A simple demo server for use. It serves some static content from a specified directory, and provides proxy to the API in a way that enables the API-key to be kept private.
# Install dependencies yarn install # Make sure the build directory is setup (has at least a copy of the template config) yarn run setup # Build the code (in build/) yarn run build # Start a dev server locally on port 8080 yarn run start-dev
Now you can visit
http://localhost:8080/ and you should get a "hello world!" page.
To configure the server, create a
server_config.json file with the following
"port": The port to run on. The default is 8081 for development and 8080 for production if this is left empty.
"staticPath": The path where the static html/css/js resources are located.
that this path is relative to the directory of the run_server.js file, which
is in you projects
"googleCloudApiKey": The API key for your google cloud project.
"toxicityAttribute": The name of the attribute to use for toxicity scores. See the PerspectiveAPI documentation for more details.
Note: The following can be done with npm as well as yarn, but yarn is recommended.
To use this server, run:
yarn add @conversationai/perspectiveapi-simple-server yarn install
This will add the server to your
Then, to run the server, call:
node node_modules/@conversationai/perspectiveapi-simple-server/build/server/run_server.js server_config.json
This is example code to help experimentation with the Perspective API; it is not an official Google product.