code-runner is a command line application that reads code as a json payload from stdin – compiles and runs the code – and writes the result as json to stdout. This is used by glot-images to run code on glot.io See the overview on how everything is connected.

code-runner requires that the compiler / interpreter for the languages you want to run is installed and is in PATH.

• Assembly
• Ats
• Bash
• C
• Clojure
• Cobol
• CoffeeScript
• Cpp
• Crystal
• Csharp
• D
• Elixir
• Elm
• Erlang
• Fsharp
• Go
• Groovy
• Haskell
• Idris
• Java
• JavaScript
• Julia
• Kotlin
• Lua
• Mercury
• Nim
• Ocaml
• Perl
• Php
• Python
• Raku
• Ruby
• Rust
• Scala
• Swift
• TypeScript

The input is required to be a json object containing the properties language and files. language must be a lowecase string matching one of the supported languages. files must be an array with at least one object containing the properties name and content. name is the name of the file and can include forward slashes to create the file in a subdirectory relative to the base directory. All files are written into the same base directory under the OS's temp dir.

In addition, one may optionally provide the stdin and command properties to provide stdin data to the running code and to run the code with a custom command. See examples below.

The output is a json object containing the properties stdout, stderr and error. stdout and stderr is captured from the output of the ran code. error is popuplated if there is a compiler / interpreter error. Note that the code-runner will exit with a non-zero code if invalid input is given or if the files cannot be written to disk (permissions, disk space, etc). No json will be written to stdout in those cases. Otherwise the exit code is 0.

{
  "language": "python",
  "files": [
    {
      "name": "main.py",
      "content": "print(42)"
    }
  ]
}

{
  "stdout": "42\n",
  "stderr": "",
  "error": ""
}

{
  "language": "python",
  "stdin": "42",
  "files": [
    {
      "name": "main.py",
      "content": "print(input('Number from stdin: '))"
    }
  ]
}

{
  "stdout": "Number from stdin: 42\n",
  "stderr": "",
  "error": ""
}

{
  "language": "bash",
  "command": "bash main.sh 42",
  "files": [
    {
      "name": "main.sh",
      "content": "echo Number from arg: $1"
    }
  ]
}

{
  "stdout": "Number from arg: 42\n",
  "stderr": "",
  "error": ""
}

{
  "language": "python",
  "stdin": "<parameters>",
  "fn_content": "<full code>",
  "class":"",
  "method":""
}

输出以下信息：

• 函数的返回值，native code，输出至控制台，rust解析后 {"type":"fn_ret","content":""}
• 统计数据：内存、运行时间 ，来自native code /rust {"type":"sta","content":"<json_string>"}
• 控制台输出，执行命令正常，控制台输出，需要过滤统计数据和函数返回
• 运行错误(stderr)，运行执行命令出错，控制台的输出
• 编译错误(error)，运行编译命令出错，控制台的输出

{
  "stdout": "42\n",
  "sta":"<time,mem>",
  "fn_result":"<result>",
  "stderr": "",
  "error": ""
}

Goal: Not only can support the running code, but also support online judges

Feature

• You can specify the method of running the function
• Convert stdin to function parameter
• Running data statistics, including running time
• The output result is the return value of the running function