Skip to main content

paque simplifies running simple workflows you want to run. It offers a few features of `make`, but removing most of its power

Project description

Paque ¯\_(ツ)_/¯

PyPI version

Paque simplifies running simple workflows you want to run. It offers a few features of make, but removing most of its power. It runs on a paquefile or paquefile.yaml (or just pass the name of the file). You can see a simple example in the root folder.

It supports Python 3.6.5+ (for no particular reason aside from being my default, it should work just fine on any relatively recent Python 3)

Installation

It should be enough to run

pip install paque

and then

paque taskname  # If you have a paquefile or paquefile.yaml file

Tab completion

This is a reminder to myself to implement tab completion (at lest for ZSH, since I have done that before)

Why?

I had a series of Dockerfile I wanted to build sequentially, and run some commands. It wasn't a right fit for docker-compose, so I wrote a makefile (here). The result was excellent, BUT if you want to add logging, or any kind of information you need to start resorting to "makefile hacks" (like passing variables down the dependency stack, or accessing subpaths of requirements...) that didn't feel right.

So, I decided to write this.

What?

You define your tasks in a YAML file with specific syntax, of the form:

taskname:
  - run: "what it runs"
  - message: "what it logs"
  - depends:
      - task_it_depends_1
      - task_it_depends_2
  - sleep: integer

otherwise:
  - run:
      - "You can run several commands"
      - "Passing them as an array"
  - message:
      - "Likewise for logging"
      - "Yes."

You can also use arguments, multiple arguments, and conditions

taskname:
  - run: "{something} {folder}"
  - message: "This does {something} on {folder}"
  - condition: "do-if-bash-says-this-is-0"

main:
  - depends:
      - taskname folder:/Users/foo/ something:rmdir

For now you can't have spaces in arguments. Sorry. Also, there is no way at the moment to pass arguments from the command line to tasks, this will be coming soon.

For usage, you would just

paque taskname

How?

YAML (following the rules above) is converted into a dictionary of task names and Tasks by a Parser. Then a simple depth-first-search planner finds an execution that satisfies all dependencies and transitive dependencies (with arguments) of taskname. Finally, the plan is passed to an executor that offloads it to the shell (or just logs it).

FAQ

Why YAML and not FOOML?

I find YAML pretty readable and writable, as long as you restrict what you can do. Since there is no nesting here, you can't shoot yourself in the foot with YAML. If you really can't stand YAML, you have two options

  • Use dhall and convert from dhall to YAML (recommended)
  • Write a parser for your favourite markup

Is this production ready?

Well… ¯\_(ツ)_/¯ I'm pretty sure there is an issue with argument replacement in a corner case, but I can't put my finger on which. For simple use cases, this should be safe. Since there is no branching, there is not much that can go wrong though.

What's with the name?

For one, it's python+make=pake, but it was taken (I should have checked before), so left it at paque. It's also a form of "pa' qué", an Spanish slang for "para qué". I.e. what for? ¯\_(ツ)_/¯

Contributing

I'm happy to receive PRs, so don't be shy. Also let me know if you used it, that could be fun. As you see from the root folder, you will need to use poetry

Future development

I will keep using it, so any bugs I find will be fixed. Likewise, I will keep improving it, although the current version is "almost enough". Currently on the "roadmap" I have:

  • Fix dry-run display of run when it is multiline
  • Generate a graphviz plot of the plan(s). This was one of the motivations to write my own thingy, after all
  • Better tests: I wrote the ones I have with a combination of TDD and "let's test and print". I want less tests of happy paths and more tests of corner cases
  • Custom exceptions, right now it's just "raise that"
  • Automatically convert paquefile.dhall into a YAML paquefile (this was supposed to be in this version but I got lazy)
  • Have a nicer CLI (probably using cleo) Moved to using click, after the great experience with motllo
  • Possibly, running tasks in parallel (this is a hard one given how the planner works, so probably won't)
  • Conditionals? Available as optional tasks. The condition is on what is run, assumes that the task has run if condition is false. So, a false condition does not stop execution of the rest of the plan
  • Fixing the bug that is likely there in argument substitution (note: I have been using paque quite regularly in other projects and I have still not hit it)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

paque-0.1.1.tar.gz (12.0 kB view details)

Uploaded Source

Built Distribution

paque-0.1.1-py3-none-any.whl (10.4 kB view details)

Uploaded Python 3

File details

Details for the file paque-0.1.1.tar.gz.

File metadata

  • Download URL: paque-0.1.1.tar.gz
  • Upload date:
  • Size: 12.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.0.5 CPython/3.7.4 Darwin/18.7.0

File hashes

Hashes for paque-0.1.1.tar.gz
Algorithm Hash digest
SHA256 9141e91f33217681099f0ab7e67c27f22a785ad191028cdeae220f82e96d91b8
MD5 890512d1596cb71411e104292be980ca
BLAKE2b-256 d4766f32a2477950ef32bb8a0015dd63218122a6d7834c9c906e6ab3cf0e5067

See more details on using hashes here.

File details

Details for the file paque-0.1.1-py3-none-any.whl.

File metadata

  • Download URL: paque-0.1.1-py3-none-any.whl
  • Upload date:
  • Size: 10.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: poetry/1.0.5 CPython/3.7.4 Darwin/18.7.0

File hashes

Hashes for paque-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 a63ed52b522663b2ce2ac48064af2036bed82d602ab901bb9a1b3fdb321310ce
MD5 f12a295d28c609aeff2a511a63d3c674
BLAKE2b-256 c4f6fee8d6e9028a10ff7e54154b8af01d47be25f1a6ece7532deee5a7feea63

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page