taskcapsule 0.2.0

A simple task runner for Python

taskcapsule

A simple framework for running commands or Python functions concrrently using templated commands/functions and dictionaries of arguments.

Warning

This has sharp edges which can only be smoothed as errors are discovered. Expect better guardrails to occur over time as issues are discovered.

CLI/Subprocess Warning

Shells and environments behave in all sorts of ways. When in doubt, use absolute paths for the items that you're calling

Features

• Support for CLI subshells and Python callables
• Output filtering, which is the functional equivilent of grep
• Minimum worker spawning. If you have fewer tasks than defined workers, the worker pool will be reduced to the number of tasks
• Task validation to ensure that either function or command is set

Future Work

• Validate function calls and arguments via inspect
• Tests
• Result gathering
• Pipelines

Examples

Subshell and Callable Example

# from examples/cli_or_callable.py
from taskcapsule import Task, TaskRunner


def hello(message: str):
    return f"Hello, {message}!"


my_tasks = []
my_tasks.append(
    Task(
        command="echo {message}",
        kwargs={"message": "Hello, CLI runner!"},
        output_filter="Hello",
    )
)
my_tasks.append(
    Task(
        function=hello,
        kwargs={"message": "callable runner!"},
        output_filter="Hello",
    )
)

runner = TaskRunner(
    tasks=my_tasks,
    workers=4,
)
runner.run()

$ python cli_or_callable.py |jq
INFO:task-runner:spawning 2 workers
{
  "function": "hello",
  "kwargs": {
    "message": "Hello, callable runner!"
  },
  "return_code": 0,
  "output": "Hello, callable runner!!",
  "error": "",
  "duration": 9.5367431640625E-7,
  "success_filter": "Hello"
}
{
  "command": "echo Hello, CLI runner!",
  "kwargs": {
    "message": "Hello, CLI runner!"
  },
  "return_code": 0,
  "output": "Hello, CLI runner!\n",
  "error": "",
  "duration": 0.016596078872680664,
  "success_filter": "Hello"
}
INFO:task-runner:all tasks completed

Nmap Example

from taskcapsule import TaskRunner, Task
# Look to see if the host is running WebLogic T3
template = "nmap -oG - -p {port} --script weblogic-t3-info {addr}"
items = [{"addr":"1.2.3.4","port": "7002"}]
my_tasks = []
for i in items:
    kwargs = items
    my_tasks.append(
        Task(
            command=template,
            kwargs=kwargs,
            target_metadata={"uuid": "fedcba09-1234-1111-bcde-1234567890fe"},
            # This is in the output if, and only if, T3 is running
            output_filter="T3",
        )
    )

tr = TaskRunner(tasks=my_tasks)
tr.run()

Output

python example.py|jq

INFO:taskcapsule:spawning 1 workers
{
  "command": "nmap -oG - -p 7002 --script weblogic-t3-info 1.2.3.4",
  "kwargs": {
    "addr": "1.2.3.4",
    "port": "7002"
  },
  "return_code": 0,
  "stdout": "# Nmap 7.95 scan initiated Thu Apr 17 09:23:50 2025 as: nmap -oG - -p 7002 --script weblogic-t3-info 1.2.3.4\nHost: 1.2.3.4 ()\tStatus: Up\nHost: 45.60.186.97 ()\tPorts: 7002/open/tcp//afs3-prserver//WebLogic application server 12.2.1.4 (T3 enabled)/\n# Nmap done at Thu Apr 17 09:23:51 2025 -- 1 IP address (1 host up) scanned in 0.70 seconds\n",
  "stderr": "",
  "duration": 0.7490861415863037,
  "success_filter": "T3",
  "target_metadata": {
    "uuid": "fedcba09-1234-1111-bcde-1234567890fe"
  }
}
INFO:task-runner:all tasks completed

Error Handling

Code

from taskcapsule import Task, TaskRunner


def hello(message: str):
    return f"Hello, {message}!"


my_tasks = []
my_tasks.append(
    Task(
        command="nosuchcommand {message}",
        kwargs={"message": "Hello, CLI runner!"},
    )
)
my_tasks.append(
    Task(
        function=hello,
        kwargs={"wrong_arg": "callable runner!"},
    )
)
runner = TaskRunner(
    tasks=my_tasks,
    workers=4,
    show_failures=True,
)
runner.run()

Output

$ python show_errors.py|jq
INFO:task-runner:spawning 2 workers
ERROR:task-runner:processing command=None function=<function hello at 0x104ec4220> kwargs={'wrong_arg': 'callable runner!'} output_filter=None target_metadata=None: hello() got an unexpected keyword argument 'wrong_arg'
INFO:task-runner:success filter is None, skipping check
{
  "function": "hello",
  "kwargs": {
    "wrong_arg": "callable runner!"
  },
  "return_code": -1,
  "output": "",
  "error": "hello() got an unexpected keyword argument 'wrong_arg'",
  "duration": 0.00011897087097167969
}
INFO:task-runner:success filter is None, skipping check
{
  "command": "nosuchcommand Hello, CLI runner!",
  "kwargs": {
    "message": "Hello, CLI runner!"
  },
  "return_code": 127,
  "output": "",
  "error": "/bin/sh: nosuchcommand: command not found\n",
  "duration": 0.021525859832763672
}