Pants v2: Fast, consistent builds for Python and more

Welcome to the Pants v2 documentation hub!

Pants v2 is a fast, scalable build system for growing codebases. It's currently focused on Python, with support for other languages coming soon.

Here you'll find guides to help you get started with Pants v2, comprehensive documentation on how to configure, run and customize Pants v2, and information on how to get help from the Pants community.

Get Started

Resources and archives

How to include resources such as images and config files in your project.

There are two ways to include resource files in your project: resources targets and files targets.

resources

A resources target is for files that are members of code packages, and are loaded via language-specific mechanisms, such as Python's pkgutil.get_data() or Java's getResource().

Pants will make resources available on the appropriate runtime path, such as Python's PYTHONPATH or the JVM classpath. Resources can be loaded directly from a binary in which they are embedded, such as a Pex file, without first unpacking it.

For example, to load resources in Python:

import pkgutil

if __name__ == "__main__":
    config = pkgutil.get_data("project", "config.json").decode("utf-8")
    print(f"Config: {config}")
python_library(
  # Pants cannot infer this dependency, so we explicitly add it.
  dependencies=[":config"],
)

resources(
  name="config",
  sources=["config.json"],
)
{"k1": "v", "k2": "v"}

Source root stripping applies to resources, just as it does for code. In the example above, Python loads the resource named project/config, rather than src/python/project/config.json.

files

A files target is for loose files that are copied into the chroot where Pants runs your code. You can then load these files through direct mechanisms like Python's open() or Java's FileInputStream. The files are not associated with a code package, and must be extracted out of a deployed archive file before they can be loaded.

For example, to load loose files in Python:

def test_open_file():
    with open("src/python/project/config.json") as f:
        content = f.read().decode()
    assert content == '{"k1": "v", "k2": "v"}'
python_tests(
  # Pants cannot infer this dependency, so we explicitly add it.
  dependencies=[":config"],
)

files(
  name="config",
  sources=["config.json"],
)
{"k1": "v", "k2": "v"}

Note that we open the file with its full path, including the src/python prefix.

🚧

files are not included with binaries like pex_binary

Pants will not include dependencies on files targets when creating binaries like pex_binary and python_awslambda via ./pants package. Filesystem APIs like Python's open() are relative to the current working directory, and they would try to read the files from where the binary is executed, rather than reading from the binary itself.

Instead, use resources targets or an archive target.

When to use resources

Use resources for files that are associated with (and typically live alongside) the code that loads them. That code's target (e.g. the python_library) should depend on the resources target, ensuring that code and data together are embedded directly in a binary package, such as a wheel, Pex file or AWS Lambda.

When to use files

Use files for files that aren't tightly coupled to any specific code, but need to be deployed alongside a binary, such as images served by a web server.

When writing tests, it is also often more convenient to open a file than to load a resource.

resources

files

Runtime path

Relative to source root

Relative to repo root

Loading mechanism

Language's package loader, relative to package

Language's file loading idioms, relative to repo root

Use with

Targets that produce binaries, such as pex_binary, python_distribution, python_awslambda.

archive targets, tests

relocated_files

When you use a files target, Pants will preserve the path to the files, relative to your build root. For example, the file src/assets/logo.png in your repo would be under this same path in the runtime chroot.

However, you may want to change the path to something else. For example, when creating an archive target and setting the files field, you might want those files to be placed at a different path in the archive; rather than src/assets/logo.png, for example, you might want the file to be at imgs/logo.png.

You can use the relocated_files target to change the path used at runtime for the files. Your other targets can then add this target to their dependencies field, rather than using the original files target:

# Original files target.
files(
  name="logo",
  sources=["logo.png"],
)

# At runtime, the file will be `imgs/logo.png`.
relocated_files(
  name="relocated_logo",
  files_targets=[":logo"],
  src="src/assets",
  dest="imgs",
)

You can use an empty string in the src to add to an existing prefix and an empty string in the dest to strip an existing prefix.

If you want multiple different re-mappings for the same original files, you can define multiple relocated_files targets.

The relocated_files target only accepts files targets in its files_targets field. To relocate where other targets like resources and python_library show up at runtime, you need to change where that code is located in your repository.

archive: create a zip or tar file

Running ./pants package on an archive target will create a zip or tar file with built packages and/or loose files included. This is often useful when you want to create a binary and bundle it with some loose config files.

For example:

archive(
    name="app_with_config",
    packages=[":app"],
    files=[":production_config"],
    format="tar.xz",
)

The format can be zip, tar, tar.xz, tar.gz, or tar.bz2.

The packages field is a list of targets that can be built using ./pants package, such as pex_binary, python_awslambda, and even other archive targets. Pants will build the packages as if you had run ./pants package. It will include the results in your archive using the same name they would normally have, but without the dist/ prefix.

The files field is a list of files and/or relocated_files targets. See resources for more details.

You can optionally set the field output_path to change the generated archive's name.

Updated about a month ago


Resources and archives


How to include resources such as images and config files in your project.

Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.