gmakstutis/openrag
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge branch 'main' into issue-502-monolithic-pages

Browse source
This commit is contained in:
April M 2025-11-26 16:30:02 -08:00
commit 0f02b5fbc0
49 changed files with 13012 additions and 5541 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

77
.github/workflows/build-multiarch.yml vendored
View file

@ -53,58 +53,63 @@ jobs:
# backend # backend
- image: backend - image: backend
file: ./Dockerfile.backend file: ./Dockerfile.backend
tag: phact/openrag-backend tag: langflowai/openrag-backend
platform: linux/amd64 platform: linux/amd64
arch: amd64 arch: amd64
runs-on: ubuntu-latest-16-cores runs-on: ubuntu-latest-16-cores
- image: backend - image: backend
file: ./Dockerfile.backend file: ./Dockerfile.backend
tag: phact/openrag-backend tag: langflowai/openrag-backend
platform: linux/arm64 platform: linux/arm64
arch: arm64 arch: arm64
runs-on: [self-hosted, linux, ARM64, langflow-ai-arm64-2] #runs-on: [self-hosted, linux, ARM64, langflow-ai-arm64-2]
runs-on: RagRunner
# frontend # frontend
- image: frontend - image: frontend
file: ./Dockerfile.frontend file: ./Dockerfile.frontend
tag: phact/openrag-frontend tag: langflowai/openrag-frontend
platform: linux/amd64 platform: linux/amd64
arch: amd64 arch: amd64
runs-on: ubuntu-latest-16-cores runs-on: ubuntu-latest-16-cores
- image: frontend - image: frontend
file: ./Dockerfile.frontend file: ./Dockerfile.frontend
tag: phact/openrag-frontend tag: langflowai/openrag-frontend
platform: linux/arm64 platform: linux/arm64
arch: arm64 arch: arm64
runs-on: [self-hosted, linux, ARM64, langflow-ai-arm64-2] #runs-on: [self-hosted, linux, ARM64, langflow-ai-arm64-2]
runs-on: RagRunner
# langflow # langflow
- image: langflow - image: langflow
file: ./Dockerfile.langflow file: ./Dockerfile.langflow
tag: phact/openrag-langflow tag: langflowai/openrag-langflow
platform: linux/amd64 platform: linux/amd64
arch: amd64 arch: amd64
runs-on: ubuntu-latest-16-cores runs-on: ubuntu-latest-16-cores
- image: langflow - image: langflow
file: ./Dockerfile.langflow file: ./Dockerfile.langflow
tag: phact/openrag-langflow tag: langflowai/openrag-langflow
platform: linux/arm64 platform: linux/arm64
arch: arm64 arch: arm64
runs-on: self-hosted #runs-on: self-hosted
runs-on: RagRunner
# opensearch # opensearch
- image: opensearch - image: opensearch
file: ./Dockerfile file: ./Dockerfile
tag: phact/openrag-opensearch tag: langflowai/openrag-opensearch
platform: linux/amd64 platform: linux/amd64
arch: amd64 arch: amd64
runs-on: ubuntu-latest-16-cores runs-on: ubuntu-latest-16-cores
- image: opensearch - image: opensearch
file: ./Dockerfile file: ./Dockerfile
tag: phact/openrag-opensearch tag: langflowai/openrag-opensearch
platform: linux/arm64 platform: linux/arm64
arch: arm64 arch: arm64
runs-on: [self-hosted, linux, ARM64, langflow-ai-arm64-2] #runs-on: [self-hosted, linux, ARM64, langflow-ai-arm64-2]
#runs-on: self-hosted
runs-on: RagRunner
runs-on: ${{ matrix.runs-on }} runs-on: ${{ matrix.runs-on }}
@ -165,40 +170,40 @@ jobs:
VERSION=${{ steps.version.outputs.version }} VERSION=${{ steps.version.outputs.version }}
# Create versioned tags # Create versioned tags
docker buildx imagetools create -t phact/openrag-backend:$VERSION \ docker buildx imagetools create -t langflowai/openrag-backend:$VERSION \
phact/openrag-backend:$VERSION-amd64 \ langflowai/openrag-backend:$VERSION-amd64 \
phact/openrag-backend:$VERSION-arm64 langflowai/openrag-backend:$VERSION-arm64
docker buildx imagetools create -t phact/openrag-frontend:$VERSION \ docker buildx imagetools create -t langflowai/openrag-frontend:$VERSION \
phact/openrag-frontend:$VERSION-amd64 \ langflowai/openrag-frontend:$VERSION-amd64 \
phact/openrag-frontend:$VERSION-arm64 langflowai/openrag-frontend:$VERSION-arm64
docker buildx imagetools create -t phact/openrag-langflow:$VERSION \ docker buildx imagetools create -t langflowai/openrag-langflow:$VERSION \
phact/openrag-langflow:$VERSION-amd64 \ langflowai/openrag-langflow:$VERSION-amd64 \
phact/openrag-langflow:$VERSION-arm64 langflowai/openrag-langflow:$VERSION-arm64
docker buildx imagetools create -t phact/openrag-opensearch:$VERSION \ docker buildx imagetools create -t langflowai/openrag-opensearch:$VERSION \
phact/openrag-opensearch:$VERSION-amd64 \ langflowai/openrag-opensearch:$VERSION-amd64 \
phact/openrag-opensearch:$VERSION-arm64 langflowai/openrag-opensearch:$VERSION-arm64
# Only update latest tags if version is numeric # Only update latest tags if version is numeric
if [[ "$VERSION" =~ ^[0-9.-]+$ ]]; then if [[ "$VERSION" =~ ^[0-9.-]+$ ]]; then
echo "Updating latest tags for production release: $VERSION" echo "Updating latest tags for production release: $VERSION"
docker buildx imagetools create -t phact/openrag-backend:latest \ docker buildx imagetools create -t langflowai/openrag-backend:latest \
phact/openrag-backend:$VERSION-amd64 \ langflowai/openrag-backend:$VERSION-amd64 \
phact/openrag-backend:$VERSION-arm64 langflowai/openrag-backend:$VERSION-arm64
docker buildx imagetools create -t phact/openrag-frontend:latest \ docker buildx imagetools create -t langflowai/openrag-frontend:latest \
phact/openrag-frontend:$VERSION-amd64 \ langflowai/openrag-frontend:$VERSION-amd64 \
phact/openrag-frontend:$VERSION-arm64 langflowai/openrag-frontend:$VERSION-arm64
docker buildx imagetools create -t phact/openrag-langflow:latest \ docker buildx imagetools create -t langflowai/openrag-langflow:latest \
phact/openrag-langflow:$VERSION-amd64 \ langflowai/openrag-langflow:$VERSION-amd64 \
phact/openrag-langflow:$VERSION-arm64 langflowai/openrag-langflow:$VERSION-arm64
docker buildx imagetools create -t phact/openrag-opensearch:latest \ docker buildx imagetools create -t langflowai/openrag-opensearch:latest \
phact/openrag-opensearch:$VERSION-amd64 \ langflowai/openrag-opensearch:$VERSION-amd64 \
phact/openrag-opensearch:$VERSION-arm64 langflowai/openrag-opensearch:$VERSION-arm64
else else
echo "Skipping latest tags - version: $VERSION (not numeric)" echo "Skipping latest tags - version: $VERSION (not numeric)"
fi fi

19
.github/workflows/test-integration.yml vendored
View file

@ -31,14 +31,23 @@ jobs:
steps: steps:
- run: df -h - run: df -h
#- name: "node-cleanup"
#run: | - name: Cleanup Docker cache
# sudo rm -rf /usr/share/dotnet /usr/local/lib/android /opt/ghc /opt/hostedtoolcache/CodeQL run: |
# sudo docker image prune --all --force docker system prune -af || true
# sudo docker builder prune -a docker builder prune -af || true
docker-compose -f docker-compose.yml down -v --remove-orphans || true
- run: df -h - run: df -h
- name: Checkout - name: Checkout
uses: actions/checkout@v4 uses: actions/checkout@v4
- name: Verify workspace
run: |
echo "Current directory: $(pwd)"
echo "Workspace: ${GITHUB_WORKSPACE}"
ls -la
- name: Set up UV - name: Set up UV
uses: astral-sh/setup-uv@v3 uses: astral-sh/setup-uv@v3

2
Dockerfile.backend
View file

@ -21,7 +21,7 @@ COPY pyproject.toml uv.lock ./
RUN uv sync RUN uv sync
# Copy sample document and warmup script for docling # Copy sample document and warmup script for docling
COPY documents/warmup_ocr.pdf ./ COPY openrag-documents/warmup_ocr.pdf ./
COPY warm_up_docling.py ./ COPY warm_up_docling.py ./
RUN uv run docling-tools models download RUN uv run docling-tools models download
RUN uv run python - <<'PY' RUN uv run python - <<'PY'

2
Dockerfile.langflow
View file

@ -1,4 +1,4 @@
FROM langflowai/langflow-nightly:1.7.0.dev5 FROM langflowai/langflow-nightly:1.7.0.dev21
EXPOSE 7860 EXPOSE 7860

10
Makefile
View file

@ -210,7 +210,7 @@ test-ci:
echo "Pulling latest images..."; \ echo "Pulling latest images..."; \
docker compose -f docker-compose-cpu.yml pull; \ docker compose -f docker-compose-cpu.yml pull; \
echo "Building OpenSearch image override..."; \ echo "Building OpenSearch image override..."; \
docker build --no-cache -t phact/openrag-opensearch:latest -f Dockerfile .; \ docker build --no-cache -t langflowai/openrag-opensearch:latest -f Dockerfile .; \
echo "Starting infra (OpenSearch + Dashboards + Langflow) with CPU containers"; \ echo "Starting infra (OpenSearch + Dashboards + Langflow) with CPU containers"; \
docker compose -f docker-compose-cpu.yml up -d opensearch dashboards langflow; \ docker compose -f docker-compose-cpu.yml up -d opensearch dashboards langflow; \
echo "Starting docling-serve..."; \ echo "Starting docling-serve..."; \
@ -288,10 +288,10 @@ test-ci-local:
echo "Cleaning up old containers and volumes..."; \ echo "Cleaning up old containers and volumes..."; \
docker compose -f docker-compose-cpu.yml down -v 2>/dev/null || true; \ docker compose -f docker-compose-cpu.yml down -v 2>/dev/null || true; \
echo "Building all images locally..."; \ echo "Building all images locally..."; \
docker build -t phact/openrag-opensearch:latest -f Dockerfile .; \ docker build -t langflowai/openrag-opensearch:latest -f Dockerfile .; \
docker build -t phact/openrag-backend:latest -f Dockerfile.backend .; \ docker build -t langflowai/openrag-backend:latest -f Dockerfile.backend .; \
docker build -t phact/openrag-frontend:latest -f Dockerfile.frontend .; \ docker build -t langflowai/openrag-frontend:latest -f Dockerfile.frontend .; \
docker build -t phact/openrag-langflow:latest -f Dockerfile.langflow .; \ docker build -t langflowai/openrag-langflow:latest -f Dockerfile.langflow .; \
echo "Starting infra (OpenSearch + Dashboards + Langflow) with CPU containers"; \ echo "Starting infra (OpenSearch + Dashboards + Langflow) with CPU containers"; \
docker compose -f docker-compose-cpu.yml up -d opensearch dashboards langflow; \ docker compose -f docker-compose-cpu.yml up -d opensearch dashboards langflow; \
echo "Starting docling-serve..."; \ echo "Starting docling-serve..."; \

42
README.md
View file

@ -18,27 +18,29 @@ OpenRAG is a comprehensive Retrieval-Augmented Generation platform that enables
</div> </div>
<div align="center"> <div align="center">
<a href="#quickstart" style="color: #0366d6;">Quickstart</a> &nbsp;&nbsp;|&nbsp;&nbsp; <a href="#quickstart" style="color: #0366d6;">Quickstart</a> &nbsp;&nbsp;|&nbsp;&nbsp;
<a href="#tui-interface" style="color: #0366d6;">TUI Interface</a> &nbsp;&nbsp;|&nbsp;&nbsp; <a href="#install-python-package" style="color: #0366d6;">Python package</a> &nbsp;&nbsp;|&nbsp;&nbsp;
<a href="#docker-deployment" style="color: #0366d6;">Docker Deployment</a> &nbsp;&nbsp;|&nbsp;&nbsp; <a href="#docker-or-podman-installation" style="color: #0366d6;">Docker or Podman</a> &nbsp;&nbsp;|&nbsp;&nbsp;
<a href="#development" style="color: #0366d6;">Development</a> &nbsp;&nbsp;|&nbsp;&nbsp; <a href="#development" style="color: #0366d6;">Development</a> &nbsp;&nbsp;|&nbsp;&nbsp;
<a href="#troubleshooting" style="color: #0366d6;">Troubleshooting</a> <a href="#troubleshooting" style="color: #0366d6;">Troubleshooting</a>
</div> </div>
## Quickstart ## Quickstart
To quickly run OpenRAG without creating or modifying any project files, use `uvx`: To run OpenRAG without creating or modifying any project files, use `uvx`:
```bash ```bash
uvx openrag uvx openrag
``` ```
This runs OpenRAG without installing it to your project or globally.
To run a specific version of OpenRAG, add the version to the command, such as: `uvx --from openrag==0.1.25 openrag`. This command runs OpenRAG without installing it to your project or globally.
To run a specific version of OpenRAG, run `uvx --from openrag==VERSION openrag`.
## Install Python package ## Install Python package
To first set up a project and then install the OpenRAG Python package, do the following: To add the OpenRAG Python package to a Python project, use `uv`:
1. Create a new project with a virtual environment using `uv init`. 1. Create a new project with a virtual environment using `uv init`:
```bash ```bash
uv init YOUR_PROJECT_NAME uv init YOUR_PROJECT_NAME
@ -48,33 +50,33 @@ To first set up a project and then install the OpenRAG Python package, do the fo
The `(venv)` prompt doesn't change, but `uv` commands will automatically use the project's virtual environment. The `(venv)` prompt doesn't change, but `uv` commands will automatically use the project's virtual environment.
For more information on virtual environments, see the [uv documentation](https://docs.astral.sh/uv/pip/environments). For more information on virtual environments, see the [uv documentation](https://docs.astral.sh/uv/pip/environments).
2. Add OpenRAG to your project. 2. Add OpenRAG to your project:
```bash ```bash
uv add openrag uv add openrag
``` ```
To add a specific version of OpenRAG: To add a specific version of OpenRAG, run `uv add openrag==VERSION`.
```bash
uv add openrag==0.1.25 3. Start the OpenRAG terminal user interface (TUI):
```
3. Start the OpenRAG TUI.
```bash ```bash
uv run openrag uv run openrag
``` ```
4. Continue with the [Quickstart](https://docs.openr.ag/quickstart). 4. Continue with the [Quickstart](https://docs.openr.ag/quickstart).
For the full TUI installation guide, see [TUI](https://docs.openr.ag/install). For all installation options, see the [OpenRAG installation guide](https://docs.openr.ag/install).
## Docker or Podman installation ## Docker or Podman installation
For more information, see [Install OpenRAG containers](https://docs.openr.ag/docker). By default, OpenRAG automatically starts the required containers and helps you manage them.
To install OpenRAG with self-managed containers, see the [OpenRAG installation guide](https://docs.openr.ag/docker).
## Troubleshooting
For common issues and fixes, see [Troubleshoot](https://docs.openr.ag/support/troubleshoot).
## Development ## Development
For developers wanting to contribute to OpenRAG or set up a development environment, see [CONTRIBUTING.md](CONTRIBUTING.md). For developers wanting to contribute to OpenRAG or set up a development environment, see [CONTRIBUTING.md](CONTRIBUTING.md).
## Troubleshooting
For common issues and fixes, see [Troubleshoot OpenRAG](https://docs.openr.ag/support/troubleshoot).

13
docker-compose-cpu.yml
View file

@ -1,6 +1,6 @@
services: services:
opensearch: opensearch:
image: phact/openrag-opensearch:${OPENRAG_VERSION:-latest} image: langflowai/openrag-opensearch:${OPENRAG_VERSION:-latest}
#build: #build:
# context: . # context: .
# dockerfile: Dockerfile # dockerfile: Dockerfile
@ -44,7 +44,7 @@ services:
- "5601:5601" - "5601:5601"
openrag-backend: openrag-backend:
image: phact/openrag-backend:${OPENRAG_VERSION:-latest} image: langflowai/openrag-backend:${OPENRAG_VERSION:-latest}
# build: # build:
# context: . # context: .
# dockerfile: Dockerfile.backend # dockerfile: Dockerfile.backend
@ -81,12 +81,12 @@ services:
- AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID} - AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID}
- AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY} - AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY}
volumes: volumes:
- ./documents:/app/documents:Z - ./openrag-documents:/app/documents:Z
- ./keys:/app/keys:Z - ./keys:/app/keys:Z
- ./flows:/app/flows:U,z - ./flows:/app/flows:U,z
openrag-frontend: openrag-frontend:
image: phact/openrag-frontend:${OPENRAG_VERSION:-latest} image: langflowai/openrag-frontend:${OPENRAG_VERSION:-latest}
# build: # build:
# context: . # context: .
# dockerfile: Dockerfile.frontend # dockerfile: Dockerfile.frontend
@ -101,7 +101,7 @@ services:
langflow: langflow:
volumes: volumes:
- ./flows:/app/flows:U,z - ./flows:/app/flows:U,z
image: phact/openrag-langflow:${LANGFLOW_VERSION:-latest} image: langflowai/openrag-langflow:${LANGFLOW_VERSION:-latest}
# build: # build:
# context: . # context: .
# dockerfile: Dockerfile.langflow # dockerfile: Dockerfile.langflow
@ -129,7 +129,8 @@ services:
- FILENAME=None - FILENAME=None
- MIMETYPE=None - MIMETYPE=None
- FILESIZE=0 - FILESIZE=0
- LANGFLOW_VARIABLES_TO_GET_FROM_ENVIRONMENT=JWT,OPENRAG-QUERY-FILTER,OPENSEARCH_PASSWORD,OWNER,OWNER_NAME,OWNER_EMAIL,CONNECTOR_TYPE,FILENAME,MIMETYPE,FILESIZE - SELECTED_EMBEDDING_MODEL=${SELECTED_EMBEDDING_MODEL:-}
- LANGFLOW_VARIABLES_TO_GET_FROM_ENVIRONMENT=JWT,OPENRAG-QUERY-FILTER,OPENSEARCH_PASSWORD,OWNER,OWNER_NAME,OWNER_EMAIL,CONNECTOR_TYPE,FILENAME,MIMETYPE,FILESIZE,SELECTED_EMBEDDING_MODEL
- LANGFLOW_LOG_LEVEL=DEBUG - LANGFLOW_LOG_LEVEL=DEBUG
- LANGFLOW_AUTO_LOGIN=${LANGFLOW_AUTO_LOGIN} - LANGFLOW_AUTO_LOGIN=${LANGFLOW_AUTO_LOGIN}
- LANGFLOW_SUPERUSER=${LANGFLOW_SUPERUSER} - LANGFLOW_SUPERUSER=${LANGFLOW_SUPERUSER}

7
docker-compose.gpu.yml Normal file
View file

@ -0,0 +1,7 @@
services:
openrag-backend:
environment:
- NVIDIA_DRIVER_CAPABILITIES=compute,utility
- NVIDIA_VISIBLE_DEVICES=all
gpus: all

47
docker-compose.yml
View file

@ -1,6 +1,6 @@
services: services:
opensearch: opensearch:
image: phact/openrag-opensearch:${OPENRAG_VERSION:-latest} image: langflowai/openrag-opensearch:${OPENRAG_VERSION:-latest}
#build: #build:
#context: . #context: .
#dockerfile: Dockerfile #dockerfile: Dockerfile
@ -44,10 +44,10 @@ services:
- "5601:5601" - "5601:5601"
openrag-backend: openrag-backend:
image: phact/openrag-backend:${OPENRAG_VERSION:-latest} image: langflowai/openrag-backend:${OPENRAG_VERSION:-latest}
# build: build:
# context: . context: .
# dockerfile: Dockerfile.backend dockerfile: Dockerfile.backend
container_name: openrag-backend container_name: openrag-backend
depends_on: depends_on:
- langflow - langflow
@ -72,8 +72,6 @@ services:
- WATSONX_ENDPOINT=${WATSONX_ENDPOINT} - WATSONX_ENDPOINT=${WATSONX_ENDPOINT}
- WATSONX_PROJECT_ID=${WATSONX_PROJECT_ID} - WATSONX_PROJECT_ID=${WATSONX_PROJECT_ID}
- OLLAMA_ENDPOINT=${OLLAMA_ENDPOINT} - OLLAMA_ENDPOINT=${OLLAMA_ENDPOINT}
- NVIDIA_DRIVER_CAPABILITIES=compute,utility
- NVIDIA_VISIBLE_DEVICES=all
- GOOGLE_OAUTH_CLIENT_ID=${GOOGLE_OAUTH_CLIENT_ID} - GOOGLE_OAUTH_CLIENT_ID=${GOOGLE_OAUTH_CLIENT_ID}
- GOOGLE_OAUTH_CLIENT_SECRET=${GOOGLE_OAUTH_CLIENT_SECRET} - GOOGLE_OAUTH_CLIENT_SECRET=${GOOGLE_OAUTH_CLIENT_SECRET}
- MICROSOFT_GRAPH_OAUTH_CLIENT_ID=${MICROSOFT_GRAPH_OAUTH_CLIENT_ID} - MICROSOFT_GRAPH_OAUTH_CLIENT_ID=${MICROSOFT_GRAPH_OAUTH_CLIENT_ID}
@ -82,16 +80,15 @@ services:
- AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID} - AWS_ACCESS_KEY_ID=${AWS_ACCESS_KEY_ID}
- AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY} - AWS_SECRET_ACCESS_KEY=${AWS_SECRET_ACCESS_KEY}
volumes: volumes:
- ./documents:/app/documents:Z - ./openrag-documents:/app/documents:Z
- ./keys:/app/keys:Z - ./keys:/app/keys:Z
- ./flows:/app/flows:U,z - ./flows:/app/flows:U,z
gpus: all
openrag-frontend: openrag-frontend:
image: phact/openrag-frontend:${OPENRAG_VERSION:-latest} image: langflowai/openrag-frontend:${OPENRAG_VERSION:-latest}
# build: build:
# context: . context: .
# dockerfile: Dockerfile.frontend dockerfile: Dockerfile.frontend
container_name: openrag-frontend container_name: openrag-frontend
depends_on: depends_on:
- openrag-backend - openrag-backend
@ -103,21 +100,21 @@ services:
langflow: langflow:
volumes: volumes:
- ./flows:/app/flows:U,z - ./flows:/app/flows:U,z
image: phact/openrag-langflow:${LANGFLOW_VERSION:-latest} image: langflowai/openrag-langflow:${LANGFLOW_VERSION:-latest}
# build: build:
# context: . context: .
# dockerfile: Dockerfile.langflow dockerfile: Dockerfile.langflow
container_name: langflow container_name: langflow
ports: ports:
- "7860:7860" - "7860:7860"
environment: environment:
- LANGFLOW_DEACTIVATE_TRACING=true - LANGFLOW_DEACTIVATE_TRACING=true
- OPENAI_API_KEY=${OPENAI_API_KEY} - OPENAI_API_KEY=${OPENAI_API_KEY:-None}
- ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY} - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-None}
- WATSONX_API_KEY=${WATSONX_API_KEY} - WATSONX_API_KEY=${WATSONX_API_KEY:-None}
- WATSONX_ENDPOINT=${WATSONX_ENDPOINT} - WATSONX_ENDPOINT=${WATSONX_ENDPOINT:-None}
- WATSONX_PROJECT_ID=${WATSONX_PROJECT_ID} - WATSONX_PROJECT_ID=${WATSONX_PROJECT_ID:-None}
- OLLAMA_BASE_URL=${OLLAMA_ENDPOINT} - OLLAMA_BASE_URL=${OLLAMA_ENDPOINT:-None}
- LANGFLOW_LOAD_FLOWS_PATH=/app/flows - LANGFLOW_LOAD_FLOWS_PATH=/app/flows
- LANGFLOW_SECRET_KEY=${LANGFLOW_SECRET_KEY} - LANGFLOW_SECRET_KEY=${LANGFLOW_SECRET_KEY}
- JWT=None - JWT=None
@ -127,11 +124,13 @@ services:
- CONNECTOR_TYPE=system - CONNECTOR_TYPE=system
- CONNECTOR_TYPE_URL=url - CONNECTOR_TYPE_URL=url
- OPENRAG-QUERY-FILTER="{}" - OPENRAG-QUERY-FILTER="{}"
- OPENSEARCH_PASSWORD=${OPENSEARCH_PASSWORD}
- FILENAME=None - FILENAME=None
- MIMETYPE=None - MIMETYPE=None
- FILESIZE=0 - FILESIZE=0
- SELECTED_EMBEDDING_MODEL=${SELECTED_EMBEDDING_MODEL:-}
- OPENSEARCH_PASSWORD=${OPENSEARCH_PASSWORD} - OPENSEARCH_PASSWORD=${OPENSEARCH_PASSWORD}
- LANGFLOW_VARIABLES_TO_GET_FROM_ENVIRONMENT=JWT,OPENRAG-QUERY-FILTER,OPENSEARCH_PASSWORD,OWNER,OWNER_NAME,OWNER_EMAIL,CONNECTOR_TYPE,FILENAME,MIMETYPE,FILESIZE - LANGFLOW_VARIABLES_TO_GET_FROM_ENVIRONMENT=JWT,OPENRAG-QUERY-FILTER,OPENSEARCH_PASSWORD,OWNER,OWNER_NAME,OWNER_EMAIL,CONNECTOR_TYPE,FILENAME,MIMETYPE,FILESIZE,SELECTED_EMBEDDING_MODEL,OPENAI_API_KEY,ANTHROPIC_API_KEY,WATSONX_API_KEY,WATSONX_ENDPOINT,WATSONX_PROJECT_ID,OLLAMA_BASE_URL
- LANGFLOW_LOG_LEVEL=DEBUG - LANGFLOW_LOG_LEVEL=DEBUG
- LANGFLOW_AUTO_LOGIN=${LANGFLOW_AUTO_LOGIN} - LANGFLOW_AUTO_LOGIN=${LANGFLOW_AUTO_LOGIN}
- LANGFLOW_SUPERUSER=${LANGFLOW_SUPERUSER} - LANGFLOW_SUPERUSER=${LANGFLOW_SUPERUSER}

4
docs/README.md
View file

@ -42,7 +42,7 @@ If you are using GitHub pages for hosting, this command is a convenient way to b
## Update the OpenRAG documentation PDF ## Update the OpenRAG documentation PDF
The documentation PDF at `openrag/documents/openrag-documentation.pdf` is used by the OpenRAG application, so keep it up to date. The documentation PDF at `openrag/openrag-documents/openrag-documentation.pdf` is used by the OpenRAG application, so keep it up to date.
To update the PDF, do the following: To update the PDF, do the following:
@ -68,7 +68,7 @@ To remove these items, give the following prompt or something similar to your ID
2. Check your `.mdx` files to confirm these elements are removed. 2. Check your `.mdx` files to confirm these elements are removed.
Don't commit the changes. Don't commit the changes.
3. From `openrag/docs`, run this command to build the site with the changes, and create a PDF at `openrag/documents`. 3. From `openrag/docs`, run this command to build the site with the changes, and create a PDF at `openrag/openrag-documents`.
``` ```
npm run build:pdf npm run build:pdf

2
docs/docs/core-components/knowledge.mdx
View file

@ -140,7 +140,7 @@ The default value is 200 characters, which represents an overlap of 20 percent i
### Set the local documents path {#set-the-local-documents-path} ### Set the local documents path {#set-the-local-documents-path}
The default path for local uploads is the `./documents` subdirectory in your OpenRAG installation directory. This is mounted to the `/app/documents/` directory inside the OpenRAG container. Files added to the host or container directory are visible in both locations. The default path for local uploads is the `./openrag-documents` subdirectory in your OpenRAG installation directory. This is mounted to the `/app/documents/` directory inside the OpenRAG container. Files added to the host or container directory are visible in both locations.
To change this location, modify the **Documents Paths** variable in either the [**Advanced Setup** menu](/install#setup) or in the `.env` used by Docker Compose. To change this location, modify the **Documents Paths** variable in either the [**Advanced Setup** menu](/install#setup) or in the `.env` used by Docker Compose.

4
docs/docs/get-started/docker.mdx
View file

@ -18,7 +18,7 @@ OpenRAG has two Docker Compose files. Both files deploy the same applications an
- Install the following: - Install the following:
- [Python](https://www.python.org/downloads/release/python-3100/) version 3.10 to 3.13. - [Python](https://www.python.org/downloads/release/python-3100/) version 3.13 or later.
- [uv](https://docs.astral.sh/uv/getting-started/installation/). - [uv](https://docs.astral.sh/uv/getting-started/installation/).
- [Podman](https://podman.io/docs/installation) (recommended) or [Docker](https://docs.docker.com/get-docker/). - [Podman](https://podman.io/docs/installation) (recommended) or [Docker](https://docs.docker.com/get-docker/).
- [`podman-compose`](https://docs.podman.io/en/latest/markdown/podman-compose.1.html) or [Docker Compose](https://docs.docker.com/compose/install/). To use Docker Compose with Podman, you must alias Docker Compose commands to Podman commands. - [`podman-compose`](https://docs.podman.io/en/latest/markdown/podman-compose.1.html) or [Docker Compose](https://docs.docker.com/compose/install/). To use Docker Compose with Podman, you must alias Docker Compose commands to Podman commands.
@ -187,7 +187,7 @@ docker compose up -d --force-recreate
Reset state by rebuilding all of your containers. Reset state by rebuilding all of your containers.
Your OpenSearch and Langflow databases will be lost. Your OpenSearch and Langflow databases will be lost.
Documents stored in the `./documents` directory will persist, since the directory is mounted as a volume in the OpenRAG backend container. Documents stored in the `./openrag-documents` directory will persist, since the directory is mounted as a volume in the OpenRAG backend container.
```bash ```bash
docker compose up --build --force-recreate --remove-orphans docker compose up --build --force-recreate --remove-orphans

2
docs/docs/get-started/install.mdx
View file

@ -22,7 +22,7 @@ If you prefer running Podman or Docker containers and manually editing `.env` fi
## Prerequisites ## Prerequisites
- All OpenRAG installations require [Python](https://www.python.org/downloads/release/python-3100/) version 3.10 to 3.13. - All OpenRAG installations require [Python](https://www.python.org/downloads/release/python-3100/) version 3.13 or later.
- If you aren't using the automatic installer script, install the following: - If you aren't using the automatic installer script, install the following:

4
docs/docs/get-started/quickstart.mdx
View file

@ -19,7 +19,7 @@ This quickstart requires the following:
This quickstart uses OpenAI for simplicity. This quickstart uses OpenAI for simplicity.
For other providers, see the complete [installation guide](/install). For other providers, see the complete [installation guide](/install).
- [Python](https://www.python.org/downloads/release/python-3100/) version 3.10 to 3.13. - [Python](https://www.python.org/downloads/release/python-3100/) version 3.13 or later.
- Microsoft Windows only: To run OpenRAG on Windows, you must use the Windows Subsystem for Linux (WSL). - Microsoft Windows only: To run OpenRAG on Windows, you must use the Windows Subsystem for Linux (WSL).
@ -102,7 +102,7 @@ You can click a document to view the chunks of the document as they are stored i
For this quickstart, use either the <Icon name="File" aria-hidden="true"/> **File** or <Icon name="Folder" aria-hidden="true"/> **Folder** upload options to load documents from your local machine. For this quickstart, use either the <Icon name="File" aria-hidden="true"/> **File** or <Icon name="Folder" aria-hidden="true"/> **Folder** upload options to load documents from your local machine.
**Folder** uploads an entire directory. **Folder** uploads an entire directory.
The default directory is the `/documents` subdirectory in your OpenRAG installation directory. The default directory is the `/openrag-documents` subdirectory in your OpenRAG installation directory.
For information about the cloud storage provider options, see [Ingest files with OAuth connectors](/ingestion#oauth-ingestion). For information about the cloud storage provider options, see [Ingest files with OAuth connectors](/ingestion#oauth-ingestion).

2
docs/docs/reference/configuration.mdx
View file

@ -80,7 +80,7 @@ Control how OpenRAG [processes and ingests documents](/ingestion) into your know
| `DISABLE_INGEST_WITH_LANGFLOW` | `false` | Disable Langflow ingestion pipeline. | | `DISABLE_INGEST_WITH_LANGFLOW` | `false` | Disable Langflow ingestion pipeline. |
| `DOCLING_OCR_ENGINE` | - | OCR engine for document processing. | | `DOCLING_OCR_ENGINE` | - | OCR engine for document processing. |
| `OCR_ENABLED` | `false` | Enable OCR for image processing. | | `OCR_ENABLED` | `false` | Enable OCR for image processing. |
| `OPENRAG_DOCUMENTS_PATHS` | `./documents` | Document paths for ingestion. | | `OPENRAG_DOCUMENTS_PATHS` | `./openrag-documents` | Document paths for ingestion. |
| `PICTURE_DESCRIPTIONS_ENABLED` | `false` | Enable picture descriptions. | | `PICTURE_DESCRIPTIONS_ENABLED` | `false` | Enable picture descriptions. |
### Langflow settings ### Langflow settings

2
docs/package.json
View file

@ -6,7 +6,7 @@
"docusaurus": "docusaurus", "docusaurus": "docusaurus",
"start": "docusaurus start", "start": "docusaurus start",
"build": "docusaurus build", "build": "docusaurus build",
"build:pdf": "rm -f ../documents/openrag-documentation.pdf && npm run build && npm run serve & sleep 10 && npx docusaurus-to-pdf && pkill -f 'docusaurus serve'", "build:pdf": "rm -f ../openrag-documents/openrag-documentation.pdf && npm run build && npm run serve & sleep 10 && npx docusaurus-to-pdf && pkill -f 'docusaurus serve'",
"swizzle": "docusaurus swizzle", "swizzle": "docusaurus swizzle",
"deploy": "docusaurus deploy", "deploy": "docusaurus deploy",
"clear": "docusaurus clear", "clear": "docusaurus clear",

2
docs/scraper.config.json
View file

@ -1,7 +1,7 @@
{ {
"baseUrl": "http://localhost:3000", "baseUrl": "http://localhost:3000",
"entryPoint": "http://localhost:3000", "entryPoint": "http://localhost:3000",
"outputDir": "../documents/openrag-documentation.pdf", "outputDir": "../openrag-documents/openrag-documentation.pdf",
"customStyles": "table { max-width: 3500px !important; } .navbar, .footer, .breadcrumbs { display: none !important; }", "customStyles": "table { max-width: 3500px !important; } .navbar, .footer, .breadcrumbs { display: none !important; }",
"forceImages": true "forceImages": true
} }

2988
flows/ingestion_flow.json
View file

File diff suppressed because one or more lines are too long

3936
flows/openrag_agent.json
View file

File diff suppressed because one or more lines are too long

3662
flows/openrag_nudges.json
View file

File diff suppressed because one or more lines are too long

3804
flows/openrag_url_mcp.json
View file

File diff suppressed because one or more lines are too long

14
frontend/app/api/queries/useGetSearchQuery.ts
View file

@ -4,6 +4,7 @@ import {
useQueryClient, useQueryClient,
} from "@tanstack/react-query"; } from "@tanstack/react-query";
import type { ParsedQueryData } from "@/contexts/knowledge-filter-context"; import type { ParsedQueryData } from "@/contexts/knowledge-filter-context";
import { SEARCH_CONSTANTS } from "@/lib/constants";
export interface SearchPayload { export interface SearchPayload {
query: string; query: string;
@ -70,13 +71,16 @@ export const useGetSearchQuery = (
async function getFiles(): Promise<File[]> { async function getFiles(): Promise<File[]> {
try { try {
// For wildcard queries, use a high limit to get all files
// Otherwise use the limit from queryData or default to 100
const isWildcardQuery = effectiveQuery.trim() === "*" || effectiveQuery.trim() === "";
const searchLimit = isWildcardQuery
? SEARCH_CONSTANTS.WILDCARD_QUERY_LIMIT
: (queryData?.limit || 100);
const searchPayload: SearchPayload = { const searchPayload: SearchPayload = {
query: effectiveQuery, query: effectiveQuery,
limit: limit: searchLimit,
queryData?.limit ||
(effectiveQuery.trim() === "*" || effectiveQuery.trim() === ""
? 10000
: 10), // Maximum allowed limit for wildcard searches
scoreThreshold: queryData?.scoreThreshold || 0, scoreThreshold: queryData?.scoreThreshold || 0,
}; };
if (queryData?.filters) { if (queryData?.filters) {

18
frontend/app/knowledge/page.tsx
View file

@ -31,6 +31,11 @@ import {
DialogTrigger, DialogTrigger,
} from "@/components/ui/dialog"; } from "@/components/ui/dialog";
import { StatusBadge } from "@/components/ui/status-badge"; import { StatusBadge } from "@/components/ui/status-badge";
import {
Tooltip,
TooltipContent,
TooltipTrigger,
} from "@/components/ui/tooltip";
import { import {
DeleteConfirmationDialog, DeleteConfirmationDialog,
formatFilesToDelete, formatFilesToDelete,
@ -156,9 +161,16 @@ function SearchPage() {
}} }}
> >
{getSourceIcon(data?.connector_type)} {getSourceIcon(data?.connector_type)}
<span className="font-medium text-foreground truncate"> <Tooltip>
{value} <TooltipTrigger asChild>
</span> <span className="font-medium text-foreground truncate">
{value}
</span>
</TooltipTrigger>
<TooltipContent side="top" align="start">
{value}
</TooltipContent>
</Tooltip>
</button> </button>
</div> </div>
); );

73
frontend/components/navigation.tsx
View file

@ -91,6 +91,10 @@ export function Navigation({
const { loading } = useLoadingStore(); const { loading } = useLoadingStore();
useEffect(() => {
console.log("loading", loading);
}, [loading]);
const [previousConversationCount, setPreviousConversationCount] = useState(0); const [previousConversationCount, setPreviousConversationCount] = useState(0);
const [deleteModalOpen, setDeleteModalOpen] = useState(false); const [deleteModalOpen, setDeleteModalOpen] = useState(false);
const [conversationToDelete, setConversationToDelete] = const [conversationToDelete, setConversationToDelete] =
@ -391,8 +395,70 @@ export function Navigation({
No conversations yet No conversations yet
</div> </div>
) : ( ) : (
conversations.map((conversation) => ( <>
<button {/* Optimistic rendering: Show placeholder conversation button while loading */}
{(() => {
// Show placeholder when:
// 1. Loading is true AND conversation doesn't exist yet (creating new conversation), OR
// 2. currentConversationId exists but isn't in conversations yet (gap between response and list update)
const conversationExists = currentConversationId
? conversations.some(
(conv) => conv.response_id === currentConversationId,
)
: false;
const shouldShowPlaceholder =
!conversationExists &&
(loading ||
(currentConversationId !== null &&
currentConversationId !== undefined));
// Use placeholderConversation if available
// Otherwise create a placeholder with currentConversationId if it exists
// Or use a temporary ID if we're loading but don't have an ID yet
const placeholderToShow =
placeholderConversation
? placeholderConversation
: currentConversationId
? {
response_id: currentConversationId,
title: "",
endpoint: endpoint,
messages: [],
total_messages: 0,
}
: loading
? {
response_id: `loading-${Date.now()}`,
title: "",
endpoint: endpoint,
messages: [],
total_messages: 0,
}
: null;
return (
shouldShowPlaceholder &&
placeholderToShow && (
<button
key={placeholderToShow.response_id}
type="button"
className="w-full px-3 h-11 rounded-lg bg-accent group relative text-left cursor-not-allowed"
disabled
>
<div className="flex items-center justify-between">
<div className="flex-1 min-w-0">
<div className="text-sm font-medium text-muted-foreground truncate">
<span className="thinking-dots"></span>
</div>
</div>
</div>
</button>
)
);
})()}
{conversations.map((conversation) => (
<button
key={conversation.response_id} key={conversation.response_id}
type="button" type="button"
className={`w-full px-3 h-11 rounded-lg group relative text-left ${ className={`w-full px-3 h-11 rounded-lg group relative text-left ${
@ -467,7 +533,8 @@ export function Navigation({
</DropdownMenu> </DropdownMenu>
</div> </div>
</button> </button>
)) ))}
</>
)} )}
</> </>
)} )}

7
frontend/lib/constants.ts
View file

@ -25,6 +25,13 @@ export const UI_CONSTANTS = {
MAX_SYSTEM_PROMPT_CHARS: 4000, MAX_SYSTEM_PROMPT_CHARS: 4000,
} as const; } as const;
/**
* Search Constants
*/
export const SEARCH_CONSTANTS = {
WILDCARD_QUERY_LIMIT: 10000, // Maximum allowed limit for wildcard searches
} as const;
export const ANIMATION_DURATION = 0.4; export const ANIMATION_DURATION = 0.4;
export const SIDEBAR_WIDTH = 280; export const SIDEBAR_WIDTH = 280;
export const HEADER_HEIGHT = 54; export const HEADER_HEIGHT = 54;

0
documents/docling.pdf → openrag-documents/docling.pdf
View file

0
documents/ibm_anthropic.pdf → openrag-documents/ibm_anthropic.pdf
View file

0
documents/openrag-documentation.pdf → openrag-documents/openrag-documentation.pdf
View file

0
documents/warmup_ocr.pdf → openrag-documents/warmup_ocr.pdf
View file

2
pyproject.toml
View file

@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
[project] [project]
name = "openrag" name = "openrag"
version = "0.1.38" version = "0.1.42"
description = "Add your description here" description = "Add your description here"
readme = "README.md" readme = "README.md"
requires-python = ">=3.13" requires-python = ">=3.13"

104
src/api/settings.py
View file

@ -14,6 +14,7 @@ from config.settings import (
clients, clients,
get_openrag_config, get_openrag_config,
config_manager, config_manager,
is_no_auth_mode,
) )
from api.provider_validation import validate_provider_setup from api.provider_validation import validate_provider_setup
@ -614,7 +615,7 @@ async def update_settings(request, session_manager):
) )
logger.info("Set OLLAMA_BASE_URL global variable in Langflow") logger.info("Set OLLAMA_BASE_URL global variable in Langflow")
# Update model values across flows if provider or model changed # Update LLM model values across flows if provider or model changed
if "llm_provider" in body or "llm_model" in body: if "llm_provider" in body or "llm_model" in body:
flows_service = _get_flows_service() flows_service = _get_flows_service()
llm_provider = current_config.agent.llm_provider.lower() llm_provider = current_config.agent.llm_provider.lower()
@ -629,19 +630,49 @@ async def update_settings(request, session_manager):
f"Successfully updated Langflow flows for LLM provider {llm_provider}" f"Successfully updated Langflow flows for LLM provider {llm_provider}"
) )
# Update SELECTED_EMBEDDING_MODEL global variable (no flow updates needed)
if "embedding_provider" in body or "embedding_model" in body: if "embedding_provider" in body or "embedding_model" in body:
flows_service = _get_flows_service() await clients._create_langflow_global_variable(
embedding_provider = current_config.knowledge.embedding_provider.lower() "SELECTED_EMBEDDING_MODEL", current_config.knowledge.embedding_model, modify=True
embedding_provider_config = current_config.get_embedding_provider_config()
embedding_endpoint = getattr(embedding_provider_config, "endpoint", None)
await flows_service.change_langflow_model_value(
embedding_provider,
embedding_model=current_config.knowledge.embedding_model,
endpoint=embedding_endpoint,
) )
logger.info( logger.info(
f"Successfully updated Langflow flows for embedding provider {embedding_provider}" f"Set SELECTED_EMBEDDING_MODEL global variable to {current_config.knowledge.embedding_model}"
) )
# Update MCP servers with provider credentials
try:
from services.langflow_mcp_service import LangflowMCPService
from utils.langflow_headers import build_mcp_global_vars_from_config
mcp_service = LangflowMCPService()
# Build global vars using utility function
mcp_global_vars = build_mcp_global_vars_from_config(current_config)
# In no-auth mode, add the anonymous JWT token and user details
if is_no_auth_mode() and session_manager:
from session_manager import AnonymousUser
# Create/get anonymous JWT for no-auth mode
anonymous_jwt = session_manager.get_effective_jwt_token(None, None)
if anonymous_jwt:
mcp_global_vars["JWT"] = anonymous_jwt
# Add anonymous user details
anonymous_user = AnonymousUser()
mcp_global_vars["OWNER"] = anonymous_user.user_id # "anonymous"
mcp_global_vars["OWNER_NAME"] = f'"{anonymous_user.name}"' # "Anonymous User" (quoted)
mcp_global_vars["OWNER_EMAIL"] = anonymous_user.email # "anonymous@localhost"
logger.debug("Added anonymous JWT and user details to MCP servers for no-auth mode")
if mcp_global_vars:
result = await mcp_service.update_mcp_servers_with_global_vars(mcp_global_vars)
logger.info("Updated MCP servers with provider credentials after settings change", **result)
except Exception as mcp_error:
logger.warning(f"Failed to update MCP servers after settings change: {str(mcp_error)}")
# Don't fail the entire settings update if MCP update fails
except Exception as e: except Exception as e:
logger.error(f"Failed to update Langflow settings: {str(e)}") logger.error(f"Failed to update Langflow settings: {str(e)}")
@ -660,7 +691,7 @@ async def update_settings(request, session_manager):
) )
async def onboarding(request, flows_service): async def onboarding(request, flows_service, session_manager=None):
"""Handle onboarding configuration setup""" """Handle onboarding configuration setup"""
try: try:
# Get current configuration # Get current configuration
@ -928,7 +959,7 @@ async def onboarding(request, flows_service):
) )
logger.info("Set OLLAMA_BASE_URL global variable in Langflow") logger.info("Set OLLAMA_BASE_URL global variable in Langflow")
# Update flows with model values # Update flows with LLM model values
if "llm_provider" in body or "llm_model" in body: if "llm_provider" in body or "llm_model" in body:
llm_provider = current_config.agent.llm_provider.lower() llm_provider = current_config.agent.llm_provider.lower()
llm_provider_config = current_config.get_llm_provider_config() llm_provider_config = current_config.get_llm_provider_config()
@ -940,16 +971,49 @@ async def onboarding(request, flows_service):
) )
logger.info(f"Updated Langflow flows for LLM provider {llm_provider}") logger.info(f"Updated Langflow flows for LLM provider {llm_provider}")
# Set SELECTED_EMBEDDING_MODEL global variable (no flow updates needed)
if "embedding_provider" in body or "embedding_model" in body: if "embedding_provider" in body or "embedding_model" in body:
embedding_provider = current_config.knowledge.embedding_provider.lower() await clients._create_langflow_global_variable(
embedding_provider_config = current_config.get_embedding_provider_config() "SELECTED_EMBEDDING_MODEL", current_config.knowledge.embedding_model, modify=True
embedding_endpoint = getattr(embedding_provider_config, "endpoint", None)
await flows_service.change_langflow_model_value(
provider=embedding_provider,
embedding_model=current_config.knowledge.embedding_model,
endpoint=embedding_endpoint,
) )
logger.info(f"Updated Langflow flows for embedding provider {embedding_provider}") logger.info(
f"Set SELECTED_EMBEDDING_MODEL global variable to {current_config.knowledge.embedding_model}"
)
# Update MCP servers with provider credentials during onboarding
try:
from services.langflow_mcp_service import LangflowMCPService
from utils.langflow_headers import build_mcp_global_vars_from_config
mcp_service = LangflowMCPService()
# Build global vars using utility function
mcp_global_vars = build_mcp_global_vars_from_config(current_config)
# In no-auth mode, add the anonymous JWT token and user details
if is_no_auth_mode() and session_manager:
from session_manager import AnonymousUser
# Create/get anonymous JWT for no-auth mode
anonymous_jwt = session_manager.get_effective_jwt_token(None, None)
if anonymous_jwt:
mcp_global_vars["JWT"] = anonymous_jwt
# Add anonymous user details
anonymous_user = AnonymousUser()
mcp_global_vars["OWNER"] = anonymous_user.user_id # "anonymous"
mcp_global_vars["OWNER_NAME"] = f'"{anonymous_user.name}"' # "Anonymous User" (quoted)
mcp_global_vars["OWNER_EMAIL"] = anonymous_user.email # "anonymous@localhost"
logger.debug("Added anonymous JWT and user details to MCP servers for no-auth mode during onboarding")
if mcp_global_vars:
result = await mcp_service.update_mcp_servers_with_global_vars(mcp_global_vars)
logger.info("Updated MCP servers with provider credentials during onboarding", **result)
except Exception as mcp_error:
logger.warning(f"Failed to update MCP servers during onboarding: {str(mcp_error)}")
# Don't fail onboarding if MCP update fails
except Exception as e: except Exception as e:
logger.error( logger.error(

79
src/main.py
View file

@ -2,6 +2,7 @@
from connectors.langflow_connector_service import LangflowConnectorService from connectors.langflow_connector_service import LangflowConnectorService
from connectors.service import ConnectorService from connectors.service import ConnectorService
from services.flows_service import FlowsService from services.flows_service import FlowsService
from utils.container_utils import detect_container_environment
from utils.embeddings import create_dynamic_index_body from utils.embeddings import create_dynamic_index_body
from utils.logging_config import configure_from_env, get_logger from utils.logging_config import configure_from_env, get_logger
@ -13,6 +14,7 @@ import atexit
import mimetypes import mimetypes
import multiprocessing import multiprocessing
import os import os
import shutil
import subprocess import subprocess
from functools import partial from functools import partial
@ -300,6 +302,21 @@ async def init_index_when_ready():
) )
def _get_documents_dir():
"""Get the documents directory path, handling both Docker and local environments."""
# In Docker, the volume is mounted at /app/documents
# Locally, we use openrag-documents
container_env = detect_container_environment()
if container_env:
path = os.path.abspath("/app/documents")
logger.debug(f"Running in {container_env}, using container path: {path}")
return path
else:
path = os.path.abspath(os.path.join(os.getcwd(), "openrag-documents"))
logger.debug(f"Running locally, using local path: {path}")
return path
async def ingest_default_documents_when_ready(services): async def ingest_default_documents_when_ready(services):
"""Scan the local documents folder and ingest files like a non-auth upload.""" """Scan the local documents folder and ingest files like a non-auth upload."""
try: try:
@ -307,7 +324,7 @@ async def ingest_default_documents_when_ready(services):
"Ingesting default documents when ready", "Ingesting default documents when ready",
disable_langflow_ingest=DISABLE_INGEST_WITH_LANGFLOW, disable_langflow_ingest=DISABLE_INGEST_WITH_LANGFLOW,
) )
base_dir = os.path.abspath(os.path.join(os.getcwd(), "documents")) base_dir = _get_documents_dir()
if not os.path.isdir(base_dir): if not os.path.isdir(base_dir):
logger.info( logger.info(
"Default documents directory not found; skipping ingestion", "Default documents directory not found; skipping ingestion",
@ -370,7 +387,7 @@ async def _ingest_default_documents_langflow(services, file_paths):
# Prepare tweaks for default documents with anonymous user metadata # Prepare tweaks for default documents with anonymous user metadata
default_tweaks = { default_tweaks = {
"OpenSearchHybrid-Ve6bS": { "OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4": {
"docs_metadata": [ "docs_metadata": [
{"key": "owner", "value": None}, {"key": "owner", "value": None},
{"key": "owner_name", "value": anonymous_user.name}, {"key": "owner_name", "value": anonymous_user.name},
@ -433,6 +450,55 @@ async def _ingest_default_documents_openrag(services, file_paths):
) )
async def _update_mcp_servers_with_provider_credentials(services):
"""Update MCP servers with provider credentials at startup.
This is especially important for no-auth mode where users don't go through
the OAuth login flow that would normally set these credentials.
"""
try:
auth_service = services.get("auth_service")
session_manager = services.get("session_manager")
if not auth_service or not auth_service.langflow_mcp_service:
logger.debug("MCP service not available, skipping credential update")
return
config = get_openrag_config()
# Build global vars with provider credentials using utility function
from utils.langflow_headers import build_mcp_global_vars_from_config
global_vars = build_mcp_global_vars_from_config(config)
# In no-auth mode, add the anonymous JWT token and user details
if is_no_auth_mode() and session_manager:
from session_manager import AnonymousUser
# Create/get anonymous JWT for no-auth mode
anonymous_jwt = session_manager.get_effective_jwt_token(None, None)
if anonymous_jwt:
global_vars["JWT"] = anonymous_jwt
# Add anonymous user details
anonymous_user = AnonymousUser()
global_vars["OWNER"] = anonymous_user.user_id # "anonymous"
global_vars["OWNER_NAME"] = f'"{anonymous_user.name}"' # "Anonymous User" (quoted for spaces)
global_vars["OWNER_EMAIL"] = anonymous_user.email # "anonymous@localhost"
logger.info("Added anonymous JWT and user details to MCP servers for no-auth mode")
if global_vars:
result = await auth_service.langflow_mcp_service.update_mcp_servers_with_global_vars(global_vars)
logger.info("Updated MCP servers with provider credentials at startup", **result)
else:
logger.debug("No provider credentials configured, skipping MCP server update")
except Exception as e:
logger.warning("Failed to update MCP servers with provider credentials at startup", error=str(e))
# Don't fail startup if MCP update fails
async def startup_tasks(services): async def startup_tasks(services):
"""Startup tasks""" """Startup tasks"""
logger.info("Starting startup tasks") logger.info("Starting startup tasks")
@ -445,6 +511,9 @@ async def startup_tasks(services):
# Configure alerting security # Configure alerting security
await configure_alerting_security() await configure_alerting_security()
# Update MCP servers with provider credentials (especially important for no-auth mode)
await _update_mcp_servers_with_provider_credentials(services)
async def initialize_services(): async def initialize_services():
@ -1052,7 +1121,11 @@ async def create_app():
Route( Route(
"/onboarding", "/onboarding",
require_auth(services["session_manager"])( require_auth(services["session_manager"])(
partial(settings.onboarding, flows_service=services["flows_service"]) partial(
settings.onboarding,
flows_service=services["flows_service"],
session_manager=services["session_manager"]
)
), ),
methods=["POST"], methods=["POST"],
), ),

6
src/models/processors.py
View file

@ -743,9 +743,9 @@ class LangflowFileProcessor(TaskProcessor):
if metadata_tweaks: if metadata_tweaks:
# Initialize the OpenSearch component tweaks if not already present # Initialize the OpenSearch component tweaks if not already present
if "OpenSearchHybrid-Ve6bS" not in final_tweaks: if "OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4" not in final_tweaks:
final_tweaks["OpenSearchHybrid-Ve6bS"] = {} final_tweaks["OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4"] = {}
final_tweaks["OpenSearchHybrid-Ve6bS"]["docs_metadata"] = metadata_tweaks final_tweaks["OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4"]["docs_metadata"] = metadata_tweaks
# Process file using langflow service # Process file using langflow service
result = await self.langflow_file_service.upload_and_ingest_file( result = await self.langflow_file_service.upload_and_ingest_file(

10
src/services/auth_service.py
View file

@ -308,6 +308,16 @@ class AuthService:
global_vars["OWNER_NAME"] = str(f"\"{owner_name}\"") global_vars["OWNER_NAME"] = str(f"\"{owner_name}\"")
if user_info.get("email"): if user_info.get("email"):
global_vars["OWNER_EMAIL"] = user_info.get("email") global_vars["OWNER_EMAIL"] = user_info.get("email")
# Add provider credentials to MCP servers using utility function
from config.settings import get_openrag_config
from utils.langflow_headers import build_mcp_global_vars_from_config
config = get_openrag_config()
provider_vars = build_mcp_global_vars_from_config(config)
# Merge provider credentials with user info
global_vars.update(provider_vars)
# Run in background to avoid delaying login flow # Run in background to avoid delaying login flow
task = asyncio.create_task( task = asyncio.create_task(

40
src/services/chat_service.py
View file

@ -60,11 +60,22 @@ class ChatService:
"LANGFLOW_URL and LANGFLOW_CHAT_FLOW_ID environment variables are required" "LANGFLOW_URL and LANGFLOW_CHAT_FLOW_ID environment variables are required"
) )
# Prepare extra headers for JWT authentication # Prepare extra headers for JWT authentication and embedding model
extra_headers = {} extra_headers = {}
if jwt_token: if jwt_token:
extra_headers["X-LANGFLOW-GLOBAL-VAR-JWT"] = jwt_token extra_headers["X-LANGFLOW-GLOBAL-VAR-JWT"] = jwt_token
# Pass the selected embedding model as a global variable
from config.settings import get_openrag_config
from utils.langflow_headers import add_provider_credentials_to_headers
config = get_openrag_config()
embedding_model = config.knowledge.embedding_model
extra_headers["X-LANGFLOW-GLOBAL-VAR-SELECTED_EMBEDDING_MODEL"] = embedding_model
# Add provider credentials to headers
add_provider_credentials_to_headers(extra_headers, config)
logger.debug(f"[LF] Extra headers {extra_headers}")
# Get context variables for filters, limit, and threshold # Get context variables for filters, limit, and threshold
from auth_context import ( from auth_context import (
get_score_threshold, get_score_threshold,
@ -169,11 +180,22 @@ class ChatService:
"LANGFLOW_URL and NUDGES_FLOW_ID environment variables are required" "LANGFLOW_URL and NUDGES_FLOW_ID environment variables are required"
) )
# Prepare extra headers for JWT authentication # Prepare extra headers for JWT authentication and embedding model
extra_headers = {} extra_headers = {}
if jwt_token: if jwt_token:
extra_headers["X-LANGFLOW-GLOBAL-VAR-JWT"] = jwt_token extra_headers["X-LANGFLOW-GLOBAL-VAR-JWT"] = jwt_token
# Pass the selected embedding model as a global variable
from config.settings import get_openrag_config
from utils.langflow_headers import add_provider_credentials_to_headers
config = get_openrag_config()
embedding_model = config.knowledge.embedding_model
extra_headers["X-LANGFLOW-GLOBAL-VAR-SELECTED_EMBEDDING_MODEL"] = embedding_model
# Add provider credentials to headers
add_provider_credentials_to_headers(extra_headers, config)
# Build the complete filter expression like the chat service does # Build the complete filter expression like the chat service does
filter_expression = {} filter_expression = {}
has_user_filters = False has_user_filters = False
@ -287,10 +309,22 @@ class ChatService:
document_prompt = f"I'm uploading a document called '{filename}'. Here is its content:\n\n{document_content}\n\nPlease confirm you've received this document and are ready to answer questions about it." document_prompt = f"I'm uploading a document called '{filename}'. Here is its content:\n\n{document_content}\n\nPlease confirm you've received this document and are ready to answer questions about it."
if endpoint == "langflow": if endpoint == "langflow":
# Prepare extra headers for JWT authentication # Prepare extra headers for JWT authentication and embedding model
extra_headers = {} extra_headers = {}
if jwt_token: if jwt_token:
extra_headers["X-LANGFLOW-GLOBAL-VAR-JWT"] = jwt_token extra_headers["X-LANGFLOW-GLOBAL-VAR-JWT"] = jwt_token
# Pass the selected embedding model as a global variable
from config.settings import get_openrag_config
from utils.langflow_headers import add_provider_credentials_to_headers
config = get_openrag_config()
embedding_model = config.knowledge.embedding_model
extra_headers["X-LANGFLOW-GLOBAL-VAR-SELECTED_EMBEDDING_MODEL"] = embedding_model
# Add provider credentials to headers
add_provider_credentials_to_headers(extra_headers, config)
# Ensure the Langflow client exists; try lazy init if needed # Ensure the Langflow client exists; try lazy init if needed
langflow_client = await clients.ensure_langflow_client() langflow_client = await clients.ensure_langflow_client()
if not langflow_client: if not langflow_client:

19
src/services/langflow_file_service.py
View file

@ -94,7 +94,7 @@ class LangflowFileService:
# Pass JWT token via tweaks using the x-langflow-global-var- pattern # Pass JWT token via tweaks using the x-langflow-global-var- pattern
if jwt_token: if jwt_token:
# Using the global variable pattern that Langflow expects for OpenSearch components # Using the global variable pattern that Langflow expects for OpenSearch components
tweaks["OpenSearchHybrid-Ve6bS"] = {"jwt_token": jwt_token} tweaks["OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4"] = {"jwt_token": jwt_token}
logger.debug("[LF] Added JWT token to tweaks for OpenSearch components") logger.debug("[LF] Added JWT token to tweaks for OpenSearch components")
else: else:
logger.warning("[LF] No JWT token provided") logger.warning("[LF] No JWT token provided")
@ -112,9 +112,9 @@ class LangflowFileService:
logger.info(f"[LF] Metadata tweaks {metadata_tweaks}") logger.info(f"[LF] Metadata tweaks {metadata_tweaks}")
# if metadata_tweaks: # if metadata_tweaks:
# # Initialize the OpenSearch component tweaks if not already present # # Initialize the OpenSearch component tweaks if not already present
# if "OpenSearchHybrid-Ve6bS" not in tweaks: # if "OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4" not in tweaks:
# tweaks["OpenSearchHybrid-Ve6bS"] = {} # tweaks["OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4"] = {}
# tweaks["OpenSearchHybrid-Ve6bS"]["docs_metadata"] = metadata_tweaks # tweaks["OpenSearchVectorStoreComponentMultimodalMultiEmbedding-By9U4"]["docs_metadata"] = metadata_tweaks
# logger.debug( # logger.debug(
# "[LF] Added metadata to tweaks", metadata_count=len(metadata_tweaks) # "[LF] Added metadata to tweaks", metadata_count=len(metadata_tweaks)
# ) # )
@ -140,6 +140,13 @@ class LangflowFileService:
filename = str(file_tuples[0][0]) if file_tuples and len(file_tuples) > 0 else "" filename = str(file_tuples[0][0]) if file_tuples and len(file_tuples) > 0 else ""
mimetype = str(file_tuples[0][2]) if file_tuples and len(file_tuples) > 0 and len(file_tuples[0]) > 2 else "" mimetype = str(file_tuples[0][2]) if file_tuples and len(file_tuples) > 0 and len(file_tuples[0]) > 2 else ""
# Get the current embedding model and provider credentials from config
from config.settings import get_openrag_config
from utils.langflow_headers import add_provider_credentials_to_headers
config = get_openrag_config()
embedding_model = config.knowledge.embedding_model
headers={ headers={
"X-Langflow-Global-Var-JWT": str(jwt_token), "X-Langflow-Global-Var-JWT": str(jwt_token),
"X-Langflow-Global-Var-OWNER": str(owner), "X-Langflow-Global-Var-OWNER": str(owner),
@ -149,7 +156,11 @@ class LangflowFileService:
"X-Langflow-Global-Var-FILENAME": filename, "X-Langflow-Global-Var-FILENAME": filename,
"X-Langflow-Global-Var-MIMETYPE": mimetype, "X-Langflow-Global-Var-MIMETYPE": mimetype,
"X-Langflow-Global-Var-FILESIZE": str(file_size_bytes), "X-Langflow-Global-Var-FILESIZE": str(file_size_bytes),
"X-Langflow-Global-Var-SELECTED_EMBEDDING_MODEL": str(embedding_model),
} }
# Add provider credentials as global variables for ingestion
add_provider_credentials_to_headers(headers, config)
logger.info(f"[LF] Headers {headers}") logger.info(f"[LF] Headers {headers}")
logger.info(f"[LF] Payload {payload}") logger.info(f"[LF] Payload {payload}")
resp = await clients.langflow_request( resp = await clients.langflow_request(

1
src/tui/_assets/docker-compose-cpu.yml
View file

@ -1 +0,0 @@
../../../docker-compose-cpu.yml

1
src/tui/_assets/docker-compose.gpu.yml Symbolic link
View file

@ -0,0 +1 @@
../../../docker-compose.gpu.yml

4
src/tui/main.py
View file

@ -455,7 +455,7 @@ def _copy_assets(resource_tree, destination: Path, allowed_suffixes: Optional[It
def copy_sample_documents(*, force: bool = False) -> None: def copy_sample_documents(*, force: bool = False) -> None:
"""Copy sample documents from package to current directory if they don't exist.""" """Copy sample documents from package to current directory if they don't exist."""
documents_dir = Path("documents") documents_dir = Path("openrag-documents")
try: try:
assets_files = files("tui._assets.documents") assets_files = files("tui._assets.documents")
@ -485,7 +485,7 @@ def copy_compose_files(*, force: bool = False) -> None:
logger.debug(f"Could not access compose assets: {e}") logger.debug(f"Could not access compose assets: {e}")
return return
for filename in ("docker-compose.yml", "docker-compose-cpu.yml"): for filename in ("docker-compose.yml", "docker-compose.gpu.yml"):
destination = Path(filename) destination = Path(filename)
if destination.exists() and not force: if destination.exists() and not force:
continue continue

258
src/tui/managers/container_manager.py
View file

@ -43,6 +43,7 @@ class ServiceInfo:
image: Optional[str] = None image: Optional[str] = None
image_digest: Optional[str] = None image_digest: Optional[str] = None
created: Optional[str] = None created: Optional[str] = None
error_message: Optional[str] = None
def __post_init__(self): def __post_init__(self):
if self.ports is None: if self.ports is None:
@ -56,15 +57,15 @@ class ContainerManager:
self.platform_detector = PlatformDetector() self.platform_detector = PlatformDetector()
self.runtime_info = self.platform_detector.detect_runtime() self.runtime_info = self.platform_detector.detect_runtime()
self.compose_file = compose_file or self._find_compose_file("docker-compose.yml") self.compose_file = compose_file or self._find_compose_file("docker-compose.yml")
self.cpu_compose_file = self._find_compose_file("docker-compose-cpu.yml") self.gpu_compose_file = self._find_compose_file("docker-compose.gpu.yml")
self.services_cache: Dict[str, ServiceInfo] = {} self.services_cache: Dict[str, ServiceInfo] = {}
self.last_status_update = 0 self.last_status_update = 0
# Auto-select CPU compose if no GPU available # Auto-select GPU override if GPU is available
try: try:
has_gpu, _ = detect_gpu_devices() has_gpu, _ = detect_gpu_devices()
self.use_cpu_compose = not has_gpu self.use_gpu_compose = has_gpu
except Exception: except Exception:
self.use_cpu_compose = True self.use_gpu_compose = False
# Expected services based on compose files # Expected services based on compose files
self.expected_services = [ self.expected_services = [
@ -135,6 +136,96 @@ class ContainerManager:
return self.platform_detector.get_compose_installation_instructions() return self.platform_detector.get_compose_installation_instructions()
return self.platform_detector.get_installation_instructions() return self.platform_detector.get_installation_instructions()
def _extract_ports_from_compose(self) -> Dict[str, List[int]]:
"""Extract port mappings from compose files.
Returns:
Dict mapping service name to list of host ports
"""
service_ports: Dict[str, List[int]] = {}
compose_files = [self.compose_file]
if hasattr(self, 'cpu_compose_file') and self.cpu_compose_file and self.cpu_compose_file.exists():
compose_files.append(self.cpu_compose_file)
for compose_file in compose_files:
if not compose_file.exists():
continue
try:
import re
content = compose_file.read_text()
current_service = None
in_ports_section = False
for line in content.splitlines():
# Detect service names
service_match = re.match(r'^ (\w[\w-]*):$', line)
if service_match:
current_service = service_match.group(1)
in_ports_section = False
if current_service not in service_ports:
service_ports[current_service] = []
continue
# Detect ports section
if current_service and re.match(r'^ ports:$', line):
in_ports_section = True
continue
# Exit ports section on new top-level key
if in_ports_section and re.match(r'^ \w+:', line):
in_ports_section = False
# Extract port mappings
if in_ports_section and current_service:
# Match patterns like: - "3000:3000", - "9200:9200", - 7860:7860
port_match = re.search(r'["\']?(\d+):\d+["\']?', line)
if port_match:
host_port = int(port_match.group(1))
if host_port not in service_ports[current_service]:
service_ports[current_service].append(host_port)
except Exception as e:
logger.debug(f"Error parsing {compose_file} for ports: {e}")
continue
return service_ports
async def check_ports_available(self) -> tuple[bool, List[tuple[str, int, str]]]:
"""Check if required ports are available.
Returns:
Tuple of (all_available, conflicts) where conflicts is a list of
(service_name, port, error_message) tuples
"""
import socket
service_ports = self._extract_ports_from_compose()
conflicts: List[tuple[str, int, str]] = []
for service_name, ports in service_ports.items():
for port in ports:
try:
# Try to bind to the port to check if it's available
sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
sock.settimeout(0.5)
result = sock.connect_ex(('127.0.0.1', port))
sock.close()
if result == 0:
# Port is in use
conflicts.append((
service_name,
port,
f"Port {port} is already in use"
))
except Exception as e:
logger.debug(f"Error checking port {port}: {e}")
continue
return (len(conflicts) == 0, conflicts)
async def _run_compose_command( async def _run_compose_command(
self, args: List[str], cpu_mode: Optional[bool] = None self, args: List[str], cpu_mode: Optional[bool] = None
) -> tuple[bool, str, str]: ) -> tuple[bool, str, str]:
@ -143,9 +234,15 @@ class ContainerManager:
return False, "", "No container runtime available" return False, "", "No container runtime available"
if cpu_mode is None: if cpu_mode is None:
cpu_mode = self.use_cpu_compose use_gpu = self.use_gpu_compose
compose_file = self.cpu_compose_file if cpu_mode else self.compose_file else:
cmd = self.runtime_info.compose_command + ["-f", str(compose_file)] + args use_gpu = not cpu_mode
# Build compose command with override pattern
cmd = self.runtime_info.compose_command + ["-f", str(self.compose_file)]
if use_gpu and self.gpu_compose_file.exists():
cmd.extend(["-f", str(self.gpu_compose_file)])
cmd.extend(args)
try: try:
process = await asyncio.create_subprocess_exec( process = await asyncio.create_subprocess_exec(
@ -179,9 +276,15 @@ class ContainerManager:
return return
if cpu_mode is None: if cpu_mode is None:
cpu_mode = self.use_cpu_compose use_gpu = self.use_gpu_compose
compose_file = self.cpu_compose_file if cpu_mode else self.compose_file else:
cmd = self.runtime_info.compose_command + ["-f", str(compose_file)] + args use_gpu = not cpu_mode
# Build compose command with override pattern
cmd = self.runtime_info.compose_command + ["-f", str(self.compose_file)]
if use_gpu and self.gpu_compose_file.exists():
cmd.extend(["-f", str(self.gpu_compose_file)])
cmd.extend(args)
try: try:
process = await asyncio.create_subprocess_exec( process = await asyncio.create_subprocess_exec(
@ -242,9 +345,15 @@ class ContainerManager:
return return
if cpu_mode is None: if cpu_mode is None:
cpu_mode = self.use_cpu_compose use_gpu = self.use_gpu_compose
compose_file = self.cpu_compose_file if cpu_mode else self.compose_file else:
cmd = self.runtime_info.compose_command + ["-f", str(compose_file)] + args use_gpu = not cpu_mode
# Build compose command with override pattern
cmd = self.runtime_info.compose_command + ["-f", str(self.compose_file)]
if use_gpu and self.gpu_compose_file.exists():
cmd.extend(["-f", str(self.gpu_compose_file)])
cmd.extend(args)
try: try:
process = await asyncio.create_subprocess_exec( process = await asyncio.create_subprocess_exec(
@ -551,44 +660,61 @@ class ContainerManager:
"""Get resolved image names from compose files using docker/podman compose, with robust fallbacks.""" """Get resolved image names from compose files using docker/podman compose, with robust fallbacks."""
images: set[str] = set() images: set[str] = set()
compose_files = [self.compose_file, self.cpu_compose_file] # Try both GPU and CPU modes to get all images
for compose_file in compose_files: for use_gpu in [True, False]:
try: try:
if not compose_file or not compose_file.exists(): # Build compose command with override pattern
continue cmd = self.runtime_info.compose_command + ["-f", str(self.compose_file)]
if use_gpu and self.gpu_compose_file.exists():
cmd.extend(["-f", str(self.gpu_compose_file)])
cmd.extend(["config", "--format", "json"])
cpu_mode = (compose_file == self.cpu_compose_file) process = await asyncio.create_subprocess_exec(
*cmd,
# Try JSON format first stdout=asyncio.subprocess.PIPE,
success, stdout, _ = await self._run_compose_command( stderr=asyncio.subprocess.PIPE,
["config", "--format", "json"], cwd=Path.cwd(),
cpu_mode=cpu_mode
) )
stdout, stderr = await process.communicate()
stdout_text = stdout.decode() if stdout else ""
if success and stdout.strip(): if process.returncode == 0 and stdout_text.strip():
from_cfg = self._extract_images_from_compose_config(stdout, tried_json=True) from_cfg = self._extract_images_from_compose_config(stdout_text, tried_json=True)
if from_cfg: if from_cfg:
images.update(from_cfg) images.update(from_cfg)
continue # this compose file succeeded; move to next file continue
# Fallback to YAML output (for older compose versions) # Fallback to YAML output (for older compose versions)
success, stdout, _ = await self._run_compose_command( cmd = self.runtime_info.compose_command + ["-f", str(self.compose_file)]
["config"], if use_gpu and self.gpu_compose_file.exists():
cpu_mode=cpu_mode cmd.extend(["-f", str(self.gpu_compose_file)])
) cmd.append("config")
if success and stdout.strip(): process = await asyncio.create_subprocess_exec(
from_cfg = self._extract_images_from_compose_config(stdout, tried_json=False) *cmd,
stdout=asyncio.subprocess.PIPE,
stderr=asyncio.subprocess.PIPE,
cwd=Path.cwd(),
)
stdout, stderr = await process.communicate()
stdout_text = stdout.decode() if stdout else ""
if process.returncode == 0 and stdout_text.strip():
from_cfg = self._extract_images_from_compose_config(stdout_text, tried_json=False)
if from_cfg: if from_cfg:
images.update(from_cfg) images.update(from_cfg)
continue continue
except Exception: except Exception:
# Keep behavior resilient—just continue to next file # Keep behavior resilient—just continue to next mode
continue continue
# Fallback: manual parsing if compose config didn't work # Fallback: manual parsing if compose config didn't work
if not images: if not images:
compose_files = [self.compose_file]
if self.gpu_compose_file.exists():
compose_files.append(self.gpu_compose_file)
for compose in compose_files: for compose in compose_files:
try: try:
if not compose.exists(): if not compose.exists():
@ -638,8 +764,11 @@ class ContainerManager:
yield False, "No container runtime available" yield False, "No container runtime available"
return return
# Diagnostic info about compose files # Determine GPU mode
compose_file = self.cpu_compose_file if (cpu_mode if cpu_mode is not None else self.use_cpu_compose) else self.compose_file if cpu_mode is None:
use_gpu = self.use_gpu_compose
else:
use_gpu = not cpu_mode
# Show the search process for debugging # Show the search process for debugging
if hasattr(self, '_compose_search_log'): if hasattr(self, '_compose_search_log'):
@ -650,9 +779,23 @@ class ContainerManager:
# Show runtime detection info # Show runtime detection info
runtime_cmd_str = " ".join(self.runtime_info.compose_command) runtime_cmd_str = " ".join(self.runtime_info.compose_command)
yield False, f"Using compose command: {runtime_cmd_str}", False yield False, f"Using compose command: {runtime_cmd_str}", False
yield False, f"Final compose file: {compose_file.absolute()}", False compose_files_str = str(self.compose_file.absolute())
if not compose_file.exists(): if use_gpu and self.gpu_compose_file.exists():
yield False, f"ERROR: Compose file not found at {compose_file.absolute()}", False compose_files_str += f" + {self.gpu_compose_file.absolute()}"
yield False, f"Compose files: {compose_files_str}", False
if not self.compose_file.exists():
yield False, f"ERROR: Base compose file not found at {self.compose_file.absolute()}", False
return
# Check for port conflicts before starting
yield False, "Checking port availability...", False
ports_available, conflicts = await self.check_ports_available()
if not ports_available:
yield False, "ERROR: Port conflicts detected:", False
for service_name, port, error_msg in conflicts:
yield False, f" - {service_name}: {error_msg}", False
yield False, "Please stop the conflicting services and try again.", False
yield False, "Services not started due to port conflicts.", False
return return
yield False, "Starting OpenRAG services...", False yield False, "Starting OpenRAG services...", False
@ -677,13 +820,37 @@ class ContainerManager:
yield False, "Creating and starting containers...", False yield False, "Creating and starting containers...", False
up_success = {"value": True} up_success = {"value": True}
error_messages = []
async for message, replace_last in self._stream_compose_command(["up", "-d"], up_success, cpu_mode): async for message, replace_last in self._stream_compose_command(["up", "-d"], up_success, cpu_mode):
# Detect error patterns in the output
import re
lower_msg = message.lower()
# Check for common error patterns
if any(pattern in lower_msg for pattern in [
"port.*already.*allocated",
"address already in use",
"bind.*address already in use",
"port is already allocated"
]):
error_messages.append("Port conflict detected")
up_success["value"] = False
elif "error" in lower_msg or "failed" in lower_msg:
# Generic error detection
if message not in error_messages:
error_messages.append(message)
yield False, message, replace_last yield False, message, replace_last
if up_success["value"]: if up_success["value"]:
yield True, "Services started successfully", False yield True, "Services started successfully", False
else: else:
yield False, "Failed to start services. See output above for details.", False yield False, "Failed to start services. See output above for details.", False
if error_messages:
yield False, "\nDetected errors:", False
for err in error_messages[:5]: # Limit to first 5 errors
yield False, f" - {err}", False
async def stop_services(self) -> AsyncIterator[tuple[bool, str]]: async def stop_services(self) -> AsyncIterator[tuple[bool, str]]:
"""Stop all services and yield progress updates.""" """Stop all services and yield progress updates."""
@ -786,16 +953,11 @@ class ContainerManager:
yield "No container runtime available" yield "No container runtime available"
return return
compose_file = ( # Build compose command with override pattern
self.cpu_compose_file if self.use_cpu_compose else self.compose_file cmd = self.runtime_info.compose_command + ["-f", str(self.compose_file)]
) if self.use_gpu_compose and self.gpu_compose_file.exists():
cmd = self.runtime_info.compose_command + [ cmd.extend(["-f", str(self.gpu_compose_file)])
"-f", cmd.extend(["logs", "-f", service_name])
str(compose_file),
"logs",
"-f",
service_name,
]
try: try:
process = await asyncio.create_subprocess_exec( process = await asyncio.create_subprocess_exec(

23
src/tui/managers/docling_manager.py
View file

@ -143,6 +143,29 @@ class DoclingManager:
self._external_process = False self._external_process = False
return False return False
def check_port_available(self) -> tuple[bool, Optional[str]]:
"""Check if the native service port is available.
Returns:
Tuple of (available, error_message)
"""
import socket
try:
sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
sock.settimeout(0.5)
result = sock.connect_ex(('127.0.0.1', self._port))
sock.close()
if result == 0:
# Port is in use
return False, f"Port {self._port} is already in use"
return True, None
except Exception as e:
logger.debug(f"Error checking port {self._port}: {e}")
# If we can't check, assume it's available
return True, None
def get_status(self) -> Dict[str, Any]: def get_status(self) -> Dict[str, Any]:
"""Get current status of docling serve.""" """Get current status of docling serve."""
# Check for starting state first # Check for starting state first

6
src/tui/managers/env_manager.py
View file

@ -64,7 +64,7 @@ class EnvConfig:
nudges_flow_id: str = "ebc01d31-1976-46ce-a385-b0240327226c" nudges_flow_id: str = "ebc01d31-1976-46ce-a385-b0240327226c"
# Document paths (comma-separated) # Document paths (comma-separated)
openrag_documents_paths: str = "./documents" openrag_documents_paths: str = "./openrag-documents"
# OpenSearch data path # OpenSearch data path
opensearch_data_path: str = "./opensearch-data" opensearch_data_path: str = "./opensearch-data"
@ -454,7 +454,7 @@ class EnvManager:
( (
"openrag_documents_paths", "openrag_documents_paths",
"Documents Paths", "Documents Paths",
"./documents,/path/to/more/docs", "./openrag-documents,/path/to/more/docs",
False, False,
), ),
] ]
@ -521,7 +521,7 @@ class EnvManager:
) )
if not is_valid: if not is_valid:
return ["./documents:/app/documents:Z"] # fallback return ["./openrag-documents:/app/documents:Z"] # fallback
volume_mounts = [] volume_mounts = []
for i, path in enumerate(validated_paths): for i, path in enumerate(validated_paths):

2
src/tui/screens/config.py
View file

@ -523,7 +523,7 @@ class ConfigScreen(Screen):
yield Label("Documents Paths") yield Label("Documents Paths")
current_value = getattr(self.env_manager.config, "openrag_documents_paths", "") current_value = getattr(self.env_manager.config, "openrag_documents_paths", "")
input_widget = Input( input_widget = Input(
placeholder="./documents,/path/to/more/docs", placeholder="./openrag-documents,/path/to/more/docs",
value=current_value, value=current_value,
validators=[DocumentsPathValidator()], validators=[DocumentsPathValidator()],
id="input-openrag_documents_paths", id="input-openrag_documents_paths",

73
src/tui/screens/monitor.py
View file

@ -33,13 +33,14 @@ class MonitorScreen(Screen):
("u", "upgrade", "Upgrade"), ("u", "upgrade", "Upgrade"),
("x", "reset", "Reset"), ("x", "reset", "Reset"),
("l", "logs", "View Logs"), ("l", "logs", "View Logs"),
("g", "toggle_mode", "Toggle GPU/CPU"),
("j", "cursor_down", "Move Down"), ("j", "cursor_down", "Move Down"),
("k", "cursor_up", "Move Up"), ("k", "cursor_up", "Move Up"),
] ]
def __init__(self): def __init__(self):
super().__init__() super().__init__()
self.container_manager = ContainerManager() self._container_manager = None # Use app's shared instance
self.docling_manager = DoclingManager() self.docling_manager = DoclingManager()
self.services_table = None self.services_table = None
self.docling_table = None self.docling_table = None
@ -52,6 +53,13 @@ class MonitorScreen(Screen):
# Track which table was last selected for mutual exclusion # Track which table was last selected for mutual exclusion
self._last_selected_table = None self._last_selected_table = None
@property
def container_manager(self) -> ContainerManager:
"""Get the shared container manager from the app."""
if self._container_manager is None:
self._container_manager = self.app.container_manager
return self._container_manager
def on_unmount(self) -> None: def on_unmount(self) -> None:
"""Clean up when the screen is unmounted.""" """Clean up when the screen is unmounted."""
if hasattr(self, 'docling_manager'): if hasattr(self, 'docling_manager'):
@ -69,10 +77,10 @@ class MonitorScreen(Screen):
def _create_services_tab(self) -> ComposeResult: def _create_services_tab(self) -> ComposeResult:
"""Create the services monitoring tab.""" """Create the services monitoring tab."""
# Current mode indicator + toggle # GPU/CPU mode section
yield Static("GPU Mode", id="mode-indicator", classes="tab-header")
yield Horizontal( yield Horizontal(
Static("", id="mode-indicator"), Button("Switch to CPU Mode", id="toggle-mode-btn"),
Button("Toggle Mode", id="toggle-mode-btn"),
classes="button-row", classes="button-row",
id="mode-row", id="mode-row",
) )
@ -311,17 +319,46 @@ class MonitorScreen(Screen):
"""Start services with progress updates.""" """Start services with progress updates."""
self.operation_in_progress = True self.operation_in_progress = True
try: try:
# Check for port conflicts before attempting to start
ports_available, conflicts = await self.container_manager.check_ports_available()
if not ports_available:
# Show error notification instead of modal
conflict_msgs = []
for service_name, port, error_msg in conflicts[:3]: # Show first 3
conflict_msgs.append(f"{service_name} (port {port})")
conflict_str = ", ".join(conflict_msgs)
if len(conflicts) > 3:
conflict_str += f" and {len(conflicts) - 3} more"
self.notify(
f"Cannot start services: Port conflicts detected for {conflict_str}. "
f"Please stop the conflicting services first.",
severity="error",
timeout=10
)
# Refresh to show current state
await self._refresh_services()
return
# Show command output in modal dialog # Show command output in modal dialog
command_generator = self.container_manager.start_services(cpu_mode) command_generator = self.container_manager.start_services(cpu_mode)
modal = CommandOutputModal( modal = CommandOutputModal(
"Starting Services", "Starting Services",
command_generator, command_generator,
on_complete=None, # We'll refresh in on_screen_resume instead on_complete=self._on_start_complete, # Refresh after completion
) )
self.app.push_screen(modal) self.app.push_screen(modal)
except Exception as e:
self.notify(f"Error starting services: {str(e)}", severity="error")
await self._refresh_services()
finally: finally:
self.operation_in_progress = False self.operation_in_progress = False
async def _on_start_complete(self) -> None:
"""Callback after service start completes."""
await self._refresh_services()
async def _stop_services(self) -> None: async def _stop_services(self) -> None:
"""Stop services with progress updates.""" """Stop services with progress updates."""
self.operation_in_progress = True self.operation_in_progress = True
@ -386,6 +423,19 @@ class MonitorScreen(Screen):
"""Start docling serve.""" """Start docling serve."""
self.operation_in_progress = True self.operation_in_progress = True
try: try:
# Check for port conflicts before attempting to start
port_available, error_msg = self.docling_manager.check_port_available()
if not port_available:
self.notify(
f"Cannot start docling serve: {error_msg}. "
f"Please stop the conflicting service first.",
severity="error",
timeout=10
)
# Refresh to show current state
await self._refresh_services()
return
# Start the service (this sets _starting = True internally at the start) # Start the service (this sets _starting = True internally at the start)
# Create task and let it begin executing (which sets the flag) # Create task and let it begin executing (which sets the flag)
start_task = asyncio.create_task(self.docling_manager.start()) start_task = asyncio.create_task(self.docling_manager.start())
@ -581,22 +631,21 @@ class MonitorScreen(Screen):
def _update_mode_row(self) -> None: def _update_mode_row(self) -> None:
"""Update the mode indicator and toggle button label.""" """Update the mode indicator and toggle button label."""
try: try:
use_cpu = getattr(self.container_manager, "use_cpu_compose", True) use_gpu = getattr(self.container_manager, "use_gpu_compose", False)
indicator = self.query_one("#mode-indicator", Static) indicator = self.query_one("#mode-indicator", Static)
mode_text = "Mode: CPU (no GPU detected)" if use_cpu else "Mode: GPU" indicator.update("GPU Mode" if use_gpu else "CPU Mode")
indicator.update(mode_text)
toggle_btn = self.query_one("#toggle-mode-btn", Button) toggle_btn = self.query_one("#toggle-mode-btn", Button)
toggle_btn.label = "Switch to GPU Mode" if use_cpu else "Switch to CPU Mode" toggle_btn.label = "Switch to CPU Mode" if use_gpu else "Switch to GPU Mode"
except Exception: except Exception:
pass pass
def action_toggle_mode(self) -> None: def action_toggle_mode(self) -> None:
"""Toggle between CPU/GPU compose files and refresh view.""" """Toggle between CPU/GPU compose files and refresh view."""
try: try:
current = getattr(self.container_manager, "use_cpu_compose", True) current = getattr(self.container_manager, "use_gpu_compose", False)
self.container_manager.use_cpu_compose = not current self.container_manager.use_gpu_compose = not current
self.notify( self.notify(
"Switched to GPU compose" if not current else "Switched to CPU compose", "Switched to GPU mode" if not current else "Switched to CPU mode",
severity="information", severity="information",
) )
self._update_mode_row() self._update_mode_row()

68
src/tui/screens/welcome.py
View file

@ -101,9 +101,22 @@ class WelcomeScreen(Screen):
except json.JSONDecodeError: except json.JSONDecodeError:
continue continue
# Check if any services are running # Check if services are running (exclude starting/created states)
running_services = [s for s in services if isinstance(s, dict) and s.get('State') == 'running'] # State can be lowercase or mixed case, so normalize it
self.services_running = len(running_services) > 0 running_services = []
starting_services = []
for s in services:
if not isinstance(s, dict):
continue
state = str(s.get('State', '')).lower()
if state == 'running':
running_services.append(s)
elif 'starting' in state or 'created' in state:
starting_services.append(s)
# Only consider services running if we have running services AND no starting services
# This prevents showing the button when containers are still coming up
self.services_running = len(running_services) > 0 and len(starting_services) == 0
else: else:
self.services_running = False self.services_running = False
except Exception: except Exception:
@ -220,7 +233,12 @@ class WelcomeScreen(Screen):
running_services = [ running_services = [
s.name for s in services.values() if s.status == ServiceStatus.RUNNING s.name for s in services.values() if s.status == ServiceStatus.RUNNING
] ]
self.services_running = len(running_services) > 0 starting_services = [
s.name for s in services.values() if s.status == ServiceStatus.STARTING
]
# Only consider services running if we have running services AND no starting services
# This prevents showing the button when containers are still coming up
self.services_running = len(running_services) > 0 and len(starting_services) == 0
else: else:
self.services_running = False self.services_running = False
@ -385,6 +403,34 @@ class WelcomeScreen(Screen):
async def _start_all_services(self) -> None: async def _start_all_services(self) -> None:
"""Start all services: containers first, then native services.""" """Start all services: containers first, then native services."""
# Check for port conflicts before attempting to start anything
conflicts = []
# Check container ports
if self.container_manager.is_available():
ports_available, port_conflicts = await self.container_manager.check_ports_available()
if not ports_available:
for service_name, port, error_msg in port_conflicts[:3]: # Show first 3
conflicts.append(f"{service_name} (port {port})")
if len(port_conflicts) > 3:
conflicts.append(f"and {len(port_conflicts) - 3} more")
# Check native service port
port_available, error_msg = self.docling_manager.check_port_available()
if not port_available:
conflicts.append(f"docling (port {self.docling_manager._port})")
# If there are any conflicts, show error and return
if conflicts:
conflict_str = ", ".join(conflicts)
self.notify(
f"Cannot start services: Port conflicts detected for {conflict_str}. "
f"Please stop the conflicting services first.",
severity="error",
timeout=10
)
return
# Step 1: Start container services first (to create the network) # Step 1: Start container services first (to create the network)
if self.container_manager.is_available(): if self.container_manager.is_available():
command_generator = self.container_manager.start_services() command_generator = self.container_manager.start_services()
@ -410,6 +456,20 @@ class WelcomeScreen(Screen):
async def _start_native_services_after_containers(self) -> None: async def _start_native_services_after_containers(self) -> None:
"""Start native services after containers have been started.""" """Start native services after containers have been started."""
if not self.docling_manager.is_running(): if not self.docling_manager.is_running():
# Check for port conflicts before attempting to start
port_available, error_msg = self.docling_manager.check_port_available()
if not port_available:
self.notify(
f"Cannot start native services: {error_msg}. "
f"Please stop the conflicting service first.",
severity="error",
timeout=10
)
# Update state and return
self.docling_running = False
await self._refresh_welcome_content()
return
self.notify("Starting native services...", severity="information") self.notify("Starting native services...", severity="information")
success, message = await self.docling_manager.start() success, message = await self.docling_manager.start()
if success: if success:

26
src/tui/widgets/command_modal.py
View file

@ -23,6 +23,7 @@ class CommandOutputModal(ModalScreen):
("p", "pause_waves", "Pause"), ("p", "pause_waves", "Pause"),
("f", "speed_up", "Faster"), ("f", "speed_up", "Faster"),
("s", "speed_down", "Slower"), ("s", "speed_down", "Slower"),
("escape", "close_modal", "Close"),
] ]
DEFAULT_CSS = """ DEFAULT_CSS = """
@ -188,6 +189,8 @@ class CommandOutputModal(ModalScreen):
self._output_lines: list[str] = [] self._output_lines: list[str] = []
self._layer_line_map: dict[str, int] = {} # Maps layer ID to line index self._layer_line_map: dict[str, int] = {} # Maps layer ID to line index
self._status_task: Optional[asyncio.Task] = None self._status_task: Optional[asyncio.Task] = None
self._error_detected = False
self._command_complete = False
def compose(self) -> ComposeResult: def compose(self) -> ComposeResult:
"""Create the modal dialog layout.""" """Create the modal dialog layout."""
@ -254,6 +257,12 @@ class CommandOutputModal(ModalScreen):
for w in waves.wavelets: for w in waves.wavelets:
w.speed = max(0.1, w.speed * 0.8) w.speed = max(0.1, w.speed * 0.8)
def action_close_modal(self) -> None:
"""Close the modal (only if error detected or command complete)."""
close_btn = self.query_one("#close-btn", Button)
if not close_btn.disabled:
self.dismiss()
async def _run_command(self) -> None: async def _run_command(self) -> None:
"""Run the command and update the output in real-time.""" """Run the command and update the output in real-time."""
output = self.query_one("#command-output", TextArea) output = self.query_one("#command-output", TextArea)
@ -273,8 +282,25 @@ class CommandOutputModal(ModalScreen):
# Move cursor to end to trigger scroll # Move cursor to end to trigger scroll
output.move_cursor((len(self._output_lines), 0)) output.move_cursor((len(self._output_lines), 0))
# Detect error patterns in messages
import re
lower_msg = message.lower() if message else ""
if not self._error_detected and any(pattern in lower_msg for pattern in [
"error:",
"failed",
"port.*already.*allocated",
"address already in use",
"not found",
"permission denied"
]):
self._error_detected = True
# Enable close button when error detected
close_btn = self.query_one("#close-btn", Button)
close_btn.disabled = False
# If command is complete, update UI # If command is complete, update UI
if is_complete: if is_complete:
self._command_complete = True
self._update_output("Command completed successfully", False) self._update_output("Command completed successfully", False)
output.text = "\n".join(self._output_lines) output.text = "\n".join(self._output_lines)
output.move_cursor((len(self._output_lines), 0)) output.move_cursor((len(self._output_lines), 0))

71
src/utils/langflow_headers.py Normal file
View file

@ -0,0 +1,71 @@
"""Utility functions for building Langflow request headers."""
from typing import Dict
from utils.container_utils import transform_localhost_url
def add_provider_credentials_to_headers(headers: Dict[str, str], config) -> None:
"""Add provider credentials to headers as Langflow global variables.
Args:
headers: Dictionary of headers to add credentials to
config: OpenRAGConfig object containing provider configurations
"""
# Add OpenAI credentials
if config.providers.openai.api_key:
headers["X-LANGFLOW-GLOBAL-VAR-OPENAI_API_KEY"] = str(config.providers.openai.api_key)
# Add Anthropic credentials
if config.providers.anthropic.api_key:
headers["X-LANGFLOW-GLOBAL-VAR-ANTHROPIC_API_KEY"] = str(config.providers.anthropic.api_key)
# Add WatsonX credentials
if config.providers.watsonx.api_key:
headers["X-LANGFLOW-GLOBAL-VAR-WATSONX_API_KEY"] = str(config.providers.watsonx.api_key)
if config.providers.watsonx.project_id:
headers["X-LANGFLOW-GLOBAL-VAR-WATSONX_PROJECT_ID"] = str(config.providers.watsonx.project_id)
# Add Ollama endpoint (with localhost transformation)
if config.providers.ollama.endpoint:
ollama_endpoint = transform_localhost_url(config.providers.ollama.endpoint)
headers["X-LANGFLOW-GLOBAL-VAR-OLLAMA_BASE_URL"] = str(ollama_endpoint)
def build_mcp_global_vars_from_config(config) -> Dict[str, str]:
"""Build MCP global variables dictionary from OpenRAG configuration.
Args:
config: OpenRAGConfig object containing provider configurations
Returns:
Dictionary of global variables for MCP servers (without X-Langflow-Global-Var prefix)
"""
global_vars = {}
# Add OpenAI credentials
if config.providers.openai.api_key:
global_vars["OPENAI_API_KEY"] = config.providers.openai.api_key
# Add Anthropic credentials
if config.providers.anthropic.api_key:
global_vars["ANTHROPIC_API_KEY"] = config.providers.anthropic.api_key
# Add WatsonX credentials
if config.providers.watsonx.api_key:
global_vars["WATSONX_API_KEY"] = config.providers.watsonx.api_key
if config.providers.watsonx.project_id:
global_vars["WATSONX_PROJECT_ID"] = config.providers.watsonx.project_id
# Add Ollama endpoint (with localhost transformation)
if config.providers.ollama.endpoint:
ollama_endpoint = transform_localhost_url(config.providers.ollama.endpoint)
global_vars["OLLAMA_BASE_URL"] = ollama_endpoint
# Add selected embedding model
if config.knowledge.embedding_model:
global_vars["SELECTED_EMBEDDING_MODEL"] = config.knowledge.embedding_model
return global_vars

2
tests/integration/test_startup_ingest.py
View file

@ -29,7 +29,7 @@ async def wait_for_ready(client: httpx.AsyncClient, timeout_s: float = 30.0):
def count_files_in_documents() -> int: def count_files_in_documents() -> int:
base_dir = Path(os.getcwd()) / "documents" base_dir = Path(os.getcwd()) / "openrag-documents"
if not base_dir.is_dir(): if not base_dir.is_dir():
return 0 return 0
return sum(1 for _ in base_dir.rglob("*") if _.is_file() and _.name not in EXCLUDED_INGESTION_FILES) return sum(1 for _ in base_dir.rglob("*") if _.is_file() and _.name not in EXCLUDED_INGESTION_FILES)

3015
uv.lock generated
View file

File diff suppressed because it is too large Load diff