AI/dify
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/langgenius/dify.git synced 2025-08-14 15:45:53 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'main' into feat/mcp

Browse Source
This commit is contained in:
Novice 2025-05-28 09:37:55 +08:00
commit 1a7242abd4
799 changed files with 22592 additions and 6640 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
.devcontainer/Dockerfile
View File

@ -1,5 +1,4 @@
FROM mcr.microsoft.com/devcontainers/python:3.12
# [Optional] Uncomment this section to install additional OS packages.
# RUN apt-get update && export DEBIAN_FRONTEND=noninteractive \
# && apt-get -y install --no-install-recommends <your-package-list-here>
RUN apt-get update && export DEBIAN_FRONTEND=noninteractive \
&& apt-get -y install libgmp-dev libmpfr-dev libmpc-dev

1
.devcontainer/post_create_command.sh
View File

@ -7,6 +7,7 @@ pipx install uv
echo 'alias start-api="cd /workspaces/dify/api && uv run python -m flask run --host 0.0.0.0 --port=5001 --debug"' >> ~/.bashrc
echo 'alias start-worker="cd /workspaces/dify/api && uv run python -m celery -A app.celery worker -P gevent -c 1 --loglevel INFO -Q dataset,generation,mail,ops_trace,app_deletion"' >> ~/.bashrc
echo 'alias start-web="cd /workspaces/dify/web && pnpm dev"' >> ~/.bashrc
echo 'alias start-web-prod="cd /workspaces/dify/web && pnpm build && pnpm start"' >> ~/.bashrc
echo 'alias start-containers="cd /workspaces/dify/docker && docker-compose -f docker-compose.middleware.yaml -p dify --env-file middleware.env up -d"' >> ~/.bashrc
echo 'alias stop-containers="cd /workspaces/dify/docker && docker-compose -f docker-compose.middleware.yaml -p dify --env-file middleware.env down"' >> ~/.bashrc

1
.github/workflows/style.yml vendored
View File

@ -139,6 +139,7 @@ jobs:
- name: Checkout code
uses: actions/checkout@v4
with:
fetch-depth: 0
persist-credentials: false
- name: Check changed files

8
.github/workflows/translate-i18n-base-on-english.yml vendored
View File

@ -31,11 +31,19 @@ jobs:
echo "FILES_CHANGED=false" >> $GITHUB_ENV
fi
- name: Install pnpm
uses: pnpm/action-setup@v4
with:
version: 10
run_install: false
- name: Set up Node.js
if: env.FILES_CHANGED == 'true'
uses: actions/setup-node@v4
with:
node-version: 'lts/*'
cache: pnpm
cache-dependency-path: ./web/package.json
- name: Install dependencies
if: env.FILES_CHANGED == 'true'

6
README.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
📌 <a href="https://dify.ai/blog/introducing-dify-workflow-file-upload-a-demo-on-ai-podcast">Introducing Dify Workflow File Upload: Recreate Google NotebookLM Podcast</a>
@ -87,8 +87,6 @@ Please refer to our [FAQ](https://docs.dify.ai/getting-started/install-self-host
**1. Workflow**:
Build and test powerful AI workflows on a visual canvas, leveraging all the following features and beyond.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Comprehensive model support**:
Seamless integration with hundreds of proprietary / open-source LLMs from dozens of inference providers and self-hosted solutions, covering GPT, Mistral, Llama3, and any OpenAI API-compatible models. A full list of supported model providers can be found [here](https://docs.dify.ai/getting-started/readme/model-providers).
@ -237,7 +235,7 @@ At the same time, please consider supporting Dify by sharing it on social media
## Community & contact
- [Github Discussion](https://github.com/langgenius/dify/discussions). Best for: sharing feedback and asking questions.
- [GitHub Discussion](https://github.com/langgenius/dify/discussions). Best for: sharing feedback and asking questions.
- [GitHub Issues](https://github.com/langgenius/dify/issues). Best for: bugs you encounter using Dify.AI, and feature proposals. See our [Contribution Guide](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md).
- [Discord](https://discord.gg/FngNHpbcY7). Best for: sharing your applications and hanging out with the community.
- [X(Twitter)](https://twitter.com/dify_ai). Best for: sharing your applications and hanging out with the community.

6
README_AR.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Cloud</a> ·
@ -54,8 +54,6 @@
**1. سير العمل**: قم ببناء واختبار سير عمل الذكاء الاصطناعي القوي على قماش بصري، مستفيدًا من جميع الميزات التالية وأكثر.
<https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa>
**2. الدعم الشامل للنماذج**: تكامل سلس مع مئات من LLMs الخاصة / مفتوحة المصدر من عشرات من موفري التحليل والحلول المستضافة ذاتيًا، مما يغطي GPT و Mistral و Llama3 وأي نماذج متوافقة مع واجهة OpenAI API. يمكن العثور على قائمة كاملة بمزودي النموذج المدعومين [هنا](https://docs.dify.ai/getting-started/readme/model-providers).
![providers-v5](https://github.com/langgenius/dify/assets/13230914/5a17bdbe-097a-4100-8363-40255b70f6e3)
@ -225,7 +223,7 @@ docker compose up -d
</a>
## المجتمع والاتصال
- [مناقشة Github](https://github.com/langgenius/dify/discussions). الأفضل لـ: مشاركة التعليقات وطرح الأسئلة.
- [مناقشة GitHub](https://github.com/langgenius/dify/discussions). الأفضل لـ: مشاركة التعليقات وطرح الأسئلة.
- [المشكلات على GitHub](https://github.com/langgenius/dify/issues). الأفضل لـ: الأخطاء التي تواجهها في استخدام Dify.AI، واقتراحات الميزات. انظر [دليل المساهمة](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md).
- [Discord](https://discord.gg/FngNHpbcY7). الأفضل لـ: مشاركة تطبيقاتك والترفيه مع المجتمع.
- [تويتر](https://twitter.com/dify_ai). الأفضل لـ: مشاركة تطبيقاتك والترفيه مع المجتمع.

6
README_BN.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
📌 <a href="https://dify.ai/blog/introducing-dify-workflow-file-upload-a-demo-on-ai-podcast">ডিফাই ওয়ার্কফ্লো ফাইল আপলোড পরিচিতি: গুগল নোটবুক-এলএম পডকাস্ট পুনর্নির্মাণ</a>
@ -84,8 +84,6 @@ docker compose up -d
**১. ওয়ার্কফ্লো**:
ভিজ্যুয়াল ক্যানভাসে AI ওয়ার্কফ্লো তৈরি এবং পরীক্ষা করুন, নিম্নলিখিত সব ফিচার এবং তার বাইরেও আরও অনেক কিছু ব্যবহার করে।
<https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa>
**২. মডেল সাপোর্ট**:
GPT, Mistral, Llama3, এবং যেকোনো OpenAI API-সামঞ্জস্যপূর্ণ মডেলসহ, কয়েক ডজন ইনফারেন্স প্রদানকারী এবং সেল্ফ-হোস্টেড সমাধান থেকে শুরু করে প্রোপ্রাইটরি/ওপেন-সোর্স LLM-এর সাথে সহজে ইন্টিগ্রেশন। সমর্থিত মডেল প্রদানকারীদের একটি সম্পূর্ণ তালিকা পাওয়া যাবে [এখানে](https://docs.dify.ai/getting-started/readme/model-providers)।
@ -236,7 +234,7 @@ GitHub-এ ডিফাইকে স্টার দিয়ে রাখুন
## কমিউনিটি এবং যোগাযোগ
- [Github Discussion](https://github.com/langgenius/dify/discussions) ফিডব্যাক এবং প্রতিক্রিয়া জানানোর মাধ্যম।
- [GitHub Discussion](https://github.com/langgenius/dify/discussions) ফিডব্যাক এবং প্রতিক্রিয়া জানানোর মাধ্যম।
- [GitHub Issues](https://github.com/langgenius/dify/issues). Dify.AI ব্যবহার করে আপনি যেসব বাগের সম্মুখীন হন এবং ফিচার প্রস্তাবনা। আমাদের [অবদান নির্দেশিকা](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md) দেখুন।
- [Discord](https://discord.gg/FngNHpbcY7) আপনার এপ্লিকেশন শেয়ার এবং কমিউনিটি আড্ডার মাধ্যম।
- [X(Twitter)](https://twitter.com/dify_ai) আপনার এপ্লিকেশন শেয়ার এবং কমিউনিটি আড্ডার মাধ্যম।

9
README_CN.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<div align="center">
<a href="https://cloud.dify.ai">Dify 云服务</a> ·
@ -61,11 +61,6 @@ Dify 是一个开源的 LLM 应用开发平台。其直观的界面结合了 AI
**1. 工作流**:
在画布上构建和测试功能强大的 AI 工作流程，利用以下所有功能以及更多功能。
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. 全面的模型支持**:
与数百种专有/开源 LLMs 以及数十种推理提供商和自托管解决方案无缝集成，涵盖 GPT、Mistral、Llama3 以及任何与 OpenAI API 兼容的模型。完整的支持模型提供商列表可在[此处](https://docs.dify.ai/getting-started/readme/model-providers)找到。
@ -248,7 +243,7 @@ docker compose up -d
我们欢迎您为 Dify 做出贡献，以帮助改善 Dify。包括：提交代码、问题、新想法，或分享您基于 Dify 创建的有趣且有用的 AI 应用程序。同时，我们也欢迎您在不同的活动、会议和社交媒体上分享 Dify。
- [Github Discussion](https://github.com/langgenius/dify/discussions). 👉：分享您的应用程序并与社区交流。
- [GitHub Discussion](https://github.com/langgenius/dify/discussions). 👉：分享您的应用程序并与社区交流。
- [GitHub Issues](https://github.com/langgenius/dify/issues)。👉：使用 Dify.AI 时遇到的错误和问题，请参阅[贡献指南](CONTRIBUTING.md)。
- [电子邮件支持](mailto:hello@dify.ai?subject=[GitHub]Questions%20About%20Dify)。👉：关于使用 Dify.AI 的问题。
- [Discord](https://discord.gg/FngNHpbcY7)。👉：分享您的应用程序并与社区交流。

9
README_DE.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
📌 <a href="https://dify.ai/blog/introducing-dify-workflow-file-upload-a-demo-on-ai-podcast">Einführung in Dify Workflow File Upload: Google NotebookLM Podcast nachbilden</a>
@ -83,11 +83,6 @@ Bitte beachten Sie unsere [FAQ](https://docs.dify.ai/getting-started/install-sel
**1. Workflow**:
Erstellen und testen Sie leistungsstarke KI-Workflows auf einer visuellen Oberfläche, wobei Sie alle der folgenden Funktionen und darüber hinaus nutzen können.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Umfassende Modellunterstützung**:
Nahtlose Integration mit Hunderten von proprietären und Open-Source-LLMs von Dutzenden Inferenzanbietern und selbstgehosteten Lösungen, die GPT, Mistral, Llama3 und alle mit der OpenAI API kompatiblen Modelle abdecken. Eine vollständige Liste der unterstützten Modellanbieter finden Sie [hier](https://docs.dify.ai/getting-started/readme/model-providers).
@ -235,7 +230,7 @@ Falls Sie Code beitragen möchten, lesen Sie bitte unseren [Contribution Guide](
## Gemeinschaft & Kontakt
* [Github Discussion](https://github.com/langgenius/dify/discussions). Am besten geeignet für: den Austausch von Feedback und das Stellen von Fragen.
* [GitHub Discussion](https://github.com/langgenius/dify/discussions). Am besten geeignet für: den Austausch von Feedback und das Stellen von Fragen.
* [GitHub Issues](https://github.com/langgenius/dify/issues). Am besten für: Fehler, auf die Sie bei der Verwendung von Dify.AI stoßen, und Funktionsvorschläge. Siehe unseren [Contribution Guide](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md).
* [Discord](https://discord.gg/FngNHpbcY7). Am besten geeignet für: den Austausch von Bewerbungen und den Austausch mit der Community.
* [X(Twitter)](https://twitter.com/dify_ai). Am besten geeignet für: den Austausch von Bewerbungen und den Austausch mit der Community.

7
README_ES.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Cloud</a> ·
@ -59,11 +59,6 @@ Dify es una plataforma de desarrollo de aplicaciones de LLM de código abierto.
**1. Flujo de trabajo**:
Construye y prueba potentes flujos de trabajo de IA en un lienzo visual, aprovechando todas las siguientes características y más.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Soporte de modelos completo**:
Integración perfecta con cientos de LLMs propietarios / de código abierto de docenas de proveedores de inferencia y soluciones auto-alojadas, que cubren GPT, Mistral, Llama3 y cualquier modelo compatible con la API de OpenAI. Se puede encontrar una lista completa de proveedores de modelos admitidos [aquí](https://docs.dify.ai/getting-started/readme/model-providers).

7
README_FR.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Cloud</a> ·
@ -59,11 +59,6 @@ Dify est une plateforme de développement d'applications LLM open source. Son in
**1. Flux de travail** :
Construisez et testez des flux de travail d'IA puissants sur un canevas visuel, en utilisant toutes les fonctionnalités suivantes et plus encore.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Prise en charge complète des modèles** :
Intégration transparente avec des centaines de LLM propriétaires / open source provenant de dizaines de fournisseurs d'inférence et de solutions auto-hébergées, couvrant GPT, Mistral, Llama3, et tous les modèles compatibles avec l'API OpenAI. Une liste complète des fournisseurs de modèles pris en charge se trouve [ici](https://docs.dify.ai/getting-started/readme/model-providers).

9
README_JA.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Cloud</a> ·
@ -60,11 +60,6 @@ DifyはオープンソースのLLMアプリケーション開発プラットフ
**1. ワークフロー**:
強力なAIワークフローをビジュアルキャンバス上で構築し、テストできます。すべての機能、および以下の機能を使用できます。
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. 総合的なモデルサポート**:
数百ものプロプライエタリ/オープンソースのLLMと、数十もの推論プロバイダーおよびセルフホスティングソリューションとのシームレスな統合を提供します。GPT、Mistral、Llama3、OpenAI APIと互換性のあるすべてのモデルを統合されています。サポートされているモデルプロバイダーの完全なリストは[こちら](https://docs.dify.ai/getting-started/readme/model-providers)をご覧ください。
@ -241,7 +236,7 @@ docker compose up -d
## コミュニティ & お問い合わせ
* [Github Discussion](https://github.com/langgenius/dify/discussions). 主に: フィードバックの共有や質問。
* [GitHub Discussion](https://github.com/langgenius/dify/discussions). 主に: フィードバックの共有や質問。
* [GitHub Issues](https://github.com/langgenius/dify/issues). 主に: Dify.AIを使用する際に発生するエラーや問題については、[貢献ガイド](CONTRIBUTING_JA.md)を参照してください
* [Discord](https://discord.gg/FngNHpbcY7). 主に: アプリケーションの共有やコミュニティとの交流。
* [X(Twitter)](https://twitter.com/dify_ai). 主に: アプリケーションの共有やコミュニティとの交流。

9
README_KL.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Cloud</a> ·
@ -59,11 +59,6 @@ Dify is an open-source LLM app development platform. Its intuitive interface com
**1. Workflow**:
Build and test powerful AI workflows on a visual canvas, leveraging all the following features and beyond.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Comprehensive model support**:
Seamless integration with hundreds of proprietary / open-source LLMs from dozens of inference providers and self-hosted solutions, covering GPT, Mistral, Llama3, and any OpenAI API-compatible models. A full list of supported model providers can be found [here](https://docs.dify.ai/getting-started/readme/model-providers).
@ -240,7 +235,7 @@ At the same time, please consider supporting Dify by sharing it on social media
## Community & Contact
* [Github Discussion](https://github.com/langgenius/dify/discussions
* [GitHub Discussion](https://github.com/langgenius/dify/discussions
). Best for: sharing feedback and asking questions.
* [GitHub Issues](https://github.com/langgenius/dify/issues). Best for: bugs you encounter using Dify.AI, and feature proposals. See our [Contribution Guide](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md).

9
README_KR.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify 클라우드</a> ·
@ -54,11 +54,6 @@
**1. 워크플로우**:
다음 기능들을 비롯한 다양한 기능을 활용하여 시각적 캔버스에서 강력한 AI 워크플로우를 구축하고 테스트하세요.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. 포괄적인 모델 지원:**:
수십 개의 추론 제공업체와 자체 호스팅 솔루션에서 제공하는 수백 개의 독점 및 오픈 소스 LLM과 원활하게 통합되며, GPT, Mistral, Llama3 및 모든 OpenAI API 호환 모델을 포함합니다. 지원되는 모델 제공업체의 전체 목록은 [여기](https://docs.dify.ai/getting-started/readme/model-providers)에서 확인할 수 있습니다.
@ -234,7 +229,7 @@ Dify를 Kubernetes에 배포하고 프리미엄 스케일링 설정을 구성했
## 커뮤니티 & 연락처
* [Github 토론](https://github.com/langgenius/dify/discussions). 피드백 공유 및 질문하기에 적합합니다.
* [GitHub 토론](https://github.com/langgenius/dify/discussions). 피드백 공유 및 질문하기에 적합합니다.
* [GitHub 이슈](https://github.com/langgenius/dify/issues). Dify.AI 사용 중 발견한 버그와 기능 제안에 적합합니다. [기여 가이드](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md)를 참조하세요.
* [디스코드](https://discord.gg/FngNHpbcY7). 애플리케이션 공유 및 커뮤니티와 소통하기에 적합합니다.
* [트위터](https://twitter.com/dify_ai). 애플리케이션 공유 및 커뮤니티와 소통하기에 적합합니다.

8
README_PT.md
View File

@ -1,5 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
📌 <a href="https://dify.ai/blog/introducing-dify-workflow-file-upload-a-demo-on-ai-podcast">Introduzindo o Dify Workflow com Upload de Arquivo: Recrie o Podcast Google NotebookLM</a>
</p>
@ -59,11 +58,6 @@ Dify é uma plataforma de desenvolvimento de aplicativos LLM de código aberto.
**1. Workflow**:
Construa e teste workflows poderosos de IA em uma interface visual, aproveitando todos os recursos a seguir e muito mais.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Suporte abrangente a modelos**:
Integração perfeita com centenas de LLMs proprietários e de código aberto de diversas provedoras e soluções auto-hospedadas, abrangendo GPT, Mistral, Llama3 e qualquer modelo compatível com a API da OpenAI. A lista completa de provedores suportados pode ser encontrada [aqui](https://docs.dify.ai/getting-started/readme/model-providers).

9
README_SI.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
📌 <a href="https://dify.ai/blog/introducing-dify-workflow-file-upload-a-demo-on-ai-podcast">Predstavljamo nalaganje datotek Dify Workflow: znova ustvarite Google NotebookLM Podcast</a>
@ -81,11 +81,6 @@ Prosimo, glejte naša pogosta vprašanja [FAQ](https://docs.dify.ai/getting-star
**1. Potek dela**:
Zgradite in preizkusite zmogljive poteke dela AI na vizualnem platnu, pri čemer izkoristite vse naslednje funkcije in več.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Celovita podpora za modele**:
Brezhibna integracija s stotinami lastniških/odprtokodnih LLM-jev ducatov ponudnikov sklepanja in samostojnih rešitev, ki pokrivajo GPT, Mistral, Llama3 in vse modele, združljive z API-jem OpenAI. Celoten seznam podprtih ponudnikov modelov najdete [tukaj](https://docs.dify.ai/getting-started/readme/model-providers).
@ -234,7 +229,7 @@ Za tiste, ki bi radi prispevali kodo, si oglejte naš vodnik za prispevke . Hkra
## Skupnost in stik
* [Github Discussion](https://github.com/langgenius/dify/discussions). Najboljše za: izmenjavo povratnih informacij in postavljanje vprašanj.
* [GitHub Discussion](https://github.com/langgenius/dify/discussions). Najboljše za: izmenjavo povratnih informacij in postavljanje vprašanj.
* [GitHub Issues](https://github.com/langgenius/dify/issues). Najboljše za: hrošče, na katere naletite pri uporabi Dify.AI, in predloge funkcij. Oglejte si naš [vodnik za prispevke](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md).
* [Discord](https://discord.gg/FngNHpbcY7). Najboljše za: deljenje vaših aplikacij in druženje s skupnostjo.
* [X(Twitter)](https://twitter.com/dify_ai). Najboljše za: deljenje vaših aplikacij in druženje s skupnostjo.

9
README_TR.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Bulut</a> ·
@ -55,11 +55,6 @@ Dify, açık kaynaklı bir LLM uygulama geliştirme platformudur. Sezgisel aray
**1. Workflow**:
Görsel bir arayüz üzerinde güçlü AI iş akışları oluşturun ve test edin, aşağıdaki tüm özellikleri ve daha fazlasını kullanarak.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Kapsamlı model desteği**:
Çok sayıda çıkarım sağlayıcısı ve kendi kendine barındırılan çözümlerden yüzlerce özel / açık kaynaklı LLM ile sorunsuz entegrasyon sağlar. GPT, Mistral, Llama3 ve OpenAI API uyumlu tüm modelleri kapsar. Desteklenen model sağlayıcılarının tam listesine [buradan](https://docs.dify.ai/getting-started/readme/model-providers) ulaşabilirsiniz.
@ -232,7 +227,7 @@ Aynı zamanda, lütfen Dify'ı sosyal medyada, etkinliklerde ve konferanslarda p
## Topluluk & iletişim
* [Github Tartışmaları](https://github.com/langgenius/dify/discussions). En uygun: geri bildirim paylaşmak ve soru sormak için.
* [GitHub Tartışmaları](https://github.com/langgenius/dify/discussions). En uygun: geri bildirim paylaşmak ve soru sormak için.
* [GitHub Sorunları](https://github.com/langgenius/dify/issues). En uygun: Dify.AI kullanırken karşılaştığınız hatalar ve özellik önerileri için. [Katkı Kılavuzumuza](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md) bakın.
* [Discord](https://discord.gg/FngNHpbcY7). En uygun: uygulamalarınızı paylaşmak ve toplulukla vakit geçirmek için.
* [X(Twitter)](https://twitter.com/dify_ai). En uygun: uygulamalarınızı paylaşmak ve toplulukla vakit geçirmek için.

6
README_TW.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
📌 <a href="https://dify.ai/blog/introducing-dify-workflow-file-upload-a-demo-on-ai-podcast">介紹 Dify 工作流程檔案上傳功能：重現 Google NotebookLM Podcast</a>
@ -86,8 +86,6 @@ docker compose up -d
**1. 工作流程**：
在視覺化畫布上建立和測試強大的 AI 工作流程，利用以下所有功能及更多。
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. 全面的模型支援**：
無縫整合來自數十個推理提供商和自託管解決方案的數百個專有/開源 LLM，涵蓋 GPT、Mistral、Llama3 和任何與 OpenAI API 兼容的模型。您可以在[此處](https://docs.dify.ai/getting-started/readme/model-providers)找到支援的模型提供商完整列表。
@ -235,7 +233,7 @@ Dify 的所有功能都提供相應的 API，因此您可以輕鬆地將 Dify
## 社群與聯絡方式
- [Github Discussion](https://github.com/langgenius/dify/discussions)：最適合分享反饋和提問。
- [GitHub Discussion](https://github.com/langgenius/dify/discussions)：最適合分享反饋和提問。
- [GitHub Issues](https://github.com/langgenius/dify/issues)：最適合報告使用 Dify.AI 時遇到的問題和提出功能建議。請參閱我們的[貢獻指南](https://github.com/langgenius/dify/blob/main/CONTRIBUTING.md)。
- [Discord](https://discord.gg/FngNHpbcY7)：最適合分享您的應用程式並與社群互動。
- [X(Twitter)](https://twitter.com/dify_ai)：最適合分享您的應用程式並與社群互動。

7
README_VI.md
View File

@ -1,4 +1,4 @@
![cover-v5-optimized](https://github.com/langgenius/dify/assets/13230914/f9e19af5-61ba-4119-b926-d10c4c06ebab)
![cover-v5-optimized](./images/GitHub_README_if.png)
<p align="center">
<a href="https://cloud.dify.ai">Dify Cloud</a> ·
@ -55,11 +55,6 @@ Dify là một nền tảng phát triển ứng dụng LLM mã nguồn mở. Gia
**1. Quy trình làm việc**:
Xây dựng và kiểm tra các quy trình làm việc AI mạnh mẽ trên một canvas trực quan, tận dụng tất cả các tính năng sau đây và hơn thế nữa.
https://github.com/langgenius/dify/assets/13230914/356df23e-1604-483d-80a6-9517ece318aa
**2. Hỗ trợ mô hình toàn diện**:
Tích hợp liền mạch với hàng trăm mô hình LLM độc quyền / mã nguồn mở từ hàng chục nhà cung cấp suy luận và giải pháp tự lưu trữ, bao gồm GPT, Mistral, Llama3, và bất kỳ mô hình tương thích API OpenAI nào. Danh sách đầy đủ các nhà cung cấp mô hình được hỗ trợ có thể được tìm thấy [tại đây](https://docs.dify.ai/getting-started/readme/model-providers).

4
api/.env.example
View File

@ -152,6 +152,7 @@ QDRANT_API_KEY=difyai123456
QDRANT_CLIENT_TIMEOUT=20
QDRANT_GRPC_ENABLED=false
QDRANT_GRPC_PORT=6334
QDRANT_REPLICATION_FACTOR=1
#Couchbase configuration
COUCHBASE_CONNECTION_STRING=127.0.0.1
@ -269,6 +270,7 @@ OPENSEARCH_PORT=9200
OPENSEARCH_USER=admin
OPENSEARCH_PASSWORD=admin
OPENSEARCH_SECURE=true
OPENSEARCH_VERIFY_CERTS=true
# Baidu configuration
BAIDU_VECTOR_DB_ENDPOINT=http://127.0.0.1:5287
@ -348,6 +350,7 @@ SENTRY_DSN=
# DEBUG
DEBUG=false
ENABLE_REQUEST_LOGGING=False
SQLALCHEMY_ECHO=false
# Notion import configuration, support public and internal
@ -476,6 +479,7 @@ LOGIN_LOCKOUT_DURATION=86400
ENABLE_OTEL=false
OTLP_BASE_ENDPOINT=http://localhost:4318
OTLP_API_KEY=
OTEL_EXPORTER_OTLP_PROTOCOL=
OTEL_EXPORTER_TYPE=otlp
OTEL_SAMPLING_RATE=0.1
OTEL_BATCH_EXPORT_SCHEDULE_DELAY=5000

4
api/app_factory.py
View File

@ -54,7 +54,7 @@ def initialize_extensions(app: DifyApp):
ext_otel,
ext_proxy_fix,
ext_redis,
ext_repositories,
ext_request_logging,
ext_sentry,
ext_set_secretkey,
ext_storage,
@ -75,7 +75,6 @@ def initialize_extensions(app: DifyApp):
ext_migrate,
ext_redis,
ext_storage,
ext_repositories,
ext_celery,
ext_login,
ext_mail,
@ -85,6 +84,7 @@ def initialize_extensions(app: DifyApp):
ext_blueprints,
ext_commands,
ext_otel,
ext_request_logging,
]
for ext in extensions:
short_name = ext.__name__.split(".")[-1]

24
api/commands.py
View File

@ -6,6 +6,7 @@ from typing import Optional
import click
from flask import current_app
from sqlalchemy import select
from werkzeug.exceptions import NotFound
from configs import dify_config
@ -297,11 +298,11 @@ def migrate_knowledge_vector_database():
page = 1
while True:
try:
datasets = (
Dataset.query.filter(Dataset.indexing_technique == "high_quality")
.order_by(Dataset.created_at.desc())
.paginate(page=page, per_page=50)
stmt = (
select(Dataset).filter(Dataset.indexing_technique == "high_quality").order_by(Dataset.created_at.desc())
)
datasets = db.paginate(select=stmt, page=page, per_page=50, max_per_page=50, error_out=False)
except NotFound:
break
@ -551,11 +552,12 @@ def old_metadata_migration():
page = 1
while True:
try:
documents = (
DatasetDocument.query.filter(DatasetDocument.doc_metadata is not None)
stmt = (
select(DatasetDocument)
.filter(DatasetDocument.doc_metadata.is_not(None))
.order_by(DatasetDocument.created_at.desc())
.paginate(page=page, per_page=50)
)
documents = db.paginate(select=stmt, page=page, per_page=50, max_per_page=50, error_out=False)
except NotFound:
break
if not documents:
@ -592,11 +594,15 @@ def old_metadata_migration():
)
db.session.add(dataset_metadata_binding)
else:
dataset_metadata_binding = DatasetMetadataBinding.query.filter(
dataset_metadata_binding = (
db.session.query(DatasetMetadataBinding) # type: ignore
.filter(
DatasetMetadataBinding.dataset_id == document.dataset_id,
DatasetMetadataBinding.document_id == document.id,
DatasetMetadataBinding.metadata_id == dataset_metadata.id,
).first()
)
.first()
)
if not dataset_metadata_binding:
dataset_metadata_binding = DatasetMetadataBinding(
tenant_id=document.tenant_id,

6
api/configs/deploy/__init__.py
View File

@ -17,6 +17,12 @@ class DeploymentConfig(BaseSettings):
default=False,
)
# Request logging configuration
ENABLE_REQUEST_LOGGING: bool = Field(
description="Enable request and response body logging",
default=False,
)
EDITION: str = Field(
description="Deployment edition of the application (e.g., 'SELF_HOSTED', 'CLOUD')",
default="SELF_HOSTED",

6
api/configs/feature/__init__.py
View File

@ -74,7 +74,7 @@ class CodeExecutionSandboxConfig(BaseSettings):
CODE_EXECUTION_ENDPOINT: HttpUrl = Field(
description="URL endpoint for the code execution service",
default="http://sandbox:8194",
default=HttpUrl("http://sandbox:8194"),
)
CODE_EXECUTION_API_KEY: str = Field(
@ -145,7 +145,7 @@ class PluginConfig(BaseSettings):
PLUGIN_DAEMON_URL: HttpUrl = Field(
description="Plugin API URL",
default="http://localhost:5002",
default=HttpUrl("http://localhost:5002"),
)
PLUGIN_DAEMON_KEY: str = Field(
@ -188,7 +188,7 @@ class MarketplaceConfig(BaseSettings):
MARKETPLACE_API_URL: HttpUrl = Field(
description="Marketplace API URL",
default="https://marketplace.dify.ai",
default=HttpUrl("https://marketplace.dify.ai"),
)

22
api/configs/middleware/__init__.py
View File

@ -1,6 +1,6 @@
import os
from typing import Any, Literal, Optional
from urllib.parse import quote_plus
from urllib.parse import parse_qsl, quote_plus
from pydantic import Field, NonNegativeInt, PositiveFloat, PositiveInt, computed_field
from pydantic_settings import BaseSettings
@ -173,17 +173,31 @@ class DatabaseConfig(BaseSettings):
RETRIEVAL_SERVICE_EXECUTORS: NonNegativeInt = Field(
description="Number of processes for the retrieval service, default to CPU cores.",
default=os.cpu_count(),
default=os.cpu_count() or 1,
)
@computed_field
@computed_field # type: ignore[misc]
@property
def SQLALCHEMY_ENGINE_OPTIONS(self) -> dict[str, Any]:
# Parse DB_EXTRAS for 'options'
db_extras_dict = dict(parse_qsl(self.DB_EXTRAS))
options = db_extras_dict.get("options", "")
# Always include timezone
timezone_opt = "-c timezone=UTC"
if options:
# Merge user options and timezone
merged_options = f"{options} {timezone_opt}"
else:
merged_options = timezone_opt
connect_args = {"options": merged_options}
return {
"pool_size": self.SQLALCHEMY_POOL_SIZE,
"max_overflow": self.SQLALCHEMY_MAX_OVERFLOW,
"pool_recycle": self.SQLALCHEMY_POOL_RECYCLE,
"pool_pre_ping": self.SQLALCHEMY_POOL_PRE_PING,
"connect_args": {"options": "-c timezone=UTC"},
"connect_args": connect_args,
}

10
api/configs/middleware/cache/redis_config.py vendored
View File

@ -83,3 +83,13 @@ class RedisConfig(BaseSettings):
description="Password for Redis Clusters authentication (if required)",
default=None,
)
REDIS_SERIALIZATION_PROTOCOL: int = Field(
description="Redis serialization protocol (RESP) version",
default=3,
)
REDIS_ENABLE_CLIENT_SIDE_CACHE: bool = Field(
description="Enable client side cache in redis",
default=False,
)

4
api/configs/middleware/storage/amazon_s3_storage_config.py
View File

@ -1,4 +1,4 @@
from typing import Optional
from typing import Literal, Optional
from pydantic import Field
from pydantic_settings import BaseSettings
@ -34,7 +34,7 @@ class S3StorageConfig(BaseSettings):
default=None,
)
S3_ADDRESS_STYLE: str = Field(
S3_ADDRESS_STYLE: Literal["auto", "virtual", "path"] = Field(
description="S3 addressing style: 'auto', 'path', or 'virtual'",
default="auto",
)

5
api/configs/middleware/vdb/opensearch_config.py
View File

@ -33,6 +33,11 @@ class OpenSearchConfig(BaseSettings):
default=False,
)
OPENSEARCH_VERIFY_CERTS: bool = Field(
description="Whether to verify SSL certificates for HTTPS connections (recommended to set True in production)",
default=True,
)
OPENSEARCH_AUTH_METHOD: AuthMethod = Field(
description="Authentication method for OpenSearch connection (default is 'basic')",
default=AuthMethod.BASIC,

5
api/configs/middleware/vdb/qdrant_config.py
View File

@ -33,3 +33,8 @@ class QdrantConfig(BaseSettings):
description="Port number for gRPC connection to Qdrant server (default is 6334)",
default=6334,
)
QDRANT_REPLICATION_FACTOR: PositiveInt = Field(
description="Replication factor for Qdrant collections (default is 1)",
default=1,
)

5
api/configs/observability/otel/otel_config.py
View File

@ -27,6 +27,11 @@ class OTelConfig(BaseSettings):
default="otlp",
)
OTEL_EXPORTER_OTLP_PROTOCOL: str = Field(
description="OTLP exporter protocol ('grpc' or 'http')",
default="http",
)
OTEL_SAMPLING_RATE: float = Field(default=0.1, description="Sampling rate for traces (0.0 to 1.0)")
OTEL_BATCH_EXPORT_SCHEDULE_DELAY: int = Field(

2
api/configs/packaging/__init__.py
View File

@ -9,7 +9,7 @@ class PackagingInfo(BaseSettings):
CURRENT_VERSION: str = Field(
description="Dify version",
default="1.3.1",
default="1.4.1",
)
COMMIT_SHA: str = Field(

4
api/contexts/__init__.py
View File

@ -11,10 +11,6 @@ if TYPE_CHECKING:
from core.workflow.entities.variable_pool import VariablePool
tenant_id: ContextVar[str] = ContextVar("tenant_id")
workflow_variable_pool: ContextVar["VariablePool"] = ContextVar("workflow_variable_pool")
"""
To avoid race-conditions caused by gunicorn thread recycling, using RecyclableContextVar to replace with
"""

19
api/controllers/common/fields.py
View File

@ -1,5 +1,7 @@
from flask_restful import fields
from libs.helper import AppIconUrlField
parameters__system_parameters = {
"image_file_size_limit": fields.Integer,
"video_file_size_limit": fields.Integer,
@ -22,3 +24,20 @@ parameters_fields = {
"file_upload": fields.Raw,
"system_parameters": fields.Nested(parameters__system_parameters),
}
site_fields = {
"title": fields.String,
"chat_color_theme": fields.String,
"chat_color_theme_inverted": fields.Boolean,
"icon_type": fields.String,
"icon": fields.String,
"icon_background": fields.String,
"icon_url": AppIconUrlField,
"description": fields.String,
"copyright": fields.String,
"privacy_policy": fields.String,
"custom_disclaimer": fields.String,
"default_language": fields.String,
"show_workflow_steps": fields.Boolean,
"use_icon_as_answer_icon": fields.Boolean,
}

24
api/controllers/console/app/app.py
View File

@ -17,15 +17,13 @@ from controllers.console.wraps import (
)
from core.ops.ops_trace_manager import OpsTraceManager
from extensions.ext_database import db
from fields.app_fields import (
app_detail_fields,
app_detail_fields_with_site,
app_pagination_fields,
)
from fields.app_fields import app_detail_fields, app_detail_fields_with_site, app_pagination_fields
from libs.login import login_required
from models import Account, App
from services.app_dsl_service import AppDslService, ImportMode
from services.app_service import AppService
from services.enterprise.enterprise_service import EnterpriseService
from services.feature_service import FeatureService
ALLOW_CREATE_APP_MODES = ["chat", "agent-chat", "advanced-chat", "workflow", "completion"]
@ -75,7 +73,17 @@ class AppListApi(Resource):
if not app_pagination:
return {"data": [], "total": 0, "page": 1, "limit": 20, "has_more": False}
return marshal(app_pagination, app_pagination_fields)
if FeatureService.get_system_features().webapp_auth.enabled:
app_ids = [str(app.id) for app in app_pagination.items]
res = EnterpriseService.WebAppAuth.batch_get_app_access_mode_by_id(app_ids=app_ids)
if len(res) != len(app_ids):
raise BadRequest("Invalid app id in webapp auth")
for app in app_pagination.items:
if str(app.id) in res:
app.access_mode = res[str(app.id)].access_mode
return marshal(app_pagination, app_pagination_fields), 200
@setup_required
@login_required
@ -119,6 +127,10 @@ class AppApi(Resource):
app_model = app_service.get_app(app_model)
if FeatureService.get_system_features().webapp_auth.enabled:
app_setting = EnterpriseService.WebAppAuth.get_app_access_mode_by_id(app_id=str(app_model.id))
app_model.access_mode = app_setting.access_mode
return app_model
@setup_required

3
api/controllers/console/app/workflow.py
View File

@ -81,8 +81,7 @@ class DraftWorkflowApi(Resource):
parser.add_argument("graph", type=dict, required=True, nullable=False, location="json")
parser.add_argument("features", type=dict, required=True, nullable=False, location="json")
parser.add_argument("hash", type=str, required=False, location="json")
# TODO: set this to required=True after frontend is updated
parser.add_argument("environment_variables", type=list, required=False, location="json")
parser.add_argument("environment_variables", type=list, required=True, location="json")
parser.add_argument("conversation_variables", type=list, required=False, location="json")
args = parser.parse_args()
elif "text/plain" in content_type:

13
api/controllers/console/app/workflow_run.py
View File

@ -1,3 +1,6 @@
from typing import cast
from flask_login import current_user
from flask_restful import Resource, marshal_with, reqparse
from flask_restful.inputs import int_range
@ -12,8 +15,7 @@ from fields.workflow_run_fields import (
)
from libs.helper import uuid_value
from libs.login import login_required
from models import App
from models.model import AppMode
from models import Account, App, AppMode, EndUser
from services.workflow_run_service import WorkflowRunService
@ -90,7 +92,12 @@ class WorkflowRunNodeExecutionListApi(Resource):
run_id = str(run_id)
workflow_run_service = WorkflowRunService()
node_executions = workflow_run_service.get_workflow_run_node_executions(app_model=app_model, run_id=run_id)
user = cast("Account | EndUser", current_user)
node_executions = workflow_run_service.get_workflow_run_node_executions(
app_model=app_model,
run_id=run_id,
user=user,
)
return {"data": node_executions}

7
api/controllers/console/auth/forgot_password.py
View File

@ -24,7 +24,7 @@ from libs.password import hash_password, valid_password
from models.account import Account
from services.account_service import AccountService, TenantService
from services.errors.account import AccountRegisterError
from services.errors.workspace import WorkSpaceNotAllowedCreateError
from services.errors.workspace import WorkSpaceNotAllowedCreateError, WorkspacesLimitExceededError
from services.feature_service import FeatureService
@ -119,6 +119,9 @@ class ForgotPasswordResetApi(Resource):
if not reset_data:
raise InvalidTokenError()
# Must use token in reset phase
if reset_data.get("phase", "") != "reset":
raise InvalidTokenError()
# Must use token in reset phase
if reset_data.get("phase", "") != "reset":
raise InvalidTokenError()
@ -168,6 +171,8 @@ class ForgotPasswordResetApi(Resource):
)
except WorkSpaceNotAllowedCreateError:
pass
except WorkspacesLimitExceededError:
pass
except AccountRegisterError:
raise AccountInFreezeError()

25
api/controllers/console/auth/login.py
View File

@ -21,6 +21,7 @@ from controllers.console.error import (
AccountNotFound,
EmailSendIpLimitError,
NotAllowedCreateWorkspace,
WorkspacesLimitExceeded,
)
from controllers.console.wraps import email_password_login_enabled, setup_required
from events.tenant_event import tenant_was_created
@ -30,7 +31,7 @@ from models.account import Account
from services.account_service import AccountService, RegisterService, TenantService
from services.billing_service import BillingService
from services.errors.account import AccountRegisterError
from services.errors.workspace import WorkSpaceNotAllowedCreateError
from services.errors.workspace import WorkSpaceNotAllowedCreateError, WorkspacesLimitExceededError
from services.feature_service import FeatureService
@ -88,6 +89,11 @@ class LoginApi(Resource):
# SELF_HOSTED only have one workspace
tenants = TenantService.get_join_tenants(account)
if len(tenants) == 0:
system_features = FeatureService.get_system_features()
if system_features.is_allow_create_workspace and not system_features.license.workspaces.is_available():
raise WorkspacesLimitExceeded()
else:
return {
"result": "fail",
"data": "workspace not found, please contact system admin to invite you to join in a workspace",
@ -196,15 +202,18 @@ class EmailCodeLoginApi(Resource):
except AccountRegisterError as are:
raise AccountInFreezeError()
if account:
tenant = TenantService.get_join_tenants(account)
if not tenant:
tenants = TenantService.get_join_tenants(account)
if not tenants:
workspaces = FeatureService.get_system_features().license.workspaces
if not workspaces.is_available():
raise WorkspacesLimitExceeded()
if not FeatureService.get_system_features().is_allow_create_workspace:
raise NotAllowedCreateWorkspace()
else:
tenant = TenantService.create_tenant(f"{account.name}'s Workspace")
TenantService.create_tenant_member(tenant, account, role="owner")
account.current_tenant = tenant
tenant_was_created.send(tenant)
new_tenant = TenantService.create_tenant(f"{account.name}'s Workspace")
TenantService.create_tenant_member(new_tenant, account, role="owner")
account.current_tenant = new_tenant
tenant_was_created.send(new_tenant)
if account is None:
try:
@ -215,6 +224,8 @@ class EmailCodeLoginApi(Resource):
return NotAllowedCreateWorkspace()
except AccountRegisterError as are:
raise AccountInFreezeError()
except WorkspacesLimitExceededError:
raise WorkspacesLimitExceeded()
token_pair = AccountService.login(account, ip_address=extract_remote_ip(request))
AccountService.reset_login_error_rate_limit(args["email"])
return {"result": "success", "data": token_pair.model_dump()}

12
api/controllers/console/auth/oauth.py
View File

@ -148,15 +148,15 @@ def _generate_account(provider: str, user_info: OAuthUserInfo):
account = _get_account_by_openid_or_email(provider, user_info)
if account:
tenant = TenantService.get_join_tenants(account)
if not tenant:
tenants = TenantService.get_join_tenants(account)
if not tenants:
if not FeatureService.get_system_features().is_allow_create_workspace:
raise WorkSpaceNotAllowedCreateError()
else:
tenant = TenantService.create_tenant(f"{account.name}'s Workspace")
TenantService.create_tenant_member(tenant, account, role="owner")
account.current_tenant = tenant
tenant_was_created.send(tenant)
new_tenant = TenantService.create_tenant(f"{account.name}'s Workspace")
TenantService.create_tenant_member(new_tenant, account, role="owner")
account.current_tenant = new_tenant
tenant_was_created.send(new_tenant)
if not account:
if not FeatureService.get_system_features().is_allow_register:

35
api/controllers/console/datasets/datasets.py
View File

@ -526,17 +526,36 @@ class DatasetIndexingStatusApi(Resource):
)
documents_status = []
for document in documents:
completed_segments = DocumentSegment.query.filter(
completed_segments = (
db.session.query(DocumentSegment)
.filter(
DocumentSegment.completed_at.isnot(None),
DocumentSegment.document_id == str(document.id),
DocumentSegment.status != "re_segment",
).count()
total_segments = DocumentSegment.query.filter(
DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment"
).count()
document.completed_segments = completed_segments
document.total_segments = total_segments
documents_status.append(marshal(document, document_status_fields))
)
.count()
)
total_segments = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
.count()
)
# Create a dictionary with document attributes and additional fields
document_dict = {
"id": document.id,
"indexing_status": document.indexing_status,
"processing_started_at": document.processing_started_at,
"parsing_completed_at": document.parsing_completed_at,
"cleaning_completed_at": document.cleaning_completed_at,
"splitting_completed_at": document.splitting_completed_at,
"completed_at": document.completed_at,
"paused_at": document.paused_at,
"error": document.error,
"stopped_at": document.stopped_at,
"completed_segments": completed_segments,
"total_segments": total_segments,
}
documents_status.append(marshal(document_dict, document_status_fields))
data = {"data": documents_status}
return data

98
api/controllers/console/datasets/datasets_document.py
View File

@ -6,7 +6,7 @@ from typing import cast
from flask import request
from flask_login import current_user
from flask_restful import Resource, fields, marshal, marshal_with, reqparse
from sqlalchemy import asc, desc
from sqlalchemy import asc, desc, select
from werkzeug.exceptions import Forbidden, NotFound
import services
@ -112,7 +112,7 @@ class GetProcessRuleApi(Resource):
limits = DocumentService.DEFAULT_RULES["limits"]
if document_id:
# get the latest process rule
document = Document.query.get_or_404(document_id)
document = db.get_or_404(Document, document_id)
dataset = DatasetService.get_dataset(document.dataset_id)
@ -175,7 +175,7 @@ class DatasetDocumentListApi(Resource):
except services.errors.account.NoPermissionError as e:
raise Forbidden(str(e))
query = Document.query.filter_by(dataset_id=str(dataset_id), tenant_id=current_user.current_tenant_id)
query = select(Document).filter_by(dataset_id=str(dataset_id), tenant_id=current_user.current_tenant_id)
if search:
search = f"%{search}%"
@ -209,18 +209,24 @@ class DatasetDocumentListApi(Resource):
desc(Document.position),
)
paginated_documents = query.paginate(page=page, per_page=limit, max_per_page=100, error_out=False)
paginated_documents = db.paginate(select=query, page=page, per_page=limit, max_per_page=100, error_out=False)
documents = paginated_documents.items
if fetch:
for document in documents:
completed_segments = DocumentSegment.query.filter(
completed_segments = (
db.session.query(DocumentSegment)
.filter(
DocumentSegment.completed_at.isnot(None),
DocumentSegment.document_id == str(document.id),
DocumentSegment.status != "re_segment",
).count()
total_segments = DocumentSegment.query.filter(
DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment"
).count()
)
.count()
)
total_segments = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
.count()
)
document.completed_segments = completed_segments
document.total_segments = total_segments
data = marshal(documents, document_with_segments_fields)
@ -563,19 +569,36 @@ class DocumentBatchIndexingStatusApi(DocumentResource):
documents = self.get_batch_documents(dataset_id, batch)
documents_status = []
for document in documents:
completed_segments = DocumentSegment.query.filter(
completed_segments = (
db.session.query(DocumentSegment)
.filter(
DocumentSegment.completed_at.isnot(None),
DocumentSegment.document_id == str(document.id),
DocumentSegment.status != "re_segment",
).count()
total_segments = DocumentSegment.query.filter(
DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment"
).count()
document.completed_segments = completed_segments
document.total_segments = total_segments
if document.is_paused:
document.indexing_status = "paused"
documents_status.append(marshal(document, document_status_fields))
)
.count()
)
total_segments = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
.count()
)
# Create a dictionary with document attributes and additional fields
document_dict = {
"id": document.id,
"indexing_status": "paused" if document.is_paused else document.indexing_status,
"processing_started_at": document.processing_started_at,
"parsing_completed_at": document.parsing_completed_at,
"cleaning_completed_at": document.cleaning_completed_at,
"splitting_completed_at": document.splitting_completed_at,
"completed_at": document.completed_at,
"paused_at": document.paused_at,
"error": document.error,
"stopped_at": document.stopped_at,
"completed_segments": completed_segments,
"total_segments": total_segments,
}
documents_status.append(marshal(document_dict, document_status_fields))
data = {"data": documents_status}
return data
@ -589,20 +612,37 @@ class DocumentIndexingStatusApi(DocumentResource):
document_id = str(document_id)
document = self.get_document(dataset_id, document_id)
completed_segments = DocumentSegment.query.filter(
completed_segments = (
db.session.query(DocumentSegment)
.filter(
DocumentSegment.completed_at.isnot(None),
DocumentSegment.document_id == str(document_id),
DocumentSegment.status != "re_segment",
).count()
total_segments = DocumentSegment.query.filter(
DocumentSegment.document_id == str(document_id), DocumentSegment.status != "re_segment"
).count()
)
.count()
)
total_segments = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.document_id == str(document_id), DocumentSegment.status != "re_segment")
.count()
)
document.completed_segments = completed_segments
document.total_segments = total_segments
if document.is_paused:
document.indexing_status = "paused"
return marshal(document, document_status_fields)
# Create a dictionary with document attributes and additional fields
document_dict = {
"id": document.id,
"indexing_status": "paused" if document.is_paused else document.indexing_status,
"processing_started_at": document.processing_started_at,
"parsing_completed_at": document.parsing_completed_at,
"cleaning_completed_at": document.cleaning_completed_at,
"splitting_completed_at": document.splitting_completed_at,
"completed_at": document.completed_at,
"paused_at": document.paused_at,
"error": document.error,
"stopped_at": document.stopped_at,
"completed_segments": completed_segments,
"total_segments": total_segments,
}
return marshal(document_dict, document_status_fields)
class DocumentDetailApi(DocumentResource):

87
api/controllers/console/datasets/datasets_segments.py
View File

@ -4,6 +4,7 @@ import pandas as pd
from flask import request
from flask_login import current_user
from flask_restful import Resource, marshal, reqparse
from sqlalchemy import select
from werkzeug.exceptions import Forbidden, NotFound
import services
@ -26,6 +27,7 @@ from controllers.console.wraps import (
from core.errors.error import LLMBadRequestError, ProviderTokenNotInitError
from core.model_manager import ModelManager
from core.model_runtime.entities.model_entities import ModelType
from extensions.ext_database import db
from extensions.ext_redis import redis_client
from fields.segment_fields import child_chunk_fields, segment_fields
from libs.login import login_required
@ -74,9 +76,14 @@ class DatasetDocumentSegmentListApi(Resource):
hit_count_gte = args["hit_count_gte"]
keyword = args["keyword"]
query = DocumentSegment.query.filter(
DocumentSegment.document_id == str(document_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).order_by(DocumentSegment.position.asc())
query = (
select(DocumentSegment)
.filter(
DocumentSegment.document_id == str(document_id),
DocumentSegment.tenant_id == current_user.current_tenant_id,
)
.order_by(DocumentSegment.position.asc())
)
if status_list:
query = query.filter(DocumentSegment.status.in_(status_list))
@ -93,7 +100,7 @@ class DatasetDocumentSegmentListApi(Resource):
elif args["enabled"].lower() == "false":
query = query.filter(DocumentSegment.enabled == False)
segments = query.paginate(page=page, per_page=limit, max_per_page=100, error_out=False)
segments = db.paginate(select=query, page=page, per_page=limit, max_per_page=100, error_out=False)
response = {
"data": marshal(segments.items, segment_fields),
@ -276,9 +283,11 @@ class DatasetDocumentSegmentUpdateApi(Resource):
raise ProviderNotInitializeError(ex.description)
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
# The role of the current user in the ta table must be admin, owner, dataset_operator, or editor
@ -320,9 +329,11 @@ class DatasetDocumentSegmentUpdateApi(Resource):
raise NotFound("Document not found.")
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
# The role of the current user in the ta table must be admin, owner, dataset_operator, or editor
@ -423,9 +434,11 @@ class ChildChunkAddApi(Resource):
raise NotFound("Document not found.")
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
if not current_user.is_dataset_editor:
@ -478,9 +491,11 @@ class ChildChunkAddApi(Resource):
raise NotFound("Document not found.")
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
parser = reqparse.RequestParser()
@ -523,9 +538,11 @@ class ChildChunkAddApi(Resource):
raise NotFound("Document not found.")
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
# The role of the current user in the ta table must be admin, owner, dataset_operator, or editor
@ -567,16 +584,20 @@ class ChildChunkUpdateApi(Resource):
raise NotFound("Document not found.")
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
# check child chunk
child_chunk_id = str(child_chunk_id)
child_chunk = ChildChunk.query.filter(
ChildChunk.id == str(child_chunk_id), ChildChunk.tenant_id == current_user.current_tenant_id
).first()
child_chunk = (
db.session.query(ChildChunk)
.filter(ChildChunk.id == str(child_chunk_id), ChildChunk.tenant_id == current_user.current_tenant_id)
.first()
)
if not child_chunk:
raise NotFound("Child chunk not found.")
# The role of the current user in the ta table must be admin, owner, dataset_operator, or editor
@ -612,16 +633,20 @@ class ChildChunkUpdateApi(Resource):
raise NotFound("Document not found.")
# check segment
segment_id = str(segment_id)
segment = DocumentSegment.query.filter(
DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id
).first()
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == str(segment_id), DocumentSegment.tenant_id == current_user.current_tenant_id)
.first()
)
if not segment:
raise NotFound("Segment not found.")
# check child chunk
child_chunk_id = str(child_chunk_id)
child_chunk = ChildChunk.query.filter(
ChildChunk.id == str(child_chunk_id), ChildChunk.tenant_id == current_user.current_tenant_id
).first()
child_chunk = (
db.session.query(ChildChunk)
.filter(ChildChunk.id == str(child_chunk_id), ChildChunk.tenant_id == current_user.current_tenant_id)
.first()
)
if not child_chunk:
raise NotFound("Child chunk not found.")
# The role of the current user in the ta table must be admin, owner, dataset_operator, or editor

2
api/controllers/console/datasets/external.py
View File

@ -209,6 +209,7 @@ class ExternalKnowledgeHitTestingApi(Resource):
parser = reqparse.RequestParser()
parser.add_argument("query", type=str, location="json")
parser.add_argument("external_retrieval_model", type=dict, required=False, location="json")
parser.add_argument("metadata_filtering_conditions", type=dict, required=False, location="json")
args = parser.parse_args()
HitTestingService.hit_testing_args_check(args)
@ -219,6 +220,7 @@ class ExternalKnowledgeHitTestingApi(Resource):
query=args["query"],
account=current_user,
external_retrieval_model=args["external_retrieval_model"],
metadata_filtering_conditions=args["metadata_filtering_conditions"],
)
return response

12
api/controllers/console/error.py
View File

@ -46,6 +46,18 @@ class NotAllowedCreateWorkspace(BaseHTTPException):
code = 400
class WorkspaceMembersLimitExceeded(BaseHTTPException):
error_code = "limit_exceeded"
description = "Unable to add member because the maximum workspace's member limit was exceeded"
code = 400
class WorkspacesLimitExceeded(BaseHTTPException):
error_code = "limit_exceeded"
description = "Unable to create workspace because the maximum workspace limit was exceeded"
code = 400
class AccountBannedError(BaseHTTPException):
error_code = "account_banned"
description = "Account is banned."

6
api/controllers/console/explore/error.py
View File

@ -23,3 +23,9 @@ class AppSuggestedQuestionsAfterAnswerDisabledError(BaseHTTPException):
error_code = "app_suggested_questions_after_answer_disabled"
description = "Function Suggested questions after answer disabled."
code = 403
class AppAccessDeniedError(BaseHTTPException):
error_code = "access_denied"
description = "App access denied."
code = 403

31
api/controllers/console/explore/installed_app.py
View File

@ -1,3 +1,4 @@
import logging
from datetime import UTC, datetime
from typing import Any
@ -15,6 +16,11 @@ from fields.installed_app_fields import installed_app_list_fields
from libs.login import login_required
from models import App, InstalledApp, RecommendedApp
from services.account_service import TenantService
from services.app_service import AppService
from services.enterprise.enterprise_service import EnterpriseService
from services.feature_service import FeatureService
logger = logging.getLogger(__name__)
class InstalledAppsListApi(Resource):
@ -48,6 +54,21 @@ class InstalledAppsListApi(Resource):
for installed_app in installed_apps
if installed_app.app is not None
]
# filter out apps that user doesn't have access to
if FeatureService.get_system_features().webapp_auth.enabled:
user_id = current_user.id
res = []
for installed_app in installed_app_list:
app_code = AppService.get_app_code_by_id(str(installed_app["app"].id))
if EnterpriseService.WebAppAuth.is_user_allowed_to_access_webapp(
user_id=user_id,
app_code=app_code,
):
res.append(installed_app)
installed_app_list = res
logger.debug(f"installed_app_list: {installed_app_list}, user_id: {user_id}")
installed_app_list.sort(
key=lambda app: (
-app["is_pinned"],
@ -66,7 +87,7 @@ class InstalledAppsListApi(Resource):
parser.add_argument("app_id", type=str, required=True, help="Invalid app_id")
args = parser.parse_args()
recommended_app = RecommendedApp.query.filter(RecommendedApp.app_id == args["app_id"]).first()
recommended_app = db.session.query(RecommendedApp).filter(RecommendedApp.app_id == args["app_id"]).first()
if recommended_app is None:
raise NotFound("App not found")
@ -79,9 +100,11 @@ class InstalledAppsListApi(Resource):
if not app.is_public:
raise Forbidden("You can't install a non-public app")
installed_app = InstalledApp.query.filter(
and_(InstalledApp.app_id == args["app_id"], InstalledApp.tenant_id == current_tenant_id)
).first()
installed_app = (
db.session.query(InstalledApp)
.filter(and_(InstalledApp.app_id == args["app_id"], InstalledApp.tenant_id == current_tenant_id))
.first()
)
if installed_app is None:
# todo: position

36
api/controllers/console/explore/wraps.py
View File

@ -4,10 +4,14 @@ from flask_login import current_user
from flask_restful import Resource
from werkzeug.exceptions import NotFound
from controllers.console.explore.error import AppAccessDeniedError
from controllers.console.wraps import account_initialization_required
from extensions.ext_database import db
from libs.login import login_required
from models import InstalledApp
from services.app_service import AppService
from services.enterprise.enterprise_service import EnterpriseService
from services.feature_service import FeatureService
def installed_app_required(view=None):
@ -48,6 +52,36 @@ def installed_app_required(view=None):
return decorator
def user_allowed_to_access_app(view=None):
def decorator(view):
@wraps(view)
def decorated(installed_app: InstalledApp, *args, **kwargs):
feature = FeatureService.get_system_features()
if feature.webapp_auth.enabled:
app_id = installed_app.app_id
app_code = AppService.get_app_code_by_id(app_id)
res = EnterpriseService.WebAppAuth.is_user_allowed_to_access_webapp(
user_id=str(current_user.id),
app_code=app_code,
)
if not res:
raise AppAccessDeniedError()
return view(installed_app, *args, **kwargs)
return decorated
if view:
return decorator(view)
return decorator
class InstalledAppResource(Resource):
# must be reversed if there are multiple decorators
method_decorators = [installed_app_required, account_initialization_required, login_required]
method_decorators = [
user_allowed_to_access_app,
installed_app_required,
account_initialization_required,
login_required,
]

9
api/controllers/console/workspace/members.py
View File

@ -6,6 +6,7 @@ from flask_restful import Resource, abort, marshal_with, reqparse
import services
from configs import dify_config
from controllers.console import api
from controllers.console.error import WorkspaceMembersLimitExceeded
from controllers.console.wraps import (
account_initialization_required,
cloud_edition_billing_resource_check,
@ -17,6 +18,7 @@ from libs.login import login_required
from models.account import Account, TenantAccountRole
from services.account_service import RegisterService, TenantService
from services.errors.account import AccountAlreadyInTenantError
from services.feature_service import FeatureService
class MemberListApi(Resource):
@ -54,6 +56,12 @@ class MemberInviteEmailApi(Resource):
inviter = current_user
invitation_results = []
console_web_url = dify_config.CONSOLE_WEB_URL
workspace_members = FeatureService.get_features(tenant_id=inviter.current_tenant.id).workspace_members
if not workspace_members.is_available(len(invitee_emails)):
raise WorkspaceMembersLimitExceeded()
for invitee_email in invitee_emails:
try:
token = RegisterService.invite_new_member(
@ -71,7 +79,6 @@ class MemberInviteEmailApi(Resource):
invitation_results.append(
{"status": "success", "email": invitee_email, "url": f"{console_web_url}/signin"}
)
break
except Exception as e:
invitation_results.append({"status": "failed", "email": invitee_email, "message": str(e)})

8
api/controllers/console/workspace/plugin.py
View File

@ -41,12 +41,16 @@ class PluginListApi(Resource):
@account_initialization_required
def get(self):
tenant_id = current_user.current_tenant_id
parser = reqparse.RequestParser()
parser.add_argument("page", type=int, required=False, location="args", default=1)
parser.add_argument("page_size", type=int, required=False, location="args", default=256)
args = parser.parse_args()
try:
plugins = PluginService.list(tenant_id)
plugins_with_total = PluginService.list_with_total(tenant_id, args["page"], args["page_size"])
except PluginDaemonClientSideError as e:
raise ValueError(e)
return jsonable_encoder({"plugins": plugins})
return jsonable_encoder({"plugins": plugins_with_total.list, "total": plugins_with_total.total})
class PluginListLatestVersionsApi(Resource):

32
api/controllers/console/workspace/workspace.py
View File

@ -3,6 +3,7 @@ import logging
from flask import request
from flask_login import current_user
from flask_restful import Resource, fields, inputs, marshal, marshal_with, reqparse
from sqlalchemy import select
from werkzeug.exceptions import Unauthorized
import services
@ -67,16 +68,24 @@ class TenantListApi(Resource):
@account_initialization_required
def get(self):
tenants = TenantService.get_join_tenants(current_user)
tenant_dicts = []
for tenant in tenants:
features = FeatureService.get_features(tenant.id)
if features.billing.enabled:
tenant.plan = features.billing.subscription.plan
else:
tenant.plan = "sandbox"
if tenant.id == current_user.current_tenant_id:
tenant.current = True # Set current=True for current tenant
return {"workspaces": marshal(tenants, tenants_fields)}, 200
# Create a dictionary with tenant attributes
tenant_dict = {
"id": tenant.id,
"name": tenant.name,
"status": tenant.status,
"created_at": tenant.created_at,
"plan": features.billing.subscription.plan if features.billing.enabled else "sandbox",
"current": tenant.id == current_user.current_tenant_id,
}
tenant_dicts.append(tenant_dict)
return {"workspaces": marshal(tenant_dicts, tenants_fields)}, 200
class WorkspaceListApi(Resource):
@ -88,9 +97,8 @@ class WorkspaceListApi(Resource):
parser.add_argument("limit", type=inputs.int_range(1, 100), required=False, default=20, location="args")
args = parser.parse_args()
tenants = Tenant.query.order_by(Tenant.created_at.desc()).paginate(
page=args["page"], per_page=args["limit"], error_out=False
)
stmt = select(Tenant).order_by(Tenant.created_at.desc())
tenants = db.paginate(select=stmt, page=args["page"], per_page=args["limit"], error_out=False)
has_more = False
if tenants.has_next:
@ -162,7 +170,7 @@ class CustomConfigWorkspaceApi(Resource):
parser.add_argument("replace_webapp_logo", type=str, location="json")
args = parser.parse_args()
tenant = Tenant.query.filter(Tenant.id == current_user.current_tenant_id).one_or_404()
tenant = db.get_or_404(Tenant, current_user.current_tenant_id)
custom_config_dict = {
"remove_webapp_brand": args["remove_webapp_brand"],
@ -226,7 +234,7 @@ class WorkspaceInfoApi(Resource):
parser.add_argument("name", type=str, required=True, location="json")
args = parser.parse_args()
tenant = Tenant.query.filter(Tenant.id == current_user.current_tenant_id).one_or_404()
tenant = db.get_or_404(Tenant, current_user.current_tenant_id)
tenant.name = args["name"]
db.session.commit()

21
api/controllers/files/upload.py
View File

@ -64,9 +64,24 @@ class PluginUploadFileApi(Resource):
extension = guess_extension(tool_file.mimetype) or ".bin"
preview_url = ToolFileManager.sign_file(tool_file_id=tool_file.id, extension=extension)
tool_file.mime_type = mimetype
tool_file.extension = extension
tool_file.preview_url = preview_url
# Create a dictionary with all the necessary attributes
result = {
"id": tool_file.id,
"user_id": tool_file.user_id,
"tenant_id": tool_file.tenant_id,
"conversation_id": tool_file.conversation_id,
"file_key": tool_file.file_key,
"mimetype": tool_file.mimetype,
"original_url": tool_file.original_url,
"name": tool_file.name,
"size": tool_file.size,
"mime_type": mimetype,
"extension": extension,
"preview_url": preview_url,
}
return result, 201
except services.errors.file.FileTooLargeError as file_too_large_error:
raise FileTooLargeError(file_too_large_error.description)
except services.errors.file.UnsupportedFileTypeError:

1
api/controllers/inner_api/__init__.py
View File

@ -5,5 +5,6 @@ from libs.external_api import ExternalApi
bp = Blueprint("inner_api", __name__, url_prefix="/inner/api")
api = ExternalApi(bp)
from . import mail
from .plugin import plugin
from .workspace import workspace

27
api/controllers/inner_api/mail.py Normal file
View File

@ -0,0 +1,27 @@
from flask_restful import (
Resource, # type: ignore
reqparse,
)
from controllers.console.wraps import setup_required
from controllers.inner_api import api
from controllers.inner_api.wraps import enterprise_inner_api_only
from services.enterprise.mail_service import DifyMail, EnterpriseMailService
class EnterpriseMail(Resource):
@setup_required
@enterprise_inner_api_only
def post(self):
parser = reqparse.RequestParser()
parser.add_argument("to", type=str, action="append", required=True)
parser.add_argument("subject", type=str, required=True)
parser.add_argument("body", type=str, required=True)
parser.add_argument("substitutions", type=dict, required=False)
args = parser.parse_args()
EnterpriseMailService.send_mail(DifyMail(**args))
return {"message": "success"}, 200
api.add_resource(EnterpriseMail, "/enterprise/mail")

2
api/controllers/service_api/__init__.py
View File

@ -6,6 +6,6 @@ bp = Blueprint("service_api", __name__, url_prefix="/v1")
api = ExternalApi(bp)
from . import index
from .app import annotation, app, audio, completion, conversation, file, message, workflow
from .app import annotation, app, audio, completion, conversation, file, message, site, workflow
from .dataset import dataset, document, hit_testing, metadata, segment, upload_file
from .workspace import models

14
api/controllers/service_api/app/annotation.py
View File

@ -3,7 +3,7 @@ from flask_restful import Resource, marshal, marshal_with, reqparse
from werkzeug.exceptions import Forbidden
from controllers.service_api import api
from controllers.service_api.wraps import FetchUserArg, WhereisUserArg, validate_app_token
from controllers.service_api.wraps import validate_app_token
from extensions.ext_redis import redis_client
from fields.annotation_fields import (
annotation_fields,
@ -14,7 +14,7 @@ from services.annotation_service import AppAnnotationService
class AnnotationReplyActionApi(Resource):
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.JSON))
@validate_app_token
def post(self, app_model: App, end_user: EndUser, action):
parser = reqparse.RequestParser()
parser.add_argument("score_threshold", required=True, type=float, location="json")
@ -31,7 +31,7 @@ class AnnotationReplyActionApi(Resource):
class AnnotationReplyActionStatusApi(Resource):
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.QUERY))
@validate_app_token
def get(self, app_model: App, end_user: EndUser, job_id, action):
job_id = str(job_id)
app_annotation_job_key = "{}_app_annotation_job_{}".format(action, str(job_id))
@ -49,7 +49,7 @@ class AnnotationReplyActionStatusApi(Resource):
class AnnotationListApi(Resource):
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.QUERY))
@validate_app_token
def get(self, app_model: App, end_user: EndUser):
page = request.args.get("page", default=1, type=int)
limit = request.args.get("limit", default=20, type=int)
@ -65,7 +65,7 @@ class AnnotationListApi(Resource):
}
return response, 200
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.JSON))
@validate_app_token
@marshal_with(annotation_fields)
def post(self, app_model: App, end_user: EndUser):
parser = reqparse.RequestParser()
@ -77,7 +77,7 @@ class AnnotationListApi(Resource):
class AnnotationUpdateDeleteApi(Resource):
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.JSON))
@validate_app_token
@marshal_with(annotation_fields)
def put(self, app_model: App, end_user: EndUser, annotation_id):
if not current_user.is_editor:
@ -91,7 +91,7 @@ class AnnotationUpdateDeleteApi(Resource):
annotation = AppAnnotationService.update_app_annotation_directly(args, app_model.id, annotation_id)
return annotation
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.QUERY))
@validate_app_token
def delete(self, app_model: App, end_user: EndUser, annotation_id):
if not current_user.is_editor:
raise Forbidden()

13
api/controllers/service_api/app/message.py
View File

@ -93,6 +93,18 @@ class MessageFeedbackApi(Resource):
return {"result": "success"}
class AppGetFeedbacksApi(Resource):
@validate_app_token
def get(self, app_model: App):
"""Get All Feedbacks of an app"""
parser = reqparse.RequestParser()
parser.add_argument("page", type=int, default=1, location="args")
parser.add_argument("limit", type=int_range(1, 101), required=False, default=20, location="args")
args = parser.parse_args()
feedbacks = MessageService.get_all_messages_feedbacks(app_model, page=args["page"], limit=args["limit"])
return {"data": feedbacks}
class MessageSuggestedApi(Resource):
@validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.QUERY, required=True))
def get(self, app_model: App, end_user: EndUser, message_id):
@ -119,3 +131,4 @@ class MessageSuggestedApi(Resource):
api.add_resource(MessageListApi, "/messages")
api.add_resource(MessageFeedbackApi, "/messages/<uuid:message_id>/feedbacks")
api.add_resource(MessageSuggestedApi, "/messages/<uuid:message_id>/suggested")
api.add_resource(AppGetFeedbacksApi, "/app/feedbacks")

30
api/controllers/service_api/app/site.py Normal file
View File

@ -0,0 +1,30 @@
from flask_restful import Resource, marshal_with
from werkzeug.exceptions import Forbidden
from controllers.common import fields
from controllers.service_api import api
from controllers.service_api.wraps import validate_app_token
from extensions.ext_database import db
from models.account import TenantStatus
from models.model import App, Site
class AppSiteApi(Resource):
"""Resource for app sites."""
@validate_app_token
@marshal_with(fields.site_fields)
def get(self, app_model: App):
"""Retrieve app site info."""
site = db.session.query(Site).filter(Site.app_id == app_model.id).first()
if not site:
raise Forbidden()
if app_model.tenant.status == TenantStatus.ARCHIVE:
raise Forbidden()
return site
api.add_resource(AppSiteApi, "/site")

2
api/controllers/service_api/dataset/dataset.py
View File

@ -313,7 +313,7 @@ class DatasetApi(DatasetApiResource):
try:
if DatasetService.delete_dataset(dataset_id_str, current_user):
DatasetPermissionService.clear_partial_member_list(dataset_id_str)
return {"result": "success"}, 204
return 204
else:
raise NotFound("Dataset not found.")
except services.errors.dataset.DatasetInUseError:

47
api/controllers/service_api/dataset/document.py
View File

@ -2,10 +2,10 @@ import json
from flask import request
from flask_restful import marshal, reqparse
from sqlalchemy import desc
from sqlalchemy import desc, select
from werkzeug.exceptions import NotFound
import services.dataset_service
import services
from controllers.common.errors import FilenameNotExistsError
from controllers.service_api import api
from controllers.service_api.app.error import (
@ -323,7 +323,7 @@ class DocumentDeleteApi(DatasetApiResource):
except services.errors.document.DocumentIndexingError:
raise DocumentIndexingError("Cannot delete document during indexing.")
return {"result": "success"}, 204
return 204
class DocumentListApi(DatasetApiResource):
@ -337,7 +337,7 @@ class DocumentListApi(DatasetApiResource):
if not dataset:
raise NotFound("Dataset not found.")
query = Document.query.filter_by(dataset_id=str(dataset_id), tenant_id=tenant_id)
query = select(Document).filter_by(dataset_id=str(dataset_id), tenant_id=tenant_id)
if search:
search = f"%{search}%"
@ -345,7 +345,7 @@ class DocumentListApi(DatasetApiResource):
query = query.order_by(desc(Document.created_at), desc(Document.position))
paginated_documents = query.paginate(page=page, per_page=limit, max_per_page=100, error_out=False)
paginated_documents = db.paginate(select=query, page=page, per_page=limit, max_per_page=100, error_out=False)
documents = paginated_documents.items
response = {
@ -374,19 +374,36 @@ class DocumentIndexingStatusApi(DatasetApiResource):
raise NotFound("Documents not found.")
documents_status = []
for document in documents:
completed_segments = DocumentSegment.query.filter(
completed_segments = (
db.session.query(DocumentSegment)
.filter(
DocumentSegment.completed_at.isnot(None),
DocumentSegment.document_id == str(document.id),
DocumentSegment.status != "re_segment",
).count()
total_segments = DocumentSegment.query.filter(
DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment"
).count()
document.completed_segments = completed_segments
document.total_segments = total_segments
if document.is_paused:
document.indexing_status = "paused"
documents_status.append(marshal(document, document_status_fields))
)
.count()
)
total_segments = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.document_id == str(document.id), DocumentSegment.status != "re_segment")
.count()
)
# Create a dictionary with document attributes and additional fields
document_dict = {
"id": document.id,
"indexing_status": "paused" if document.is_paused else document.indexing_status,
"processing_started_at": document.processing_started_at,
"parsing_completed_at": document.parsing_completed_at,
"cleaning_completed_at": document.cleaning_completed_at,
"splitting_completed_at": document.splitting_completed_at,
"completed_at": document.completed_at,
"paused_at": document.paused_at,
"error": document.error,
"stopped_at": document.stopped_at,
"completed_segments": completed_segments,
"total_segments": total_segments,
}
documents_status.append(marshal(document_dict, document_status_fields))
data = {"data": documents_status}
return data

4
api/controllers/service_api/dataset/segment.py
View File

@ -159,7 +159,7 @@ class DatasetSegmentApi(DatasetApiResource):
if not segment:
raise NotFound("Segment not found.")
SegmentService.delete_segment(segment, document, dataset)
return {"result": "success"}, 204
return 204
@cloud_edition_billing_resource_check("vector_space", "dataset")
def post(self, tenant_id, dataset_id, document_id, segment_id):
@ -344,7 +344,7 @@ class DatasetChildChunkApi(DatasetApiResource):
except ChildChunkDeleteIndexServiceError as e:
raise ChildChunkDeleteIndexError(str(e))
return {"result": "success"}, 204
return 204
@cloud_edition_billing_resource_check("vector_space", "dataset")
@cloud_edition_billing_knowledge_limit_check("add_segment", "dataset")

7
api/controllers/service_api/wraps.py
View File

@ -99,7 +99,12 @@ def validate_app_token(view: Optional[Callable] = None, *, fetch_user_arg: Optio
if user_id:
user_id = str(user_id)
kwargs["end_user"] = create_or_update_end_user_for_user_id(app_model, user_id)
end_user = create_or_update_end_user_for_user_id(app_model, user_id)
kwargs["end_user"] = end_user
# Set EndUser as current logged-in user for flask_login.current_user
current_app.login_manager._update_request_context_with_user(end_user) # type: ignore
user_logged_in.send(current_app._get_current_object(), user=end_user) # type: ignore
return view_func(*args, **kwargs)

51
api/controllers/web/app.py
View File

@ -1,12 +1,15 @@
from flask_restful import marshal_with
from flask import request
from flask_restful import Resource, marshal_with, reqparse
from controllers.common import fields
from controllers.web import api
from controllers.web.error import AppUnavailableError
from controllers.web.wraps import WebApiResource
from core.app.app_config.common.parameters_mapping import get_parameters_from_feature_dict
from libs.passport import PassportService
from models.model import App, AppMode
from services.app_service import AppService
from services.enterprise.enterprise_service import EnterpriseService
class AppParameterApi(WebApiResource):
@ -40,5 +43,51 @@ class AppMeta(WebApiResource):
return AppService().get_app_meta(app_model)
class AppAccessMode(Resource):
def get(self):
parser = reqparse.RequestParser()
parser.add_argument("appId", type=str, required=True, location="args")
args = parser.parse_args()
app_id = args["appId"]
res = EnterpriseService.WebAppAuth.get_app_access_mode_by_id(app_id)
return {"accessMode": res.access_mode}
class AppWebAuthPermission(Resource):
def get(self):
user_id = "visitor"
try:
auth_header = request.headers.get("Authorization")
if auth_header is None:
raise
if " " not in auth_header:
raise
auth_scheme, tk = auth_header.split(None, 1)
auth_scheme = auth_scheme.lower()
if auth_scheme != "bearer":
raise
decoded = PassportService().verify(tk)
user_id = decoded.get("user_id", "visitor")
except Exception as e:
pass
parser = reqparse.RequestParser()
parser.add_argument("appId", type=str, required=True, location="args")
args = parser.parse_args()
app_id = args["appId"]
app_code = AppService.get_app_code_by_id(app_id)
res = EnterpriseService.WebAppAuth.is_user_allowed_to_access_webapp(str(user_id), app_code)
return {"result": res}
api.add_resource(AppParameterApi, "/parameters")
api.add_resource(AppMeta, "/meta")
# webapp auth apis
api.add_resource(AppAccessMode, "/webapp/access-mode")
api.add_resource(AppWebAuthPermission, "/webapp/permission")

10
api/controllers/web/error.py
View File

@ -121,9 +121,15 @@ class UnsupportedFileTypeError(BaseHTTPException):
code = 415
class WebSSOAuthRequiredError(BaseHTTPException):
class WebAppAuthRequiredError(BaseHTTPException):
error_code = "web_sso_auth_required"
description = "Web SSO authentication required."
description = "Web app authentication required."
code = 401
class WebAppAuthAccessDeniedError(BaseHTTPException):
error_code = "web_app_access_denied"
description = "You do not have permission to access this web app."
code = 401

120
api/controllers/web/login.py Normal file
View File

@ -0,0 +1,120 @@
from flask import request
from flask_restful import Resource, reqparse
from jwt import InvalidTokenError # type: ignore
from werkzeug.exceptions import BadRequest
import services
from controllers.console.auth.error import EmailCodeError, EmailOrPasswordMismatchError, InvalidEmailError
from controllers.console.error import AccountBannedError, AccountNotFound
from controllers.console.wraps import setup_required
from libs.helper import email
from libs.password import valid_password
from services.account_service import AccountService
from services.webapp_auth_service import WebAppAuthService
class LoginApi(Resource):
"""Resource for web app email/password login."""
def post(self):
"""Authenticate user and login."""
parser = reqparse.RequestParser()
parser.add_argument("email", type=email, required=True, location="json")
parser.add_argument("password", type=valid_password, required=True, location="json")
args = parser.parse_args()
app_code = request.headers.get("X-App-Code")
if app_code is None:
raise BadRequest("X-App-Code header is missing.")
try:
account = WebAppAuthService.authenticate(args["email"], args["password"])
except services.errors.account.AccountLoginError:
raise AccountBannedError()
except services.errors.account.AccountPasswordError:
raise EmailOrPasswordMismatchError()
except services.errors.account.AccountNotFoundError:
raise AccountNotFound()
WebAppAuthService._validate_user_accessibility(account=account, app_code=app_code)
end_user = WebAppAuthService.create_end_user(email=args["email"], app_code=app_code)
token = WebAppAuthService.login(account=account, app_code=app_code, end_user_id=end_user.id)
return {"result": "success", "token": token}
# class LogoutApi(Resource):
# @setup_required
# def get(self):
# account = cast(Account, flask_login.current_user)
# if isinstance(account, flask_login.AnonymousUserMixin):
# return {"result": "success"}
# flask_login.logout_user()
# return {"result": "success"}
class EmailCodeLoginSendEmailApi(Resource):
@setup_required
def post(self):
parser = reqparse.RequestParser()
parser.add_argument("email", type=email, required=True, location="json")
parser.add_argument("language", type=str, required=False, location="json")
args = parser.parse_args()
if args["language"] is not None and args["language"] == "zh-Hans":
language = "zh-Hans"
else:
language = "en-US"
account = WebAppAuthService.get_user_through_email(args["email"])
if account is None:
raise AccountNotFound()
else:
token = WebAppAuthService.send_email_code_login_email(account=account, language=language)
return {"result": "success", "data": token}
class EmailCodeLoginApi(Resource):
@setup_required
def post(self):
parser = reqparse.RequestParser()
parser.add_argument("email", type=str, required=True, location="json")
parser.add_argument("code", type=str, required=True, location="json")
parser.add_argument("token", type=str, required=True, location="json")
args = parser.parse_args()
user_email = args["email"]
app_code = request.headers.get("X-App-Code")
if app_code is None:
raise BadRequest("X-App-Code header is missing.")
token_data = WebAppAuthService.get_email_code_login_data(args["token"])
if token_data is None:
raise InvalidTokenError()
if token_data["email"] != args["email"]:
raise InvalidEmailError()
if token_data["code"] != args["code"]:
raise EmailCodeError()
WebAppAuthService.revoke_email_code_login_token(args["token"])
account = WebAppAuthService.get_user_through_email(user_email)
if not account:
raise AccountNotFound()
WebAppAuthService._validate_user_accessibility(account=account, app_code=app_code)
end_user = WebAppAuthService.create_end_user(email=user_email, app_code=app_code)
token = WebAppAuthService.login(account=account, app_code=app_code, end_user_id=end_user.id)
AccountService.reset_login_error_rate_limit(args["email"])
return {"result": "success", "token": token}
# api.add_resource(LoginApi, "/login")
# api.add_resource(LogoutApi, "/logout")
# api.add_resource(EmailCodeLoginSendEmailApi, "/email-code-login")
# api.add_resource(EmailCodeLoginApi, "/email-code-login/validity")

10
api/controllers/web/passport.py
View File

@ -5,7 +5,7 @@ from flask_restful import Resource
from werkzeug.exceptions import NotFound, Unauthorized
from controllers.web import api
from controllers.web.error import WebSSOAuthRequiredError
from controllers.web.error import WebAppAuthRequiredError
from extensions.ext_database import db
from libs.passport import PassportService
from models.model import App, EndUser, Site
@ -24,10 +24,10 @@ class PassportResource(Resource):
if app_code is None:
raise Unauthorized("X-App-Code header is missing.")
if system_features.sso_enforced_for_web:
app_web_sso_enabled = EnterpriseService.get_app_web_sso_enabled(app_code).get("enabled", False)
if app_web_sso_enabled:
raise WebSSOAuthRequiredError()
if system_features.webapp_auth.enabled:
app_settings = EnterpriseService.WebAppAuth.get_app_access_mode_by_code(app_code=app_code)
if not app_settings or not app_settings.access_mode == "public":
raise WebAppAuthRequiredError()
# get site from db and check if it is normal
site = db.session.query(Site).filter(Site.code == app_code, Site.status == "normal").first()

58
api/controllers/web/wraps.py
View File

@ -4,7 +4,7 @@ from flask import request
from flask_restful import Resource
from werkzeug.exceptions import BadRequest, NotFound, Unauthorized
from controllers.web.error import WebSSOAuthRequiredError
from controllers.web.error import WebAppAuthAccessDeniedError, WebAppAuthRequiredError
from extensions.ext_database import db
from libs.passport import PassportService
from models.model import App, EndUser, Site
@ -29,7 +29,7 @@ def validate_jwt_token(view=None):
def decode_jwt_token():
system_features = FeatureService.get_system_features()
app_code = request.headers.get("X-App-Code")
app_code = str(request.headers.get("X-App-Code"))
try:
auth_header = request.headers.get("Authorization")
if auth_header is None:
@ -57,35 +57,53 @@ def decode_jwt_token():
if not end_user:
raise NotFound()
_validate_web_sso_token(decoded, system_features, app_code)
# for enterprise webapp auth
app_web_auth_enabled = False
if system_features.webapp_auth.enabled:
app_web_auth_enabled = (
EnterpriseService.WebAppAuth.get_app_access_mode_by_code(app_code=app_code).access_mode != "public"
)
_validate_webapp_token(decoded, app_web_auth_enabled, system_features.webapp_auth.enabled)
_validate_user_accessibility(decoded, app_code, app_web_auth_enabled, system_features.webapp_auth.enabled)
return app_model, end_user
except Unauthorized as e:
if system_features.sso_enforced_for_web:
app_web_sso_enabled = EnterpriseService.get_app_web_sso_enabled(app_code).get("enabled", False)
if app_web_sso_enabled:
raise WebSSOAuthRequiredError()
if system_features.webapp_auth.enabled:
app_web_auth_enabled = (
EnterpriseService.WebAppAuth.get_app_access_mode_by_code(app_code=str(app_code)).access_mode != "public"
)
if app_web_auth_enabled:
raise WebAppAuthRequiredError()
raise Unauthorized(e.description)
def _validate_web_sso_token(decoded, system_features, app_code):
app_web_sso_enabled = False
# Check if SSO is enforced for web, and if the token source is not SSO, raise an error and redirect to SSO login
if system_features.sso_enforced_for_web:
app_web_sso_enabled = EnterpriseService.get_app_web_sso_enabled(app_code).get("enabled", False)
if app_web_sso_enabled:
def _validate_webapp_token(decoded, app_web_auth_enabled: bool, system_webapp_auth_enabled: bool):
# Check if authentication is enforced for web app, and if the token source is not webapp,
# raise an error and redirect to login
if system_webapp_auth_enabled and app_web_auth_enabled:
source = decoded.get("token_source")
if not source or source != "sso":
raise WebSSOAuthRequiredError()
if not source or source != "webapp":
raise WebAppAuthRequiredError()
# Check if SSO is not enforced for web, and if the token source is SSO,
# Check if authentication is not enforced for web, and if the token source is webapp,
# raise an error and redirect to normal passport login
if not system_features.sso_enforced_for_web or not app_web_sso_enabled:
if not system_webapp_auth_enabled or not app_web_auth_enabled:
source = decoded.get("token_source")
if source and source == "sso":
raise Unauthorized("sso token expired.")
if source and source == "webapp":
raise Unauthorized("webapp token expired.")
def _validate_user_accessibility(decoded, app_code, app_web_auth_enabled: bool, system_webapp_auth_enabled: bool):
if system_webapp_auth_enabled and app_web_auth_enabled:
# Check if the user is allowed to access the web app
user_id = decoded.get("user_id")
if not user_id:
raise WebAppAuthRequiredError()
if not EnterpriseService.WebAppAuth.is_user_allowed_to_access_webapp(user_id, app_code=app_code):
raise WebAppAuthAccessDeniedError()
class WebApiResource(Resource):

2
api/core/agent/base_agent_runner.py
View File

@ -91,6 +91,8 @@ class BaseAgentRunner(AppRunner):
return_resource=app_config.additional_features.show_retrieve_source,
invoke_from=application_generate_entity.invoke_from,
hit_callback=hit_callback,
user_id=user_id,
inputs=cast(dict, application_generate_entity.inputs),
)
# get how many agent thoughts have been created
self.agent_thought_count = (

8
api/core/agent/cot_agent_runner.py
View File

@ -69,13 +69,6 @@ class CotAgentRunner(BaseAgentRunner, ABC):
tool_instances, prompt_messages_tools = self._init_prompt_tools()
self._prompt_messages_tools = prompt_messages_tools
# fix metadata filter not work
if app_config.dataset is not None:
metadata_filtering_conditions = app_config.dataset.retrieve_config.metadata_filtering_conditions
for key, dataset_retriever_tool in tool_instances.items():
if hasattr(dataset_retriever_tool, "retrieval_tool"):
dataset_retriever_tool.retrieval_tool.metadata_filtering_conditions = metadata_filtering_conditions
function_call_state = True
llm_usage: dict[str, Optional[LLMUsage]] = {"usage": None}
final_answer = ""
@ -87,6 +80,7 @@ class CotAgentRunner(BaseAgentRunner, ABC):
llm_usage = final_llm_usage_dict["usage"]
llm_usage.prompt_tokens += usage.prompt_tokens
llm_usage.completion_tokens += usage.completion_tokens
llm_usage.total_tokens += usage.total_tokens
llm_usage.prompt_price += usage.prompt_price
llm_usage.completion_price += usage.completion_price
llm_usage.total_price += usage.total_price

8
api/core/agent/fc_agent_runner.py
View File

@ -45,13 +45,6 @@ class FunctionCallAgentRunner(BaseAgentRunner):
# convert tools into ModelRuntime Tool format
tool_instances, prompt_messages_tools = self._init_prompt_tools()
# fix metadata filter not work
if app_config.dataset is not None:
metadata_filtering_conditions = app_config.dataset.retrieve_config.metadata_filtering_conditions
for key, dataset_retriever_tool in tool_instances.items():
if hasattr(dataset_retriever_tool, "retrieval_tool"):
dataset_retriever_tool.retrieval_tool.metadata_filtering_conditions = metadata_filtering_conditions
assert app_config.agent
iteration_step = 1
@ -72,6 +65,7 @@ class FunctionCallAgentRunner(BaseAgentRunner):
llm_usage = final_llm_usage_dict["usage"]
llm_usage.prompt_tokens += usage.prompt_tokens
llm_usage.completion_tokens += usage.completion_tokens
llm_usage.total_tokens += usage.total_tokens
llm_usage.prompt_price += usage.prompt_price
llm_usage.completion_price += usage.completion_price
llm_usage.total_price += usage.total_price

1
api/core/app/app_config/entities.py
View File

@ -109,6 +109,7 @@ class VariableEntity(BaseModel):
description: str = ""
type: VariableEntityType
required: bool = False
hide: bool = False
max_length: Optional[int] = None
options: Sequence[str] = Field(default_factory=list)
allowed_file_types: Sequence[FileType] = Field(default_factory=list)

137
api/core/app/apps/advanced_chat/app_generator.py
View File

@ -5,7 +5,7 @@ import uuid
from collections.abc import Generator, Mapping
from typing import Any, Literal, Optional, Union, overload
from flask import Flask, current_app
from flask import Flask, copy_current_request_context, current_app, has_request_context
from pydantic import ValidationError
from sqlalchemy.orm import sessionmaker
@ -25,13 +25,14 @@ from core.app.entities.task_entities import ChatbotAppBlockingResponse, ChatbotA
from core.model_runtime.errors.invoke import InvokeAuthorizationError
from core.ops.ops_trace_manager import TraceQueueManager
from core.prompt.utils.get_thread_messages_length import get_thread_messages_length
from core.workflow.repository import RepositoryFactory
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.repositories.sqlalchemy_workflow_execution_repository import SQLAlchemyWorkflowExecutionRepository
from core.workflow.repository.workflow_execution_repository import WorkflowExecutionRepository
from core.workflow.repository.workflow_node_execution_repository import WorkflowNodeExecutionRepository
from extensions.ext_database import db
from factories import file_factory
from models.account import Account
from models.model import App, Conversation, EndUser, Message
from models.workflow import Workflow
from models import Account, App, Conversation, EndUser, Message, Workflow, WorkflowNodeExecutionTriggeredFrom
from models.enums import WorkflowRunTriggeredFrom
from services.conversation_service import ConversationService
from services.errors.message import MessageNotExistsError
@ -157,18 +158,30 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
trace_manager=trace_manager,
workflow_run_id=workflow_run_id,
)
contexts.tenant_id.set(application_generate_entity.app_config.tenant_id)
contexts.plugin_tool_providers.set({})
contexts.plugin_tool_providers_lock.set(threading.Lock())
# Create workflow node execution repository
# Create repositories
#
# Create session factory
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": application_generate_entity.app_config.tenant_id,
"app_id": application_generate_entity.app_config.app_id,
"session_factory": session_factory,
}
# Create workflow execution(aka workflow run) repository
if invoke_from == InvokeFrom.DEBUGGER:
workflow_triggered_from = WorkflowRunTriggeredFrom.DEBUGGING
else:
workflow_triggered_from = WorkflowRunTriggeredFrom.APP_RUN
workflow_execution_repository = SQLAlchemyWorkflowExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=workflow_triggered_from,
)
# Create workflow node execution repository
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
return self._generate(
@ -176,6 +189,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
user=user,
invoke_from=invoke_from,
application_generate_entity=application_generate_entity,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
conversation=conversation,
stream=streaming,
@ -225,18 +239,26 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
node_id=node_id, inputs=args["inputs"]
),
)
contexts.tenant_id.set(application_generate_entity.app_config.tenant_id)
contexts.plugin_tool_providers.set({})
contexts.plugin_tool_providers_lock.set(threading.Lock())
# Create workflow node execution repository
# Create repositories
#
# Create session factory
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": application_generate_entity.app_config.tenant_id,
"app_id": application_generate_entity.app_config.app_id,
"session_factory": session_factory,
}
# Create workflow execution(aka workflow run) repository
workflow_execution_repository = SQLAlchemyWorkflowExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowRunTriggeredFrom.DEBUGGING,
)
# Create workflow node execution repository
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowNodeExecutionTriggeredFrom.SINGLE_STEP,
)
return self._generate(
@ -244,6 +266,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
user=user,
invoke_from=InvokeFrom.DEBUGGER,
application_generate_entity=application_generate_entity,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
conversation=None,
stream=streaming,
@ -291,18 +314,26 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
extras={"auto_generate_conversation_name": False},
single_loop_run=AdvancedChatAppGenerateEntity.SingleLoopRunEntity(node_id=node_id, inputs=args["inputs"]),
)
contexts.tenant_id.set(application_generate_entity.app_config.tenant_id)
contexts.plugin_tool_providers.set({})
contexts.plugin_tool_providers_lock.set(threading.Lock())
# Create workflow node execution repository
# Create repositories
#
# Create session factory
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": application_generate_entity.app_config.tenant_id,
"app_id": application_generate_entity.app_config.app_id,
"session_factory": session_factory,
}
# Create workflow execution(aka workflow run) repository
workflow_execution_repository = SQLAlchemyWorkflowExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowRunTriggeredFrom.DEBUGGING,
)
# Create workflow node execution repository
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowNodeExecutionTriggeredFrom.SINGLE_STEP,
)
return self._generate(
@ -310,6 +341,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
user=user,
invoke_from=InvokeFrom.DEBUGGER,
application_generate_entity=application_generate_entity,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
conversation=None,
stream=streaming,
@ -322,6 +354,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
user: Union[Account, EndUser],
invoke_from: InvokeFrom,
application_generate_entity: AdvancedChatAppGenerateEntity,
workflow_execution_repository: WorkflowExecutionRepository,
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
conversation: Optional[Conversation] = None,
stream: bool = True,
@ -363,19 +396,24 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
message_id=message.id,
)
# new thread
worker_thread = threading.Thread(
target=self._generate_worker,
kwargs={
"flask_app": current_app._get_current_object(), # type: ignore
"application_generate_entity": application_generate_entity,
"queue_manager": queue_manager,
"conversation_id": conversation.id,
"message_id": message.id,
"context": contextvars.copy_context(),
},
# new thread with request context and contextvars
context = contextvars.copy_context()
@copy_current_request_context
def worker_with_context():
# Run the worker within the copied context
return context.run(
self._generate_worker,
flask_app=current_app._get_current_object(), # type: ignore
application_generate_entity=application_generate_entity,
queue_manager=queue_manager,
conversation_id=conversation.id,
message_id=message.id,
context=context,
)
worker_thread = threading.Thread(target=worker_with_context)
worker_thread.start()
# return response or stream generator
@ -386,6 +424,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
conversation=conversation,
message=message,
user=user,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
stream=stream,
)
@ -412,8 +451,22 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
"""
for var, val in context.items():
var.set(val)
# FIXME(-LAN-): Save current user before entering new app context
from flask import g
saved_user = None
if has_request_context() and hasattr(g, "_login_user"):
saved_user = g._login_user
with flask_app.app_context():
try:
# Restore user in new app context
if saved_user is not None:
from flask import g
g._login_user = saved_user
# get conversation and message
conversation = self._get_conversation(conversation_id)
message = self._get_message(message_id)
@ -458,6 +511,7 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
conversation: Conversation,
message: Message,
user: Union[Account, EndUser],
workflow_execution_repository: WorkflowExecutionRepository,
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
stream: bool = False,
) -> Union[ChatbotAppBlockingResponse, Generator[ChatbotAppStreamResponse, None, None]]:
@ -481,9 +535,10 @@ class AdvancedChatAppGenerator(MessageBasedAppGenerator):
conversation=conversation,
message=message,
user=user,
stream=stream,
dialogue_count=self._dialogue_count,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
stream=stream,
)
try:

188
api/core/app/apps/advanced_chat/generate_task_pipeline.py
View File

@ -9,8 +9,8 @@ from sqlalchemy import select
from sqlalchemy.orm import Session
from constants.tts_auto_play_timeout import TTS_AUTO_PLAY_TIMEOUT, TTS_AUTO_PLAY_YIELD_CPU_TIME
from core.app.apps.advanced_chat.app_generator_tts_publisher import AppGeneratorTTSPublisher, AudioTrunk
from core.app.apps.base_app_queue_manager import AppQueueManager, PublishFrom
from core.app.apps.common.workflow_response_converter import WorkflowResponseConverter
from core.app.entities.app_invoke_entities import (
AdvancedChatAppGenerateEntity,
InvokeFrom,
@ -58,19 +58,21 @@ from core.app.entities.task_entities import (
)
from core.app.task_pipeline.based_generate_task_pipeline import BasedGenerateTaskPipeline
from core.app.task_pipeline.message_cycle_manage import MessageCycleManage
from core.app.task_pipeline.workflow_cycle_manage import WorkflowCycleManage
from core.base.tts import AppGeneratorTTSPublisher, AudioTrunk
from core.model_runtime.entities.llm_entities import LLMUsage
from core.model_runtime.utils.encoders import jsonable_encoder
from core.ops.ops_trace_manager import TraceQueueManager
from core.workflow.enums import SystemVariableKey
from core.workflow.graph_engine.entities.graph_runtime_state import GraphRuntimeState
from core.workflow.nodes import NodeType
from core.workflow.repository.workflow_execution_repository import WorkflowExecutionRepository
from core.workflow.repository.workflow_node_execution_repository import WorkflowNodeExecutionRepository
from core.workflow.workflow_cycle_manager import WorkflowCycleManager
from events.message_event import message_was_created
from extensions.ext_database import db
from models import Conversation, EndUser, Message, MessageFile
from models.account import Account
from models.enums import CreatedByRole
from models.enums import CreatorUserRole
from models.workflow import (
Workflow,
WorkflowRunStatus,
@ -94,6 +96,7 @@ class AdvancedChatAppGenerateTaskPipeline:
user: Union[Account, EndUser],
stream: bool,
dialogue_count: int,
workflow_execution_repository: WorkflowExecutionRepository,
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
) -> None:
self._base_task_pipeline = BasedGenerateTaskPipeline(
@ -105,15 +108,15 @@ class AdvancedChatAppGenerateTaskPipeline:
if isinstance(user, EndUser):
self._user_id = user.id
user_session_id = user.session_id
self._created_by_role = CreatedByRole.END_USER
self._created_by_role = CreatorUserRole.END_USER
elif isinstance(user, Account):
self._user_id = user.id
user_session_id = user.id
self._created_by_role = CreatedByRole.ACCOUNT
self._created_by_role = CreatorUserRole.ACCOUNT
else:
raise NotImplementedError(f"User type not supported: {type(user)}")
self._workflow_cycle_manager = WorkflowCycleManage(
self._workflow_cycle_manager = WorkflowCycleManager(
application_generate_entity=application_generate_entity,
workflow_system_variables={
SystemVariableKey.QUERY: message.query,
@ -125,9 +128,14 @@ class AdvancedChatAppGenerateTaskPipeline:
SystemVariableKey.WORKFLOW_ID: workflow.id,
SystemVariableKey.WORKFLOW_RUN_ID: application_generate_entity.workflow_run_id,
},
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
)
self._workflow_response_converter = WorkflowResponseConverter(
application_generate_entity=application_generate_entity,
)
self._task_state = WorkflowTaskState()
self._message_cycle_manager = MessageCycleManage(
application_generate_entity=application_generate_entity, task_state=self._task_state
@ -294,19 +302,18 @@ class AdvancedChatAppGenerateTaskPipeline:
with Session(db.engine, expire_on_commit=False) as session:
# init workflow run
workflow_run = self._workflow_cycle_manager._handle_workflow_run_start(
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_start(
session=session,
workflow_id=self._workflow_id,
user_id=self._user_id,
created_by_role=self._created_by_role,
)
self._workflow_run_id = workflow_run.id
self._workflow_run_id = workflow_execution.id
message = self._get_message(session=session)
if not message:
raise ValueError(f"Message not found: {self._message_id}")
message.workflow_run_id = workflow_run.id
workflow_start_resp = self._workflow_cycle_manager._workflow_start_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
message.workflow_run_id = workflow_execution.id
workflow_start_resp = self._workflow_response_converter.workflow_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
session.commit()
@ -319,13 +326,10 @@ class AdvancedChatAppGenerateTaskPipeline:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
workflow_node_execution = self._workflow_cycle_manager.handle_workflow_node_execution_retried(
workflow_execution_id=self._workflow_run_id, event=event
)
workflow_node_execution = self._workflow_cycle_manager._handle_workflow_node_execution_retried(
workflow_run=workflow_run, event=event
)
node_retry_resp = self._workflow_cycle_manager._workflow_node_retry_to_stream_response(
node_retry_resp = self._workflow_response_converter.workflow_node_retry_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
@ -338,20 +342,15 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
workflow_node_execution = self._workflow_cycle_manager._handle_node_execution_start(
workflow_run=workflow_run, event=event
workflow_node_execution = self._workflow_cycle_manager.handle_node_execution_start(
workflow_execution_id=self._workflow_run_id, event=event
)
node_start_resp = self._workflow_cycle_manager._workflow_node_start_to_stream_response(
node_start_resp = self._workflow_response_converter.workflow_node_start_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
)
session.commit()
if node_start_resp:
yield node_start_resp
@ -359,15 +358,15 @@ class AdvancedChatAppGenerateTaskPipeline:
# Record files if it's an answer node or end node
if event.node_type in [NodeType.ANSWER, NodeType.END]:
self._recorded_files.extend(
self._workflow_cycle_manager._fetch_files_from_node_outputs(event.outputs or {})
self._workflow_response_converter.fetch_files_from_node_outputs(event.outputs or {})
)
with Session(db.engine, expire_on_commit=False) as session:
workflow_node_execution = self._workflow_cycle_manager._handle_workflow_node_execution_success(
workflow_node_execution = self._workflow_cycle_manager.handle_workflow_node_execution_success(
event=event
)
node_finish_resp = self._workflow_cycle_manager._workflow_node_finish_to_stream_response(
node_finish_resp = self._workflow_response_converter.workflow_node_finish_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
@ -383,11 +382,11 @@ class AdvancedChatAppGenerateTaskPipeline:
| QueueNodeInLoopFailedEvent
| QueueNodeExceptionEvent,
):
workflow_node_execution = self._workflow_cycle_manager._handle_workflow_node_execution_failed(
workflow_node_execution = self._workflow_cycle_manager.handle_workflow_node_execution_failed(
event=event
)
node_finish_resp = self._workflow_cycle_manager._workflow_node_finish_to_stream_response(
node_finish_resp = self._workflow_response_converter.workflow_node_finish_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
@ -399,15 +398,10 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
parallel_start_resp = (
self._workflow_cycle_manager._workflow_parallel_branch_start_to_stream_response(
session=session,
self._workflow_response_converter.workflow_parallel_branch_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
)
@ -417,15 +411,10 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
parallel_finish_resp = (
self._workflow_cycle_manager._workflow_parallel_branch_finished_to_stream_response(
session=session,
self._workflow_response_converter.workflow_parallel_branch_finished_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
)
@ -435,14 +424,9 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
iter_start_resp = self._workflow_cycle_manager._workflow_iteration_start_to_stream_response(
session=session,
iter_start_resp = self._workflow_response_converter.workflow_iteration_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -451,14 +435,9 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
iter_next_resp = self._workflow_cycle_manager._workflow_iteration_next_to_stream_response(
session=session,
iter_next_resp = self._workflow_response_converter.workflow_iteration_next_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -467,14 +446,9 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
iter_finish_resp = self._workflow_cycle_manager._workflow_iteration_completed_to_stream_response(
session=session,
iter_finish_resp = self._workflow_response_converter.workflow_iteration_completed_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -483,14 +457,9 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
loop_start_resp = self._workflow_cycle_manager._workflow_loop_start_to_stream_response(
session=session,
loop_start_resp = self._workflow_response_converter.workflow_loop_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -499,14 +468,9 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
loop_next_resp = self._workflow_cycle_manager._workflow_loop_next_to_stream_response(
session=session,
loop_next_resp = self._workflow_response_converter.workflow_loop_next_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -515,14 +479,9 @@ class AdvancedChatAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
loop_finish_resp = self._workflow_cycle_manager._workflow_loop_completed_to_stream_response(
session=session,
loop_finish_resp = self._workflow_response_converter.workflow_loop_completed_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -535,10 +494,8 @@ class AdvancedChatAppGenerateTaskPipeline:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_success(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_success(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
outputs=event.outputs,
@ -546,10 +503,11 @@ class AdvancedChatAppGenerateTaskPipeline:
trace_manager=trace_manager,
)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
session.commit()
yield workflow_finish_resp
self._base_task_pipeline._queue_manager.publish(
@ -562,10 +520,8 @@ class AdvancedChatAppGenerateTaskPipeline:
raise ValueError("graph runtime state not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_partial_success(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_partial_success(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
outputs=event.outputs,
@ -573,10 +529,11 @@ class AdvancedChatAppGenerateTaskPipeline:
conversation_id=None,
trace_manager=trace_manager,
)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
session.commit()
yield workflow_finish_resp
self._base_task_pipeline._queue_manager.publish(
@ -589,26 +546,25 @@ class AdvancedChatAppGenerateTaskPipeline:
raise ValueError("graph runtime state not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_failed(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_failed(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
status=WorkflowRunStatus.FAILED,
error=event.error,
error_message=event.error,
conversation_id=self._conversation_id,
trace_manager=trace_manager,
exceptions_count=event.exceptions_count,
)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
err_event = QueueErrorEvent(error=ValueError(f"Run failed: {workflow_run.error}"))
err_event = QueueErrorEvent(error=ValueError(f"Run failed: {workflow_execution.error_message}"))
err = self._base_task_pipeline._handle_error(
event=err_event, session=session, message_id=self._message_id
)
session.commit()
yield workflow_finish_resp
yield self._base_task_pipeline._error_to_stream_response(err)
@ -616,21 +572,19 @@ class AdvancedChatAppGenerateTaskPipeline:
elif isinstance(event, QueueStopEvent):
if self._workflow_run_id and graph_runtime_state:
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_failed(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_failed(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
status=WorkflowRunStatus.STOPPED,
error=event.get_stop_reason(),
error_message=event.get_stop_reason(),
conversation_id=self._conversation_id,
trace_manager=trace_manager,
)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution=workflow_execution,
)
# Save message
self._save_message(session=session, graph_runtime_state=graph_runtime_state)
@ -711,7 +665,7 @@ class AdvancedChatAppGenerateTaskPipeline:
yield self._message_end_to_stream_response()
elif isinstance(event, QueueAgentLogEvent):
yield self._workflow_cycle_manager._handle_agent_log(
yield self._workflow_response_converter.handle_agent_log(
task_id=self._application_generate_entity.task_id, event=event
)
else:
@ -739,9 +693,9 @@ class AdvancedChatAppGenerateTaskPipeline:
url=file["remote_url"],
belongs_to="assistant",
upload_file_id=file["related_id"],
created_by_role=CreatedByRole.ACCOUNT
created_by_role=CreatorUserRole.ACCOUNT
if message.invoke_from in {InvokeFrom.EXPLORE, InvokeFrom.DEBUGGER}
else CreatedByRole.END_USER,
else CreatorUserRole.END_USER,
created_by=message.from_account_id or message.from_end_user_id or "",
)
for file in self._recorded_files

42
api/core/app/apps/agent_chat/app_generator.py
View File

@ -5,7 +5,7 @@ import uuid
from collections.abc import Generator, Mapping
from typing import Any, Literal, Union, overload
from flask import Flask, current_app
from flask import Flask, copy_current_request_context, current_app, has_request_context
from pydantic import ValidationError
from configs import dify_config
@ -179,19 +179,24 @@ class AgentChatAppGenerator(MessageBasedAppGenerator):
message_id=message.id,
)
# new thread
worker_thread = threading.Thread(
target=self._generate_worker,
kwargs={
"flask_app": current_app._get_current_object(), # type: ignore
"context": contextvars.copy_context(),
"application_generate_entity": application_generate_entity,
"queue_manager": queue_manager,
"conversation_id": conversation.id,
"message_id": message.id,
},
# new thread with request context and contextvars
context = contextvars.copy_context()
@copy_current_request_context
def worker_with_context():
# Run the worker within the copied context
return context.run(
self._generate_worker,
flask_app=current_app._get_current_object(), # type: ignore
context=context,
application_generate_entity=application_generate_entity,
queue_manager=queue_manager,
conversation_id=conversation.id,
message_id=message.id,
)
worker_thread = threading.Thread(target=worker_with_context)
worker_thread.start()
# return response or stream generator
@ -227,8 +232,21 @@ class AgentChatAppGenerator(MessageBasedAppGenerator):
for var, val in context.items():
var.set(val)
# FIXME(-LAN-): Save current user before entering new app context
from flask import g
saved_user = None
if has_request_context() and hasattr(g, "_login_user"):
saved_user = g._login_user
with flask_app.app_context():
try:
# Restore user in new app context
if saved_user is not None:
from flask import g
g._login_user = saved_user
# get conversation and message
conversation = self._get_conversation(conversation_id)
message = self._get_message(message_id)

23
api/core/app/apps/chat/app_generator.py
View File

@ -4,7 +4,7 @@ import uuid
from collections.abc import Generator, Mapping
from typing import Any, Literal, Union, overload
from flask import Flask, current_app
from flask import Flask, copy_current_request_context, current_app
from pydantic import ValidationError
from configs import dify_config
@ -170,18 +170,19 @@ class ChatAppGenerator(MessageBasedAppGenerator):
message_id=message.id,
)
# new thread
worker_thread = threading.Thread(
target=self._generate_worker,
kwargs={
"flask_app": current_app._get_current_object(), # type: ignore
"application_generate_entity": application_generate_entity,
"queue_manager": queue_manager,
"conversation_id": conversation.id,
"message_id": message.id,
},
# new thread with request context
@copy_current_request_context
def worker_with_context():
return self._generate_worker(
flask_app=current_app._get_current_object(), # type: ignore
application_generate_entity=application_generate_entity,
queue_manager=queue_manager,
conversation_id=conversation.id,
message_id=message.id,
)
worker_thread = threading.Thread(target=worker_with_context)
worker_thread.start()
# return response or stream generator

0
api/core/app/apps/common/__init__.py Normal file
View File

564
api/core/app/apps/common/workflow_response_converter.py Normal file
View File

@ -0,0 +1,564 @@
import time
from collections.abc import Mapping, Sequence
from datetime import UTC, datetime
from typing import Any, Optional, Union, cast
from sqlalchemy import select
from sqlalchemy.orm import Session
from core.app.entities.app_invoke_entities import AdvancedChatAppGenerateEntity, WorkflowAppGenerateEntity
from core.app.entities.queue_entities import (
QueueAgentLogEvent,
QueueIterationCompletedEvent,
QueueIterationNextEvent,
QueueIterationStartEvent,
QueueLoopCompletedEvent,
QueueLoopNextEvent,
QueueLoopStartEvent,
QueueNodeExceptionEvent,
QueueNodeFailedEvent,
QueueNodeInIterationFailedEvent,
QueueNodeInLoopFailedEvent,
QueueNodeRetryEvent,
QueueNodeStartedEvent,
QueueNodeSucceededEvent,
QueueParallelBranchRunFailedEvent,
QueueParallelBranchRunStartedEvent,
QueueParallelBranchRunSucceededEvent,
)
from core.app.entities.task_entities import (
AgentLogStreamResponse,
IterationNodeCompletedStreamResponse,
IterationNodeNextStreamResponse,
IterationNodeStartStreamResponse,
LoopNodeCompletedStreamResponse,
LoopNodeNextStreamResponse,
LoopNodeStartStreamResponse,
NodeFinishStreamResponse,
NodeRetryStreamResponse,
NodeStartStreamResponse,
ParallelBranchFinishedStreamResponse,
ParallelBranchStartStreamResponse,
WorkflowFinishStreamResponse,
WorkflowStartStreamResponse,
)
from core.file import FILE_MODEL_IDENTITY, File
from core.tools.tool_manager import ToolManager
from core.workflow.entities.node_execution_entities import NodeExecution
from core.workflow.entities.workflow_execution_entities import WorkflowExecution
from core.workflow.nodes import NodeType
from core.workflow.nodes.tool.entities import ToolNodeData
from models import (
Account,
CreatorUserRole,
EndUser,
WorkflowNodeExecutionStatus,
WorkflowRun,
)
class WorkflowResponseConverter:
def __init__(
self,
*,
application_generate_entity: Union[AdvancedChatAppGenerateEntity, WorkflowAppGenerateEntity],
) -> None:
self._application_generate_entity = application_generate_entity
def workflow_start_to_stream_response(
self,
*,
task_id: str,
workflow_execution: WorkflowExecution,
) -> WorkflowStartStreamResponse:
return WorkflowStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution.id,
data=WorkflowStartStreamResponse.Data(
id=workflow_execution.id,
workflow_id=workflow_execution.workflow_id,
sequence_number=workflow_execution.sequence_number,
inputs=workflow_execution.inputs,
created_at=int(workflow_execution.started_at.timestamp()),
),
)
def workflow_finish_to_stream_response(
self,
*,
session: Session,
task_id: str,
workflow_execution: WorkflowExecution,
) -> WorkflowFinishStreamResponse:
created_by = None
workflow_run = session.scalar(select(WorkflowRun).where(WorkflowRun.id == workflow_execution.id))
assert workflow_run is not None
if workflow_run.created_by_role == CreatorUserRole.ACCOUNT:
stmt = select(Account).where(Account.id == workflow_run.created_by)
account = session.scalar(stmt)
if account:
created_by = {
"id": account.id,
"name": account.name,
"email": account.email,
}
elif workflow_run.created_by_role == CreatorUserRole.END_USER:
stmt = select(EndUser).where(EndUser.id == workflow_run.created_by)
end_user = session.scalar(stmt)
if end_user:
created_by = {
"id": end_user.id,
"user": end_user.session_id,
}
else:
raise NotImplementedError(f"unknown created_by_role: {workflow_run.created_by_role}")
# Handle the case where finished_at is None by using current time as default
finished_at_timestamp = (
int(workflow_execution.finished_at.timestamp())
if workflow_execution.finished_at
else int(datetime.now(UTC).timestamp())
)
return WorkflowFinishStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution.id,
data=WorkflowFinishStreamResponse.Data(
id=workflow_execution.id,
workflow_id=workflow_execution.workflow_id,
sequence_number=workflow_execution.sequence_number,
status=workflow_execution.status,
outputs=workflow_execution.outputs,
error=workflow_execution.error_message,
elapsed_time=workflow_execution.elapsed_time,
total_tokens=workflow_execution.total_tokens,
total_steps=workflow_execution.total_steps,
created_by=created_by,
created_at=int(workflow_execution.started_at.timestamp()),
finished_at=finished_at_timestamp,
files=self.fetch_files_from_node_outputs(workflow_execution.outputs),
exceptions_count=workflow_execution.exceptions_count,
),
)
def workflow_node_start_to_stream_response(
self,
*,
event: QueueNodeStartedEvent,
task_id: str,
workflow_node_execution: NodeExecution,
) -> Optional[NodeStartStreamResponse]:
if workflow_node_execution.node_type in {NodeType.ITERATION, NodeType.LOOP}:
return None
if not workflow_node_execution.workflow_run_id:
return None
response = NodeStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_node_execution.workflow_run_id,
data=NodeStartStreamResponse.Data(
id=workflow_node_execution.id,
node_id=workflow_node_execution.node_id,
node_type=workflow_node_execution.node_type,
title=workflow_node_execution.title,
index=workflow_node_execution.index,
predecessor_node_id=workflow_node_execution.predecessor_node_id,
inputs=workflow_node_execution.inputs,
created_at=int(workflow_node_execution.created_at.timestamp()),
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
parallel_run_id=event.parallel_mode_run_id,
agent_strategy=event.agent_strategy,
),
)
# extras logic
if event.node_type == NodeType.TOOL:
node_data = cast(ToolNodeData, event.node_data)
response.data.extras["icon"] = ToolManager.get_tool_icon(
tenant_id=self._application_generate_entity.app_config.tenant_id,
provider_type=node_data.provider_type,
provider_id=node_data.provider_id,
)
return response
def workflow_node_finish_to_stream_response(
self,
*,
event: QueueNodeSucceededEvent
| QueueNodeFailedEvent
| QueueNodeInIterationFailedEvent
| QueueNodeInLoopFailedEvent
| QueueNodeExceptionEvent,
task_id: str,
workflow_node_execution: NodeExecution,
) -> Optional[NodeFinishStreamResponse]:
if workflow_node_execution.node_type in {NodeType.ITERATION, NodeType.LOOP}:
return None
if not workflow_node_execution.workflow_run_id:
return None
if not workflow_node_execution.finished_at:
return None
return NodeFinishStreamResponse(
task_id=task_id,
workflow_run_id=workflow_node_execution.workflow_run_id,
data=NodeFinishStreamResponse.Data(
id=workflow_node_execution.id,
node_id=workflow_node_execution.node_id,
node_type=workflow_node_execution.node_type,
index=workflow_node_execution.index,
title=workflow_node_execution.title,
predecessor_node_id=workflow_node_execution.predecessor_node_id,
inputs=workflow_node_execution.inputs,
process_data=workflow_node_execution.process_data,
outputs=workflow_node_execution.outputs,
status=workflow_node_execution.status,
error=workflow_node_execution.error,
elapsed_time=workflow_node_execution.elapsed_time,
execution_metadata=workflow_node_execution.metadata,
created_at=int(workflow_node_execution.created_at.timestamp()),
finished_at=int(workflow_node_execution.finished_at.timestamp()),
files=self.fetch_files_from_node_outputs(workflow_node_execution.outputs or {}),
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
),
)
def workflow_node_retry_to_stream_response(
self,
*,
event: QueueNodeRetryEvent,
task_id: str,
workflow_node_execution: NodeExecution,
) -> Optional[Union[NodeRetryStreamResponse, NodeFinishStreamResponse]]:
if workflow_node_execution.node_type in {NodeType.ITERATION, NodeType.LOOP}:
return None
if not workflow_node_execution.workflow_run_id:
return None
if not workflow_node_execution.finished_at:
return None
return NodeRetryStreamResponse(
task_id=task_id,
workflow_run_id=workflow_node_execution.workflow_run_id,
data=NodeRetryStreamResponse.Data(
id=workflow_node_execution.id,
node_id=workflow_node_execution.node_id,
node_type=workflow_node_execution.node_type,
index=workflow_node_execution.index,
title=workflow_node_execution.title,
predecessor_node_id=workflow_node_execution.predecessor_node_id,
inputs=workflow_node_execution.inputs,
process_data=workflow_node_execution.process_data,
outputs=workflow_node_execution.outputs,
status=workflow_node_execution.status,
error=workflow_node_execution.error,
elapsed_time=workflow_node_execution.elapsed_time,
execution_metadata=workflow_node_execution.metadata,
created_at=int(workflow_node_execution.created_at.timestamp()),
finished_at=int(workflow_node_execution.finished_at.timestamp()),
files=self.fetch_files_from_node_outputs(workflow_node_execution.outputs or {}),
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
retry_index=event.retry_index,
),
)
def workflow_parallel_branch_start_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueParallelBranchRunStartedEvent,
) -> ParallelBranchStartStreamResponse:
return ParallelBranchStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=ParallelBranchStartStreamResponse.Data(
parallel_id=event.parallel_id,
parallel_branch_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
created_at=int(time.time()),
),
)
def workflow_parallel_branch_finished_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueParallelBranchRunSucceededEvent | QueueParallelBranchRunFailedEvent,
) -> ParallelBranchFinishedStreamResponse:
return ParallelBranchFinishedStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=ParallelBranchFinishedStreamResponse.Data(
parallel_id=event.parallel_id,
parallel_branch_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
status="succeeded" if isinstance(event, QueueParallelBranchRunSucceededEvent) else "failed",
error=event.error if isinstance(event, QueueParallelBranchRunFailedEvent) else None,
created_at=int(time.time()),
),
)
def workflow_iteration_start_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueIterationStartEvent,
) -> IterationNodeStartStreamResponse:
return IterationNodeStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=IterationNodeStartStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
metadata=event.metadata or {},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def workflow_iteration_next_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueIterationNextEvent,
) -> IterationNodeNextStreamResponse:
return IterationNodeNextStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=IterationNodeNextStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
index=event.index,
pre_iteration_output=event.output,
created_at=int(time.time()),
extras={},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parallel_mode_run_id=event.parallel_mode_run_id,
duration=event.duration,
),
)
def workflow_iteration_completed_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueIterationCompletedEvent,
) -> IterationNodeCompletedStreamResponse:
return IterationNodeCompletedStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=IterationNodeCompletedStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
outputs=event.outputs,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
status=WorkflowNodeExecutionStatus.SUCCEEDED
if event.error is None
else WorkflowNodeExecutionStatus.FAILED,
error=None,
elapsed_time=(datetime.now(UTC).replace(tzinfo=None) - event.start_at).total_seconds(),
total_tokens=event.metadata.get("total_tokens", 0) if event.metadata else 0,
execution_metadata=event.metadata,
finished_at=int(time.time()),
steps=event.steps,
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def workflow_loop_start_to_stream_response(
self, *, task_id: str, workflow_execution_id: str, event: QueueLoopStartEvent
) -> LoopNodeStartStreamResponse:
return LoopNodeStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=LoopNodeStartStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
metadata=event.metadata or {},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def workflow_loop_next_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueLoopNextEvent,
) -> LoopNodeNextStreamResponse:
return LoopNodeNextStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=LoopNodeNextStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
index=event.index,
pre_loop_output=event.output,
created_at=int(time.time()),
extras={},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parallel_mode_run_id=event.parallel_mode_run_id,
duration=event.duration,
),
)
def workflow_loop_completed_to_stream_response(
self,
*,
task_id: str,
workflow_execution_id: str,
event: QueueLoopCompletedEvent,
) -> LoopNodeCompletedStreamResponse:
return LoopNodeCompletedStreamResponse(
task_id=task_id,
workflow_run_id=workflow_execution_id,
data=LoopNodeCompletedStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
outputs=event.outputs,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
status=WorkflowNodeExecutionStatus.SUCCEEDED
if event.error is None
else WorkflowNodeExecutionStatus.FAILED,
error=None,
elapsed_time=(datetime.now(UTC).replace(tzinfo=None) - event.start_at).total_seconds(),
total_tokens=event.metadata.get("total_tokens", 0) if event.metadata else 0,
execution_metadata=event.metadata,
finished_at=int(time.time()),
steps=event.steps,
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def fetch_files_from_node_outputs(self, outputs_dict: Mapping[str, Any] | None) -> Sequence[Mapping[str, Any]]:
"""
Fetch files from node outputs
:param outputs_dict: node outputs dict
:return:
"""
if not outputs_dict:
return []
files = [self._fetch_files_from_variable_value(output_value) for output_value in outputs_dict.values()]
# Remove None
files = [file for file in files if file]
# Flatten list
# Flatten the list of sequences into a single list of mappings
flattened_files = [file for sublist in files if sublist for file in sublist]
# Convert to tuple to match Sequence type
return tuple(flattened_files)
def _fetch_files_from_variable_value(self, value: Union[dict, list]) -> Sequence[Mapping[str, Any]]:
"""
Fetch files from variable value
:param value: variable value
:return:
"""
if not value:
return []
files = []
if isinstance(value, list):
for item in value:
file = self._get_file_var_from_value(item)
if file:
files.append(file)
elif isinstance(value, dict):
file = self._get_file_var_from_value(value)
if file:
files.append(file)
return files
def _get_file_var_from_value(self, value: Union[dict, list]) -> Mapping[str, Any] | None:
"""
Get file var from value
:param value: variable value
:return:
"""
if not value:
return None
if isinstance(value, dict) and value.get("dify_model_identity") == FILE_MODEL_IDENTITY:
return value
elif isinstance(value, File):
return value.to_dict()
return None
def handle_agent_log(self, task_id: str, event: QueueAgentLogEvent) -> AgentLogStreamResponse:
"""
Handle agent log
:param task_id: task id
:param event: agent log event
:return:
"""
return AgentLogStreamResponse(
task_id=task_id,
data=AgentLogStreamResponse.Data(
node_execution_id=event.node_execution_id,
id=event.id,
parent_id=event.parent_id,
label=event.label,
error=event.error,
status=event.status,
data=event.data,
metadata=event.metadata,
node_id=event.node_id,
),
)

40
api/core/app/apps/completion/app_generator.py
View File

@ -4,7 +4,7 @@ import uuid
from collections.abc import Generator, Mapping
from typing import Any, Literal, Union, overload
from flask import Flask, current_app
from flask import Flask, copy_current_request_context, current_app
from pydantic import ValidationError
from configs import dify_config
@ -151,17 +151,18 @@ class CompletionAppGenerator(MessageBasedAppGenerator):
message_id=message.id,
)
# new thread
worker_thread = threading.Thread(
target=self._generate_worker,
kwargs={
"flask_app": current_app._get_current_object(), # type: ignore
"application_generate_entity": application_generate_entity,
"queue_manager": queue_manager,
"message_id": message.id,
},
# new thread with request context
@copy_current_request_context
def worker_with_context():
return self._generate_worker(
flask_app=current_app._get_current_object(), # type: ignore
application_generate_entity=application_generate_entity,
queue_manager=queue_manager,
message_id=message.id,
)
worker_thread = threading.Thread(target=worker_with_context)
worker_thread.start()
# return response or stream generator
@ -313,17 +314,18 @@ class CompletionAppGenerator(MessageBasedAppGenerator):
message_id=message.id,
)
# new thread
worker_thread = threading.Thread(
target=self._generate_worker,
kwargs={
"flask_app": current_app._get_current_object(), # type: ignore
"application_generate_entity": application_generate_entity,
"queue_manager": queue_manager,
"message_id": message.id,
},
# new thread with request context
@copy_current_request_context
def worker_with_context():
return self._generate_worker(
flask_app=current_app._get_current_object(), # type: ignore
application_generate_entity=application_generate_entity,
queue_manager=queue_manager,
message_id=message.id,
)
worker_thread = threading.Thread(target=worker_with_context)
worker_thread.start()
# return response or stream generator

4
api/core/app/apps/message_based_app_generator.py
View File

@ -25,7 +25,7 @@ from core.app.task_pipeline.easy_ui_based_generate_task_pipeline import EasyUIBa
from core.prompt.utils.prompt_template_parser import PromptTemplateParser
from extensions.ext_database import db
from models import Account
from models.enums import CreatedByRole
from models.enums import CreatorUserRole
from models.model import App, AppMode, AppModelConfig, Conversation, EndUser, Message, MessageFile
from services.errors.app_model_config import AppModelConfigBrokenError
from services.errors.conversation import ConversationNotExistsError
@ -223,7 +223,7 @@ class MessageBasedAppGenerator(BaseAppGenerator):
belongs_to="user",
url=file.remote_url,
upload_file_id=file.related_id,
created_by_role=(CreatedByRole.ACCOUNT if account_id else CreatedByRole.END_USER),
created_by_role=(CreatorUserRole.ACCOUNT if account_id else CreatorUserRole.END_USER),
created_by=account_id or end_user_id or "",
)
db.session.add(message_file)

133
api/core/app/apps/workflow/app_generator.py
View File

@ -5,7 +5,7 @@ import uuid
from collections.abc import Generator, Mapping, Sequence
from typing import Any, Literal, Optional, Union, overload
from flask import Flask, current_app
from flask import Flask, copy_current_request_context, current_app, has_request_context
from pydantic import ValidationError
from sqlalchemy.orm import sessionmaker
@ -23,11 +23,14 @@ from core.app.entities.app_invoke_entities import InvokeFrom, WorkflowAppGenerat
from core.app.entities.task_entities import WorkflowAppBlockingResponse, WorkflowAppStreamResponse
from core.model_runtime.errors.invoke import InvokeAuthorizationError
from core.ops.ops_trace_manager import TraceQueueManager
from core.workflow.repository import RepositoryFactory
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.repositories.sqlalchemy_workflow_execution_repository import SQLAlchemyWorkflowExecutionRepository
from core.workflow.repository.workflow_execution_repository import WorkflowExecutionRepository
from core.workflow.repository.workflow_node_execution_repository import WorkflowNodeExecutionRepository
from extensions.ext_database import db
from factories import file_factory
from models import Account, App, EndUser, Workflow
from models import Account, App, EndUser, Workflow, WorkflowNodeExecutionTriggeredFrom
from models.enums import WorkflowRunTriggeredFrom
logger = logging.getLogger(__name__)
@ -132,18 +135,30 @@ class WorkflowAppGenerator(BaseAppGenerator):
workflow_run_id=workflow_run_id,
)
contexts.tenant_id.set(application_generate_entity.app_config.tenant_id)
contexts.plugin_tool_providers.set({})
contexts.plugin_tool_providers_lock.set(threading.Lock())
# Create workflow node execution repository
# Create repositories
#
# Create session factory
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": application_generate_entity.app_config.tenant_id,
"app_id": application_generate_entity.app_config.app_id,
"session_factory": session_factory,
}
# Create workflow execution(aka workflow run) repository
if invoke_from == InvokeFrom.DEBUGGER:
workflow_triggered_from = WorkflowRunTriggeredFrom.DEBUGGING
else:
workflow_triggered_from = WorkflowRunTriggeredFrom.APP_RUN
workflow_execution_repository = SQLAlchemyWorkflowExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=workflow_triggered_from,
)
# Create workflow node execution repository
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
return self._generate(
@ -152,6 +167,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
user=user,
application_generate_entity=application_generate_entity,
invoke_from=invoke_from,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
streaming=streaming,
workflow_thread_pool_id=workflow_thread_pool_id,
@ -165,6 +181,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
user: Union[Account, EndUser],
application_generate_entity: WorkflowAppGenerateEntity,
invoke_from: InvokeFrom,
workflow_execution_repository: WorkflowExecutionRepository,
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
streaming: bool = True,
workflow_thread_pool_id: Optional[str] = None,
@ -189,18 +206,23 @@ class WorkflowAppGenerator(BaseAppGenerator):
app_mode=app_model.mode,
)
# new thread
worker_thread = threading.Thread(
target=self._generate_worker,
kwargs={
"flask_app": current_app._get_current_object(), # type: ignore
"application_generate_entity": application_generate_entity,
"queue_manager": queue_manager,
"context": contextvars.copy_context(),
"workflow_thread_pool_id": workflow_thread_pool_id,
},
# new thread with request context and contextvars
context = contextvars.copy_context()
@copy_current_request_context
def worker_with_context():
# Run the worker within the copied context
return context.run(
self._generate_worker,
flask_app=current_app._get_current_object(), # type: ignore
application_generate_entity=application_generate_entity,
queue_manager=queue_manager,
context=context,
workflow_thread_pool_id=workflow_thread_pool_id,
)
worker_thread = threading.Thread(target=worker_with_context)
worker_thread.start()
# return response or stream generator
@ -209,6 +231,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
workflow=workflow,
queue_manager=queue_manager,
user=user,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
stream=streaming,
)
@ -258,18 +281,28 @@ class WorkflowAppGenerator(BaseAppGenerator):
),
workflow_run_id=str(uuid.uuid4()),
)
contexts.tenant_id.set(application_generate_entity.app_config.tenant_id)
contexts.plugin_tool_providers.set({})
contexts.plugin_tool_providers_lock.set(threading.Lock())
# Create repositories
#
# Create session factory
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
# Create workflow execution(aka workflow run) repository
workflow_execution_repository = SQLAlchemyWorkflowExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowRunTriggeredFrom.DEBUGGING,
)
# Create workflow node execution repository
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": application_generate_entity.app_config.tenant_id,
"app_id": application_generate_entity.app_config.app_id,
"session_factory": session_factory,
}
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowNodeExecutionTriggeredFrom.SINGLE_STEP,
)
return self._generate(
@ -278,6 +311,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
user=user,
invoke_from=InvokeFrom.DEBUGGER,
application_generate_entity=application_generate_entity,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
streaming=streaming,
)
@ -323,18 +357,28 @@ class WorkflowAppGenerator(BaseAppGenerator):
single_loop_run=WorkflowAppGenerateEntity.SingleLoopRunEntity(node_id=node_id, inputs=args["inputs"]),
workflow_run_id=str(uuid.uuid4()),
)
contexts.tenant_id.set(application_generate_entity.app_config.tenant_id)
contexts.plugin_tool_providers.set({})
contexts.plugin_tool_providers_lock.set(threading.Lock())
# Create repositories
#
# Create session factory
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
# Create workflow execution(aka workflow run) repository
workflow_execution_repository = SQLAlchemyWorkflowExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowRunTriggeredFrom.DEBUGGING,
)
# Create workflow node execution repository
session_factory = sessionmaker(bind=db.engine, expire_on_commit=False)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": application_generate_entity.app_config.tenant_id,
"app_id": application_generate_entity.app_config.app_id,
"session_factory": session_factory,
}
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=user,
app_id=application_generate_entity.app_config.app_id,
triggered_from=WorkflowNodeExecutionTriggeredFrom.SINGLE_STEP,
)
return self._generate(
@ -343,6 +387,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
user=user,
invoke_from=InvokeFrom.DEBUGGER,
application_generate_entity=application_generate_entity,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
streaming=streaming,
)
@ -365,8 +410,22 @@ class WorkflowAppGenerator(BaseAppGenerator):
"""
for var, val in context.items():
var.set(val)
# FIXME(-LAN-): Save current user before entering new app context
from flask import g
saved_user = None
if has_request_context() and hasattr(g, "_login_user"):
saved_user = g._login_user
with flask_app.app_context():
try:
# Restore user in new app context
if saved_user is not None:
from flask import g
g._login_user = saved_user
# workflow app
runner = WorkflowAppRunner(
application_generate_entity=application_generate_entity,
@ -400,6 +459,7 @@ class WorkflowAppGenerator(BaseAppGenerator):
workflow: Workflow,
queue_manager: AppQueueManager,
user: Union[Account, EndUser],
workflow_execution_repository: WorkflowExecutionRepository,
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
stream: bool = False,
) -> Union[WorkflowAppBlockingResponse, Generator[WorkflowAppStreamResponse, None, None]]:
@ -419,8 +479,9 @@ class WorkflowAppGenerator(BaseAppGenerator):
workflow=workflow,
queue_manager=queue_manager,
user=user,
stream=stream,
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
stream=stream,
)
try:

183
api/core/app/apps/workflow/generate_task_pipeline.py
View File

@ -3,11 +3,12 @@ import time
from collections.abc import Generator
from typing import Optional, Union
from sqlalchemy import select
from sqlalchemy.orm import Session
from constants.tts_auto_play_timeout import TTS_AUTO_PLAY_TIMEOUT, TTS_AUTO_PLAY_YIELD_CPU_TIME
from core.app.apps.advanced_chat.app_generator_tts_publisher import AppGeneratorTTSPublisher, AudioTrunk
from core.app.apps.base_app_queue_manager import AppQueueManager
from core.app.apps.common.workflow_response_converter import WorkflowResponseConverter
from core.app.entities.app_invoke_entities import (
InvokeFrom,
WorkflowAppGenerateEntity,
@ -52,13 +53,16 @@ from core.app.entities.task_entities import (
WorkflowTaskState,
)
from core.app.task_pipeline.based_generate_task_pipeline import BasedGenerateTaskPipeline
from core.app.task_pipeline.workflow_cycle_manage import WorkflowCycleManage
from core.base.tts import AppGeneratorTTSPublisher, AudioTrunk
from core.ops.ops_trace_manager import TraceQueueManager
from core.workflow.entities.workflow_execution_entities import WorkflowExecution
from core.workflow.enums import SystemVariableKey
from core.workflow.repository.workflow_execution_repository import WorkflowExecutionRepository
from core.workflow.repository.workflow_node_execution_repository import WorkflowNodeExecutionRepository
from core.workflow.workflow_cycle_manager import WorkflowCycleManager
from extensions.ext_database import db
from models.account import Account
from models.enums import CreatedByRole
from models.enums import CreatorUserRole
from models.model import EndUser
from models.workflow import (
Workflow,
@ -83,6 +87,7 @@ class WorkflowAppGenerateTaskPipeline:
queue_manager: AppQueueManager,
user: Union[Account, EndUser],
stream: bool,
workflow_execution_repository: WorkflowExecutionRepository,
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
) -> None:
self._base_task_pipeline = BasedGenerateTaskPipeline(
@ -94,15 +99,15 @@ class WorkflowAppGenerateTaskPipeline:
if isinstance(user, EndUser):
self._user_id = user.id
user_session_id = user.session_id
self._created_by_role = CreatedByRole.END_USER
self._created_by_role = CreatorUserRole.END_USER
elif isinstance(user, Account):
self._user_id = user.id
user_session_id = user.id
self._created_by_role = CreatedByRole.ACCOUNT
self._created_by_role = CreatorUserRole.ACCOUNT
else:
raise ValueError(f"Invalid user type: {type(user)}")
self._workflow_cycle_manager = WorkflowCycleManage(
self._workflow_cycle_manager = WorkflowCycleManager(
application_generate_entity=application_generate_entity,
workflow_system_variables={
SystemVariableKey.FILES: application_generate_entity.files,
@ -111,9 +116,14 @@ class WorkflowAppGenerateTaskPipeline:
SystemVariableKey.WORKFLOW_ID: workflow.id,
SystemVariableKey.WORKFLOW_RUN_ID: application_generate_entity.workflow_run_id,
},
workflow_execution_repository=workflow_execution_repository,
workflow_node_execution_repository=workflow_node_execution_repository,
)
self._workflow_response_converter = WorkflowResponseConverter(
application_generate_entity=application_generate_entity,
)
self._application_generate_entity = application_generate_entity
self._workflow_id = workflow.id
self._workflow_features_dict = workflow.features_dict
@ -258,17 +268,15 @@ class WorkflowAppGenerateTaskPipeline:
with Session(db.engine, expire_on_commit=False) as session:
# init workflow run
workflow_run = self._workflow_cycle_manager._handle_workflow_run_start(
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_start(
session=session,
workflow_id=self._workflow_id,
user_id=self._user_id,
created_by_role=self._created_by_role,
)
self._workflow_run_id = workflow_run.id
start_resp = self._workflow_cycle_manager._workflow_start_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
self._workflow_run_id = workflow_execution.id
start_resp = self._workflow_response_converter.workflow_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
session.commit()
yield start_resp
elif isinstance(
@ -278,13 +286,11 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
workflow_node_execution = self._workflow_cycle_manager.handle_workflow_node_execution_retried(
workflow_execution_id=self._workflow_run_id,
event=event,
)
workflow_node_execution = self._workflow_cycle_manager._handle_workflow_node_execution_retried(
workflow_run=workflow_run, event=event
)
response = self._workflow_cycle_manager._workflow_node_retry_to_stream_response(
response = self._workflow_response_converter.workflow_node_retry_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
@ -297,27 +303,22 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
workflow_node_execution = self._workflow_cycle_manager.handle_node_execution_start(
workflow_execution_id=self._workflow_run_id, event=event
)
workflow_node_execution = self._workflow_cycle_manager._handle_node_execution_start(
workflow_run=workflow_run, event=event
)
node_start_response = self._workflow_cycle_manager._workflow_node_start_to_stream_response(
node_start_response = self._workflow_response_converter.workflow_node_start_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
)
session.commit()
if node_start_response:
yield node_start_response
elif isinstance(event, QueueNodeSucceededEvent):
workflow_node_execution = self._workflow_cycle_manager._handle_workflow_node_execution_success(
workflow_node_execution = self._workflow_cycle_manager.handle_workflow_node_execution_success(
event=event
)
node_success_response = self._workflow_cycle_manager._workflow_node_finish_to_stream_response(
node_success_response = self._workflow_response_converter.workflow_node_finish_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
@ -332,10 +333,10 @@ class WorkflowAppGenerateTaskPipeline:
| QueueNodeInLoopFailedEvent
| QueueNodeExceptionEvent,
):
workflow_node_execution = self._workflow_cycle_manager._handle_workflow_node_execution_failed(
workflow_node_execution = self._workflow_cycle_manager.handle_workflow_node_execution_failed(
event=event,
)
node_failed_response = self._workflow_cycle_manager._workflow_node_finish_to_stream_response(
node_failed_response = self._workflow_response_converter.workflow_node_finish_to_stream_response(
event=event,
task_id=self._application_generate_entity.task_id,
workflow_node_execution=workflow_node_execution,
@ -348,15 +349,10 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
parallel_start_resp = (
self._workflow_cycle_manager._workflow_parallel_branch_start_to_stream_response(
session=session,
self._workflow_response_converter.workflow_parallel_branch_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
)
@ -367,15 +363,10 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
parallel_finish_resp = (
self._workflow_cycle_manager._workflow_parallel_branch_finished_to_stream_response(
session=session,
self._workflow_response_converter.workflow_parallel_branch_finished_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
)
@ -386,14 +377,9 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
iter_start_resp = self._workflow_cycle_manager._workflow_iteration_start_to_stream_response(
session=session,
iter_start_resp = self._workflow_response_converter.workflow_iteration_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -403,14 +389,9 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
iter_next_resp = self._workflow_cycle_manager._workflow_iteration_next_to_stream_response(
session=session,
iter_next_resp = self._workflow_response_converter.workflow_iteration_next_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -420,14 +401,9 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
iter_finish_resp = self._workflow_cycle_manager._workflow_iteration_completed_to_stream_response(
session=session,
iter_finish_resp = self._workflow_response_converter.workflow_iteration_completed_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -437,14 +413,9 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
loop_start_resp = self._workflow_cycle_manager._workflow_loop_start_to_stream_response(
session=session,
loop_start_resp = self._workflow_response_converter.workflow_loop_start_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -454,14 +425,9 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
loop_next_resp = self._workflow_cycle_manager._workflow_loop_next_to_stream_response(
session=session,
loop_next_resp = self._workflow_response_converter.workflow_loop_next_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -471,14 +437,9 @@ class WorkflowAppGenerateTaskPipeline:
if not self._workflow_run_id:
raise ValueError("workflow run not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._get_workflow_run(
session=session, workflow_run_id=self._workflow_run_id
)
loop_finish_resp = self._workflow_cycle_manager._workflow_loop_completed_to_stream_response(
session=session,
loop_finish_resp = self._workflow_response_converter.workflow_loop_completed_to_stream_response(
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution_id=self._workflow_run_id,
event=event,
)
@ -491,10 +452,8 @@ class WorkflowAppGenerateTaskPipeline:
raise ValueError("graph runtime state not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_success(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_success(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
outputs=event.outputs,
@ -503,12 +462,12 @@ class WorkflowAppGenerateTaskPipeline:
)
# save workflow app log
self._save_workflow_app_log(session=session, workflow_run=workflow_run)
self._save_workflow_app_log(session=session, workflow_execution=workflow_execution)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_run=workflow_run,
workflow_execution=workflow_execution,
)
session.commit()
@ -520,10 +479,8 @@ class WorkflowAppGenerateTaskPipeline:
raise ValueError("graph runtime state not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_partial_success(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_partial_success(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
outputs=event.outputs,
@ -533,10 +490,12 @@ class WorkflowAppGenerateTaskPipeline:
)
# save workflow app log
self._save_workflow_app_log(session=session, workflow_run=workflow_run)
self._save_workflow_app_log(session=session, workflow_execution=workflow_execution)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
session.commit()
@ -548,26 +507,28 @@ class WorkflowAppGenerateTaskPipeline:
raise ValueError("graph runtime state not initialized.")
with Session(db.engine, expire_on_commit=False) as session:
workflow_run = self._workflow_cycle_manager._handle_workflow_run_failed(
session=session,
workflow_execution = self._workflow_cycle_manager.handle_workflow_run_failed(
workflow_run_id=self._workflow_run_id,
start_at=graph_runtime_state.start_at,
total_tokens=graph_runtime_state.total_tokens,
total_steps=graph_runtime_state.node_run_steps,
status=WorkflowRunStatus.FAILED
if isinstance(event, QueueWorkflowFailedEvent)
else WorkflowRunStatus.STOPPED,
error=event.error if isinstance(event, QueueWorkflowFailedEvent) else event.get_stop_reason(),
error_message=event.error
if isinstance(event, QueueWorkflowFailedEvent)
else event.get_stop_reason(),
conversation_id=None,
trace_manager=trace_manager,
exceptions_count=event.exceptions_count if isinstance(event, QueueWorkflowFailedEvent) else 0,
)
# save workflow app log
self._save_workflow_app_log(session=session, workflow_run=workflow_run)
self._save_workflow_app_log(session=session, workflow_execution=workflow_execution)
workflow_finish_resp = self._workflow_cycle_manager._workflow_finish_to_stream_response(
session=session, task_id=self._application_generate_entity.task_id, workflow_run=workflow_run
workflow_finish_resp = self._workflow_response_converter.workflow_finish_to_stream_response(
session=session,
task_id=self._application_generate_entity.task_id,
workflow_execution=workflow_execution,
)
session.commit()
@ -586,7 +547,7 @@ class WorkflowAppGenerateTaskPipeline:
delta_text, from_variable_selector=event.from_variable_selector
)
elif isinstance(event, QueueAgentLogEvent):
yield self._workflow_cycle_manager._handle_agent_log(
yield self._workflow_response_converter.handle_agent_log(
task_id=self._application_generate_entity.task_id, event=event
)
else:
@ -595,11 +556,9 @@ class WorkflowAppGenerateTaskPipeline:
if tts_publisher:
tts_publisher.publish(None)
def _save_workflow_app_log(self, *, session: Session, workflow_run: WorkflowRun) -> None:
"""
Save workflow app log.
:return:
"""
def _save_workflow_app_log(self, *, session: Session, workflow_execution: WorkflowExecution) -> None:
workflow_run = session.scalar(select(WorkflowRun).where(WorkflowRun.id == workflow_execution.id))
assert workflow_run is not None
invoke_from = self._application_generate_entity.invoke_from
if invoke_from == InvokeFrom.SERVICE_API:
created_from = WorkflowAppLogCreatedFrom.SERVICE_API

26
api/core/app/entities/task_entities.py
View File

@ -6,7 +6,7 @@ from pydantic import BaseModel, ConfigDict
from core.model_runtime.entities.llm_entities import LLMResult
from core.model_runtime.utils.encoders import jsonable_encoder
from core.workflow.entities.node_entities import AgentNodeStrategyInit
from core.workflow.entities.node_entities import AgentNodeStrategyInit, NodeRunMetadataKey
from models.workflow import WorkflowNodeExecutionStatus
@ -190,7 +190,7 @@ class WorkflowStartStreamResponse(StreamResponse):
id: str
workflow_id: str
sequence_number: int
inputs: dict
inputs: Mapping[str, Any]
created_at: int
event: StreamEvent = StreamEvent.WORKFLOW_STARTED
@ -212,7 +212,7 @@ class WorkflowFinishStreamResponse(StreamResponse):
workflow_id: str
sequence_number: int
status: str
outputs: Optional[dict] = None
outputs: Optional[Mapping[str, Any]] = None
error: Optional[str] = None
elapsed_time: float
total_tokens: int
@ -244,7 +244,7 @@ class NodeStartStreamResponse(StreamResponse):
title: str
index: int
predecessor_node_id: Optional[str] = None
inputs: Optional[dict] = None
inputs: Optional[Mapping[str, Any]] = None
created_at: int
extras: dict = {}
parallel_id: Optional[str] = None
@ -301,13 +301,13 @@ class NodeFinishStreamResponse(StreamResponse):
title: str
index: int
predecessor_node_id: Optional[str] = None
inputs: Optional[dict] = None
process_data: Optional[dict] = None
outputs: Optional[dict] = None
inputs: Optional[Mapping[str, Any]] = None
process_data: Optional[Mapping[str, Any]] = None
outputs: Optional[Mapping[str, Any]] = None
status: str
error: Optional[str] = None
elapsed_time: float
execution_metadata: Optional[dict] = None
execution_metadata: Optional[Mapping[NodeRunMetadataKey, Any]] = None
created_at: int
finished_at: int
files: Optional[Sequence[Mapping[str, Any]]] = []
@ -370,13 +370,13 @@ class NodeRetryStreamResponse(StreamResponse):
title: str
index: int
predecessor_node_id: Optional[str] = None
inputs: Optional[dict] = None
process_data: Optional[dict] = None
outputs: Optional[dict] = None
inputs: Optional[Mapping[str, Any]] = None
process_data: Optional[Mapping[str, Any]] = None
outputs: Optional[Mapping[str, Any]] = None
status: str
error: Optional[str] = None
elapsed_time: float
execution_metadata: Optional[dict] = None
execution_metadata: Optional[Mapping[NodeRunMetadataKey, Any]] = None
created_at: int
finished_at: int
files: Optional[Sequence[Mapping[str, Any]]] = []
@ -788,7 +788,7 @@ class WorkflowAppBlockingResponse(AppBlockingResponse):
id: str
workflow_id: str
status: str
outputs: Optional[dict] = None
outputs: Optional[Mapping[str, Any]] = None
error: Optional[str] = None
elapsed_time: float
total_tokens: int

2
api/core/app/task_pipeline/easy_ui_based_generate_task_pipeline.py
View File

@ -9,7 +9,6 @@ from sqlalchemy import select
from sqlalchemy.orm import Session
from constants.tts_auto_play_timeout import TTS_AUTO_PLAY_TIMEOUT, TTS_AUTO_PLAY_YIELD_CPU_TIME
from core.app.apps.advanced_chat.app_generator_tts_publisher import AppGeneratorTTSPublisher, AudioTrunk
from core.app.apps.base_app_queue_manager import AppQueueManager, PublishFrom
from core.app.entities.app_invoke_entities import (
AgentChatAppGenerateEntity,
@ -45,6 +44,7 @@ from core.app.entities.task_entities import (
)
from core.app.task_pipeline.based_generate_task_pipeline import BasedGenerateTaskPipeline
from core.app.task_pipeline.message_cycle_manage import MessageCycleManage
from core.base.tts import AppGeneratorTTSPublisher, AudioTrunk
from core.model_manager import ModelInstance
from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk, LLMResultChunkDelta, LLMUsage
from core.model_runtime.entities.message_entities import (

948
api/core/app/task_pipeline/workflow_cycle_manage.py
View File

@ -1,948 +0,0 @@
import json
import time
from collections.abc import Mapping, Sequence
from datetime import UTC, datetime
from typing import Any, Optional, Union, cast
from uuid import uuid4
from sqlalchemy import func, select
from sqlalchemy.orm import Session
from core.app.entities.app_invoke_entities import AdvancedChatAppGenerateEntity, InvokeFrom, WorkflowAppGenerateEntity
from core.app.entities.queue_entities import (
QueueAgentLogEvent,
QueueIterationCompletedEvent,
QueueIterationNextEvent,
QueueIterationStartEvent,
QueueLoopCompletedEvent,
QueueLoopNextEvent,
QueueLoopStartEvent,
QueueNodeExceptionEvent,
QueueNodeFailedEvent,
QueueNodeInIterationFailedEvent,
QueueNodeInLoopFailedEvent,
QueueNodeRetryEvent,
QueueNodeStartedEvent,
QueueNodeSucceededEvent,
QueueParallelBranchRunFailedEvent,
QueueParallelBranchRunStartedEvent,
QueueParallelBranchRunSucceededEvent,
)
from core.app.entities.task_entities import (
AgentLogStreamResponse,
IterationNodeCompletedStreamResponse,
IterationNodeNextStreamResponse,
IterationNodeStartStreamResponse,
LoopNodeCompletedStreamResponse,
LoopNodeNextStreamResponse,
LoopNodeStartStreamResponse,
NodeFinishStreamResponse,
NodeRetryStreamResponse,
NodeStartStreamResponse,
ParallelBranchFinishedStreamResponse,
ParallelBranchStartStreamResponse,
WorkflowFinishStreamResponse,
WorkflowStartStreamResponse,
)
from core.app.task_pipeline.exc import WorkflowRunNotFoundError
from core.file import FILE_MODEL_IDENTITY, File
from core.model_runtime.utils.encoders import jsonable_encoder
from core.ops.entities.trace_entity import TraceTaskName
from core.ops.ops_trace_manager import TraceQueueManager, TraceTask
from core.tools.tool_manager import ToolManager
from core.workflow.entities.node_entities import NodeRunMetadataKey
from core.workflow.enums import SystemVariableKey
from core.workflow.nodes import NodeType
from core.workflow.nodes.tool.entities import ToolNodeData
from core.workflow.repository.workflow_node_execution_repository import WorkflowNodeExecutionRepository
from core.workflow.workflow_entry import WorkflowEntry
from models.account import Account
from models.enums import CreatedByRole, WorkflowRunTriggeredFrom
from models.model import EndUser
from models.workflow import (
Workflow,
WorkflowNodeExecution,
WorkflowNodeExecutionStatus,
WorkflowNodeExecutionTriggeredFrom,
WorkflowRun,
WorkflowRunStatus,
)
class WorkflowCycleManage:
def __init__(
self,
*,
application_generate_entity: Union[AdvancedChatAppGenerateEntity, WorkflowAppGenerateEntity],
workflow_system_variables: dict[SystemVariableKey, Any],
workflow_node_execution_repository: WorkflowNodeExecutionRepository,
) -> None:
self._workflow_run: WorkflowRun | None = None
self._workflow_node_executions: dict[str, WorkflowNodeExecution] = {}
self._application_generate_entity = application_generate_entity
self._workflow_system_variables = workflow_system_variables
self._workflow_node_execution_repository = workflow_node_execution_repository
def _handle_workflow_run_start(
self,
*,
session: Session,
workflow_id: str,
user_id: str,
created_by_role: CreatedByRole,
) -> WorkflowRun:
workflow_stmt = select(Workflow).where(Workflow.id == workflow_id)
workflow = session.scalar(workflow_stmt)
if not workflow:
raise ValueError(f"Workflow not found: {workflow_id}")
max_sequence_stmt = select(func.max(WorkflowRun.sequence_number)).where(
WorkflowRun.tenant_id == workflow.tenant_id,
WorkflowRun.app_id == workflow.app_id,
)
max_sequence = session.scalar(max_sequence_stmt) or 0
new_sequence_number = max_sequence + 1
inputs = {**self._application_generate_entity.inputs}
for key, value in (self._workflow_system_variables or {}).items():
if key.value == "conversation":
continue
inputs[f"sys.{key.value}"] = value
triggered_from = (
WorkflowRunTriggeredFrom.DEBUGGING
if self._application_generate_entity.invoke_from == InvokeFrom.DEBUGGER
else WorkflowRunTriggeredFrom.APP_RUN
)
# handle special values
inputs = dict(WorkflowEntry.handle_special_values(inputs) or {})
# init workflow run
# TODO: This workflow_run_id should always not be None, maybe we can use a more elegant way to handle this
workflow_run_id = str(self._workflow_system_variables.get(SystemVariableKey.WORKFLOW_RUN_ID) or uuid4())
workflow_run = WorkflowRun()
workflow_run.id = workflow_run_id
workflow_run.tenant_id = workflow.tenant_id
workflow_run.app_id = workflow.app_id
workflow_run.sequence_number = new_sequence_number
workflow_run.workflow_id = workflow.id
workflow_run.type = workflow.type
workflow_run.triggered_from = triggered_from.value
workflow_run.version = workflow.version
workflow_run.graph = workflow.graph
workflow_run.inputs = json.dumps(inputs)
workflow_run.status = WorkflowRunStatus.RUNNING
workflow_run.created_by_role = created_by_role
workflow_run.created_by = user_id
workflow_run.created_at = datetime.now(UTC).replace(tzinfo=None)
session.add(workflow_run)
return workflow_run
def _handle_workflow_run_success(
self,
*,
session: Session,
workflow_run_id: str,
start_at: float,
total_tokens: int,
total_steps: int,
outputs: Mapping[str, Any] | None = None,
conversation_id: Optional[str] = None,
trace_manager: Optional[TraceQueueManager] = None,
) -> WorkflowRun:
"""
Workflow run success
:param workflow_run_id: workflow run id
:param start_at: start time
:param total_tokens: total tokens
:param total_steps: total steps
:param outputs: outputs
:param conversation_id: conversation id
:return:
"""
workflow_run = self._get_workflow_run(session=session, workflow_run_id=workflow_run_id)
outputs = WorkflowEntry.handle_special_values(outputs)
workflow_run.status = WorkflowRunStatus.SUCCEEDED
workflow_run.outputs = json.dumps(outputs or {})
workflow_run.elapsed_time = time.perf_counter() - start_at
workflow_run.total_tokens = total_tokens
workflow_run.total_steps = total_steps
workflow_run.finished_at = datetime.now(UTC).replace(tzinfo=None)
if trace_manager:
trace_manager.add_trace_task(
TraceTask(
TraceTaskName.WORKFLOW_TRACE,
workflow_run=workflow_run,
conversation_id=conversation_id,
user_id=trace_manager.user_id,
)
)
return workflow_run
def _handle_workflow_run_partial_success(
self,
*,
session: Session,
workflow_run_id: str,
start_at: float,
total_tokens: int,
total_steps: int,
outputs: Mapping[str, Any] | None = None,
exceptions_count: int = 0,
conversation_id: Optional[str] = None,
trace_manager: Optional[TraceQueueManager] = None,
) -> WorkflowRun:
workflow_run = self._get_workflow_run(session=session, workflow_run_id=workflow_run_id)
outputs = WorkflowEntry.handle_special_values(dict(outputs) if outputs else None)
workflow_run.status = WorkflowRunStatus.PARTIAL_SUCCEEDED.value
workflow_run.outputs = json.dumps(outputs or {})
workflow_run.elapsed_time = time.perf_counter() - start_at
workflow_run.total_tokens = total_tokens
workflow_run.total_steps = total_steps
workflow_run.finished_at = datetime.now(UTC).replace(tzinfo=None)
workflow_run.exceptions_count = exceptions_count
if trace_manager:
trace_manager.add_trace_task(
TraceTask(
TraceTaskName.WORKFLOW_TRACE,
workflow_run=workflow_run,
conversation_id=conversation_id,
user_id=trace_manager.user_id,
)
)
return workflow_run
def _handle_workflow_run_failed(
self,
*,
session: Session,
workflow_run_id: str,
start_at: float,
total_tokens: int,
total_steps: int,
status: WorkflowRunStatus,
error: str,
conversation_id: Optional[str] = None,
trace_manager: Optional[TraceQueueManager] = None,
exceptions_count: int = 0,
) -> WorkflowRun:
"""
Workflow run failed
:param workflow_run_id: workflow run id
:param start_at: start time
:param total_tokens: total tokens
:param total_steps: total steps
:param status: status
:param error: error message
:return:
"""
workflow_run = self._get_workflow_run(session=session, workflow_run_id=workflow_run_id)
workflow_run.status = status.value
workflow_run.error = error
workflow_run.elapsed_time = time.perf_counter() - start_at
workflow_run.total_tokens = total_tokens
workflow_run.total_steps = total_steps
workflow_run.finished_at = datetime.now(UTC).replace(tzinfo=None)
workflow_run.exceptions_count = exceptions_count
# Use the instance repository to find running executions for a workflow run
running_workflow_node_executions = self._workflow_node_execution_repository.get_running_executions(
workflow_run_id=workflow_run.id
)
# Update the cache with the retrieved executions
for execution in running_workflow_node_executions:
if execution.node_execution_id:
self._workflow_node_executions[execution.node_execution_id] = execution
for workflow_node_execution in running_workflow_node_executions:
now = datetime.now(UTC).replace(tzinfo=None)
workflow_node_execution.status = WorkflowNodeExecutionStatus.FAILED.value
workflow_node_execution.error = error
workflow_node_execution.finished_at = now
workflow_node_execution.elapsed_time = (now - workflow_node_execution.created_at).total_seconds()
if trace_manager:
trace_manager.add_trace_task(
TraceTask(
TraceTaskName.WORKFLOW_TRACE,
workflow_run=workflow_run,
conversation_id=conversation_id,
user_id=trace_manager.user_id,
)
)
return workflow_run
def _handle_node_execution_start(
self, *, workflow_run: WorkflowRun, event: QueueNodeStartedEvent
) -> WorkflowNodeExecution:
workflow_node_execution = WorkflowNodeExecution()
workflow_node_execution.id = str(uuid4())
workflow_node_execution.tenant_id = workflow_run.tenant_id
workflow_node_execution.app_id = workflow_run.app_id
workflow_node_execution.workflow_id = workflow_run.workflow_id
workflow_node_execution.triggered_from = WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN.value
workflow_node_execution.workflow_run_id = workflow_run.id
workflow_node_execution.predecessor_node_id = event.predecessor_node_id
workflow_node_execution.index = event.node_run_index
workflow_node_execution.node_execution_id = event.node_execution_id
workflow_node_execution.node_id = event.node_id
workflow_node_execution.node_type = event.node_type.value
workflow_node_execution.title = event.node_data.title
workflow_node_execution.status = WorkflowNodeExecutionStatus.RUNNING.value
workflow_node_execution.created_by_role = workflow_run.created_by_role
workflow_node_execution.created_by = workflow_run.created_by
workflow_node_execution.execution_metadata = json.dumps(
{
NodeRunMetadataKey.PARALLEL_MODE_RUN_ID: event.parallel_mode_run_id,
NodeRunMetadataKey.ITERATION_ID: event.in_iteration_id,
NodeRunMetadataKey.LOOP_ID: event.in_loop_id,
}
)
workflow_node_execution.created_at = datetime.now(UTC).replace(tzinfo=None)
# Use the instance repository to save the workflow node execution
self._workflow_node_execution_repository.save(workflow_node_execution)
self._workflow_node_executions[event.node_execution_id] = workflow_node_execution
return workflow_node_execution
def _handle_workflow_node_execution_success(self, *, event: QueueNodeSucceededEvent) -> WorkflowNodeExecution:
workflow_node_execution = self._get_workflow_node_execution(node_execution_id=event.node_execution_id)
inputs = WorkflowEntry.handle_special_values(event.inputs)
process_data = WorkflowEntry.handle_special_values(event.process_data)
outputs = WorkflowEntry.handle_special_values(event.outputs)
execution_metadata_dict = dict(event.execution_metadata or {})
execution_metadata = json.dumps(jsonable_encoder(execution_metadata_dict)) if execution_metadata_dict else None
finished_at = datetime.now(UTC).replace(tzinfo=None)
elapsed_time = (finished_at - event.start_at).total_seconds()
process_data = WorkflowEntry.handle_special_values(event.process_data)
workflow_node_execution.status = WorkflowNodeExecutionStatus.SUCCEEDED.value
workflow_node_execution.inputs = json.dumps(inputs) if inputs else None
workflow_node_execution.process_data = json.dumps(process_data) if process_data else None
workflow_node_execution.outputs = json.dumps(outputs) if outputs else None
workflow_node_execution.execution_metadata = execution_metadata
workflow_node_execution.finished_at = finished_at
workflow_node_execution.elapsed_time = elapsed_time
# Use the instance repository to update the workflow node execution
self._workflow_node_execution_repository.update(workflow_node_execution)
return workflow_node_execution
def _handle_workflow_node_execution_failed(
self,
*,
event: QueueNodeFailedEvent
| QueueNodeInIterationFailedEvent
| QueueNodeInLoopFailedEvent
| QueueNodeExceptionEvent,
) -> WorkflowNodeExecution:
"""
Workflow node execution failed
:param event: queue node failed event
:return:
"""
workflow_node_execution = self._get_workflow_node_execution(node_execution_id=event.node_execution_id)
inputs = WorkflowEntry.handle_special_values(event.inputs)
process_data = WorkflowEntry.handle_special_values(event.process_data)
outputs = WorkflowEntry.handle_special_values(event.outputs)
finished_at = datetime.now(UTC).replace(tzinfo=None)
elapsed_time = (finished_at - event.start_at).total_seconds()
execution_metadata = (
json.dumps(jsonable_encoder(event.execution_metadata)) if event.execution_metadata else None
)
process_data = WorkflowEntry.handle_special_values(event.process_data)
workflow_node_execution.status = (
WorkflowNodeExecutionStatus.FAILED.value
if not isinstance(event, QueueNodeExceptionEvent)
else WorkflowNodeExecutionStatus.EXCEPTION.value
)
workflow_node_execution.error = event.error
workflow_node_execution.inputs = json.dumps(inputs) if inputs else None
workflow_node_execution.process_data = json.dumps(process_data) if process_data else None
workflow_node_execution.outputs = json.dumps(outputs) if outputs else None
workflow_node_execution.finished_at = finished_at
workflow_node_execution.elapsed_time = elapsed_time
workflow_node_execution.execution_metadata = execution_metadata
self._workflow_node_execution_repository.update(workflow_node_execution)
return workflow_node_execution
def _handle_workflow_node_execution_retried(
self, *, workflow_run: WorkflowRun, event: QueueNodeRetryEvent
) -> WorkflowNodeExecution:
"""
Workflow node execution failed
:param workflow_run: workflow run
:param event: queue node failed event
:return:
"""
created_at = event.start_at
finished_at = datetime.now(UTC).replace(tzinfo=None)
elapsed_time = (finished_at - created_at).total_seconds()
inputs = WorkflowEntry.handle_special_values(event.inputs)
outputs = WorkflowEntry.handle_special_values(event.outputs)
origin_metadata = {
NodeRunMetadataKey.ITERATION_ID: event.in_iteration_id,
NodeRunMetadataKey.PARALLEL_MODE_RUN_ID: event.parallel_mode_run_id,
NodeRunMetadataKey.LOOP_ID: event.in_loop_id,
}
merged_metadata = (
{**jsonable_encoder(event.execution_metadata), **origin_metadata}
if event.execution_metadata is not None
else origin_metadata
)
execution_metadata = json.dumps(merged_metadata)
workflow_node_execution = WorkflowNodeExecution()
workflow_node_execution.id = str(uuid4())
workflow_node_execution.tenant_id = workflow_run.tenant_id
workflow_node_execution.app_id = workflow_run.app_id
workflow_node_execution.workflow_id = workflow_run.workflow_id
workflow_node_execution.triggered_from = WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN.value
workflow_node_execution.workflow_run_id = workflow_run.id
workflow_node_execution.predecessor_node_id = event.predecessor_node_id
workflow_node_execution.node_execution_id = event.node_execution_id
workflow_node_execution.node_id = event.node_id
workflow_node_execution.node_type = event.node_type.value
workflow_node_execution.title = event.node_data.title
workflow_node_execution.status = WorkflowNodeExecutionStatus.RETRY.value
workflow_node_execution.created_by_role = workflow_run.created_by_role
workflow_node_execution.created_by = workflow_run.created_by
workflow_node_execution.created_at = created_at
workflow_node_execution.finished_at = finished_at
workflow_node_execution.elapsed_time = elapsed_time
workflow_node_execution.error = event.error
workflow_node_execution.inputs = json.dumps(inputs) if inputs else None
workflow_node_execution.outputs = json.dumps(outputs) if outputs else None
workflow_node_execution.execution_metadata = execution_metadata
workflow_node_execution.index = event.node_run_index
# Use the instance repository to save the workflow node execution
self._workflow_node_execution_repository.save(workflow_node_execution)
self._workflow_node_executions[event.node_execution_id] = workflow_node_execution
return workflow_node_execution
def _workflow_start_to_stream_response(
self,
*,
session: Session,
task_id: str,
workflow_run: WorkflowRun,
) -> WorkflowStartStreamResponse:
_ = session
return WorkflowStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=WorkflowStartStreamResponse.Data(
id=workflow_run.id,
workflow_id=workflow_run.workflow_id,
sequence_number=workflow_run.sequence_number,
inputs=dict(workflow_run.inputs_dict or {}),
created_at=int(workflow_run.created_at.timestamp()),
),
)
def _workflow_finish_to_stream_response(
self,
*,
session: Session,
task_id: str,
workflow_run: WorkflowRun,
) -> WorkflowFinishStreamResponse:
created_by = None
if workflow_run.created_by_role == CreatedByRole.ACCOUNT:
stmt = select(Account).where(Account.id == workflow_run.created_by)
account = session.scalar(stmt)
if account:
created_by = {
"id": account.id,
"name": account.name,
"email": account.email,
}
elif workflow_run.created_by_role == CreatedByRole.END_USER:
stmt = select(EndUser).where(EndUser.id == workflow_run.created_by)
end_user = session.scalar(stmt)
if end_user:
created_by = {
"id": end_user.id,
"user": end_user.session_id,
}
else:
raise NotImplementedError(f"unknown created_by_role: {workflow_run.created_by_role}")
return WorkflowFinishStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=WorkflowFinishStreamResponse.Data(
id=workflow_run.id,
workflow_id=workflow_run.workflow_id,
sequence_number=workflow_run.sequence_number,
status=workflow_run.status,
outputs=dict(workflow_run.outputs_dict) if workflow_run.outputs_dict else None,
error=workflow_run.error,
elapsed_time=workflow_run.elapsed_time,
total_tokens=workflow_run.total_tokens,
total_steps=workflow_run.total_steps,
created_by=created_by,
created_at=int(workflow_run.created_at.timestamp()),
finished_at=int(workflow_run.finished_at.timestamp()),
files=self._fetch_files_from_node_outputs(dict(workflow_run.outputs_dict)),
exceptions_count=workflow_run.exceptions_count,
),
)
def _workflow_node_start_to_stream_response(
self,
*,
event: QueueNodeStartedEvent,
task_id: str,
workflow_node_execution: WorkflowNodeExecution,
) -> Optional[NodeStartStreamResponse]:
if workflow_node_execution.node_type in {NodeType.ITERATION.value, NodeType.LOOP.value}:
return None
if not workflow_node_execution.workflow_run_id:
return None
response = NodeStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_node_execution.workflow_run_id,
data=NodeStartStreamResponse.Data(
id=workflow_node_execution.id,
node_id=workflow_node_execution.node_id,
node_type=workflow_node_execution.node_type,
title=workflow_node_execution.title,
index=workflow_node_execution.index,
predecessor_node_id=workflow_node_execution.predecessor_node_id,
inputs=workflow_node_execution.inputs_dict,
created_at=int(workflow_node_execution.created_at.timestamp()),
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
parallel_run_id=event.parallel_mode_run_id,
agent_strategy=event.agent_strategy,
),
)
# extras logic
if event.node_type == NodeType.TOOL:
node_data = cast(ToolNodeData, event.node_data)
response.data.extras["icon"] = ToolManager.get_tool_icon(
tenant_id=self._application_generate_entity.app_config.tenant_id,
provider_type=node_data.provider_type,
provider_id=node_data.provider_id,
)
return response
def _workflow_node_finish_to_stream_response(
self,
*,
event: QueueNodeSucceededEvent
| QueueNodeFailedEvent
| QueueNodeInIterationFailedEvent
| QueueNodeInLoopFailedEvent
| QueueNodeExceptionEvent,
task_id: str,
workflow_node_execution: WorkflowNodeExecution,
) -> Optional[NodeFinishStreamResponse]:
if workflow_node_execution.node_type in {NodeType.ITERATION.value, NodeType.LOOP.value}:
return None
if not workflow_node_execution.workflow_run_id:
return None
if not workflow_node_execution.finished_at:
return None
return NodeFinishStreamResponse(
task_id=task_id,
workflow_run_id=workflow_node_execution.workflow_run_id,
data=NodeFinishStreamResponse.Data(
id=workflow_node_execution.id,
node_id=workflow_node_execution.node_id,
node_type=workflow_node_execution.node_type,
index=workflow_node_execution.index,
title=workflow_node_execution.title,
predecessor_node_id=workflow_node_execution.predecessor_node_id,
inputs=workflow_node_execution.inputs_dict,
process_data=workflow_node_execution.process_data_dict,
outputs=workflow_node_execution.outputs_dict,
status=workflow_node_execution.status,
error=workflow_node_execution.error,
elapsed_time=workflow_node_execution.elapsed_time,
execution_metadata=workflow_node_execution.execution_metadata_dict,
created_at=int(workflow_node_execution.created_at.timestamp()),
finished_at=int(workflow_node_execution.finished_at.timestamp()),
files=self._fetch_files_from_node_outputs(workflow_node_execution.outputs_dict or {}),
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
),
)
def _workflow_node_retry_to_stream_response(
self,
*,
event: QueueNodeRetryEvent,
task_id: str,
workflow_node_execution: WorkflowNodeExecution,
) -> Optional[Union[NodeRetryStreamResponse, NodeFinishStreamResponse]]:
if workflow_node_execution.node_type in {NodeType.ITERATION.value, NodeType.LOOP.value}:
return None
if not workflow_node_execution.workflow_run_id:
return None
if not workflow_node_execution.finished_at:
return None
return NodeRetryStreamResponse(
task_id=task_id,
workflow_run_id=workflow_node_execution.workflow_run_id,
data=NodeRetryStreamResponse.Data(
id=workflow_node_execution.id,
node_id=workflow_node_execution.node_id,
node_type=workflow_node_execution.node_type,
index=workflow_node_execution.index,
title=workflow_node_execution.title,
predecessor_node_id=workflow_node_execution.predecessor_node_id,
inputs=workflow_node_execution.inputs_dict,
process_data=workflow_node_execution.process_data_dict,
outputs=workflow_node_execution.outputs_dict,
status=workflow_node_execution.status,
error=workflow_node_execution.error,
elapsed_time=workflow_node_execution.elapsed_time,
execution_metadata=workflow_node_execution.execution_metadata_dict,
created_at=int(workflow_node_execution.created_at.timestamp()),
finished_at=int(workflow_node_execution.finished_at.timestamp()),
files=self._fetch_files_from_node_outputs(workflow_node_execution.outputs_dict or {}),
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
retry_index=event.retry_index,
),
)
def _workflow_parallel_branch_start_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueParallelBranchRunStartedEvent
) -> ParallelBranchStartStreamResponse:
_ = session
return ParallelBranchStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=ParallelBranchStartStreamResponse.Data(
parallel_id=event.parallel_id,
parallel_branch_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
created_at=int(time.time()),
),
)
def _workflow_parallel_branch_finished_to_stream_response(
self,
*,
session: Session,
task_id: str,
workflow_run: WorkflowRun,
event: QueueParallelBranchRunSucceededEvent | QueueParallelBranchRunFailedEvent,
) -> ParallelBranchFinishedStreamResponse:
_ = session
return ParallelBranchFinishedStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=ParallelBranchFinishedStreamResponse.Data(
parallel_id=event.parallel_id,
parallel_branch_id=event.parallel_start_node_id,
parent_parallel_id=event.parent_parallel_id,
parent_parallel_start_node_id=event.parent_parallel_start_node_id,
iteration_id=event.in_iteration_id,
loop_id=event.in_loop_id,
status="succeeded" if isinstance(event, QueueParallelBranchRunSucceededEvent) else "failed",
error=event.error if isinstance(event, QueueParallelBranchRunFailedEvent) else None,
created_at=int(time.time()),
),
)
def _workflow_iteration_start_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueIterationStartEvent
) -> IterationNodeStartStreamResponse:
_ = session
return IterationNodeStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=IterationNodeStartStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
metadata=event.metadata or {},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def _workflow_iteration_next_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueIterationNextEvent
) -> IterationNodeNextStreamResponse:
_ = session
return IterationNodeNextStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=IterationNodeNextStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
index=event.index,
pre_iteration_output=event.output,
created_at=int(time.time()),
extras={},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parallel_mode_run_id=event.parallel_mode_run_id,
duration=event.duration,
),
)
def _workflow_iteration_completed_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueIterationCompletedEvent
) -> IterationNodeCompletedStreamResponse:
_ = session
return IterationNodeCompletedStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=IterationNodeCompletedStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
outputs=event.outputs,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
status=WorkflowNodeExecutionStatus.SUCCEEDED
if event.error is None
else WorkflowNodeExecutionStatus.FAILED,
error=None,
elapsed_time=(datetime.now(UTC).replace(tzinfo=None) - event.start_at).total_seconds(),
total_tokens=event.metadata.get("total_tokens", 0) if event.metadata else 0,
execution_metadata=event.metadata,
finished_at=int(time.time()),
steps=event.steps,
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def _workflow_loop_start_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueLoopStartEvent
) -> LoopNodeStartStreamResponse:
_ = session
return LoopNodeStartStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=LoopNodeStartStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
metadata=event.metadata or {},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def _workflow_loop_next_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueLoopNextEvent
) -> LoopNodeNextStreamResponse:
_ = session
return LoopNodeNextStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=LoopNodeNextStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
index=event.index,
pre_loop_output=event.output,
created_at=int(time.time()),
extras={},
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
parallel_mode_run_id=event.parallel_mode_run_id,
duration=event.duration,
),
)
def _workflow_loop_completed_to_stream_response(
self, *, session: Session, task_id: str, workflow_run: WorkflowRun, event: QueueLoopCompletedEvent
) -> LoopNodeCompletedStreamResponse:
_ = session
return LoopNodeCompletedStreamResponse(
task_id=task_id,
workflow_run_id=workflow_run.id,
data=LoopNodeCompletedStreamResponse.Data(
id=event.node_id,
node_id=event.node_id,
node_type=event.node_type.value,
title=event.node_data.title,
outputs=event.outputs,
created_at=int(time.time()),
extras={},
inputs=event.inputs or {},
status=WorkflowNodeExecutionStatus.SUCCEEDED
if event.error is None
else WorkflowNodeExecutionStatus.FAILED,
error=None,
elapsed_time=(datetime.now(UTC).replace(tzinfo=None) - event.start_at).total_seconds(),
total_tokens=event.metadata.get("total_tokens", 0) if event.metadata else 0,
execution_metadata=event.metadata,
finished_at=int(time.time()),
steps=event.steps,
parallel_id=event.parallel_id,
parallel_start_node_id=event.parallel_start_node_id,
),
)
def _fetch_files_from_node_outputs(self, outputs_dict: Mapping[str, Any]) -> Sequence[Mapping[str, Any]]:
"""
Fetch files from node outputs
:param outputs_dict: node outputs dict
:return:
"""
if not outputs_dict:
return []
files = [self._fetch_files_from_variable_value(output_value) for output_value in outputs_dict.values()]
# Remove None
files = [file for file in files if file]
# Flatten list
# Flatten the list of sequences into a single list of mappings
flattened_files = [file for sublist in files if sublist for file in sublist]
# Convert to tuple to match Sequence type
return tuple(flattened_files)
def _fetch_files_from_variable_value(self, value: Union[dict, list]) -> Sequence[Mapping[str, Any]]:
"""
Fetch files from variable value
:param value: variable value
:return:
"""
if not value:
return []
files = []
if isinstance(value, list):
for item in value:
file = self._get_file_var_from_value(item)
if file:
files.append(file)
elif isinstance(value, dict):
file = self._get_file_var_from_value(value)
if file:
files.append(file)
return files
def _get_file_var_from_value(self, value: Union[dict, list]) -> Mapping[str, Any] | None:
"""
Get file var from value
:param value: variable value
:return:
"""
if not value:
return None
if isinstance(value, dict) and value.get("dify_model_identity") == FILE_MODEL_IDENTITY:
return value
elif isinstance(value, File):
return value.to_dict()
return None
def _get_workflow_run(self, *, session: Session, workflow_run_id: str) -> WorkflowRun:
if self._workflow_run and self._workflow_run.id == workflow_run_id:
cached_workflow_run = self._workflow_run
cached_workflow_run = session.merge(cached_workflow_run)
return cached_workflow_run
stmt = select(WorkflowRun).where(WorkflowRun.id == workflow_run_id)
workflow_run = session.scalar(stmt)
if not workflow_run:
raise WorkflowRunNotFoundError(workflow_run_id)
self._workflow_run = workflow_run
return workflow_run
def _get_workflow_node_execution(self, node_execution_id: str) -> WorkflowNodeExecution:
# First check the cache for performance
if node_execution_id in self._workflow_node_executions:
cached_execution = self._workflow_node_executions[node_execution_id]
# No need to merge with session since expire_on_commit=False
return cached_execution
# If not in cache, use the instance repository to get by node_execution_id
execution = self._workflow_node_execution_repository.get_by_node_execution_id(node_execution_id)
if not execution:
raise ValueError(f"Workflow node execution not found: {node_execution_id}")
# Update cache
self._workflow_node_executions[node_execution_id] = execution
return execution
def _handle_agent_log(self, task_id: str, event: QueueAgentLogEvent) -> AgentLogStreamResponse:
"""
Handle agent log
:param task_id: task id
:param event: agent log event
:return:
"""
return AgentLogStreamResponse(
task_id=task_id,
data=AgentLogStreamResponse.Data(
node_execution_id=event.node_execution_id,
id=event.id,
parent_id=event.parent_id,
label=event.label,
error=event.error,
status=event.status,
data=event.data,
metadata=event.metadata,
node_id=event.node_id,
),
)

1
api/core/base/__init__.py Normal file
View File

@ -0,0 +1 @@
# Core base package

6
api/core/base/tts/__init__.py Normal file
View File

@ -0,0 +1,6 @@
from core.base.tts.app_generator_tts_publisher import AppGeneratorTTSPublisher, AudioTrunk
__all__ = [
"AppGeneratorTTSPublisher",
"AudioTrunk",
]

0
api/core/app/apps/advanced_chat/app_generator_tts_publisher.py → api/core/base/tts/app_generator_tts_publisher.py
View File

29
api/core/callback_handler/index_tool_callback_handler.py
View File

@ -1,3 +1,5 @@
import logging
from core.app.apps.base_app_queue_manager import AppQueueManager, PublishFrom
from core.app.entities.app_invoke_entities import InvokeFrom
from core.app.entities.queue_entities import QueueRetrieverResourcesEvent
@ -7,6 +9,8 @@ from extensions.ext_database import db
from models.dataset import ChildChunk, DatasetQuery, DocumentSegment
from models.dataset import Document as DatasetDocument
_logger = logging.getLogger(__name__)
class DatasetIndexToolCallbackHandler:
"""Callback handler for dataset tool."""
@ -42,19 +46,32 @@ class DatasetIndexToolCallbackHandler:
"""Handle tool end."""
for document in documents:
if document.metadata is not None:
dataset_document = DatasetDocument.query.filter(
DatasetDocument.id == document.metadata["document_id"]
).first()
document_id = document.metadata["document_id"]
dataset_document = db.session.query(DatasetDocument).filter(DatasetDocument.id == document_id).first()
if not dataset_document:
_logger.warning(
"Expected DatasetDocument record to exist, but none was found, document_id=%s",
document_id,
)
continue
if dataset_document.doc_form == IndexType.PARENT_CHILD_INDEX:
child_chunk = ChildChunk.query.filter(
child_chunk = (
db.session.query(ChildChunk)
.filter(
ChildChunk.index_node_id == document.metadata["doc_id"],
ChildChunk.dataset_id == dataset_document.dataset_id,
ChildChunk.document_id == dataset_document.id,
).first()
)
.first()
)
if child_chunk:
segment = DocumentSegment.query.filter(DocumentSegment.id == child_chunk.segment_id).update(
segment = (
db.session.query(DocumentSegment)
.filter(DocumentSegment.id == child_chunk.segment_id)
.update(
{DocumentSegment.hit_count: DocumentSegment.hit_count + 1}, synchronize_session=False
)
)
else:
query = db.session.query(DocumentSegment).filter(
DocumentSegment.index_node_id == document.metadata["doc_id"]

19
api/core/entities/provider_configuration.py
View File

@ -754,7 +754,7 @@ class ProviderConfiguration(BaseModel):
:param only_active: return active model only
:return:
"""
provider_models = self.get_provider_models(model_type, only_active)
provider_models = self.get_provider_models(model_type, only_active, model)
for provider_model in provider_models:
if provider_model.model == model:
@ -763,12 +763,13 @@ class ProviderConfiguration(BaseModel):
return None
def get_provider_models(
self, model_type: Optional[ModelType] = None, only_active: bool = False
self, model_type: Optional[ModelType] = None, only_active: bool = False, model: Optional[str] = None
) -> list[ModelWithProviderEntity]:
"""
Get provider models.
:param model_type: model type
:param only_active: only active models
:param model: model name
:return:
"""
model_provider_factory = ModelProviderFactory(self.tenant_id)
@ -791,7 +792,10 @@ class ProviderConfiguration(BaseModel):
)
else:
provider_models = self._get_custom_provider_models(
model_types=model_types, provider_schema=provider_schema, model_setting_map=model_setting_map
model_types=model_types,
provider_schema=provider_schema,
model_setting_map=model_setting_map,
model=model,
)
if only_active:
@ -897,6 +901,7 @@ class ProviderConfiguration(BaseModel):
)
except Exception as ex:
logger.warning(f"get custom model schema failed, {ex}")
continue
if not custom_model_schema:
continue
@ -909,9 +914,7 @@ class ProviderConfiguration(BaseModel):
custom_model_schema.model_type in model_setting_map
and custom_model_schema.model in model_setting_map[custom_model_schema.model_type]
):
model_setting = model_setting_map[custom_model_schema.model_type][
custom_model_schema.model
]
model_setting = model_setting_map[custom_model_schema.model_type][custom_model_schema.model]
if model_setting.enabled is False:
status = ModelStatus.DISABLED
@ -944,6 +947,7 @@ class ProviderConfiguration(BaseModel):
model_types: Sequence[ModelType],
provider_schema: ProviderEntity,
model_setting_map: dict[ModelType, dict[str, ModelSettings]],
model: Optional[str] = None,
) -> list[ModelWithProviderEntity]:
"""
Get custom provider models.
@ -996,7 +1000,8 @@ class ProviderConfiguration(BaseModel):
for model_configuration in self.custom_configuration.models:
if model_configuration.model_type not in model_types:
continue
if model and model != model_configuration.model:
continue
try:
custom_model_schema = self.get_model_schema(
model_type=model_configuration.model_type,

34
api/core/indexing_runner.py
View File

@ -51,7 +51,7 @@ class IndexingRunner:
for dataset_document in dataset_documents:
try:
# get dataset
dataset = Dataset.query.filter_by(id=dataset_document.dataset_id).first()
dataset = db.session.query(Dataset).filter_by(id=dataset_document.dataset_id).first()
if not dataset:
raise ValueError("no dataset found")
@ -103,15 +103,17 @@ class IndexingRunner:
"""Run the indexing process when the index_status is splitting."""
try:
# get dataset
dataset = Dataset.query.filter_by(id=dataset_document.dataset_id).first()
dataset = db.session.query(Dataset).filter_by(id=dataset_document.dataset_id).first()
if not dataset:
raise ValueError("no dataset found")
# get exist document_segment list and delete
document_segments = DocumentSegment.query.filter_by(
dataset_id=dataset.id, document_id=dataset_document.id
).all()
document_segments = (
db.session.query(DocumentSegment)
.filter_by(dataset_id=dataset.id, document_id=dataset_document.id)
.all()
)
for document_segment in document_segments:
db.session.delete(document_segment)
@ -162,15 +164,17 @@ class IndexingRunner:
"""Run the indexing process when the index_status is indexing."""
try:
# get dataset
dataset = Dataset.query.filter_by(id=dataset_document.dataset_id).first()
dataset = db.session.query(Dataset).filter_by(id=dataset_document.dataset_id).first()
if not dataset:
raise ValueError("no dataset found")
# get exist document_segment list and delete
document_segments = DocumentSegment.query.filter_by(
dataset_id=dataset.id, document_id=dataset_document.id
).all()
document_segments = (
db.session.query(DocumentSegment)
.filter_by(dataset_id=dataset.id, document_id=dataset_document.id)
.all()
)
documents = []
if document_segments:
@ -254,7 +258,7 @@ class IndexingRunner:
embedding_model_instance = None
if dataset_id:
dataset = Dataset.query.filter_by(id=dataset_id).first()
dataset = db.session.query(Dataset).filter_by(id=dataset_id).first()
if not dataset:
raise ValueError("Dataset not found.")
if dataset.indexing_technique == "high_quality" or indexing_technique == "high_quality":
@ -587,7 +591,7 @@ class IndexingRunner:
@staticmethod
def _process_keyword_index(flask_app, dataset_id, document_id, documents):
with flask_app.app_context():
dataset = Dataset.query.filter_by(id=dataset_id).first()
dataset = db.session.query(Dataset).filter_by(id=dataset_id).first()
if not dataset:
raise ValueError("no dataset found")
keyword = Keyword(dataset)
@ -656,10 +660,10 @@ class IndexingRunner:
"""
Update the document indexing status.
"""
count = DatasetDocument.query.filter_by(id=document_id, is_paused=True).count()
count = db.session.query(DatasetDocument).filter_by(id=document_id, is_paused=True).count()
if count > 0:
raise DocumentIsPausedError()
document = DatasetDocument.query.filter_by(id=document_id).first()
document = db.session.query(DatasetDocument).filter_by(id=document_id).first()
if not document:
raise DocumentIsDeletedPausedError()
@ -668,7 +672,7 @@ class IndexingRunner:
if extra_update_params:
update_params.update(extra_update_params)
DatasetDocument.query.filter_by(id=document_id).update(update_params)
db.session.query(DatasetDocument).filter_by(id=document_id).update(update_params)
db.session.commit()
@staticmethod
@ -676,7 +680,7 @@ class IndexingRunner:
"""
Update the document segment by document id.
"""
DocumentSegment.query.filter_by(document_id=dataset_document_id).update(update_params)
db.session.query(DocumentSegment).filter_by(document_id=dataset_document_id).update(update_params)
db.session.commit()
def _transform(

6
api/core/llm_generator/llm_generator.py
View File

@ -51,15 +51,19 @@ class LLMGenerator:
response = cast(
LLMResult,
model_instance.invoke_llm(
prompt_messages=list(prompts), model_parameters={"max_tokens": 100, "temperature": 1}, stream=False
prompt_messages=list(prompts), model_parameters={"max_tokens": 500, "temperature": 1}, stream=False
),
)
answer = cast(str, response.message.content)
cleaned_answer = re.sub(r"^.*(\{.*\}).*$", r"\1", answer, flags=re.DOTALL)
if cleaned_answer is None:
return ""
try:
result_dict = json.loads(cleaned_answer)
answer = result_dict["Your Output"]
except json.JSONDecodeError as e:
logging.exception("Failed to generate name after answer, use query instead")
answer = query
name = answer.strip()
if len(name) > 75:

4
api/core/ops/entities/config_entity.py
View File

@ -1,9 +1,9 @@
from enum import Enum
from enum import StrEnum
from pydantic import BaseModel, ValidationInfo, field_validator
class TracingProviderEnum(Enum):
class TracingProviderEnum(StrEnum):
LANGFUSE = "langfuse"
LANGSMITH = "langsmith"
OPIK = "opik"

5
api/core/ops/langfuse_trace/entities/langfuse_trace_entity.py
View File

@ -1,3 +1,4 @@
from collections.abc import Mapping
from datetime import datetime
from enum import StrEnum
from typing import Any, Optional, Union
@ -155,10 +156,10 @@ class LangfuseSpan(BaseModel):
description="The status message of the span. Additional field for context of the event. E.g. the error "
"message of an error event.",
)
input: Optional[Union[str, dict[str, Any], list, None]] = Field(
input: Optional[Union[str, Mapping[str, Any], list, None]] = Field(
default=None, description="The input of the span. Can be any JSON object."
)
output: Optional[Union[str, dict[str, Any], list, None]] = Field(
output: Optional[Union[str, Mapping[str, Any], list, None]] = Field(
default=None, description="The output of the span. Can be any JSON object."
)
version: Optional[str] = Field(

52
api/core/ops/langfuse_trace/langfuse_trace.py
View File

@ -1,11 +1,10 @@
import json
import logging
import os
from datetime import datetime, timedelta
from typing import Optional
from langfuse import Langfuse # type: ignore
from sqlalchemy.orm import sessionmaker
from sqlalchemy.orm import Session, sessionmaker
from core.ops.base_trace_instance import BaseTraceInstance
from core.ops.entities.config_entity import LangfuseConfig
@ -29,9 +28,10 @@ from core.ops.langfuse_trace.entities.langfuse_trace_entity import (
UnitEnum,
)
from core.ops.utils import filter_none_values
from core.workflow.repository.repository_factory import RepositoryFactory
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.workflow.nodes.enums import NodeType
from extensions.ext_database import db
from models.model import EndUser
from models import Account, App, EndUser, WorkflowNodeExecutionTriggeredFrom
logger = logging.getLogger(__name__)
@ -113,8 +113,29 @@ class LangFuseDataTrace(BaseTraceInstance):
# through workflow_run_id get all_nodes_execution using repository
session_factory = sessionmaker(bind=db.engine)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={"tenant_id": trace_info.tenant_id, "session_factory": session_factory},
# Find the app's creator account
with Session(db.engine, expire_on_commit=False) as session:
# Get the app to find its creator
app_id = trace_info.metadata.get("app_id")
if not app_id:
raise ValueError("No app_id found in trace_info metadata")
app = session.query(App).filter(App.id == app_id).first()
if not app:
raise ValueError(f"App with id {app_id} not found")
if not app.created_by:
raise ValueError(f"App with id {app_id} has no creator (created_by is None)")
service_account = session.query(Account).filter(Account.id == app.created_by).first()
if not service_account:
raise ValueError(f"Creator account with id {app.created_by} not found for app {app_id}")
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=service_account,
app_id=trace_info.metadata.get("app_id"),
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
# Get all executions for this workflow run
@ -124,23 +145,22 @@ class LangFuseDataTrace(BaseTraceInstance):
for node_execution in workflow_node_executions:
node_execution_id = node_execution.id
tenant_id = node_execution.tenant_id
app_id = node_execution.app_id
tenant_id = trace_info.tenant_id # Use from trace_info instead
app_id = trace_info.metadata.get("app_id") # Use from trace_info instead
node_name = node_execution.title
node_type = node_execution.node_type
status = node_execution.status
if node_type == "llm":
inputs = (
json.loads(node_execution.process_data).get("prompts", {}) if node_execution.process_data else {}
)
if node_type == NodeType.LLM:
inputs = node_execution.process_data.get("prompts", {}) if node_execution.process_data else {}
else:
inputs = json.loads(node_execution.inputs) if node_execution.inputs else {}
outputs = json.loads(node_execution.outputs) if node_execution.outputs else {}
inputs = node_execution.inputs if node_execution.inputs else {}
outputs = node_execution.outputs if node_execution.outputs else {}
created_at = node_execution.created_at or datetime.now()
elapsed_time = node_execution.elapsed_time
finished_at = created_at + timedelta(seconds=elapsed_time)
metadata = json.loads(node_execution.execution_metadata) if node_execution.execution_metadata else {}
execution_metadata = node_execution.metadata if node_execution.metadata else {}
metadata = {str(k): v for k, v in execution_metadata.items()}
metadata.update(
{
"workflow_run_id": trace_info.workflow_run_id,
@ -152,7 +172,7 @@ class LangFuseDataTrace(BaseTraceInstance):
"status": status,
}
)
process_data = json.loads(node_execution.process_data) if node_execution.process_data else {}
process_data = node_execution.process_data if node_execution.process_data else {}
model_provider = process_data.get("model_provider", None)
model_name = process_data.get("model_name", None)
if model_provider is not None and model_name is not None:

5
api/core/ops/langsmith_trace/entities/langsmith_trace_entity.py
View File

@ -1,3 +1,4 @@
from collections.abc import Mapping
from datetime import datetime
from enum import StrEnum
from typing import Any, Optional, Union
@ -30,8 +31,8 @@ class LangSmithMultiModel(BaseModel):
class LangSmithRunModel(LangSmithTokenUsage, LangSmithMultiModel):
name: Optional[str] = Field(..., description="Name of the run")
inputs: Optional[Union[str, dict[str, Any], list, None]] = Field(None, description="Inputs of the run")
outputs: Optional[Union[str, dict[str, Any], list, None]] = Field(None, description="Outputs of the run")
inputs: Optional[Union[str, Mapping[str, Any], list, None]] = Field(None, description="Inputs of the run")
outputs: Optional[Union[str, Mapping[str, Any], list, None]] = Field(None, description="Outputs of the run")
run_type: LangSmithRunType = Field(..., description="Type of the run")
start_time: Optional[datetime | str] = Field(None, description="Start time of the run")
end_time: Optional[datetime | str] = Field(None, description="End time of the run")

64
api/core/ops/langsmith_trace/langsmith_trace.py
View File

@ -1,4 +1,3 @@
import json
import logging
import os
import uuid
@ -7,7 +6,7 @@ from typing import Optional, cast
from langsmith import Client
from langsmith.schemas import RunBase
from sqlalchemy.orm import sessionmaker
from sqlalchemy.orm import Session, sessionmaker
from core.ops.base_trace_instance import BaseTraceInstance
from core.ops.entities.config_entity import LangSmithConfig
@ -28,9 +27,11 @@ from core.ops.langsmith_trace.entities.langsmith_trace_entity import (
LangSmithRunUpdateModel,
)
from core.ops.utils import filter_none_values, generate_dotted_order
from core.workflow.repository.repository_factory import RepositoryFactory
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.workflow.entities.node_entities import NodeRunMetadataKey
from core.workflow.nodes.enums import NodeType
from extensions.ext_database import db
from models.model import EndUser, MessageFile
from models import Account, App, EndUser, MessageFile, WorkflowNodeExecutionTriggeredFrom
logger = logging.getLogger(__name__)
@ -137,12 +138,29 @@ class LangSmithDataTrace(BaseTraceInstance):
# through workflow_run_id get all_nodes_execution using repository
session_factory = sessionmaker(bind=db.engine)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": trace_info.tenant_id,
"app_id": trace_info.metadata.get("app_id"),
"session_factory": session_factory,
},
# Find the app's creator account
with Session(db.engine, expire_on_commit=False) as session:
# Get the app to find its creator
app_id = trace_info.metadata.get("app_id")
if not app_id:
raise ValueError("No app_id found in trace_info metadata")
app = session.query(App).filter(App.id == app_id).first()
if not app:
raise ValueError(f"App with id {app_id} not found")
if not app.created_by:
raise ValueError(f"App with id {app_id} has no creator (created_by is None)")
service_account = session.query(Account).filter(Account.id == app.created_by).first()
if not service_account:
raise ValueError(f"Creator account with id {app.created_by} not found for app {app_id}")
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=service_account,
app_id=trace_info.metadata.get("app_id"),
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
# Get all executions for this workflow run
@ -152,27 +170,23 @@ class LangSmithDataTrace(BaseTraceInstance):
for node_execution in workflow_node_executions:
node_execution_id = node_execution.id
tenant_id = node_execution.tenant_id
app_id = node_execution.app_id
tenant_id = trace_info.tenant_id # Use from trace_info instead
app_id = trace_info.metadata.get("app_id") # Use from trace_info instead
node_name = node_execution.title
node_type = node_execution.node_type
status = node_execution.status
if node_type == "llm":
inputs = (
json.loads(node_execution.process_data).get("prompts", {}) if node_execution.process_data else {}
)
if node_type == NodeType.LLM:
inputs = node_execution.process_data.get("prompts", {}) if node_execution.process_data else {}
else:
inputs = json.loads(node_execution.inputs) if node_execution.inputs else {}
outputs = json.loads(node_execution.outputs) if node_execution.outputs else {}
inputs = node_execution.inputs if node_execution.inputs else {}
outputs = node_execution.outputs if node_execution.outputs else {}
created_at = node_execution.created_at or datetime.now()
elapsed_time = node_execution.elapsed_time
finished_at = created_at + timedelta(seconds=elapsed_time)
execution_metadata = (
json.loads(node_execution.execution_metadata) if node_execution.execution_metadata else {}
)
node_total_tokens = execution_metadata.get("total_tokens", 0)
metadata = execution_metadata.copy()
execution_metadata = node_execution.metadata if node_execution.metadata else {}
node_total_tokens = execution_metadata.get(NodeRunMetadataKey.TOTAL_TOKENS) or 0
metadata = {str(key): value for key, value in execution_metadata.items()}
metadata.update(
{
"workflow_run_id": trace_info.workflow_run_id,
@ -185,7 +199,7 @@ class LangSmithDataTrace(BaseTraceInstance):
}
)
process_data = json.loads(node_execution.process_data) if node_execution.process_data else {}
process_data = node_execution.process_data if node_execution.process_data else {}
if process_data and process_data.get("model_mode") == "chat":
run_type = LangSmithRunType.llm
@ -195,7 +209,7 @@ class LangSmithDataTrace(BaseTraceInstance):
"ls_model_name": process_data.get("model_name", ""),
}
)
elif node_type == "knowledge-retrieval":
elif node_type == NodeType.KNOWLEDGE_RETRIEVAL:
run_type = LangSmithRunType.retriever
else:
run_type = LangSmithRunType.tool

66
api/core/ops/opik_trace/opik_trace.py
View File

@ -1,4 +1,3 @@
import json
import logging
import os
import uuid
@ -7,7 +6,7 @@ from typing import Optional, cast
from opik import Opik, Trace
from opik.id_helpers import uuid4_to_uuid7
from sqlalchemy.orm import sessionmaker
from sqlalchemy.orm import Session, sessionmaker
from core.ops.base_trace_instance import BaseTraceInstance
from core.ops.entities.config_entity import OpikConfig
@ -22,9 +21,11 @@ from core.ops.entities.trace_entity import (
TraceTaskName,
WorkflowTraceInfo,
)
from core.workflow.repository.repository_factory import RepositoryFactory
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.workflow.entities.node_entities import NodeRunMetadataKey
from core.workflow.nodes.enums import NodeType
from extensions.ext_database import db
from models.model import EndUser, MessageFile
from models import Account, App, EndUser, MessageFile, WorkflowNodeExecutionTriggeredFrom
logger = logging.getLogger(__name__)
@ -114,6 +115,7 @@ class OpikDataTrace(BaseTraceInstance):
"metadata": workflow_metadata,
"input": wrap_dict("input", trace_info.workflow_run_inputs),
"output": wrap_dict("output", trace_info.workflow_run_outputs),
"thread_id": trace_info.conversation_id,
"tags": ["message", "workflow"],
"project_name": self.project,
}
@ -143,6 +145,7 @@ class OpikDataTrace(BaseTraceInstance):
"metadata": workflow_metadata,
"input": wrap_dict("input", trace_info.workflow_run_inputs),
"output": wrap_dict("output", trace_info.workflow_run_outputs),
"thread_id": trace_info.conversation_id,
"tags": ["workflow"],
"project_name": self.project,
}
@ -150,12 +153,29 @@ class OpikDataTrace(BaseTraceInstance):
# through workflow_run_id get all_nodes_execution using repository
session_factory = sessionmaker(bind=db.engine)
workflow_node_execution_repository = RepositoryFactory.create_workflow_node_execution_repository(
params={
"tenant_id": trace_info.tenant_id,
"app_id": trace_info.metadata.get("app_id"),
"session_factory": session_factory,
},
# Find the app's creator account
with Session(db.engine, expire_on_commit=False) as session:
# Get the app to find its creator
app_id = trace_info.metadata.get("app_id")
if not app_id:
raise ValueError("No app_id found in trace_info metadata")
app = session.query(App).filter(App.id == app_id).first()
if not app:
raise ValueError(f"App with id {app_id} not found")
if not app.created_by:
raise ValueError(f"App with id {app_id} has no creator (created_by is None)")
service_account = session.query(Account).filter(Account.id == app.created_by).first()
if not service_account:
raise ValueError(f"Creator account with id {app.created_by} not found for app {app_id}")
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=service_account,
app_id=trace_info.metadata.get("app_id"),
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
# Get all executions for this workflow run
@ -165,26 +185,22 @@ class OpikDataTrace(BaseTraceInstance):
for node_execution in workflow_node_executions:
node_execution_id = node_execution.id
tenant_id = node_execution.tenant_id
app_id = node_execution.app_id
tenant_id = trace_info.tenant_id # Use from trace_info instead
app_id = trace_info.metadata.get("app_id") # Use from trace_info instead
node_name = node_execution.title
node_type = node_execution.node_type
status = node_execution.status
if node_type == "llm":
inputs = (
json.loads(node_execution.process_data).get("prompts", {}) if node_execution.process_data else {}
)
if node_type == NodeType.LLM:
inputs = node_execution.process_data.get("prompts", {}) if node_execution.process_data else {}
else:
inputs = json.loads(node_execution.inputs) if node_execution.inputs else {}
outputs = json.loads(node_execution.outputs) if node_execution.outputs else {}
inputs = node_execution.inputs if node_execution.inputs else {}
outputs = node_execution.outputs if node_execution.outputs else {}
created_at = node_execution.created_at or datetime.now()
elapsed_time = node_execution.elapsed_time
finished_at = created_at + timedelta(seconds=elapsed_time)
execution_metadata = (
json.loads(node_execution.execution_metadata) if node_execution.execution_metadata else {}
)
metadata = execution_metadata.copy()
execution_metadata = node_execution.metadata if node_execution.metadata else {}
metadata = {str(k): v for k, v in execution_metadata.items()}
metadata.update(
{
"workflow_run_id": trace_info.workflow_run_id,
@ -197,7 +213,7 @@ class OpikDataTrace(BaseTraceInstance):
}
)
process_data = json.loads(node_execution.process_data) if node_execution.process_data else {}
process_data = node_execution.process_data if node_execution.process_data else {}
provider = None
model = None
@ -230,7 +246,7 @@ class OpikDataTrace(BaseTraceInstance):
parent_span_id = trace_info.workflow_app_log_id or trace_info.workflow_run_id
if not total_tokens:
total_tokens = execution_metadata.get("total_tokens", 0)
total_tokens = execution_metadata.get(NodeRunMetadataKey.TOTAL_TOKENS) or 0
span_data = {
"trace_id": opik_trace_id,
@ -292,6 +308,7 @@ class OpikDataTrace(BaseTraceInstance):
"metadata": wrap_metadata(metadata),
"input": trace_info.inputs,
"output": message_data.answer,
"thread_id": message_data.conversation_id,
"tags": ["message", str(trace_info.conversation_mode)],
"project_name": self.project,
}
@ -406,6 +423,7 @@ class OpikDataTrace(BaseTraceInstance):
"metadata": wrap_metadata(trace_info.metadata),
"input": trace_info.inputs,
"output": trace_info.outputs,
"thread_id": trace_info.conversation_id,
"tags": ["generate_name"],
"project_name": self.project,
}

75
api/core/ops/ops_trace_manager.py
View File

@ -16,11 +16,7 @@ from sqlalchemy.orm import Session
from core.helper.encrypter import decrypt_token, encrypt_token, obfuscated_token
from core.ops.entities.config_entity import (
OPS_FILE_PATH,
LangfuseConfig,
LangSmithConfig,
OpikConfig,
TracingProviderEnum,
WeaveConfig,
)
from core.ops.entities.trace_entity import (
DatasetRetrievalTraceInfo,
@ -33,11 +29,8 @@ from core.ops.entities.trace_entity import (
TraceTaskName,
WorkflowTraceInfo,
)
from core.ops.langfuse_trace.langfuse_trace import LangFuseDataTrace
from core.ops.langsmith_trace.langsmith_trace import LangSmithDataTrace
from core.ops.opik_trace.opik_trace import OpikDataTrace
from core.ops.utils import get_message_data
from core.ops.weave_trace.weave_trace import WeaveDataTrace
from core.workflow.entities.workflow_execution_entities import WorkflowExecution
from extensions.ext_database import db
from extensions.ext_storage import storage
from models.model import App, AppModelConfig, Conversation, Message, MessageFile, TraceAppConfig
@ -45,37 +38,59 @@ from models.workflow import WorkflowAppLog, WorkflowRun
from tasks.ops_trace_task import process_trace_tasks
def build_opik_trace_instance(config: OpikConfig):
return OpikDataTrace(config)
class OpsTraceProviderConfigMap(dict[str, dict[str, Any]]):
def __getitem__(self, provider: str) -> dict[str, Any]:
match provider:
case TracingProviderEnum.LANGFUSE:
from core.ops.entities.config_entity import LangfuseConfig
from core.ops.langfuse_trace.langfuse_trace import LangFuseDataTrace
provider_config_map: dict[str, dict[str, Any]] = {
TracingProviderEnum.LANGFUSE.value: {
return {
"config_class": LangfuseConfig,
"secret_keys": ["public_key", "secret_key"],
"other_keys": ["host", "project_key"],
"trace_instance": LangFuseDataTrace,
},
TracingProviderEnum.LANGSMITH.value: {
}
case TracingProviderEnum.LANGSMITH:
from core.ops.entities.config_entity import LangSmithConfig
from core.ops.langsmith_trace.langsmith_trace import LangSmithDataTrace
return {
"config_class": LangSmithConfig,
"secret_keys": ["api_key"],
"other_keys": ["project", "endpoint"],
"trace_instance": LangSmithDataTrace,
},
TracingProviderEnum.OPIK.value: {
}
case TracingProviderEnum.OPIK:
from core.ops.entities.config_entity import OpikConfig
from core.ops.opik_trace.opik_trace import OpikDataTrace
return {
"config_class": OpikConfig,
"secret_keys": ["api_key"],
"other_keys": ["project", "url", "workspace"],
"trace_instance": lambda config: build_opik_trace_instance(config),
},
TracingProviderEnum.WEAVE.value: {
"trace_instance": OpikDataTrace,
}
case TracingProviderEnum.WEAVE:
from core.ops.entities.config_entity import WeaveConfig
from core.ops.weave_trace.weave_trace import WeaveDataTrace
return {
"config_class": WeaveConfig,
"secret_keys": ["api_key"],
"other_keys": ["project", "entity", "endpoint"],
"trace_instance": WeaveDataTrace,
},
}
case _:
raise KeyError(f"Unsupported tracing provider: {provider}")
provider_config_map: dict[str, dict[str, Any]] = OpsTraceProviderConfigMap()
class OpsTraceManager:
ops_trace_instances_cache: LRUCache = LRUCache(maxsize=128)
@ -220,7 +235,11 @@ class OpsTraceManager:
return None
tracing_provider = app_ops_trace_config.get("tracing_provider")
if tracing_provider is None or tracing_provider not in provider_config_map:
if tracing_provider is None:
return None
try:
provider_config_map[tracing_provider]
except KeyError:
return None
# decrypt_token
@ -273,7 +292,13 @@ class OpsTraceManager:
:return:
"""
# auth check
if tracing_provider not in provider_config_map and tracing_provider is not None:
if enabled == True:
try:
provider_config_map[tracing_provider]
except KeyError:
raise ValueError(f"Invalid tracing provider: {tracing_provider}")
else:
if tracing_provider is not None:
raise ValueError(f"Invalid tracing provider: {tracing_provider}")
app_config: Optional[App] = db.session.query(App).filter(App.id == app_id).first()
@ -353,7 +378,7 @@ class TraceTask:
self,
trace_type: Any,
message_id: Optional[str] = None,
workflow_run: Optional[WorkflowRun] = None,
workflow_execution: Optional[WorkflowExecution] = None,
conversation_id: Optional[str] = None,
user_id: Optional[str] = None,
timer: Optional[Any] = None,
@ -361,7 +386,7 @@ class TraceTask:
):
self.trace_type = trace_type
self.message_id = message_id
self.workflow_run_id = workflow_run.id if workflow_run else None
self.workflow_run_id = workflow_execution.id if workflow_execution else None
self.conversation_id = conversation_id
self.user_id = user_id
self.timer = timer

5
api/core/ops/weave_trace/entities/weave_trace_entity.py
View File

@ -1,3 +1,4 @@
from collections.abc import Mapping
from typing import Any, Optional, Union
from pydantic import BaseModel, Field, field_validator
@ -19,8 +20,8 @@ class WeaveMultiModel(BaseModel):
class WeaveTraceModel(WeaveTokenUsage, WeaveMultiModel):
id: str = Field(..., description="ID of the trace")
op: str = Field(..., description="Name of the operation")
inputs: Optional[Union[str, dict[str, Any], list, None]] = Field(None, description="Inputs of the trace")
outputs: Optional[Union[str, dict[str, Any], list, None]] = Field(None, description="Outputs of the trace")
inputs: Optional[Union[str, Mapping[str, Any], list, None]] = Field(None, description="Inputs of the trace")
outputs: Optional[Union[str, Mapping[str, Any], list, None]] = Field(None, description="Outputs of the trace")
attributes: Optional[Union[str, dict[str, Any], list, None]] = Field(
None, description="Metadata and attributes associated with trace"
)

87
api/core/ops/weave_trace/weave_trace.py
View File

@ -1,4 +1,3 @@
import json
import logging
import os
import uuid
@ -7,6 +6,7 @@ from typing import Any, Optional, cast
import wandb
import weave
from sqlalchemy.orm import Session, sessionmaker
from core.ops.base_trace_instance import BaseTraceInstance
from core.ops.entities.config_entity import WeaveConfig
@ -22,9 +22,11 @@ from core.ops.entities.trace_entity import (
WorkflowTraceInfo,
)
from core.ops.weave_trace.entities.weave_trace_entity import WeaveTraceModel
from core.repositories import SQLAlchemyWorkflowNodeExecutionRepository
from core.workflow.entities.node_entities import NodeRunMetadataKey
from core.workflow.nodes.enums import NodeType
from extensions.ext_database import db
from models.model import EndUser, MessageFile
from models.workflow import WorkflowNodeExecution
from models import Account, App, EndUser, MessageFile, WorkflowNodeExecutionTriggeredFrom
logger = logging.getLogger(__name__)
@ -128,58 +130,57 @@ class WeaveDataTrace(BaseTraceInstance):
self.start_call(workflow_run, parent_run_id=trace_info.message_id)
# through workflow_run_id get all_nodes_execution
workflow_nodes_execution_id_records = (
db.session.query(WorkflowNodeExecution.id)
.filter(WorkflowNodeExecution.workflow_run_id == trace_info.workflow_run_id)
.all()
# through workflow_run_id get all_nodes_execution using repository
session_factory = sessionmaker(bind=db.engine)
# Find the app's creator account
with Session(db.engine, expire_on_commit=False) as session:
# Get the app to find its creator
app_id = trace_info.metadata.get("app_id")
if not app_id:
raise ValueError("No app_id found in trace_info metadata")
app = session.query(App).filter(App.id == app_id).first()
if not app:
raise ValueError(f"App with id {app_id} not found")
if not app.created_by:
raise ValueError(f"App with id {app_id} has no creator (created_by is None)")
service_account = session.query(Account).filter(Account.id == app.created_by).first()
if not service_account:
raise ValueError(f"Creator account with id {app.created_by} not found for app {app_id}")
workflow_node_execution_repository = SQLAlchemyWorkflowNodeExecutionRepository(
session_factory=session_factory,
user=service_account,
app_id=trace_info.metadata.get("app_id"),
triggered_from=WorkflowNodeExecutionTriggeredFrom.WORKFLOW_RUN,
)
for node_execution_id_record in workflow_nodes_execution_id_records:
node_execution = (
db.session.query(
WorkflowNodeExecution.id,
WorkflowNodeExecution.tenant_id,
WorkflowNodeExecution.app_id,
WorkflowNodeExecution.title,
WorkflowNodeExecution.node_type,
WorkflowNodeExecution.status,
WorkflowNodeExecution.inputs,
WorkflowNodeExecution.outputs,
WorkflowNodeExecution.created_at,
WorkflowNodeExecution.elapsed_time,
WorkflowNodeExecution.process_data,
WorkflowNodeExecution.execution_metadata,
)
.filter(WorkflowNodeExecution.id == node_execution_id_record.id)
.first()
# Get all executions for this workflow run
workflow_node_executions = workflow_node_execution_repository.get_by_workflow_run(
workflow_run_id=trace_info.workflow_run_id
)
if not node_execution:
continue
for node_execution in workflow_node_executions:
node_execution_id = node_execution.id
tenant_id = node_execution.tenant_id
app_id = node_execution.app_id
tenant_id = trace_info.tenant_id # Use from trace_info instead
app_id = trace_info.metadata.get("app_id") # Use from trace_info instead
node_name = node_execution.title
node_type = node_execution.node_type
status = node_execution.status
if node_type == "llm":
inputs = (
json.loads(node_execution.process_data).get("prompts", {}) if node_execution.process_data else {}
)
if node_type == NodeType.LLM:
inputs = node_execution.process_data.get("prompts", {}) if node_execution.process_data else {}
else:
inputs = json.loads(node_execution.inputs) if node_execution.inputs else {}
outputs = json.loads(node_execution.outputs) if node_execution.outputs else {}
inputs = node_execution.inputs if node_execution.inputs else {}
outputs = node_execution.outputs if node_execution.outputs else {}
created_at = node_execution.created_at or datetime.now()
elapsed_time = node_execution.elapsed_time
finished_at = created_at + timedelta(seconds=elapsed_time)
execution_metadata = (
json.loads(node_execution.execution_metadata) if node_execution.execution_metadata else {}
)
node_total_tokens = execution_metadata.get("total_tokens", 0)
attributes = execution_metadata.copy()
execution_metadata = node_execution.metadata if node_execution.metadata else {}
node_total_tokens = execution_metadata.get(NodeRunMetadataKey.TOTAL_TOKENS) or 0
attributes = {str(k): v for k, v in execution_metadata.items()}
attributes.update(
{
"workflow_run_id": trace_info.workflow_run_id,
@ -192,7 +193,7 @@ class WeaveDataTrace(BaseTraceInstance):
}
)
process_data = json.loads(node_execution.process_data) if node_execution.process_data else {}
process_data = node_execution.process_data if node_execution.process_data else {}
if process_data and process_data.get("model_mode") == "chat":
attributes.update(
{

12
api/core/plugin/backwards_invocation/node.py
View File

@ -64,9 +64,9 @@ class PluginNodeBackwardsInvocation(BaseBackwardsInvocation):
)
return {
"inputs": execution.inputs_dict,
"outputs": execution.outputs_dict,
"process_data": execution.process_data_dict,
"inputs": execution.inputs,
"outputs": execution.outputs,
"process_data": execution.process_data,
}
@classmethod
@ -113,7 +113,7 @@ class PluginNodeBackwardsInvocation(BaseBackwardsInvocation):
)
return {
"inputs": execution.inputs_dict,
"outputs": execution.outputs_dict,
"process_data": execution.process_data_dict,
"inputs": execution.inputs,
"outputs": execution.outputs,
"process_data": execution.process_data,
}

Some files were not shown because too many files have changed in this diff Show More