Solomon/code execute

[solomonng2001]

Execution Service

Installation

1. Install dependencies:

go mod tidy

2. Create the .env file from copying the .env.example, and copy the firebase JSON file into execution-service/ fill in the FIREBASE_CREDENTIAL_PATH with the path of the firebase credential JSON file.

Running the Application

To start the server, run the following command:

go run main.go

The server will be available at http://localhost:8083.

Running the Application via Docker

To run the application via Docker, run the following command:

docker build -t execution-service .

docker run -p 8083:8083 --env-file .env -d execution-service

The server will be available at http://localhost:8083.

API Endpoints

• POST /tests/populate
• GET /tests/{questionDocRefId}/
• POST /tests/{questionDocRefId}/execute
• POST /tests/{questionDocRefId}/submit

Managing Firebase

To reset and repopulate the database, run the following command:

go run main.go

Repopulate test cases

To repopulate test cases, you need to repopulate the questions in the question-service, which will automatically call the execution-service to populate the test cases.

In question-service, run the following command:

go run main.go -populate

API Documentation

GET /tests/{questionDocRefId}/

To read visible test cases via a question ID, run the following command:

curl -X GET http://localhost:8083/tests/{questioinDocRefId}/ \
-H "Content-Type: application/json"

The following json format will be returned:

[
   {
      "input":"hello",
      "expected":"olleh"
   }
]

POST /tests/{questionDocRefId}/execute

To execute test cases via a question ID without custom test cases, run the following command, with custom code and language:

curl -X POST http://localhost:8083/tests/{questioinDocRefId}/execute \
-H "Content-Type: application/json" \
-d '{
"code": "name = input()\nprint(name[::-1])",
"language": "Python"
}'

The following json format will be returned:

{
  "visibleTestResults":[
    {
      "input":"hello",
      "expected":"olleh",
      "actual":"olleh",
      "passed":true,
      "error":""
    }
  ],
  "customTestResults":null
}

To execute visible and custom test cases via a question ID with custom test cases, run the following command, with custom code, language and custom test cases:

curl -X POST http://localhost:8083/tests/{questioinDocRefId}/execute \
-H "Content-Type: application/json" \
-d '{
"code": "name = input()\nprint(name[::-1])",
"language": "Python",
"customTestCases": "2\nHannah\nhannaH\nabcdefg\ngfedcba\n"
}'

The following json format will be returned:

{
  "visibleTestResults":[
    {
      "input":"hello",
      "expected":"olleh",
      "actual":"olleh",
      "passed":true,
      "error":""
    }
  ],
  "customTestResults":[
    {
      "input":"Hannah",
      "expected":"hannaH",
      "actual":"hannaH",
      "passed":true,
      "error":""
    },
    {
      "input":"abcdefg",
      "expected":"gfedcba",
      "actual":"gfedcba",
      "passed":true,
      "error":""
    }
  ]
}

POST /tests/{questionDocRefId}/submit

To submit a solution and execute visible and hidden test cases via a question ID, run the following command, with custom code and language:

curl -X POST http://localhost:8083/tests/{questioinDocRefId}/submit \
-H "Content-Type: application/json" \
-d '{
"title": "Example Title",
"code": "name = input()\nprint(name[::-1])",
"language": "Python",
"user": "user123",
"matchedUser": "user456",
"matchedTopics": ["topic1", "topic2"],
"questionDifficulty": "Medium",
"questionTopics": ["Loops", "Strings"]
}'

The following json format will be returned:

{
  "visibleTestResults":[
    {
      "input":"hello",
      "expected":"olleh",
      "actual":"olleh",
      "passed":true,
      "error":""
    }
  ],
  "hiddenTestResults":{
    "passed":2,
    "total":2
  },
  "status":"Accepted"
}

If compilation error exists or any of the tests (visible and hidden) fails, status "Attempted" will be returned:

{
  "visibleTestResults":[
    {
      "input":"hello",
      "expected":"olleh",
      "actual":"",
      "passed":false,
      "error":"Command execution failed: Traceback (most recent call last):\n  File \"/tmp/4149249165.py\", line 2, in \u003cmodule\u003e\n    prit(name[::-1])\n    ^^^^\nNameError: name 'prit' is not defined. Did you mean: 'print'?\n: %!w(*exec.ExitError=\u0026{0x4000364678 []})"
    }
  ],
  "hiddenTestResults":{
    "passed":0,
    "total":2
  },
  "status":"Attempted"
}

TODOs

• Add validation to API endpoints
• Add validation to test cases in database which will be used to validate the custom test cases

[tituschewxj]

LGTM! Thanks for implementing the code execution service. Once the comments are addressed you can merge this PR, as we still need to integrate the other changes together.

We probably can work on the message queuing after this week's demo as we don't want to cause potential issues with deployment.

However, there may be an issue when repopulating the questions as it overwrites the questionDocRefId with random values each time, which affects other services as they depend on this value being the same.

[tituschewxj]

There may be a potential problem with questionDocRefId, which is whenever the function in apps/question-service/utils/populate.go is run, all the questionDocRefId are reinitialized to random values, which affects the retrieval of records that utilize this field as an index.

We can either ensure that questionDocRefId remains the same when updating the DB, or use something else to identify a question.