Python back end documentation

Python API

This API will check correctness of the provided questions in python. There are specific types of question models that are defined. This wiki page will outline what the endpoints are expecting and what will be returned.

For all these following endpoints:

• If not POST, 405 is returned
• If Content-Type header is not application/json, 415 is returned

/checker/writeCode

This endpoint will run userAnswer and testCode and compare the standard output of each program.
Expected body:

{
    userAnswer?: string,
    testCode?: string
}

Where userAnswer is the student's provided answer and testCode is the code that should be ran. If the variable is not provided, it will become an empty string.

Response:

{
    pass: boolean,
    failMessage?: string
}

/checker/multipleChoice

This endpoint will compare userAnswer to expectedAnswer where these two variables are both strings.

Expected body:

{
    userAnswer?: string,
    expectedAnswer?: string
}

Where userAnswer is the student's provided answer and expectedAnswer is the correct choice. If the variable is not provided, it will become an empty string.

Response:

{
    pass: boolean,
    failMessage?: string
}

/checker/shortAnswer

This endpoint will compare the userAnswer to expectedAnswer as a string if questionCode isn't provided, otherwise it will run the code and compare the outputs.

Expected body:

{
    userAnswer?: string,
    questionCode?: string,
    expectedAnswer?: string
}

If questionCode is provided, it will run both the userAnswer and questionCode then compare the outputs.

If questionCode is not provided, it will just compare the userAnswer to expectedAnswer as a string.

Response:

{
    pass: boolean,
    failMessage?: string
}

/checker/checkboxQuestion

This endpoint will compare the userAnswer to the expectedAnswer. Given the data type of these two variables described below, it checks for the entire contents to be similar, otherwise it fails. This can be expanded on in the future in order to check each individual wrong solution to determine customized responses, but for now it just checks for simple correctness.

Note that the variables below are required.

Expected body:

{
    userAnswer: string[],
    expectedAnswer: string[]
}

Response:

{
    pass: boolean,
    failMessage?: string
}

/checker/table

This endpoint compares a 2d array of answers (m x n) to a 2d array of questions (m x n), and returns a 2d array with the results (m x n)

Expected body and interface:

{
    answer: Answer[n][m],
    questions: Question[n][m]
}

interface Answer = string[] | string
interface Question = {
    type: "fillBlank" | "multipleChoice" | "writeCode" | "checkboxQuestion",
    code?: string
    answer?: string[] | string
}

Answer is a string or string array. The string array is for checkbox questions.

A single question should correspond to the answer, and they should be similar to the previous endpoints. For example, a string[] in the Question's answer field is for a checkboxQuestion, and therefore the corresponding Answer is expected to also be a string[].

Question.answer is not needed in all cases because if you recall in some of the other endpoints, we compared the output of the code rather than the expectedAnswer, same thing in this case.

Response and interface:

Response[n][m]

interface Response = {
    pass?: boolean,
    failMessage?: string,
    blank?: true 
}

If the question type does not match, it will result in the response for that cell just being {blank: true}

/checker/memoryTable

This endpoint compares a memory table's expected outputs to the user's input and determines if they were correct or not

Expected body and interface:

{
    expectedAnswer: {
        [key: string]: string[]
    },
    userAnswer: {
        [key: string]: string[]
    }
}

It will basically compare the expectedAnswer object to the userAnswer object and return a pass true if these are deeply equivalent, otherwise it returns false

Response:

{
    pass: boolean,
    failMessage?: string
}