315,7 @@ def summarize_transcription_internals(llm_observability = LLMObservability(user_has_tracing_consent=user_has_tracing_consent,session_id=session_id,- user_id=owner_id,+ user_id=distinct_id,)llm_service = LLMService(llm_observability=llm_observability)@@ -400,6 +372,52 @@ def summarize_transcription_internals(##################################################################################+def _should_push_to_docs(+ payload: TranscribeTaskJob | SummarizeTaskJob,+) -> bool:+ """Determines if the transcription should be pushed to docs.++ Based on the payload and settings.+ """+ if not payload.push_to_docs_config:+ reason = "Push to docs is not requested in the payload"+ elif not settings.is_lasuite_docs_integration_enabled:+ reason = "Docs integration is disabled"+ elif not settings.get_authorized_tenant(+ tenant_id=payload.tenant_id+ ).allowed_push_to_docs:+ reason = "Tenant is not allowed to push to docs"+ else:+ return True++ logger.info("Push to docs is not requested: %s", reason)+ return False+++def _should_auto_create_summary(payload: TranscribeTaskJob) -> bool:+ """Determines if the transcription should have an auto-created summary.++ Based on the payload and settings.+ """+ if (+ payload.push_to_docs_config is None+ or not payload.push_to_docs_config.auto_create_summary+ ):+ reason = "Auto create summary is not requested in the payload"+ elif not settings.is_summary_enabled:+ reason = "Summary feature is disabled"+ elif not analytics.is_feature_enabled(+ "summary-enabled",+ distinct_id=payload.user_sub,+ ):+ reason = "Summary feature flag return false"+ else:+ return True++ logger.info("Auto create summary is not requested: %s", reason)+ return False++@celery.task(max_retries=3,queue=settings.call_webhook_queue_v2,@@ -436,7 +454,7 @@ def process_audio_transcribe_v2_task(self: Celery task instance (passed on with bind=True)payload: Serialized dictionary of TranscribeSummarizeTaskCreationV2"""- payload = TranscribeTaskV2Payload.model_validate(payload)+ payload = TranscribeTaskJob.model_validate(payload)logger.info("Transcribing for object received | Owner: %s",payload.user_sub,@@ -463,6 +481,60 @@ def process_audio_transcribe_v2_task()return failure_payload.model_dump()+ # Assign speakers and rewrite transcription/diarization output+ if settings.is_resolve_speaker_identities_enabled and payload.metadata is not None:+ try:+ transcription_res = resolve_speaker_identities_and_apply_to(+ transcription=transcription_res,+ recording_metadata=payload.metadata,+ task_id=job_id,+ )+ except Exception as e:+ logger.error(f"Failed to resolve speaker identities, skipping: {e}")++ should_push_to_docs = _should_push_to_docs(payload)+ # We do it synchronously for now+ if should_push_to_docs:+ if payload.push_to_docs_config is None:+ raise ValueError("Push to docs config is missing")++ # Format output+ content = format_transcript(+ transcription_res.model_dump(),+ payload.context_language,+ payload.language,+ payload.push_to_docs_config.download_link,+ payload.push_to_docs_config.form_link,+ )++ create_document_in_lasuite_docs(+ content=content,+ title=payload.push_to_docs_config.title,+ email=payload.push_to_docs_config.user_email,+ sub=payload.user_sub,+ )++ if _should_auto_create_summary(payload):+ locale = get_locale(payload.context_language, payload.language)++ summarize_v2_task.apply_async(+ args=[+ SummarizeTaskJob(+ received_at=datetime.now(timezone.utc),+ tenant_id=payload.tenant_id,+ user_sub=payload.user_sub,+ user_email=payload.user_email,+ push_to_docs_config=PushToDocsBaseConfig(+ user_email=payload.push_to_docs_config.user_email,+ title=locale.summary_title_template.format(+ title=payload.push_to_docs_config.title+ ),+ ),+ content=content,+ ).model_dump()+ ],+ )+file_service.store_transcript(transcript=transcription_res,job_id=job_id,@@ -475,9 +547,30 @@ def process_audio_transcribe_v2_task(call_webhook_v2_task.apply_async(args=[success_payload.model_dump(), payload.tenant_id])+ metadata_manager.capture(job_id, settings.posthog_event_success)+return success_payload.model_dump()[email protected]_prerun.connect(sender=process_audio_transcribe_v2_task)+def task_started(task_id=None, task=None, args=None, **kwargs):+ """Signal handler called before task execution begins."""+ if args:+ metadata_manager.create(task_id, TranscribeTaskJob.model_validate(args[0]))++[email protected]_retry.connect(sender=process_audio_transcribe_v2_task)+def task_retry_handler(request=None, reason=None, einfo=None, **kwargs):+ """Signal handler called when task execution retries."""+ metadata_manager.retry(request.id)++[email protected]_failure.connect(sender=process_audio_transcribe_v2_task)+def task_failure_handler(task_id, exception=None, **kwargs):+ """Signal handler called when task execution fails permanently."""+ metadata_manager.capture(task_id, settings.posthog_event_failure)++@signals.task_failure.connect(sender=process_audio_transcribe_v2_task)def handle_transcribe_v2_failed(sender,@@ -533,15 +626,26 @@ def summarize_v2_task(1. Run summary internals2. Sends the final summary via webhook."""- payload = SummarizeTaskV2Payload.model_validate(payload)+ payload = SummarizeTaskJob.model_validate(payload)summary = summarize_transcription_internals(- owner_id=payload.user_sub,+ distinct_id=payload.user_sub,transcript=payload.content,session_id=self.request.id,)job_id = self.request.idfile_service.store_summary(summary=summary, job_id=job_id)+ if _should_push_to_docs(payload):+ if payload.push_to_docs_config is None:+ raise ValueError("Push to docs config is missing")++ create_document_in_lasuite_docs(+ content=summary,+ title=payload.push_to_docs_config.title,+ email=payload.push_to_docs_config.user_email,+ sub=payload.user_sub,+ )+success_payload = SummarizeWebhookSuccessPayload(job_id=job_id,summary_data_url=file_service.get_summary_signed_url(job_id),diff --git a/src/summary/summary/core/config.py b/src/summary/summary/core/config.pyindex c3089cf8..8ee4eabf 100644--- a/src/summary/summary/core/config.py+++ b/src/summary/summary/core/config.py@@ -1,9 +1,8 @@"""Application configuration and settings."""import logging-import osfrom functools import cached_property, lru_cache-from typing import Annotated, Any, List, Literal, Mapping, Optional, Set+from typing import Annotated, List, Mapping, Optional, Setfrom fastapi import Dependsfrom pydantic import (@@ -34,6 +33,12 @@ class AuthorizedTenant(BaseModel):title="Webhook API Key",description="The api_key to authenticate the webhook request.",)+ allowed_push_to_docs: bool = Field(+ title="Allow Push to Docs",+ description="Whether to allow pushing transcript"+ " and summaries to docs for this tenant.",+ default=False,+ )class Settings(BaseSettings):@@ -83,7 +88,7 @@ class Settings(BaseSettings):# AI-related settingswhisperx_api_key: SecretStr- whisperx_base_url: str = "https://api.openai.com/v1"+ whisperx_base_url: Url = "https://api.openai.com/v1"whisperx_asr_model: str = "whisper-1"# ISO 639-1 language code (e.g., "en", "fr", "es")whisperx_default_language: Optional[str] = None@@ -105,10 +110,17 @@ class Settings(BaseSettings):webhook_max_retries: int = 2webhook_status_forcelist: List[int] = [502, 503, 504]webhook_backoff_factor: float = 0.1+ app_external_user_agent: str = "summary"# Summary related settingsis_summary_enabled: bool = True- transcription_satisfaction_form_base_url: Optional[str] = None++ # Docs service configuration+ is_lasuite_docs_integration_enabled: bool = False+ lasuite_docs_base_url: Url = "https://example.com"+ lasuite_docs_server_to_server_api_key: SecretStr = Field(+ title="API key for using docs server to server api", default="NO_API_KEY"+ )# Sentrysentry_is_enabled: bool = False@@ -120,6 +132,7 @@ class Settings(BaseSettings):posthog_api_host: Optional[str] = "https://eu.i.posthog.com"posthog_event_failure: str = "transcript-failure"posthog_event_success: str = "transcript-success"+ posthog_event_request: str = "transcript-request"# Langfuse (LLM Observability)langfuse_enabled: bool = False@@ -148,6 +161,25 @@ class Settings(BaseSettings):raise ValueError("Duplicate application API api_keys are not allowed")return self+ @model_validator(mode="after")+ def validate_docs_config(self):+ """Validate docs integration configuration."""+ if not self.is_lasuite_docs_integration_enabled:+ return self++ if not self.lasuite_docs_base_url:+ raise ValueError(+ "lasuite_docs_base_url is required when docs integration is enabled"+ )+ if self.lasuite_docs_server_to_server_api_key.get_secret_value() in {+ "",+ "NO_API_KEY",+ }:+ raise ValueError(+ "Valid lasuite_docs_server_to_server_api_key is required when"+ " docs integration is enabled"+ )+ return self@cached_propertydef authorized_tenant_api_keys(self) -> frozenset[str]:diff --git a/src/summary/summary/core/docs_service.py b/src/summary/summary/core/docs_service.pynew file mode 100644index 00000000..96c5f130--- /dev/null+++ b/src/summary/summary/core/docs_service.py@@ -0,0 +1,94 @@+"""Service for delivering content to external destinations."""++import json+import logging+from urllib.parse import urljoin++from requests import Session+from requests.adapters import HTTPAdapter+from urllib3.util import Retry++from summary.core.config import get_settings++settings = get_settings()++logger = logging.getLogger(__name__)+++def _create_retry_session(api_key: str | None = None):+ """Create an HTTP session configured with retry logic."""+ session = Session()+ retries = Retry(+ total=settings.webhook_max_retries,+ backoff_factor=settings.webhook_backoff_factor,+ status_forcelist=settings.webhook_status_forcelist,+ allowed_methods={"POST"},+ )+ session.mount("https://", HTTPAdapter(max_retries=retries))+ if api_key:+ session.headers.update(+ {+ "Authorization": f"Bearer {api_key}",+ "User-Agent": settings.app_external_user_agent,+ }+ )+ return session+++def _post_with_retries(*, url, data, api_key: str | None = None):+ """Send POST request with automatic retries."""+ session = _create_retry_session(api_key=api_key)++ try:+ response = session.post(url, json=data, timeout=(20, 3 * 60))+ response.raise_for_status()+ return response+ finally:+ session.close()+++def create_document_in_lasuite_docs(+ *, content: str, title: str, email: str, sub: str+) -> None:+ """Call the Docs API to create a document on behalf of the user there.++ Builds the payload, sends it with retries, and logs the outcome.+ """+ data = {+ "title": title,+ "content": content,+ "email": email,+ "sub": sub,+ }++ logger.debug(+ "Submitting document to docs endpoint: %s", settings.lasuite_docs_base_url+ )+ logger.debug(+ "Docs payload metadata | title_len=%s content_len=%s has_email=%s has_sub=%s",+ len(title),+ len(content),+ bool(email),+ bool(sub),+ )++ response = _post_with_retries(+ url=urljoin(+ settings.lasuite_docs_base_url, "/api/v1.0/documents/create-for-owner/"+ ),+ api_key=settings.lasuite_docs_server_to_server_api_key.get_secret_value(),+ data=data,+ )++ try:+ response_data = response.json()+ document_id = response_data.get("id", "N/A")+ except (json.JSONDecodeError, AttributeError):+ document_id = "Unable to parse response"++ logger.info(+ "Delivery success | Document %s submitted (HTTP %s)",+ document_id,+ response.status_code,+ )+ logger.debug("Docs response received (body omitted)")diff --git a/src/summary/summary/core/locales/de.py b/src/summary/summary/core/locales/de.pyindex 10a8d2c7..985c7390 100644--- a/src/summary/summary/core/locales/de.py+++ b/src/summary/summary/core/locales/de.py@@ -30,8 +30,5 @@ Einige Punkte, die wir Ihnen empfehlen zu überprüfen:"(externer Link)]({form_link})*\n"),hallucination_replacement_text="[Text konnte nicht transkribiert werden]",- document_default_title="Transkription",- document_title_template=(- 'Besprechung "{room}" am {room_recording_date} um {room_recording_time}'- ),+ summary_title_template="Zusammenfassung von {title}",)diff --git a/src/summary/summary/core/locales/en.py b/src/summary/summary/core/locales/en.pyindex 22841d84..c0d51279 100644--- a/src/summary/summary/core/locales/en.py+++ b/src/summary/summary/core/locales/en.py@@ -30,8 +30,5 @@ A few things we recommend you check:"(external link)]({form_link})*\n"),hallucination_replacement_text="[Unable to transcribe text]",- document_default_title="Transcription",- document_title_template=(- 'Meeting "{room}" on {room_recording_date} at {room_recording_time}'- ),+ summary_title_template="Summary of {title}",)diff --git a/src/summary/summary/core/locales/fr.py b/src/summary/summary/core/locales/fr.pyindex aab29617..4eec89ab 100644--- a/src/summary/summary/core/locales/fr.py+++ b/src/summary/summary/core/locales/fr.py@@ -29,8 +29,5 @@ Quelques points que nous vous conseillons de vérifier :"\n\n*[Donnez nous votre avis sur cette transcription]({form_link})*\n"),hallucination_replacement_text="[Texte impossible à transcrire]",- document_default_title="Transcription",- document_title_template=(- 'Réunion "{room}" du {room_recording_date} à {room_recording_time}'- ),+ summary_title_template="Résumé de {title}",)diff --git a/src/summary/summary/core/locales/nl.py b/src/summary/summary/core/locales/nl.pyindex 3da0de7e..ac2cfab9 100644--- a/src/summary/summary/core/locales/nl.py+++ b/src/summary/summary/core/locales/nl.py@@ -30,8 +30,5 @@ Een paar punten die wij u aanraden te controleren:"(externe link)]({form_link})*\n"),hallucination_replacement_text="[Tekst kon niet worden getranscribeerd]",- document_default_title="Transcriptie",- document_title_template=(- 'Vergadering "{room}" op {room_recording_date} om {room_recording_time}'- ),+ summary_title_template="Samenvatting van {title}",)diff --git a/src/summary/summary/core/locales/strings.py b/src/summary/summary/core/locales/strings.pyindex 77a20979..5ae27cc6 100644--- a/src/summary/summary/core/locales/strings.py+++ b/src/summary/summary/core/locales/strings.py@@ -12,5 +12,4 @@ class LocaleStrings:download_header_template: strform_footer_template: strhallucination_replacement_text: str- document_default_title: str- document_title_template: str+ summary_title_template: strdiff --git a/src/summary/summary/core/models.py b/src/summary/summary/core/models.pyindex 11983d15..e91f100b 100644--- a/src/summary/summary/core/models.py+++ b/src/summary/summary/core/models.py@@ -1,6 +1,8 @@"""Models for the API & Celery tasks creation."""-from pydantic import BaseModel, Field, field_validator+from datetime import datetime++from pydantic import AwareDatetime, BaseModel, EmailStr, Field, field_validatorfrom summary.core.config import get_settingsfrom summary.core.types import Url@@ -12,10 +14,60 @@ class SharedV2TaskCreation(BaseModel):"""Model that holds basic information for task creation."""user_sub: str = Field(title="User Sub", description="The user's sub.")+ user_email: EmailStr | None = Field(+ default=None,+ title="User Email",+ description="The user's email for analytics purposes.",+ )+++class RecordingMetadata(BaseModel):+ """Model for recording metadata."""++ cloud_storage_url: Url = Field(+ title="Cloud Storage URL",+ description="The URL of the metadata file for speaker assignement.",+ )+ started_at: AwareDatetime = Field(title="Start time of the recording to transcribe")+ ended_at: AwareDatetime = Field(title="End time of the recording to transcribe")+++class PushToDocsBaseConfig(BaseModel):+ """Model containing information for pushing transcript and summaries to docs."""++ user_email: EmailStr = Field(+ title="User Email", description="The user's email, future owner of the docs."+ )+ title: str = Field(title="Title", description="The title for the created document.")+++class PushToDocsTranscriptConfig(PushToDocsBaseConfig):+ """Model for push to docs information for transcripts."""++ download_link: Url | None = Field(+ default=None,+ title="Download Link",+ description="The link to download the recording.",+ )+ form_link: Url | None = Field(+ default=None,+ title="Form Link",+ description="The link to fill out a form for the recording.",+ )+ auto_create_summary: bool = Field(+ title="Auto Create Summary Docs",+ description="Whether to automatically create a summary "+ "for the transcription task and push it to docs.",+ default=False,+ )++class PushToDocsSummaryConfig(PushToDocsBaseConfig):+ """Model for push to docs information for summaries."""-class TranscribeTaskV2Request(SharedV2TaskCreation):- """Model for creating a transcribe and summarize task (used for API request)."""++class TranscribeTaskApiRequest(SharedV2TaskCreation):+ """Model for creating a transcribe task (used for API request)."""cloud_storage_url: Url = Field(title="Cloud storage URL",@@ -27,7 +79,17 @@ class TranscribeTaskV2Request(SharedV2TaskCreation):description="The language of the context text.",)language: str = Field(- title="Language", description="The language of the content to summarize."+ title="Language", description="The language of the content to transcribe."+ )+ metadata: RecordingMetadata | None = Field(+ title="Metadata",+ description="The metadata for the transcribe task.",+ default=None,+ )+ push_to_docs_config: PushToDocsTranscriptConfig | None = Field(+ title="Push to Docs info",+ description="If set, configuration for pushing to docs",+ default=None,)@field_validator("language")@@ -42,19 +104,30 @@ class TranscribeTaskV2Request(SharedV2TaskCreation):return v-class TranscribeTaskV2Payload(TranscribeTaskV2Request):- """Model for creating a transcribe and summarize task (used for actual task creation).""" # noqa: E501+class TranscribeTaskJob(TranscribeTaskApiRequest):+ """Model for creating a transcribe task (used for actual task creation)."""tenant_id: str = Field(title="Tenant ID", description="The ID of the tenant.")+ received_at: datetime = Field(+ title="Received At", description="The time the task was received."+ )-class SummarizeTaskV2Request(SharedV2TaskCreation):+class SummarizeTaskApiRequest(SharedV2TaskCreation):"""Model for creating a summarize task (used for API request)."""content: str = Field(title="Content", description="The content to summarize.")-class SummarizeTaskV2Payload(SummarizeTaskV2Request):+class SummarizeTaskJob(SummarizeTaskApiRequest):"""Model for creating a summarize task (used for actual task creation)."""tenant_id: str = Field(title="Tenant ID", description="The ID of the tenant.")+ received_at: datetime = Field(+ title="Received At", description="The time the task was received."+ )+ push_to_docs_config: PushToDocsBaseConfig | None = Field(+ title="Push to Docs info",+ description="If set, configuration for pushing to docs",+ default=None,+ )diff --git a/src/summary/summary/core/shared_models.py b/src/summary/summary/core/shared_models.pyindex b18c81cc..2e80896c 100644--- a/src/summary/summary/core/shared_models.py+++ b/src/summary/summary/core/shared_models.py@@ -167,4 +167,5 @@ __all__ = ["SummarizeWebhookPayloads","WebhookPayloads","WhisperXResponse",+ "webhook_payload_adapter",]diff --git a/src/summary/summary/core/transcript_formatter.py b/src/summary/summary/core/transcript_formatter.pyindex be1ff53e..b9d3e541 100644--- a/src/summary/summary/core/transcript_formatter.py+++ b/src/summary/summary/core/transcript_formatter.py@@ -1,9 +1,6 @@"""Transcript formatting into readable conversation format with speaker labels."""import logging-from datetime import datetime-from typing import Tuple-from zoneinfo import ZoneInfofrom summary.core.config import get_settingsfrom summary.core.locales import LocaleStrings@@ -38,16 +35,13 @@ class TranscriptFormatter:return None- def format( # noqa: PLR0913+ def format(self,transcription,- room: str | None = None,- recording_datetime: str | None = None,- owner_timezone: str | None = None,download_link: str | None = None,form_link: str | None = None,- ) -> Tuple[str, str]:- """Format transcription into the final document and its title."""+ ) -> str:+ """Format transcription into the final document."""segments = self._get_segments(transcription)if not segments:@@ -59,9 +53,7 @@ class TranscriptFormatter:if form_link:content = self._add_footer(content, form_link)- title = self._generate_title(room, recording_datetime, owner_timezone)-- return content, title+ return contentdef _remove_hallucinations(self, content: str) -> str:"""Remove hallucination patterns from content."""@@ -104,23 +96,3 @@ class TranscriptFormatter:footer = self._locale.form_footer_template.format(form_link=form_link)return content + footer-- def _generate_title(- self,- room: str | None = None,- recording_datetime: str | None = None,- owner_timezone: str | None = None,- ) -> str:- """Generate title from context or return default."""- if not room or not recording_datetime:- return self._locale.document_default_title-- dt = datetime.fromisoformat(recording_datetime)- if owner_timezone:- dt = dt.astimezone(ZoneInfo(owner_timezone))-- return self._locale.document_title_template.format(- room=room,- room_recording_date=dt.strftime("%Y-%m-%d"),- room_recording_time=dt.strftime("%H:%M"),- )diff --git a/src/summary/summary/core/webhook_service.py b/src/summary/summary/core/webhook_service.pyindex 0e196cb7..2c132e01 100644--- a/src/summary/summary/core/webhook_service.py+++ b/src/summary/summary/core/webhook_service.py@@ -4,9 +4,6 @@ import jsonimport loggingimport requests-from requests import Session-from requests.adapters import HTTPAdapter-from urllib3.util import Retryfrom summary.core.config import get_settingsfrom summary.core.shared_models import (@@ -18,83 +15,6 @@ settings = get_settings()logger = logging.getLogger(__name__)-def _create_retry_session(api_key: str | None = None):- """Create an HTTP session configured with retry logic."""- session = Session()- retries = Retry(- total=settings.webhook_max_retries,- backoff_factor=settings.webhook_backoff_factor,- status_forcelist=settings.webhook_status_forcelist,- allowed_methods={"POST"},- )- session.mount("https://", HTTPAdapter(max_retries=retries))- if api_key:- session.headers.update({"Authorization": f"Bearer {api_key}"})-- return session---def _post_with_retries(*, url, data, api_key: str | None = None):- """Send POST request with automatic retries."""- session = _create_retry_session(api_key=api_key)-- try:- response = session.post(url, json=data)- response.raise_for_status()- return response- finally:- session.close()---def call_webhook_v1(*, tenant_id: str, payload: dict) -> None:- """Call webhook with payload a payload and optional token."""- tenant = settings.get_authorized_tenant(tenant_id=tenant_id)-- logger.debug("Submitting to %s", tenant.webhook_url)- logger.debug("Request payload: %s", json.dumps(payload, indent=2))-- response = _post_with_retries(- url=tenant.webhook_url,- api_key=tenant.webhook_api_key.get_secret_value(),- data=payload,- )-- try:- response_data = response.json()- document_id = response_data.get("id", "N/A")- except (json.JSONDecodeError, AttributeError):- document_id = "Unable to parse response"- response_data = response.text-- logger.info(- "Delivery success | Document %s submitted (HTTP %s)",- document_id,- response.status_code,- )- logger.debug("Full response: %s", response_data)---def submit_content(content: str, title: str, email: str, sub: str) -> None:- """Submit content to the configured webhook destination.-- Builds the payload, sends it with retries, and logs the outcome.-- Notes:- Deprecated: Use call_webhook_v2 directly instead.-- Deprecated:- This will route content to the v1 default tenant- """- data = {- "title": title,- "content": content,- "email": email,- "sub": sub,- }-- call_webhook_v1(payload=data, tenant_id=settings.v1_tenant_id)--def call_webhook_v2(*,tenant_id: str,@@ -114,6 +34,7 @@ def call_webhook_v2(json=payload.model_dump(),headers={"Authorization": f"Bearer {tenant.webhook_api_key.get_secret_value()}",+ "User-Agent": settings.app_external_user_agent,},timeout=(10, 20),)diff --git a/src/summary/tests/api/test_api_tasks_v2.py b/src/summary/tests/api/test_api_tasks_v2.pyindex 0fac599e..75837bf8 100644--- a/src/summary/tests/api/test_api_tasks_v2.py+++ b/src/summary/tests/api/test_api_tasks_v2.py@@ -1,5 +1,6 @@"""Integration tests for the V2 task API endpoints."""+from datetime import datetimefrom unittest.mock import ANY, MagicMock, patch@@ -31,6 +32,8 @@ class TestTasksV2:}args = mock_apply_async.call_args.kwargs["args"]+ received_at = args[0].pop("received_at")+ assert isinstance(received_at, datetime)assert args == [{"user_sub": "remote-001",@@ -38,6 +41,9 @@ class TestTasksV2:"language": "en","context_language": "fr","tenant_id": "test-tenant",+ "user_email": None,+ "metadata": None,+ "push_to_docs_config": None,}]@@ -64,11 +70,14 @@ class TestTasksV2:}args = mock_apply_async.call_args.kwargs["args"]+ received_at = args[0].pop("received_at")+ assert isinstance(received_at, datetime)assert args == [{"user_sub": "remote-002","content": "This is a long meeting transcript to summarize.","tenant_id": "test-tenant",+ "user_email": None,}]