This code runs fine, but pylint thinks I am working with an arbitrary generator, not a Fluent. Is there a return annotation somewhere that can be changed to Fluent?

P.S. This is a nice project.

Previously in in this MR

@olirice mentioned that

Any chance you'd be willing to take a look at Fluent.collect? it currently returns a Collection[T] but I'd love to have it pick up the container type from the container_type argument and return List[T] or Set[T] depending on the callable that gets passed. So far I haven't been able to figure that out.

I messed around, and i think unfortunately its not totally possible without support for higher-kind types.

See this example here and here. Both would require users to install a custom mypy plugin. Which probably goes against the philosophy of this being a light weight library :)
Guido did reply to the issue created on mypy so there is some hope of it coming one day to standard myppy

We could instead support some standard library types with the overloads operator, such as Tuple, List, Set. That should be nice enough for most people.

But we will lose the type information of the Collection type for the other non-defined types

Hi @olirice very cool project. I've been looking for something like this for some time! Transforming collections in this way wasn't a pleasure for me at all

sum(map(lambda, filter(lambda, list)))

I would be glad to see one more function that would be useful - foldLeft or foldRight or reduce as in scala.

Ex.

from flupy import flu
class Collector:
    __init__(self):
        self.nums = []

collector = flu([1,2,3]). \
                 filter(lambda x: x>1). \
                 foldLeft(Collector(), lambda x, acc: acc.nums.append(x))  

assert collector.nums == [2,3]

Btw I think that there is a small error in the documentation https://flupy.readthedocs.io/en/latest/api.html#flupy.Fluent.group_by - Shouldn't there be 2 and 3 instead of 4 and 9?

Ran into this issue while in the process of adding flupy (and alembic-utils) to conda-forge here: conda-forge/staged-recipes#26023.

Seems to be due to this bit here:

Traceback:

>flu --help 
Traceback (most recent call last):
  File "C:\bld\flupy_1712887240528\_test_env\Scripts\flu-script.py", line 5, in <module>
    from flupy.cli.cli import main
  File "C:\bld\flupy_1712887240528\_test_env\Lib\site-packages\flupy\cli\cli.py", line 4, in <module>
    from signal import SIG_DFL, SIGPIPE, signal
ImportError: cannot import name 'SIGPIPE' from 'signal' (C:\bld\flupy_1712887240528\_test_env\Lib\signal.py). Did you mean: 'SIGFPE'?

You know, because lazy pipelines provide the fundamental paradigm around which shell programming was designed, decades before Python--let alone flupy--existed.

Not a bit deal, of course. I just had to chuckle when I read that line. 🤷‍♂️

pip install flupy in Python 3.10 causes output like this below. Is this as simple as adding 3.10 to setup.py, or is there some nasty bug which has prevented you from doing that?

(venv) PS ...> pip install flupy
Collecting flupy
Using cached flupy-1.1.8.tar.gz (12 kB)
Installing build dependencies ... done
Getting requirements to build wheel ... error
ERROR: Command errored out with exit status 1:
command: ...\python.exe' '...\venv\lib\site-packages\pip_vendor\pep517\in_process_in_process.py' get_requires_for_build_wheel '...\Temp\tmpaqtyz1ok'
cwd: ...\Temp\pip-install-opd7zhl6\flupy_7a930ec7e41d487391822b70fd7bef1e
Complete output (1 lines):
Python 3.6+ is required.

WARNING: Discarding https://files.pythonhosted.org/packages/b9/28/34591f80faa91a517650a1537f6faca99308dafd49de5e4edf400cadc777/flupy-1.1.8.tar.gz#sha256=6be9b46f737053d4058414477220dc35339fc71a2a7ea53a6498b5add4642e8f (from https://pypi.org/simple/flupy/). Command errored out with exit status 1: '...\temp\venv\Scripts\python.exe' 'C:\Users\magnus.lycka\work\temp\venv\lib\site-packages\pip_vendor\pep517\in_process_in_process.py' get_requires_for_build_wheel '...\Temp\tmpaqtyz1ok' Check the logs for full command output.

__version__ = "1.1.9", docs say there's to_list and to_set, repository source has both methods, however pypi package doesn't have them.

I think foldleft summarizing method could be added, because reduce requires a function with two same-type parameters. foldleft/foldright method is more generalizing. The behavior of this API likes what reduce does in functools module.

## return ["1", "2", "3", "4"]
flu([1, 2, 3, 4])\
 .foldleft(lambda x, y: x + str(y), '')

## same as `functools.reduce`
reduce(lambda x, y: x + str(y), [1, 2, 3, 4], '')

Summary

python 3.10 introduced type guards similar to that of typescript. however, flupy does not seem to support this feature.

from typing import Any, TypedDict, TypeGuard

from flupy import flu


class Person(TypedDict):
    name: str
    age: int


def is_person(val: Any) -> TypeGuard[Person]:
    try:
        return (
            isinstance(val, dict)
            and isinstance(val["name"], str)
            and isinstance(val["age"], int)
        )
    except KeyError:
        return False


def get_age(val: Person):
    return f"Age: {val['age']}"


result = (
    flu([Person(name="Alice", age=20), 3, "afds", {"name": "Eve"}])
    .filter(is_person)
    .map(get_age)
    .collect()
)

Doc string for enumerate shows "zip_longest"

https://flupy.readthedocs.io/en/latest/api.html#grouping

both empty on firefox and chrome.

Amazing project :-)

Why it can't be python 3.5? I could give it a try in the following weeks in some py3.5 projects I have, just to provide some use cases.

When toying around with your library and testing some snippets of your API reference I noticed that the first example of group_by results in an error when comparing str with int:

>>> from flupy import flu
>>> flu.group_by([1, 'a', 'a', 1, 1]).collect()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/bjoosten/.pyenv/versions/3.6.3/lib/python3.6/site-packages/flupy/fluent.py", line 37, in wrapper
    return func(*args, **kwargs)
  File "/Users/bjoosten/.pyenv/versions/3.6.3/lib/python3.6/site-packages/flupy/fluent.py", line 235, in group_by
    gen = self.sort(key) if sort else self
  File "/Users/bjoosten/.pyenv/versions/3.6.3/lib/python3.6/site-packages/flupy/fluent.py", line 37, in wrapper
    return func(*args, **kwargs)
  File "/Users/bjoosten/.pyenv/versions/3.6.3/lib/python3.6/site-packages/flupy/fluent.py", line 204, in sort
    return Fluent(sorted(self, key=key, reverse=reverse))
TypeError: '<' not supported between instances of 'str' and 'int'