pyochain 0.5.0__py3-none-any.whl

pyochain/_iter/_maps.py

@@ -0,0 +1,358 @@
+from __future__ import annotations
+
+import itertools
+from collections.abc import Callable, Collection, Generator, Iterable, Iterator, Mapping
+from functools import partial
+from typing import TYPE_CHECKING, Any, Concatenate, Self, overload
+
+import cytoolz as cz
+import more_itertools as mit
+
+from .._core import IterWrapper
+
+if TYPE_CHECKING:
+    from ._main import Iter
+
+
+class BaseMap[T](IterWrapper[T]):
+    def for_each[**P](
+        self,
+        func: Callable[Concatenate[T, P], Any],
+        *args: P.args,
+        **kwargs: P.kwargs,
+    ) -> Self:
+        """
+        Apply a function to each element in the iterable.
+
+        Args:
+            func: Function to apply to each element.
+            args: Positional arguments for the function.
+            kwargs: Keyword arguments for the function.
+
+        Can be used for side effects such as printing or logging.
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1, 2, 3]).for_each(lambda x: print(x)).collect().unwrap()
+        1
+        2
+        3
+        []
+
+        ```
+        """
+        for v in self.unwrap():
+            func(v, *args, **kwargs)
+        return self
+
+    def map[R](self, func: Callable[[T], R]) -> Iter[R]:
+        """
+        Map each element through func and return a Iter of results.
+
+        Args:
+            func: Function to apply to each element.
+
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1, 2]).map(lambda x: x + 1).into(list)
+        [2, 3]
+
+        ```
+        """
+        return self.apply(partial(map, func))
+
+    @overload
+    def flat_map[U, R](
+        self: IterWrapper[Iterable[Iterable[Iterable[U]]]],
+        func: Callable[[T], Iterable[Iterable[R]]],
+    ) -> Iter[Iterable[Iterable[R]]]: ...
+    @overload
+    def flat_map[U, R](
+        self: IterWrapper[Iterable[Iterable[U]]], func: Callable[[T], Iterable[R]]
+    ) -> Iter[Iterable[R]]: ...
+    @overload
+    def flat_map[U, R](
+        self: IterWrapper[Iterable[U]], func: Callable[[T], R]
+    ) -> Iter[R]: ...
+    def flat_map[U: Iterable[Any], R](
+        self: IterWrapper[U], func: Callable[[T], R]
+    ) -> Iter[Any]:
+        """
+        Map each element through func and flatten the result by one level.
+
+        Args:
+            func: Function to apply to each element.
+
+        ```python
+        >>> import pyochain as pc
+        >>> data = [[1, 2], [3, 4]]
+        >>> pc.Iter.from_(data).flat_map(lambda x: x + 10).into(list)
+        [11, 12, 13, 14]
+
+        ```
+        """
+
+        def _flat_map(data: Iterable[U]) -> map[R]:
+            return map(func, itertools.chain.from_iterable(data))
+
+        return self.apply(_flat_map)
+
+    def map_star[U: Iterable[Any], R](
+        self: IterWrapper[U], func: Callable[..., R]
+    ) -> Iter[R]:
+        """
+        Applies a function to each element, where each element is an iterable.
+
+        Args:
+            func: Function to apply to unpacked elements.
+
+        Unlike `.map()`, which passes each element as a single argument, `.starmap()` unpacks each element into positional arguments for the function.
+
+        In short, for each `element` in the sequence, it computes `func(*element)`.
+        ```python
+        >>> import pyochain as pc
+        >>> def make_sku(color, size):
+        ...     return f"{color}-{size}"
+        >>> data = pc.Seq(["blue", "red"])
+        >>> data.iter().product(["S", "M"]).map_star(make_sku).into(list)
+        ['blue-S', 'blue-M', 'red-S', 'red-M']
+
+        ```
+        This is equivalent to:
+        ```python
+        >>> data.iter().product(["S", "M"]).map(lambda x: make_sku(*x)).into(list)
+        ['blue-S', 'blue-M', 'red-S', 'red-M']
+
+        ```
+        - Use map_star when the performance matters (it is faster).
+        - Use map with unpacking when readability matters (the types can be inferred).
+        """
+
+        return self.apply(partial(itertools.starmap, func))
+
+    def map_if[R](
+        self,
+        predicate: Callable[[T], bool],
+        func: Callable[[T], R],
+        func_else: Callable[[T], R] | None = None,
+    ) -> Iter[R]:
+        """
+        Evaluate each item from iterable using pred.
+
+        Args:
+            predicate: Function to evaluate each item.
+            func: Function to apply if predicate is True.
+            func_else: Function to apply if predicate is False.
+
+        - If the result is equivalent to True, transform the item with func and yield it.
+        - Otherwise, transform the item with func_else and yield it.
+        - Predicate, func, and func_else should each be functions that accept one argument.
+
+        By default, func_else is the identity function.
+        ```python
+        >>> import pyochain as pc
+        >>> from math import sqrt
+        >>> iterable = pc.Iter.from_(range(-5, 5)).collect()
+        >>> iterable.into(list)
+        [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4]
+        >>> iterable.iter().map_if(lambda x: x > 3, lambda x: "toobig").into(list)
+        [-5, -4, -3, -2, -1, 0, 1, 2, 3, 'toobig']
+        >>> iterable.iter().map_if(
+        ...     lambda x: x >= 0,
+        ...     lambda x: f"{sqrt(x):.2f}",
+        ...     lambda x: None,
+        ... ).into(list)
+        [None, None, None, None, None, '0.00', '1.00', '1.41', '1.73', '2.00']
+
+        ```
+        """
+        return self.apply(mit.map_if, predicate, func, func_else=func_else)
+
+    def map_except[R](
+        self, func: Callable[[T], R], *exceptions: type[BaseException]
+    ) -> Iter[R]:
+        """
+        Transform each item from iterable with function and yield the result, unless function raises one of the specified exceptions.
+
+        Args:
+            func: Function to apply to each item.
+            exceptions: Exceptions to catch and ignore.
+
+        The function is called to transform each item in iterable
+
+        If an exception other than one given by exceptions is raised by function, it is raised like normal.
+        ```python
+        >>> import pyochain as pc
+        >>> iterable = ["1", "2", "three", "4", None]
+        >>> pc.Iter.from_(iterable).map_except(int, ValueError, TypeError).into(list)
+        [1, 2, 4]
+
+        ```
+        """
+
+        def _map_except(data: Iterable[T]) -> Iterator[R]:
+            return mit.map_except(func, data, *exceptions)
+
+        return self.apply(_map_except)
+
+    def repeat(
+        self, n: int, factory: Callable[[Iterable[T]], Collection[T]] = tuple
+    ) -> Iter[Iterable[T]]:
+        """
+        Repeat the entire iterable n times (as elements) and return Iter.
+
+        Args:
+            n: Number of repetitions.
+            factory: Factory to create the repeated collection (default: tuple).
+
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1, 2]).repeat(2).collect().unwrap()
+        [(1, 2), (1, 2)]
+        >>> pc.Iter.from_([1, 2]).repeat(3, list).collect().unwrap()
+        [[1, 2], [1, 2], [1, 2]]
+
+        ```
+        """
+
+        def _repeat(data: Iterable[T]) -> Iterator[Iterable[T]]:
+            return itertools.repeat(factory(data), n)
+
+        return self.apply(_repeat)
+
+    @overload
+    def repeat_last(self, default: T) -> Iter[T]: ...
+    @overload
+    def repeat_last[U](self, default: U) -> Iter[T | U]: ...
+    def repeat_last[U](self, default: U = None) -> Iter[T | U]:
+        """
+        After the iterable is exhausted, keep yielding its last element.
+
+        **Warning** ⚠️
+            This creates an infinite iterator.
+            Be sure to use `Iter.take()` or `Iter.slice()` to limit the number of items taken.
+
+        Args:
+            default: Value to yield if the iterable is empty.
+
+        Example:
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_(range(3)).repeat_last().take(5).into(list)
+        [0, 1, 2, 2, 2]
+
+        If the iterable is empty, yield default forever:
+        ```python
+        >>> pc.Iter.from_(range(0)).repeat_last(42).take(5).into(list)
+        [42, 42, 42, 42, 42]
+
+        ```
+        """
+        return self.apply(mit.repeat_last, default)
+
+    def ichunked(self, n: int) -> Iter[Iterator[T]]:
+        """
+
+        Break *iterable* into sub-iterables with *n* elements each.
+
+        Args:
+            n: Number of elements in each chunk.
+
+        If the sub-iterables are read in order, the elements of *iterable*
+        won't be stored in memory.
+
+        If they are read out of order, :func:`itertools.tee` is used to cache
+        elements as necessary.
+        ```python
+        >>> import pyochain as pc
+        >>> all_chunks = pc.Iter.from_count().ichunked(4).unwrap()
+        >>> c_1, c_2, c_3 = next(all_chunks), next(all_chunks), next(all_chunks)
+        >>> list(c_2)  # c_1's elements have been cached; c_3's haven't been
+        [4, 5, 6, 7]
+        >>> list(c_1)
+        [0, 1, 2, 3]
+        >>> list(c_3)
+        [8, 9, 10, 11]
+
+        ```
+        """
+        return self.apply(mit.ichunked, n)
+
+    @overload
+    def flatten[U](
+        self: IterWrapper[Iterable[Iterable[Iterable[U]]]],
+    ) -> Iter[Iterable[Iterable[U]]]: ...
+    @overload
+    def flatten[U](self: IterWrapper[Iterable[Iterable[U]]]) -> Iter[Iterable[U]]: ...
+    @overload
+    def flatten[U](self: IterWrapper[Iterable[U]]) -> Iter[U]: ...
+    def flatten(self: IterWrapper[Iterable[Any]]) -> Iter[Any]:
+        """
+        Flatten one level of nesting and return a new Iterable wrapper.
+
+        This is a shortcut for `.apply(itertools.chain.from_iterable)`.
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([[1, 2], [3]]).flatten().into(list)
+        [1, 2, 3]
+
+        ```
+        """
+        return self.apply(itertools.chain.from_iterable)
+
+    def pluck[U: Mapping[Any, Any]](
+        self: IterWrapper[U], *keys: str | int
+    ) -> Iter[Any]:
+        """
+        Get an element from each item in a sequence using a nested key path.
+
+        Args:
+            keys: Nested keys to extract values.
+
+        ```python
+        >>> import pyochain as pc
+        >>> data = pc.Seq(
+        ...     [
+        ...         {"id": 1, "info": {"name": "Alice", "age": 30}},
+        ...         {"id": 2, "info": {"name": "Bob", "age": 25}},
+        ...     ]
+        ... )
+        >>> data.iter().pluck("info").into(list)
+        [{'name': 'Alice', 'age': 30}, {'name': 'Bob', 'age': 25}]
+        >>> data.iter().pluck("info", "name").into(list)
+        ['Alice', 'Bob']
+
+        ```
+        Example: get the maximum age along with the corresponding id)
+        ```python
+        >>> data.iter().pluck("info", "age").zip(
+        ...     data.iter().pluck("id").into(list)
+        ... ).max()
+        (30, 1)
+
+        ```
+        """
+
+        getter = partial(cz.dicttoolz.get_in, keys)
+        return self.apply(partial(map, getter))
+
+    def round[U: float | int](
+        self: IterWrapper[U], ndigits: int | None = None
+    ) -> Iter[float]:
+        """
+        Round each element in the iterable to the given number of decimal places and return Iter.
+
+        Args:
+            ndigits: Number of decimal places to round to.
+
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1.2345, 2.3456, 3.4567]).round(2).into(list)
+        [1.23, 2.35, 3.46]
+
+        ```
+        """
+
+        def _round(data: Iterable[U]) -> Generator[float | int, None, None]:
+            return (round(x, ndigits) for x in data)
+
+        return self.apply(_round)

pyochain/_iter/_partitions.py

@@ -0,0 +1,148 @@
+from __future__ import annotations
+
+import itertools
+from collections.abc import Callable
+from functools import partial
+from typing import TYPE_CHECKING, Literal, overload
+
+import cytoolz as cz
+
+from .._core import IterWrapper
+
+if TYPE_CHECKING:
+    from ._main import Iter
+
+
+class BasePartitions[T](IterWrapper[T]):
+    @overload
+    def windows(self, length: Literal[1]) -> Iter[tuple[T]]: ...
+    @overload
+    def windows(self, length: Literal[2]) -> Iter[tuple[T, T]]: ...
+    @overload
+    def windows(self, length: Literal[3]) -> Iter[tuple[T, T, T]]: ...
+    @overload
+    def windows(self, length: Literal[4]) -> Iter[tuple[T, T, T, T]]: ...
+    @overload
+    def windows(self, length: Literal[5]) -> Iter[tuple[T, T, T, T, T]]: ...
+
+    def windows(self, length: int) -> Iter[tuple[T, ...]]:
+        """
+        A sequence of overlapping subsequences of the given length.
+
+        Args:
+            length: The length of each window.
+
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1, 2, 3, 4]).windows(2).into(list)
+        [(1, 2), (2, 3), (3, 4)]
+
+        ```
+        This function allows you to apply custom function not available in the rolling namespace.
+        ```python
+        >>> def moving_average(seq: tuple[int, ...]) -> float:
+        ...     return float(sum(seq)) / len(seq)
+        >>> pc.Iter.from_([1, 2, 3, 4]).windows(2).map(moving_average).into(list)
+        [1.5, 2.5, 3.5]
+
+        ```
+        """
+        return self.apply(partial(cz.itertoolz.sliding_window, length))
+
+    @overload
+    def partition(self, n: Literal[1], pad: None = None) -> Iter[tuple[T]]: ...
+    @overload
+    def partition(self, n: Literal[2], pad: None = None) -> Iter[tuple[T, T]]: ...
+    @overload
+    def partition(self, n: Literal[3], pad: None = None) -> Iter[tuple[T, T, T]]: ...
+    @overload
+    def partition(self, n: Literal[4], pad: None = None) -> Iter[tuple[T, T, T, T]]: ...
+    @overload
+    def partition(
+        self, n: Literal[5], pad: None = None
+    ) -> Iter[tuple[T, T, T, T, T]]: ...
+    @overload
+    def partition(self, n: int, pad: int) -> Iter[tuple[T, ...]]: ...
+    def partition(self, n: int, pad: int | None = None) -> Iter[tuple[T, ...]]:
+        """
+        Partition sequence into tuples of length n.
+
+        Args:
+            n: Length of each partition.
+            pad: Value to pad the last partition if needed.
+
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1, 2, 3, 4]).partition(2).into(list)
+        [(1, 2), (3, 4)]
+
+        ```
+        If the length of seq is not evenly divisible by n, the final tuple is dropped if pad is not specified, or filled to length n by pad:
+        ```python
+        >>> pc.Iter.from_([1, 2, 3, 4, 5]).partition(2).into(list)
+        [(1, 2), (3, 4), (5, None)]
+
+        ```
+        """
+
+        return self.apply(partial(cz.itertoolz.partition, n, pad=pad))
+
+    def partition_all(self, n: int) -> Iter[tuple[T, ...]]:
+        """
+        Partition all elements of sequence into tuples of length at most n.
+
+        Args:
+            n: Maximum length of each partition.
+
+        The final tuple may be shorter to accommodate extra elements.
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_([1, 2, 3, 4]).partition_all(2).into(list)
+        [(1, 2), (3, 4)]
+        >>> pc.Iter.from_([1, 2, 3, 4, 5]).partition_all(2).into(list)
+        [(1, 2), (3, 4), (5,)]
+
+        ```
+        """
+        return self.apply(partial(cz.itertoolz.partition_all, n))
+
+    def partition_by(self, predicate: Callable[[T], bool]) -> Iter[tuple[T, ...]]:
+        """
+        Partition the `iterable` into a sequence of `tuples` according to a predicate function.
+
+        Args:
+            predicate: Function to determine partition boundaries.
+
+        Every time the output of `predicate` changes, a new `tuple` is started,
+        and subsequent items are collected into that `tuple`.
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_("I have space").partition_by(lambda c: c == " ").into(list)
+        [('I',), (' ',), ('h', 'a', 'v', 'e'), (' ',), ('s', 'p', 'a', 'c', 'e')]
+        >>>
+        >>> data = [1, 2, 1, 99, 88, 33, 99, -1, 5]
+        >>> pc.Iter.from_(data).partition_by(lambda x: x > 10).into(list)
+        [(1, 2, 1), (99, 88, 33, 99), (-1, 5)]
+
+        ```
+        """
+        return self.apply(partial(cz.recipes.partitionby, predicate))
+
+    def batch(self, n: int) -> Iter[tuple[T, ...]]:
+        """
+        Batch elements into tuples of length n and return a new Iter.
+
+        Args:
+            n: Number of elements in each batch.
+
+        - The last batch may be shorter than n.
+        - The data is consumed lazily, just enough to fill a batch.
+        - The result is yielded as soon as a batch is full or when the input iterable is exhausted.
+        ```python
+        >>> import pyochain as pc
+        >>> pc.Iter.from_("ABCDEFG").batch(3).into(list)
+        [('A', 'B', 'C'), ('D', 'E', 'F'), ('G',)]
+
+        ```
+        """
+        return self.apply(itertools.batched, n)