Custom Checker for Approximate Solution Questions

Last updated: July 9, 2026

A custom checker evaluates a candidate submission by using your validation logic instead of relying only on exact output matching.

You can use a custom checker when:

Multiple valid answers are possible.
Output formatting can vary.
Answers can be approximate, such as when you allow numerical tolerance.
You want to assign partial credit.
You need domain-specific validation rules.

How a custom checker works

When you use a custom checker, HackerRank evaluates submissions at the test case level.

For each test case:

The platform runs the candidate’s code.
The platform passes the generated output to the custom checker.
The custom checker applies your scoring logic.
The custom checker returns a result, score, and message for the test case.

To add a custom checker to an approximate solution question, see Step 5: Custom Checker.

Custom checker structure

The custom checker contains three sections:

Head section (non-editable)
Body section (editable)
Tail section (non-editable)

You can write your validation logic in the body section.

The checker body must define the following method:

run_custom_checker (TestStructure t_obj, ResultStructure r_obj)

Note: Do not print anything to STDOUT inside run_custom_checker(), as it can interfere with the evaluation process.

TestStructure object

The TestStructure object provides access to test case data and execution details.

Field	Type	Description
`testcase_id`	Integer	ID of the test case
`testcase_input_path`	String	Path to the test case input file
`testcase_output_path`	String	Path to the candidate-generated output file
`testcase_expected_output_path`	String	Path to the expected output file that the system uses to compare against the candidate’s output.
`metadata_file_paths`	List of strings	Paths to metadata files (for example, training datasets)
`submission_code_path`	String	Path to the candidate’s source code
`testcase_result`	Boolean	Indicates whether the candidate’s output matches the expected output. The system performs a line-by-line comparison.
`testcase_signal`	Integer	Exit code of the test case process
`testcase_time`	Floating	Time taken by the test case process, in seconds.
`testcase_memory`	Integer	Peak memory usage of the test case process, in bytes.
`data`	String	Additional metadata associated with the test

ResultStructure fields

You can use the ResultStructure object to define the outcome for a test case.

Field	Type	Description
`result`	Boolean	Indicates whether the test case passes. Set to `True` if the submission meets the required criteria. Set to `False` if the submission does not meet the criteria.
`score`	Floating	Specifies the normalized score for the test case. The value must be between `0` and `1`.
`message`	String	Custom message that explains the test case result. The system displays this message to the candidate. For example, Failed because the cutoff was not reached.

Example

The following example checks whether the candidate outputs five unique prime numbers under 100.

Scoring logic:

Score = number of valid primes / 5
The candidate passes if at least four valid primes are correct

import re
# Function to determine if the given integer is a prime numberdef is_prime_number(x):
    if x >= 2:
        for y in range(2, x):
            if not x % y:
                return False
    else:
        return False
    return True

def run_custom_checker(t_obj, r_obj):
    result_data = ''
    try:
        result_data = open(t_obj.testcase_output_path, 'r').read()
    except IOError:
        r_obj.result = False
        r_obj.score = 0
        r_obj.message = 'Error reading result file'
        return

    # Read contents of the result file
    values = re.split(' |\n', result_data)

    # Make sure all the values are unique
    uniq_values = set(values)

    # Count the number of primes
    correct_values = 0
    for value in uniq_values:
        if value and is_prime_number(int(value)):
                correct_values = correct_values + 1

    # Cutoff score to determine success
    if correct_values > 3:
        r_obj.result = True
    else:
        r_obj.result = False
    r_obj.score = correct_values / 5
    r_obj.message = str(correct_values) + ' out of 5 values are correct'

Result

You can also view a similar custom checker example in Java.