oj-docs/judgeserver/testcase.md

Testcase format

Each testcase directory contains its testcase files, the conetent of input files will be redirected to stdin, and stdout of process will be compared with output files(normal problem) or judged by your special judge code(special judge).

Testcase meta data is in info file.

Normal problem

Both input and output files are required.

root@JudgeServer:~/test_case/normal# tree
.
|-- 1.in
|-- 1.out
`-- info

0 directories, 3 files
root@JudgeServer:~/test_case/normal# cat 1.in
1 2
root@JudgeServer:~/test_case/normal# cat 1.out
3

Example of info file

• spj: for normal problem, it is false
• test_cases: name - data dict
  • stripped_output_md5: md5 of output file which trailing empty characters has been removed
  • input_name and output_name: name of input and output file
  • input_size and output_size: size of input and output file

{
    "spj": false,
    "test_cases": {
        "1": {
            "stripped_output_md5": "eccbc87e4b5ce2fe28308fd9f2a7baf3",
            "output_size": 2,
            "input_name": "1.in",
            "input_size": 4,
            "output_name": "1.out"
        }
    }
}

Special Judge

You only need input files.

root@JudgeServer:~/test_case/spj# tree
.
|-- 1.in
`-- info

0 directories, 3 files
root@JudgeServer:~/test_case/normal# cat 1.in
1 2

Example of info file

• spj: for special judge problem, it is true
• test_cases: name - data dict
  • input_name: name of input file
  • input_size: size of input file

{
    "spj": true,
    "test_cases": {
        "1": {
            "input_name": "1.in",
            "input_size": 4
        }
    }
}