Valkeyrie is a python GraphQL client which serves as the evalhalla backend, providing simple REST routes to create, update and respond to surveys. If you are a curious on looker this repo is probably not as much valuable to you. Here is the Evalhalla application repo
In case you're interested. Valkyrie uses the following main components
- Google Natural Language API: Used to get sentiment values for text
- Flask: Micro web framework for python
- sgqlc: GraphQL client for python with schema introspection to generate native python object
- Flask-RESTful: Utility to efficiently create API views
The GraphQL server is not included in this repo but the apollo server can be found in this repo
Most variables can be set as either environment variables or variables in either development.py, production.py, or default.py . As the names imply default.py applies for the development environment, production.py for the production environment and default.py for variables that exist across both environments. default.py variables will be overriden if the same variable exists in the development.py or the production.py in their respective environments.
APPLICATION_NAME
Explanation: I am using a flask boilerplate that I created. So rather than hardcode the application name for some configs. I am using this config. This is also used as a means to ensure there are no environment variable naming collisions if I am running multiple applications on the same host. This name will be the pre-cursor to any environment variables i.e. ${APPLICATION_NAME}_WHATEVER_ENVIRONMENT_VARIABLE . This does not apply to the builtin Flask variables
Value: VALKYRIE
Has Environment Variable: No
RESTFUL_JSON
Explanation: variable used to configure the Flask-RESTful utility such as enabling non ascii characters in the API.
Value:
{
"ensure_ascii": False
}Has Environment Variable: No
SEND_FILE_MAX_AGE_DEFAULT
Explanation: builtin Flask variable which sets the caching time.
Value: 0
Has Environment Variable: No
VARIABLES
Explanation: Rather than hard-coding the checks for different environment variables use JSON to dynamically check for requirements
Value:
{
"GRAPHQL_ENDPOINT": {
"description": "The graphql endpoint the client should call",
"required": True,
"default": "http://localhost:4000/graphql"
},
"GRAPHQL_HEADERS":{
"description": "HTTP Headers if any to be used for requests to the graphQL endpoint",
"required": False
},
"SLACK_URL":{
"description": "The slack webhook endpoint to post logs to according to configuration",
"required": False
},
"ENABLE_NATURAL_LANGUAGE_PROCESSING":{
"description": "Flag to enable or disable natural language processing",
"required": True,
"default": True
}
}Note : Variables with environment variables are in the form ${APPLICATION_NAME}_VARIABLE_NAME
GRAPHQL_ENDPOINT
Explanation: The graphql endpoint the client should call
Required: Yes
Development Default: http://localhost:4000/graphql
Has Environment Variable: Yes
GRAPHQL_HEADERS
Explanation: HTTP Headers if any to be used for requests to the graphQL endpoint
Required: No
Has Environment Variable: No
SLACK_URL
Explanation: The slack webhook used to post logs to custom slack bot according to the logging slack handler configuration
Required: Yes only if logging handler is defined and configured
Has Environment Variable: Yes
ENABLE_NATURAL_LANGUAGE_PROCESSING
Explanation: Flag to enable or disable natural language processing
Required: No
Has Environment Variable: Yes
GOOGLE_APPLICATION_CREDENTIALS
ENVIRONMENT VARIABLE ONLY
Explanation: The path to the JSON credentials file for the Google Natural Language API
Required: Yes, if ENABLE_NATURAL_LANGUAGE_PROCESSING is set to True