Problem examples

This repository contains real-world problem examples. See problem_examples on GitHub for details.

List of examples

Grading type	Directory	Description
Standard grading	standard/aplusb	A simple A+B problem
Batched grading	batched/hungry	A problem with multiple subtasks
Custom Grading	grader/shortest1	A problem that needs special grading logic
Interactive Grading	interactive/seed2	An interactive problem (Python)
Native Interactive	interactive/seed2native	An interactive problem (C++)
Signature Grading	signature/fastbit	A function-implementation problem (IOI-style)
Generator	generator/ds3	A problem that uses a generator to create test data

Directory structure

problem_examples/
├── standard/          # Standard grading problems
│   └── aplusb/
├── batched/          # Problems with batches/subtasks
│   └── hungry/
├── grader/           # Problems using a custom grader
│   └── shortest1/
├── interactive/      # Interactive problems
│   ├── seed2/        # Python interactor
│   └── seed2native/  # C++ interactor
├── signature/        # IOI-style problems
│   └── fastbit/
└── generator/        # Problems using a generator
    └── ds3/

Usage

Each example directory contains:

• init.yml - The problem configuration file
• README.md - Detailed instructions
• Test data or a generator
• Checker/grader (if any)

Details

1. Standard grading

The simplest kind of problem: the contestant reads input from stdin and prints output to stdout.

Example: Compute A + B

archive: aplusb.zip
test_cases:
- {in: aplusb.1.in, out: aplusb.1.out, points: 50}
- {in: aplusb.2.in, out: aplusb.2.out, points: 50}

2. Batched grading

A problem with multiple subtasks, each containing multiple test cases. All test cases in a subtask must pass to earn its points.

Example:

test_cases:
- points: 30
  batched:
  - {in: test1.1.in, out: test1.1.out}
  - {in: test1.2.in, out: test1.2.out}
- points: 70
  batched:
  - {in: test2.1.in, out: test2.1.out}
  - {in: test2.2.in, out: test2.2.out}
  - {in: test2.3.in, out: test2.3.out}

3. Custom Grading

A problem that needs special grading logic beyond comparing output.

Example: A shortest-path problem with multiple correct answers.

custom_judge: grader.py
test_cases:
- {in: test1.in, out: test1.out, points: 100}

4. Interactive Grading

An interactive problem, where the contestant's program exchanges data with the grader.

Example: A number-guessing problem

custom_judge: interactor.py
unbuffered: true
test_cases:
- {in: test1.in, points: 100}

5. Native Interactive Grading

Like interactive grading, but the interactor is written in C/C++ for high performance.

unbuffered: true
interactive:
  files: interactor.cpp
  type: testlib
test_cases:
- {in: test1.in, points: 100}

6. Signature Grading (IOI-style)

The contestant implements a function and does not need to read input or write output.

Example: Implement the function is_valid(n)

signature_grader:
  entry: handler.c
  header: header.h
test_cases:
- {in: test1.in, out: test1.out, points: 100}

7. Generator

Use a generator to create test data automatically.

generator: gen.cpp
test_cases:
- {generator_args: [10], points: 30}
- {generator_args: [100], points: 30}
- {generator_args: [1000], points: 40}

References

• DMOJ Problem Examples
• Testlib Documentation
• Polygon System

Notes

• The full source code for all of the examples above is available in the DMOJ repository
• You can combine multiple techniques in a single problem
• Test thoroughly before publishing a problem on the production system