Skip to content

add Dependency.set_extras()#373

Closed
dimbleby wants to merge 1 commit intopython-poetry:mainfrom
dimbleby:set-extras
Closed

add Dependency.set_extras()#373
dimbleby wants to merge 1 commit intopython-poetry:mainfrom
dimbleby:set-extras

Conversation

@dimbleby
Copy link
Copy Markdown
Contributor

Will allow tidying of python-poetry/poetry#5688, which currently is manipulating _extras directly.

Again the split into multiple repositories makes it complicated to figure out a merge sequence that will work...

@sonarqubecloud
Copy link
Copy Markdown

Kudos, SonarCloud Quality Gate passed!    Quality Gate passed

Bug A 0 Bugs
Vulnerability A 0 Vulnerabilities
Security Hotspot A 0 Security Hotspots
Code Smell A 0 Code Smells

No Coverage information No Coverage information
No Duplication information No Duplication information

def extras(self) -> frozenset[str]:
return self._extras

def set_extras(self, extras: Iterable[str]) -> None:
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I wonder if it would be better to follow the with_extras() pattern similar to package.with_features(). The whole extras / features thing is a bit confusing at present.

Because you are doing this anyway, I would suggest you roll in the removal of the Dependency._extra attribute in favour of the PacakgeSpecification._feature change from https://github.com/python-poetry/poetry-core/pull/370/files#diff-93744eaedc27b829eefbde40b5d45732da68e0716af35c30bf010e66ea72b1f7R84-R245 here too.

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I wondered about with_extras(), but then we ought to do with_constraint() too and it all seemed like a lot of trouble for no very clear benefit.

#370 looks a bit hairy, happy to wait for the dust to settle on it before pushing forward on this one but I don't think I want to get into doing both at once...

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should really be careful that we don't tear down right away what will be cleaned up in #370. If we decide to use an attribute in the hash, it should not be changed anywhere. Thus, the with_attribute() pattern is required.

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The attribute change can be cleanly split out of #370. I can do it in another PR if you do not want to add it to this one, easy enough.

We can handle with_constraint() implementation in another PR as the need arises or once we settle #370.

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You say required, but we already don't follow that in set_constraint(), and the single place that calls that (and would call this new setter) is safe.

So I only sort of agree...

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Re: attribute removal #375

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

"Required" for clean code I assume.

That we don't follow the pattern somewhere does not mean that this is alright. Of course, we could allow setting attributes that are used in hash and evaluate each place that sets the attribute. But you don't have to be a prophet to see that eventually someone will forget that it's dangerous to set this attribute... Thus, in my opinion it may be better not to provide a setter but just a with_attribute() (or don't use the attribute in __hash__()).

dimbleby added a commit to dimbleby/poetry that referenced this pull request May 25, 2022
@dimbleby
Copy link
Copy Markdown
Contributor Author

Actually I might be changing my mind about the best way to fix this...

The current approach is vulnerable to foo[bar] being required on Windows but only foo elsewhere. Then the exporter would show foo[bar] everywhere.

I previously rejected a simpler fix that resulted in an export like:

localstack-ext==1.0.0 ; python_version >= "3.6" and python_version < "4.0"
localstack-ext[bar]==1.0.0 ; python_version >= "3.6" and python_version < "4.0"

but actually while that's cosmetically awkward, it seems like it should give the right answers and be more reliably right.

Rework coming, and in that case #373 will be redundant anyway

@abn
Copy link
Copy Markdown
Member

abn commented May 25, 2022

That resembles the original fix I was playing with that led me down the hash rabbit hole. But yeah I think that is more accurate anyway. It is upto whoevever consumes the export to decide how to handle it I reckon.

@dimbleby dimbleby closed this May 25, 2022
@dimbleby dimbleby deleted the set-extras branch May 25, 2022 20:53
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants