QDUOJ Problem Creator

Create QDUOJ problems in the QDUOJ_import/ directory following the official import format.

CRITICAL: Template Fields Cannot Be Empty

ALL template fields (prepend, template, append) MUST be non-empty strings.

Empty strings like "prepend": "" will cause a server error on import because QDUOJ's TemplateSerializer requires non-empty CharField values.

• •Correct: "prepend": "# No imports needed"
• •Wrong: "prepend": ""

Single Problem Creation

Instructions

1. •

   Find the next available sequential folder number in QDUOJ_import/ (e.g., if 1/, 2/, 3/ exist, create 4/)

2. •

   Create the folder structure:

   code

   QDUOJ_import/N/
   ├── problem.json
   └── testcase/
       ├── 1.in
       ├── 1.out
       ├── 2.in
       ├── 2.out
       └── ...
   

3. •

   Create problem.json with this structure:

   json

   {
       "display_id": "UniqueID001",
       "title": "Problem Title",
       "description": {"format": "html", "value": "<p>Description</p>"},
       "input_description": {"format": "html", "value": "<p>Input format</p>"},
       "output_description": {"format": "html", "value": "<p>Output format</p>"},
       "hint": {"format": "html", "value": ""},
       "samples": [{"input": "sample input", "output": "sample output"}],
       "test_case_score": [
           {"input_name": "1.in", "output_name": "1.out", "score": 10}
       ],
       "time_limit": 1000,
       "memory_limit": 128,
       "difficulty": "Low",
       "template": {
           "Python3": {
               "prepend": "# No imports needed",
               "template": "# Your code here\npass",
               "append": "# End of submission"
           }
       },
       "spj": null,
       "rule_type": "ACM",
       "source": "",
       "answers": [],
       "tags": ["beginner"]
   }
   

4. •

   Create test cases in testcase/ directory (scores should sum to 100)

5. •

   Create the import ZIP after all files are ready (see ZIP Creation section)

Code Templates

Templates allow you to wrap user submissions with boilerplate code.

When to Use Templates

If the source problem markdown has a ## Template: section, use that template in QDUOJ with I/O handled in the append section.

How Templates Work

When a submission is judged, QDUOJ combines the code like this:

code = f"{template['prepend']}\n{user_code}\n{template['append']}"

• •prepend: Code added BEFORE the user's submission (hidden from user)
• •template: The starter code shown to the user in the editor
• •append: Code added AFTER the user's submission (hidden from user)

Template Structure

"template": {
    "Python3": {
        "prepend": "# Prepend",
        "template": "def solve(n: int) -> int:\n    # Your code here\n    pass",
        "append": "# Append"
    }
}

Critical: prepend and append CANNOT be empty strings. Use "# Prepend" and "# Append" as placeholders.

Template Example: Function Implementation

Student implements a function, template handles I/O:

"template": {
    "Python3": {
        "prepend": "# Prepend",
        "template": "def is_leap_year(year: int) -> bool:\n    # Your code here\n    pass",
        "append": "if __name__ == \"__main__\":\n    year = int(input())\n    if is_leap_year(year):\n        print(\"Leap year\")\n    else:\n        print(\"Not leap year\")"
    }
}

In this example:

• •User only sees and edits the is_leap_year function
• •The append section handles reading input and printing output
• •Test cases use standard I/O but user focuses on logic

Template Example: Hidden Test Harness

"template": {
    "Python3": {
        "prepend": "import sys\nfrom io import StringIO",
        "template": "def process(data: list[int]) -> int:\n    # Your code here\n    pass",
        "append": "# Test harness\nresult = process([int(x) for x in input().split()])\nprint(result)"
    }
}

Batch Problem Creation

When creating multiple problems at once:

1. •

   Plan the problems first: List all problems with their folder numbers before creating any files

2. •

   Create sequential folders: Each problem gets its own numbered folder

   code

   QDUOJ_import/
   ├── 1/
   │   ├── problem.json
   │   └── testcase/
   ├── 2/
   │   ├── problem.json
   │   └── testcase/
   └── 3/
       ├── problem.json
       └── testcase/
   

3. •

   Use unique display_ids: Each problem needs a unique display_id (max 24 characters)

   • •Pattern: {Topic}{Number} (e.g., LeapYear001, PrintName001)
4. •

   Batch ZIP creation: Include all problems in a single ZIP for atomic import

   python

   import os
   import zipfile
   from pathlib import Path
   
   def create_batch_qduoj_zip(source_folders: list[str], output_filename: str):
       """Create QDUOJ import ZIP with sequential numbering starting from 1."""
       with zipfile.ZipFile(output_filename, "w", zipfile.ZIP_DEFLATED) as zipf:
           for idx, source_dir in enumerate(source_folders, start=1):
               source_path = Path(source_dir)
               for root, dirs, files in os.walk(source_path):
                   for file in files:
                       file_path = os.path.join(root, file)
                       rel_path = os.path.relpath(file_path, source_path)
                       arcname = os.path.join(str(idx), rel_path)
                       zipf.write(file_path, arcname)
       print(f"Created {output_filename} with {len(source_folders)} problems")
   
   # Example: batch import of problems 1-5
   folders = [f"QDUOJ_import/{i}" for i in range(1, 6)]
   create_batch_qduoj_zip(folders, "QDUOJ_import/batch_import.zip")
   

ZIP Creation

Single Problem ZIP

import os
import zipfile
from pathlib import Path

def create_qduoj_zip(source_folder: str, output_filename: str):
    """Create QDUOJ import ZIP with folder as 1/."""
    with zipfile.ZipFile(output_filename, "w", zipfile.ZIP_DEFLATED) as zipf:
        source_path = Path(source_folder)
        for root, dirs, files in os.walk(source_path):
            for file in files:
                file_path = os.path.join(root, file)
                rel_path = os.path.relpath(file_path, source_path)
                arcname = os.path.join("1", rel_path)
                zipf.write(file_path, arcname)
    print(f"Created {output_filename}")

# Example: create ZIP from folder N
create_qduoj_zip("QDUOJ_import/N", "QDUOJ_import/problem_name_import.zip")

Expected ZIP Structure

The QDUOJ importer expects this exact structure inside the ZIP:

1/
├── problem.json
└── testcase/
    ├── 1.in
    ├── 1.out
    └── ...
2/
├── problem.json
└── testcase/
    └── ...

Critical: The importer iterates for i in range(1, count+1) - folders MUST be numbered sequentially starting from 1 with no gaps.

Critical Rules

• •No hints by default: Leave hint value empty ("") unless explicitly requested
• •Sequential numbering: ZIP folders MUST be 1/, 2/, 3/, etc. - no gaps, always start from 1
• •Template fields: If using templates, prepend and append CANNOT be empty strings - use "# Prepend" and "# Append" as placeholders
• •Format objects: Description fields must use {"format": "html", "value": "..."}
• •Difficulty: Use "Low", "Mid", or "High"
• •Rule type: Use "ACM" (pass/fail) or "OI" (partial scoring)
• •Display ID: Max 24 characters, must be unique per problem
• •Test case scores: Must sum to 100
• •Always create ZIP: After creating problem folder(s), generate the import ZIP file

Import Steps (for reference)

1. •Go to QDUOJ Admin -> Problem -> Import QDUOJ Problems (beta)
2. •Upload the ZIP file
3. •All problems are imported atomically (within a transaction)
4. •Problems are created as hidden by default
5. •Edit each problem to make it visible and restrict languages

Ask the user for problem details before creating.