Are there extra steps to get publish working for docker images using amazon-ecr-credentials-helper ? https://github.com/awslabs/amazon-ecr-credential-helper#configuration

I'm getting an authentication error with

./pants publish

but I can push and pull manually with no problems

yes, you have to plumb through any env vars used by the helper to locate your AWS creds

one sec…

in our

pants.toml

we set:

[docker]
tools = ["docker-credential-ecr-login", "sh"]
env_vars = [
    "HOME",
    # AWS auth vars.
    "AWS_PROFILE",
    "AWS_DEFAULT_REGION",
    "AWS_ACCESS_KEY_ID",
    "AWS_SECRET_ACCESS_KEY",
    "AWS_SESSION_TOKEN",
    "AWS_ROLE_ARN",
    "AWS_WEB_IDENTITY_TOKEN_FILE",
]

oh awesome, thanks Dan

tools

seems to be an invalid option, might have to upgrade pants I guess

ah, yes. added in 2.11 I think(?)

I just ran into an interesting error when running the

package

goal after upgrading from

2.11.1

to

2.13.0

. I’ve searched through my config and build files, but I’m really not sure where this is coming from. No other code changes besides updating the version number in

pants.toml

.

NoSourceRootError: No source root found for `3rdparty`

Any suggestions how to troubleshoot this?

I am trying to run package goal for python_awslambda target which has dependency on img2pdf library and I am getting this error. running the cmd on macOS Big Sur M1.

Failed to resolve all requirements for abbreviated platform cp39-cp39-linux_x86_64 from 3rdparty/python/default.lock:

Configured with:
    build: False
    use_wheel: True

Dependency on img2pdf not satisfied, 1 incompatible candidate found:
1.) img2pdf 0.4.4 (via: img2pdf==0.4.4) does not have any compatible artifacts:
    <https://files.pythonhosted.org/packages/95/b5/f933f482a811fb9a7b3707f60e28f2925fed84726e5a6283ba07fdd54f49/img2pdf-0.4.4.tar.gz>

Hi guys, I’m new here. Just start with the pantsbuild. Might be a dump question, is that possible to delete a directory in pantsbuild? I can only find the

CreateDigest

toned down the code for reprod. using the latest version img2pdf==0.4.4 BUILD:

python_sources()

python_awslambda(
    name="img2pdf-lambda",
    runtime="python3.9",
    handler="./handler.py:lambda_handler",
)

handler.py

import img2pdf

def lambda_handler(event, context):
    print('hello world')

Where does this option come from?

```Configured with:

build: False

use_wheel: True```

To me it seems like since pants is not allowing the build and the wheels don’t exist, it fails But I didn’t know you can configure this. EDIT: This is from pants logs

Yes, this is from the pants log. And these configuration options are probably default cuz I am not configuring this explicitly.

What pants version is this? I have seen in past pants building wheels when it couldn’t find the correct wheel

2.12

[GLOBAL]
pants_version = "2.12.0"
backend_packages.add = [
  'pants.backend.python',
  "pants.backend.python.mixed_interpreter_constraints",
  "pants.backend.awslambda.python",
  'pants.backend.python.lint.docformatter',
  'pants.backend.python.lint.black',
  'pants.backend.python.lint.flake8',
  'pants.backend.python.lint.isort',
  'pants.backend.python.typecheck.mypy',
]

[anonymous-telemetry]
enabled = false

[source]
root_patterns = ["/", "projects"]

[python]
interpreter_constraints = ["CPython>=3.9"]
enable_resolves = true
lockfile_generator = "pex"

[python-infer]
string_imports = true

[pytest]
lockfile = "3rdparty/python/pytest.txt"
version = "pytest>=7.1.2,<7.2"
extra_requirements.add = [
  "pytest-django>=4.5.2,<5",
]
execution_slot_var = "PANTS_EXECUTION_SLOT" # (controlling (test) parallelism) If a non-empty string, the process execution slot id (an integer) will be exposed to tests under this environment variable name.
args = ["-s"]

Hi Folks I've noticed Javascript/typescript support is on the roadmap. I think there was some discussion about getting the linters and formatter support in first. Could someone please let me know how this is progressing and any rough idea when we can expect to see some releases 🙂 ? As always thanks for making this tool and helping out the community through this slack!! Greatly appreciated!

question regarding `experimental_shell_command`s: is it possible to pass in arguments when calling

./pants run

on a shell command? the docs recommend the following template:

experimental_shell_command(
    name="my-script",
    command="./my-script.sh --flag",
    dependencies=[":scripts"],
)

which i can the call with e.g.

./pants run src:my-script

. but the

command

here appears to be static, what if I wanted to pass in a variable?

./pants run src:my-script -- foo

doesn’t seem to work

a dump question sounds like the precursor to a shower thought 😂

There are no bad questions!

It sounds like you’re writing a plugin?

Yes! Plugins to generate something, but I need to cleanup that output directory before I start generating

That’s surprising, since typically you’d write into an output directory under the process sandbox, which is fresh every time and so should always be empty

This is not Pants configuring things evne though it reads like that might be the case:

Configured with:
    build: False
    use_wheel: True

That's really just Pex's reaction to pants passing

--platform linux_x86_64-cp-39-cp39

which is what it does to tell Pex - "Hey, I know we're on Mac, but try to build a venv that works on Linux". Since we're on Mac, we cannot build sdists into Linux wheels without cross-building. Pex does not support cross-building and so it only allows pre-built wheels. The resolve fails since there is no pre-built wheel for

img2pdf

. You need to build one for Linux and then make it available in a

--find-links

repo for Pex to find. When using Pants you configure this with `[python-repos] repos = ...`: https://www.pantsbuild.org/docs/reference-python-repos#repos

Lots of Toolchain engineers are currently working on bringing support for a form of cross-building by running Pants actions in local Docker containers. That will nicely solve cases like this, but that feature work though is still in progress.

We’re having a problem with pants where we have a library (let’s say it’s named

extras

) and within that library, a file imports another module from the same library plus some 3rd party dependencies like:

from extras.foo import bar
from pyspark.sql import functions as F

pylint

correctly identifies this as

wrong-import-order

since 3rd party libraries should come first, but if we manually change the import order,

isort

incorrectly identifies both as 3rd party libraries and puts

extras

ahead of

pyspark

. It seems like

isort

does not know when something is a local import.

I guess if we use output_directories in Process it can clear the directory automatically, but if we use output_files it will only write to those specific files?

Not quite.

output_directories

and

output_files

tell the Process where to gather outputs from when it finishes running and has to return outputs to the engine.