ocr-api-consumer

Service to consume requests for extraction of text from images and manage the requests to the OCR API.

When this application starts up it will consume messages on both the Main and Retry Topic.

There is no error topic processing so far since all clients require this data in a timely manner and so errors need to be sent back to the calling system in this case

Requirements

• Java 11
• Maven
• Docker

Usage without Docker

• Run make dev to build JAR (versioned in target and unversioned in top level d) and run the unit tests
• Run java -jar ocr-api-consumer.jar to run the application
• Alternatively, export OCR_API_CONSUMER_PORT={your-chosen-port-number} and run ./start.sh to run the application on your chosen port number.

Usage with Docker

The service can be run using docker, with the addition of the jib maven plugin.

• Run mvn clean
• Run mvn compile jib:dockerBuild to compile the project into a docker image
• Run docker run -e OCR_API_URL -t -i -p 8080:8080 ocr-api-consumer:unversioned to run the service on port 8080.

Environment Variables

The following is a list of mandatory environment variables for the service to run:

Testing Locally (dev)

Testing with curl

curl --noproxy "*" --header "Content-Type: application/json" --request POST --data '{"app_id":"CURL_TEST_X", "response_id":"9616660641","image_endpoint":"http://test.image/endpoint","converted_text_endpoint":"http://test.converted.text/endpoint"}' --write-out '%{http_code}' http://kafka-url/

Running tests with Maven

You are able to test with Maven by running:

mvn test

To test just integration tests or just unit tests run:

mvn test -Dincluded.tests=integration-test

mvn test -Dincluded.tests=unit-test