ENH: persisting tasks on a per-run basis

未关闭
#403 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

还没有人认领这个 Issue。

评估

难度
5/5
预计耗时
一周以上
新手友好度
35/100
Issue 类型
功能
描述清晰度
基本清楚
活跃度
停滞
技术栈
python
领域
tooling

调研方向

Start with the making tasks persist tutorial and the existing @pytask.mark.persist behavior. Trace how task freshness and database updates are determined, then define the per-run override options and their effects on later executions. Done means the runtime API is documented, preserves existing defaults, and has tests for rerunning, marking updated, and leaving a task stale.

由索引模型根据 Issue 内容生成。

描述

enhancement
Is your feature request related to a problem?

I'm taking about https://pytask-dev.readthedocs.io/en/stable/tutorials/making_tasks_persist.html

If I format my tasks.py file or update an unrelated task in there, then pytask will re-run the tasks. But I don't want this. I also don't want this to be a permanent feature of the task, as marked with a decorator.

Describe the solution you'd like

I want to be able to specify on a per-run basis whether or not to re-run a task if the source file changed. Ideally it should be something like

  • Keep the @pytask.mark.persist as marking default behavior: If its not present, the default is to re-run on src changes. If it is present, then the default is to NOT re-run on source changes. Therefore, we don't break the old API.
  • But, add a flag/API at runtime to override this default: On a per-task level be able to change the behavior. I the options should be "rerun", "don't rerun, and mark as updated", and "don't rerun, but don't mark as updated, so next time you run then the task will still appear stale". Maybe I'm not breaking down these options along the right axes, it could use more work.
API breaking implications

Per above, it shouldn't be breaking

Describe alternatives you've considered

Perhaps this warrants re-thinking the API for determining what is "up to date". Per those docs above:

Internally, the state of the dependencies, the source file, and the products are updated in the database such that the subsequent execution will skip the task successfully.

If we accumulate many more edge cases for when a task is up to date, things are going to get complicated. If there was a more official API when you define a task, eg @pytask.Task(func, ins, outs, *, should_run = None) and you could pass in (really brainstorming here)

  • None (keep default behavior)
  • True (rerun every time)
  • A Callable[[Not sure what the args should be], bool]
  • "on_dependency_change"
  • "on_source_change"
  • An Iterable of any of the above. If any are true, then rerun.
主要语言
Python
星标
146
派生
14
平均合并
3 天 16 小时
30 天内合并 PR
34

贡献指南

这个仓库没有索引到贡献指南

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

pytask-dev/pytask 的其他 Issue

查看 pytask-dev/pytask 的全部 Issue

相似的 Issue

更多 Python Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。