photo-service-gui
webserver photo-service-gui
Hovedfunksjonen til photo-service-gui er å analysere video og generere events på passeringer. Events inkuderer både skjermbilder og analyse av innholdet som deles på PubSub.
Slik går du fram for å kjøre dette lokalt
Utvikle og kjøre lokalt
Krav til programvare
- pyenv (recommended)
- poetry
- nox
- nox-poetry
Installere programvare og sette miljøvariable
% git clone https://github.com/langrenn-sprint/photo-service-gui.git
% cd photo-service-gui
% pyenv install 3.11
% pyenv local 3.11
% pipx install poetry
% pipx install nox
% pipx inject nox nox-poetry
% poetry install
oppdatere
% poetry update / poetry add <module>
Miljøvariable
Du må sette opp ei .env fil med miljøvariable. Eksempel:
JWT_SECRET=secret
ADMIN_USERNAME=admin
ADMIN_PASSWORD=password
DB_USER=admin
DB_PASSWORD=password
EVENTS_HOST_SERVER=localhost
EVENTS_HOST_PORT=8082
PHOTOS_HOST_SERVER=localhost
PHOTOS_HOST_PORT=8092
FERNET_KEY=23EHUWpP_MyKey_MyKeyhxndWqyc0vO-MyKeySMyKey=
GOOGLE_APPLICATION_CREDENTIALS=/Users/name/github/secrets/application_default_credentials.json
GOOGLE_CLOUD_PROJECT=sigma-celerity-257719
GOOGLE_PUBSUB_NUM_MESSAGES=1
GOOGLE_PUBSUB_TOPIC_ID=langrenn-sprint
GOOGLE_PUBSUB_SUBSCRIPTION_ID=langrenn-sprint-sub
GOOGLE_STORAGE_BUCKET=langrenn-sprint
GOOGLE_STORAGE_SERVER=https://storage.googleapis.com
GOOGLE_OAUTH_CLIENT_ID=12345My-ClientId12345.apps.googleusercontent.com
SERVICEBUS_NAMESPACE_CONNECTION_STR=connection string
JWT_EXP_DELTA_SECONDS=3600
LOGGING_LEVEL=INFO
USERS_HOST_SERVER=localhost
USERS_HOST_PORT=8086
### Rydde opp i docker stuff
docker system prune
### Starte services i docker
sudo docker-compose pull #oppdatere images
sudo docker-compose up --build #bygge og debug modus
sudo docker-compose up -d #kjøre-modus
### Oppdatere services i docker
sudo docker-compose stop #oppdatere images
sudo docker-compose pull #oppdatere images
sudo git pull #photo-service-gui
sudo docker-compose up --build #bygge og debug modus
sudo docker-compose stop #oppdatere images
sudo docker-compose up -d #kjøre-modus
### Innstillinger i google cloud
- Create OAuth2.0 client Id: <https://console.cloud.google.com/apis/credentials>
- Hints1: Javascript origins: http://localhost:8080 and http://localhost
- Hints2: Redirect URI: http://localhost:8080/photos_adm and http://localhost/photos_adm
- Download client_secret.json and save it in secrets folder, remember to add it to .env file
- Set up conset screen
- PUBSUB: Create topic and subscription
- Install python libraries: pip install --upgrade google-cloud-pubsub
- Set upp application default credentials: https://cloud.google.com/docs/authentication/provide-credentials-adc#how-to
- Cloud storage: Bucket - https://storage.googleapis.com/langrenn-photo/result2.jpg
- Hint: Set to publicly available and allUsers principals, role Viewer
Denne fila _skal_ ligge i .dockerignore og .gitignore
### Kjøre webserver lokalt
Start lokal webserver mha aiohttp-devtools(adev)
source .env
poetry run adev runserver -p 8096 photo_service_gui
poetry run adev runserver --no-livereload -p 8096 photo_service_gui
docker-compose up photo-service user-service event-service competition-format-service mongodb
start server uten reload (for live kjøring - unngår reload når bilder lagres)
poetry run adev runserver –no-livereload -p 8096 photo_service_gui
Dokumentasjon: https://langrenn-sprint.github.io/docs/
Når du endrer koden i photo_service_gui, vil webserveren laste applikasjonen på nytt autoamtisk ved lagring.
Referanser
aiohttp: https://docs.aiohttp.org/
Googel OAuth2: https://developers.google.com/identity/protocols/oauth2 Google Photos API: https://developers.google.com/photos/library/guides/get-started
testfiler for video
VIDEO_URL=http://10.0.0.6:8080/video VIDEO_URL=https://harnaes.no/maalfoto/2023SkiMaal.mp4 VIDEO_URL=20240113_105758_KJ.mp4 VIDEO_URL=https://harnaes.no/maalfoto/2023SkiMaal.mp4 VIDEO_URL=/Users/t520834/github/vision-ai-service/20240113_105758_KJ.mp4 VIDEO_URL=rtsp://stream:Video1@10.0.0.18:88/videoMain VIDEO_URL=rtsp://stream:Video1@192.168.1.160:88/videoSub VIDEO_URL=http://192.168.1.152:8080/video