Release 2.0

[samer-hamood]

This PR includes multiple improvements and a new function, with others included in separate PRs linked below (for ease of review). There will be a small change in the API, referenced in the CHANGELOG.md, which warrants the release to increment its major version to 2.

The core of this PR is:

• Renamed head_option and last_option to head_not_none and last_not_none respectively,
• Added first_not_none to match head_not_none
• Added parametrize for parameterized tests
• Fixed tests not running from anywhere in the project or via Pycharm (or other IDEs)
• Added run-test.sh for running checks locally
• Added support for Python 3.12 and 3.13 by using ruff instead of Pylint due to issues for those Python versions
• Upgraded pre-commit-hooks and ruff-pre-commit repo revisions
• Reference to repurposing of two functions, head_option/first_option and last_option, in CHANGELOG.md
• Improved language, correcting any consistency, grammar and punctuation in some of the docs
• Linked references readme
• Upgraded Poetry to 2.0.0
• Replaced deprecated tool.poetry.dev-dependencies
• Created partial function in ParallelExecutionEngine only once per instance and added symmetrical code from it to ExecutionEngine (minor refactoring)
• Fixed linter errors/warnings

Functions this release

• plus - add single or iterable to sequence
• zip_with_next - zip adjacent elements
• head_option/first_option, last_option - return head/first and last element wrapped in an Option to handle None

Boolean functions

• any and all taking functions to be equivalent to exists and for_all respectively
• none - return true if no element satisfies the given predicate function or all elements are falsy if no function is supplied

Indexed functions

• map_indexed - map with an index argument
• filter_indexed - filter with an index argument
• drop_at - drop by index/indices

Not-none functions

• filter_not_none - filter excluding None values
• map_not_none - map excluding None values

[samer-hamood]

Pylint seems to have issues with Python versions 3.12 and 3.13 - can add another PR for Ruff to more closely match Pylint's rules

[EntilZha]

Hi, thanks for the PR!

The main thing I'll mention is that I definitely have a strong preference for 1 Change -> 1 PR instead of many changes -> 1 big PR. Could you break things down?

As with prior PRs, for changes that make not functional changes and are strict improvements, I tend to merge those pretty quickly. Examples in your PR probably include things like improving tests, CI/tooling, fixing lint errors, docs, etc.

This makes reviewing code easier/faster, makes landing "easy" changes faster, and get more changes in, in the case I don't want a particular part of a huge PR. In this specific case as well, should there be a need to have a 2.0, it will also allow all the changes not needed for that land first and thus be available <2.0.

For functional changes, I usually like to see a good explanation for why they are needed. For example, the renames for head_option and last_option seem unnecessary, but I'm open to discussion. I can see the motivation for the name, but the naming comes from the more general idea of option types and isn't python specific (and one original motivation for the names is consistency with scala/similar libraries with those kinds of semantics). This is another reason to break the big PR down, even if I don't want to change the name, its not blocking any of the other changes that are ready to merge.

For new functions, these should also be in their own self contained PR, along with at least some discussion of why they are needed, particularly given the ability to register custom functions.

Again thanks for the PR, it would just help me a lot in getting changes in if PRs are broken into 1 semantic change -> 1 PR.

[samer-hamood]

I'll break up the PR.

For new functions, these should also be in their own self contained PR, along with at least some discussion of why they are needed, particularly given the ability to register custom functions.

Could you point to where that (register custom functions) is in the docs or explain how that's done?

For functional changes, I usually like to see a good explanation for why they are needed. For example, the renames for head_option and last_option seem unnecessary, but I'm open to discussion. I can see the motivation for the name, but the naming comes from the more general idea of option types and isn't python specific (and one original motivation for the names is consistency with scala/similar libraries with those kinds of semantics). This is another reason to break the big PR down, even if I don't want to change the name, its not blocking any of the other changes that are ready to merge.

I'll describe the reason for the name change in a separate PR.

[EntilZha]

@samer-hamood I misremembered and its called extend, which was merged here https://github.com/EntilZha/PyFunctional/pull/144/files

FWIW, the feature isn't documented outside of the docstring, so that would be a pretty easy improvement to make. My general thinking around adding functions is that if its something that would have somewhat general use, it might be worth adding to the library, but if its relatively niche, then extend is a good way to go.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.