Pyproject.toml support for DockerSettings #3292

schustmi · 2025-01-17T15:59:01Z

Describe changes

This PR adds the following functionality:

support for pyproject.toml files to be specified for installing requirements in Docker images.
ZenML will extract the requirements from the pyproject.toml using one of the default or a custom user-provided command.
install packages from a requirements.txt or pyproject.toml file in the source root if no requirements or pyproject file was specified in the DockerSettings.
an attribute to disable the above automatic installation.
a way to specify a command that installs a local python package.

Pre-requisites

Please ensure you have done the following:

I have read the CONTRIBUTING.md document.
I have added tests to cover my changes.
I have based my new branch on develop and the open PR is targeting develop. If your branch wasn't based on develop read Contribution guide on rebasing branch to develop.
IMPORTANT: I made sure that my changes are reflected properly in the following resources:
- ZenML Docs
- Dashboard: Needs to be communicated to the frontend team.
- Templates: Might need adjustments (that are not reflected in the template tests) in case of non-breaking changes and deprecations.
- Projects: Depending on the version dependencies, different projects might get affected.

Types of changes

Bug fix (non-breaking change which fixes an issue)
New feature (non-breaking change which adds functionality)
Breaking change (fix or feature that would cause existing functionality to change)
Other (add details above)

coderabbitai · 2025-01-17T15:59:08Z

Important

Review skipped

Auto reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Generate unit testing code for this file.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
- @coderabbitai generate unit testing code for this file.
- @coderabbitai modularize this function.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read src/utils.ts and generate unit testing code.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
- @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai full review to do a full review from scratch and review all the files again.
@coderabbitai summary to regenerate the summary of the PR.
@coderabbitai generate docstrings to generate docstrings for this PR.
@coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai configuration to show the current CodeRabbit configuration for the repository.
@coderabbitai help to get help.

Other keywords and placeholders

Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

schustmi · 2025-01-20T13:43:03Z

@htahir1 Any ideas for the general docs structure of this? Or do we just leave it as-is for now?

htahir1 · 2025-01-20T14:19:18Z

So I think the docs need to change in these places too:

In production guide on this page

Now that we have our pipeline up and running in the cloud, you might be wondering how ZenML figured out what sort of dependencies to install in the Docker image that we just ran on the VM. The answer lies in the runner script we executed (i.e. run.py), in particular, these lines:

Probably i'd include the default behavior of trying to find requirements.txt and pyproject.toml implicitly here in the top page: https://docs.zenml.io/how-to/customize-docker-builds
Probably makes sense to add the implicit behavior here as we mention a lot about allow_download_from_code_repository and allow_download_from_artifact_store? https://docs.zenml.io/how-to/customize-docker-builds/which-files-are-built-into-the-image
Maybe we need to add it somewhere herE? https://docs.zenml.io/how-to/pipeline-development/configure-python-environments/handling-dependencies#use-a-tool-like-pip-compile-for-reproducibility Not sure exactly
Finally, I would include it here as a note that pyproject.toml behavior is not recommended but if you include it this is what happens: https://docs.zenml.io/how-to/project-setup-and-management/setting-up-a-project-repository/set-up-repository

bcdurak

Leaving some initial comments.

bcdurak · 2025-01-30T12:46:05Z

src/zenml/utils/pipeline_docker_image_builder.py

@@ -68,6 +68,12 @@
 }
 UV_DEFAULT_ARGS = {"no-cache-dir": None}

+# TODO: these don't actually install any extras. Should we include all extras? No extras? Exclude dev extras?


Just a reminder for the TODO.

Is this still relevant for this PR?

bcdurak · 2025-01-30T13:14:38Z

src/zenml/config/docker_settings.py

    - The packages specified via the `required_integrations`
+    - The packages defined inside a pyproject.toml file given by the


I think there is a misalignment with the docs here. In the docs, we say the priority is as follows:

Packages in the local env

Packages for the stack

Required Integrations

Requirements (could be a list or a path to a requirements.txt)

Pyproject.toml next

In the docstring here, 4 and 5 are switched.

I see that the docstring here is correct in the code and the docs need to be updated.

You're right, I fixed the error in the docs.

In general, there are two scenarios:

You don't explicitly specify any requirements using docker settings: In this case we first look for a requirements.txt (-> highest priority), and if it doesn't exist we look for a pyproject.toml

You do explicitly specify some requirements (using e.g. DockerSettings.required_integrations or DockerSettings.requirements): In this case the order in the docstring is the one that is being used. Requirements being last means it is the highest priority (it will override values from all previous steps), which is in line with the first case.

Do you think these make sense?

I left a comment about the confusion that I had with the interaction between the implicit and explicit definition of requirements but outside of that, the explanation looks good here.

src/zenml/utils/pipeline_docker_image_builder.py

bcdurak · 2025-01-30T13:24:04Z

src/zenml/utils/pipeline_docker_image_builder.py

+        if not any(
+            [
+                docker_settings.replicate_local_python_environment,
+                docker_settings.required_integrations,


Does this mean that adding an extra integration (or a package) to be installed disables the discovery of the implicit requirements? So, if I was not using any of these settings before and now want to add sklearn to my docker settings, suddenly the requirements.txt will not be detected. I am not sure if this is the optimal behavior here.

Correct. I'm not super convinced by it myself to be honest, but what would be a better alternative in your opinion?

Yeah, I guess I do not have a super clear alternative either.

The reason why I was a bit hesitant here was this. We have a default behavior that decides whether we look for these files or not. By adding an additional integration, (that I intuitively thought would not change this behavior), I change the way that it looks for these files. I guess the replicate_local_python_environment flag at least allows people to control the behavior. Perhaps we can write a sentence about this in the docs?

bcdurak · 2025-01-30T13:48:28Z

src/zenml/config/docker_settings.py

    ] = Field(default=None, union_mode="left_to_right")
+    pyproject_path: Optional[str] = None
+    pyproject_export_command: Optional[List[str]] = None


As far as I can see, it is possible to set this as follows: DockerSettings(pyproject_export_command=[......]) without an explicit pyproject_path as long as there is a pyproject.toml file to work with, in the current source root. But as soon as someone puts an integration, requirement or the replicate flag, this becomes unused.

Perhaps a validator can be helpful to detect this early.

What would you validate here? I don't see the problem of this being unused if there is no pyproject.toml to be honest

I would validate the following two scenarios:

If pyproject_export_command is set, there needs to be a pyproject.toml in the directory.

If required_integrations are set, pyproject_export_command is unused.

However, both are not so critical, I just thought early detection might help us improve the UX a bit.

github-actions · 2025-01-31T12:22:14Z

🔍 Broken Links Report

Summary

📁 Files with broken links: 1
🔗 Total broken links: 1
📄 Broken markdown links: 1
🖼️ Broken image links: 0
⚠️ Broken reference placeholders: 0

Details

File	Link Type	Link Text	Broken Path
`containerization/containerization.md`	📄	"source root"	`./which-files-are-built-into-the-image.md`

📂 Full file paths

/home/runner/work/zenml/zenml/scripts/../docs/book/how-to/containerization/containerization.md

avishniakov · 2025-03-28T07:47:52Z

@michael-zenml @htahir1 May I add my 5 cents here? Does adding pyproject.toml support mean that it will also rely on lock file or it will lock each time? In our setting it will be crucial to have it bounded to locks, since normally locks are also committed for consistency and reproducibility.
Aside from that - great addition to have in ZenML!

P.S. Just noticed that it is a long runner already - I would love to see it priotorised, as it will ease the use of Dockers for our case, without the need to rely on external build process.

avishniakov · 2025-03-28T08:32:33Z

A small companion PR to that Poetry activity: #3470

github-actions · 2025-05-06T12:20:13Z

Documentation Link Check Results

❌ Absolute links check failed
There are broken absolute links in the documentation. See workflow logs for details
❌ Relative links check failed
There are broken relative links in the documentation. See workflow logs for details
_{Last checked: 2025-05-09 14:42:52 UTC}

schustmi · 2025-05-07T11:35:57Z

@avishniakov Users will have full control to specify a command that will be run to extract those requirements from the pyproject.toml. This will depend on what package manager you're using, for example in the case of uv, you could specify uv export --frozen which exports the requirements from the current lock file. Does that clarify, or do you think we need additional changes in this PR to make it work for your case?

avishniakov · 2025-05-07T11:46:15Z

@avishniakov Users will have full control to specify a command that will be run to extract those requirements from the pyproject.toml. This will depend on what package manager you're using, for example in the case of uv, you could specify uv export --frozen which exports the requirements from the current lock file. Does that clarify, or do you think we need additional changes in this PR to make it work for your case?

Yeah, if we can do something like poetry export --without-hashes --format=requirements.txt to specify - just good.

schustmi · 2025-05-07T12:01:25Z

@avishniakov Yep, we will use exactly this command as one of the default values even. I wasn't sure however if poetry will reuse existing lockfiles or if it will regenerate it.

AlexejPenner

🦞

AlexejPenner · 2025-05-08T15:06:06Z

docs/book/how-to/containerization/containerization.md

+        ...
+    ```
+
+    This will run `pip freeze` to get a list of your local packages, and allows you to easily replicate your local Python environment in the Docker container, ensuring that your pipeline runs with the same dependencies.


If this does not pick up locally installed libraries we should say so here

AlexejPenner · 2025-05-08T15:14:17Z

src/zenml/config/docker_settings.py

+                "Files must be included in the Docker image when trying to "
+                "install a local python package."


Suggested change

"Files must be included in the Docker image when trying to "

"install a local python package."

"Files must be included in the Docker image when trying to "

"install a local python package. You can do so by using the `allow_including_files_in_images` attribute of your docker settings: `DockerSettings(..., allow_including_files_in_images=True)`"

Lets save the user an additional odyssey to find out how to do this.

Isn't this literally in 100 code snippets on the same docs page?

schustmi added 2 commits January 17, 2025 16:51

WIP pyproject toml support

c8aef90

Docstrings

416043e

schustmi changed the title ~~pyproject.toml for DockerSettings~~ Pyproject.toml support for DockerSettings Jan 17, 2025

github-actions bot added internal To filter out internal PRs and issues enhancement New feature or request labels Jan 17, 2025

schustmi added 4 commits January 17, 2025 17:31

Some docs

9b55b10

Tests

f922209

More tests

080fb97

Add link

ac9b879

schustmi requested a review from stefannica January 20, 2025 13:43

schustmi marked this pull request as ready for review January 20, 2025 13:43

schustmi requested a review from bcdurak January 28, 2025 08:02

bcdurak reviewed Jan 30, 2025

View reviewed changes

schustmi added 2 commits January 31, 2025 13:14

Merge branch 'develop' into feature/PRD-689-pyproject-toml

71c2c35

Fix install order mistake in docs

3467f1b

schustmi requested a review from bcdurak February 10, 2025 08:09

Merge branch 'develop' into feature/PRD-689-pyproject-toml

9b0ebe9

Add support for installing local projects

d636553

schustmi added 2 commits May 7, 2025 14:10

Merge branch 'develop' into feature/PRD-689-pyproject-toml

19724ba

Add option to disable automatic requirements detection

e943bea

schustmi added 2 commits May 8, 2025 11:27

Fix some more conditions

8b98d22

Local project installation docs

f15e783

AlexejPenner reviewed May 8, 2025

View reviewed changes

Merge branch 'develop' into feature/PRD-689-pyproject-toml

edb9660

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Pyproject.toml support for DockerSettings #3292

Pyproject.toml support for DockerSettings #3292

schustmi commented Jan 17, 2025 •

edited

Loading

coderabbitai bot commented Jan 17, 2025 •

edited

Loading

Review skipped

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

schustmi commented Jan 20, 2025

htahir1 commented Jan 20, 2025

bcdurak left a comment

bcdurak Jan 30, 2025

bcdurak Feb 13, 2025

bcdurak Jan 30, 2025

bcdurak Jan 30, 2025 •

edited

Loading

schustmi Jan 31, 2025

bcdurak Feb 13, 2025

bcdurak Jan 30, 2025

schustmi Jan 31, 2025

bcdurak Feb 13, 2025

bcdurak Jan 30, 2025

bcdurak Jan 30, 2025

schustmi Jan 31, 2025

bcdurak Feb 13, 2025

github-actions bot commented Jan 31, 2025 •

edited

Loading

avishniakov commented Mar 28, 2025 •

edited

Loading

avishniakov commented Mar 28, 2025

github-actions bot commented May 6, 2025 •

edited

Loading

schustmi commented May 7, 2025

avishniakov commented May 7, 2025

schustmi commented May 7, 2025

AlexejPenner left a comment

AlexejPenner May 8, 2025

AlexejPenner May 8, 2025

schustmi May 8, 2025

		- The packages specified via the `required_integrations`
		- The packages defined inside a pyproject.toml file given by the

		"Files must be included in the Docker image when trying to "
		"install a local python package."

Pyproject.toml support for DockerSettings #3292

Are you sure you want to change the base?

Pyproject.toml support for DockerSettings #3292

Conversation

schustmi commented Jan 17, 2025 • edited Loading

Describe changes

Pre-requisites

Types of changes

coderabbitai bot commented Jan 17, 2025 • edited Loading

Review skipped

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

schustmi commented Jan 20, 2025

htahir1 commented Jan 20, 2025

bcdurak left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bcdurak Jan 30, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

github-actions bot commented Jan 31, 2025 • edited Loading

🔍 Broken Links Report

Summary

Details

avishniakov commented Mar 28, 2025 • edited Loading

avishniakov commented Mar 28, 2025

github-actions bot commented May 6, 2025 • edited Loading

Documentation Link Check Results

schustmi commented May 7, 2025

avishniakov commented May 7, 2025

schustmi commented May 7, 2025

AlexejPenner left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

schustmi commented Jan 17, 2025 •

edited

Loading

coderabbitai bot commented Jan 17, 2025 •

edited

Loading

bcdurak Jan 30, 2025 •

edited

Loading

github-actions bot commented Jan 31, 2025 •

edited

Loading

avishniakov commented Mar 28, 2025 •

edited

Loading

github-actions bot commented May 6, 2025 •

edited

Loading