mixpeek 0.8.41__py3-none-any.whl → 0.10.0__py3-none-any.whl

mixpeek/_utils/_reflection.py

@@ -0,0 +1,42 @@
+from __future__ import annotations
+
+import inspect
+from typing import Any, Callable
+
+
+def function_has_argument(func: Callable[..., Any], arg_name: str) -> bool:
+    """Returns whether or not the given function has a specific parameter"""
+    sig = inspect.signature(func)
+    return arg_name in sig.parameters
+
+
+def assert_signatures_in_sync(
+    source_func: Callable[..., Any],
+    check_func: Callable[..., Any],
+    *,
+    exclude_params: set[str] = set(),
+) -> None:
+    """Ensure that the signature of the second function matches the first."""
+
+    check_sig = inspect.signature(check_func)
+    source_sig = inspect.signature(source_func)
+
+    errors: list[str] = []
+
+    for name, source_param in source_sig.parameters.items():
+        if name in exclude_params:
+            continue
+
+        custom_param = check_sig.parameters.get(name)
+        if not custom_param:
+            errors.append(f"the `{name}` param is missing")
+            continue
+
+        if custom_param.annotation != source_param.annotation:
+            errors.append(
+                f"types for the `{name}` param are do not match; source={repr(source_param.annotation)} checking={repr(custom_param.annotation)}"
+            )
+            continue
+
+    if errors:
+        raise AssertionError(f"{len(errors)} errors encountered when comparing signatures:\n\n" + "\n\n".join(errors))

mixpeek/_utils/_streams.py

@@ -0,0 +1,12 @@
+from typing import Any
+from typing_extensions import Iterator, AsyncIterator
+
+
+def consume_sync_iterator(iterator: Iterator[Any]) -> None:
+    for _ in iterator:
+        ...
+
+
+async def consume_async_iterator(iterator: AsyncIterator[Any]) -> None:
+    async for _ in iterator:
+        ...

mixpeek/_utils/_sync.py

@@ -0,0 +1,81 @@
+from __future__ import annotations
+
+import functools
+from typing import TypeVar, Callable, Awaitable
+from typing_extensions import ParamSpec
+
+import anyio
+import anyio.to_thread
+
+from ._reflection import function_has_argument
+
+T_Retval = TypeVar("T_Retval")
+T_ParamSpec = ParamSpec("T_ParamSpec")
+
+
+# copied from `asyncer`, https://github.com/tiangolo/asyncer
+def asyncify(
+    function: Callable[T_ParamSpec, T_Retval],
+    *,
+    cancellable: bool = False,
+    limiter: anyio.CapacityLimiter | None = None,
+) -> Callable[T_ParamSpec, Awaitable[T_Retval]]:
+    """
+    Take a blocking function and create an async one that receives the same
+    positional and keyword arguments, and that when called, calls the original function
+    in a worker thread using `anyio.to_thread.run_sync()`. Internally,
+    `asyncer.asyncify()` uses the same `anyio.to_thread.run_sync()`, but it supports
+    keyword arguments additional to positional arguments and it adds better support for
+    autocompletion and inline errors for the arguments of the function called and the
+    return value.
+
+    If the `cancellable` option is enabled and the task waiting for its completion is
+    cancelled, the thread will still run its course but its return value (or any raised
+    exception) will be ignored.
+
+    Use it like this:
+
+    ```Python
+    def do_work(arg1, arg2, kwarg1="", kwarg2="") -> str:
+        # Do work
+        return "Some result"
+
+
+    result = await to_thread.asyncify(do_work)("spam", "ham", kwarg1="a", kwarg2="b")
+    print(result)
+    ```
+
+    ## Arguments
+
+    `function`: a blocking regular callable (e.g. a function)
+    `cancellable`: `True` to allow cancellation of the operation
+    `limiter`: capacity limiter to use to limit the total amount of threads running
+        (if omitted, the default limiter is used)
+
+    ## Return
+
+    An async function that takes the same positional and keyword arguments as the
+    original one, that when called runs the same original function in a thread worker
+    and returns the result.
+    """
+
+    async def wrapper(*args: T_ParamSpec.args, **kwargs: T_ParamSpec.kwargs) -> T_Retval:
+        partial_f = functools.partial(function, *args, **kwargs)
+
+        # In `v4.1.0` anyio added the `abandon_on_cancel` argument and deprecated the old
+        # `cancellable` argument, so we need to use the new `abandon_on_cancel` to avoid
+        # surfacing deprecation warnings.
+        if function_has_argument(anyio.to_thread.run_sync, "abandon_on_cancel"):
+            return await anyio.to_thread.run_sync(
+                partial_f,
+                abandon_on_cancel=cancellable,
+                limiter=limiter,
+            )
+
+        return await anyio.to_thread.run_sync(
+            partial_f,
+            cancellable=cancellable,
+            limiter=limiter,
+        )
+
+    return wrapper

mixpeek/_utils/_transform.py

@@ -0,0 +1,382 @@
+from __future__ import annotations
+
+import io
+import base64
+import pathlib
+from typing import Any, Mapping, TypeVar, cast
+from datetime import date, datetime
+from typing_extensions import Literal, get_args, override, get_type_hints
+
+import anyio
+import pydantic
+
+from ._utils import (
+    is_list,
+    is_mapping,
+    is_iterable,
+)
+from .._files import is_base64_file_input
+from ._typing import (
+    is_list_type,
+    is_union_type,
+    extract_type_arg,
+    is_iterable_type,
+    is_required_type,
+    is_annotated_type,
+    strip_annotated_type,
+)
+from .._compat import model_dump, is_typeddict
+
+_T = TypeVar("_T")
+
+
+# TODO: support for drilling globals() and locals()
+# TODO: ensure works correctly with forward references in all cases
+
+
+PropertyFormat = Literal["iso8601", "base64", "custom"]
+
+
+class PropertyInfo:
+    """Metadata class to be used in Annotated types to provide information about a given type.
+
+    For example:
+
+    class MyParams(TypedDict):
+        account_holder_name: Annotated[str, PropertyInfo(alias='accountHolderName')]
+
+    This means that {'account_holder_name': 'Robert'} will be transformed to {'accountHolderName': 'Robert'} before being sent to the API.
+    """
+
+    alias: str | None
+    format: PropertyFormat | None
+    format_template: str | None
+    discriminator: str | None
+
+    def __init__(
+        self,
+        *,
+        alias: str | None = None,
+        format: PropertyFormat | None = None,
+        format_template: str | None = None,
+        discriminator: str | None = None,
+    ) -> None:
+        self.alias = alias
+        self.format = format
+        self.format_template = format_template
+        self.discriminator = discriminator
+
+    @override
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(alias='{self.alias}', format={self.format}, format_template='{self.format_template}', discriminator='{self.discriminator}')"
+
+
+def maybe_transform(
+    data: object,
+    expected_type: object,
+) -> Any | None:
+    """Wrapper over `transform()` that allows `None` to be passed.
+
+    See `transform()` for more details.
+    """
+    if data is None:
+        return None
+    return transform(data, expected_type)
+
+
+# Wrapper over _transform_recursive providing fake types
+def transform(
+    data: _T,
+    expected_type: object,
+) -> _T:
+    """Transform dictionaries based off of type information from the given type, for example:
+
+    ```py
+    class Params(TypedDict, total=False):
+        card_id: Required[Annotated[str, PropertyInfo(alias="cardID")]]
+
+
+    transformed = transform({"card_id": "<my card ID>"}, Params)
+    # {'cardID': '<my card ID>'}
+    ```
+
+    Any keys / data that does not have type information given will be included as is.
+
+    It should be noted that the transformations that this function does are not represented in the type system.
+    """
+    transformed = _transform_recursive(data, annotation=cast(type, expected_type))
+    return cast(_T, transformed)
+
+
+def _get_annotated_type(type_: type) -> type | None:
+    """If the given type is an `Annotated` type then it is returned, if not `None` is returned.
+
+    This also unwraps the type when applicable, e.g. `Required[Annotated[T, ...]]`
+    """
+    if is_required_type(type_):
+        # Unwrap `Required[Annotated[T, ...]]` to `Annotated[T, ...]`
+        type_ = get_args(type_)[0]
+
+    if is_annotated_type(type_):
+        return type_
+
+    return None
+
+
+def _maybe_transform_key(key: str, type_: type) -> str:
+    """Transform the given `data` based on the annotations provided in `type_`.
+
+    Note: this function only looks at `Annotated` types that contain `PropertInfo` metadata.
+    """
+    annotated_type = _get_annotated_type(type_)
+    if annotated_type is None:
+        # no `Annotated` definition for this type, no transformation needed
+        return key
+
+    # ignore the first argument as it is the actual type
+    annotations = get_args(annotated_type)[1:]
+    for annotation in annotations:
+        if isinstance(annotation, PropertyInfo) and annotation.alias is not None:
+            return annotation.alias
+
+    return key
+
+
+def _transform_recursive(
+    data: object,
+    *,
+    annotation: type,
+    inner_type: type | None = None,
+) -> object:
+    """Transform the given data against the expected type.
+
+    Args:
+        annotation: The direct type annotation given to the particular piece of data.
+            This may or may not be wrapped in metadata types, e.g. `Required[T]`, `Annotated[T, ...]` etc
+
+        inner_type: If applicable, this is the "inside" type. This is useful in certain cases where the outside type
+            is a container type such as `List[T]`. In that case `inner_type` should be set to `T` so that each entry in
+            the list can be transformed using the metadata from the container type.
+
+            Defaults to the same value as the `annotation` argument.
+    """
+    if inner_type is None:
+        inner_type = annotation
+
+    stripped_type = strip_annotated_type(inner_type)
+    if is_typeddict(stripped_type) and is_mapping(data):
+        return _transform_typeddict(data, stripped_type)
+
+    if (
+        # List[T]
+        (is_list_type(stripped_type) and is_list(data))
+        # Iterable[T]
+        or (is_iterable_type(stripped_type) and is_iterable(data) and not isinstance(data, str))
+    ):
+        inner_type = extract_type_arg(stripped_type, 0)
+        return [_transform_recursive(d, annotation=annotation, inner_type=inner_type) for d in data]
+
+    if is_union_type(stripped_type):
+        # For union types we run the transformation against all subtypes to ensure that everything is transformed.
+        #
+        # TODO: there may be edge cases where the same normalized field name will transform to two different names
+        # in different subtypes.
+        for subtype in get_args(stripped_type):
+            data = _transform_recursive(data, annotation=annotation, inner_type=subtype)
+        return data
+
+    if isinstance(data, pydantic.BaseModel):
+        return model_dump(data, exclude_unset=True)
+
+    annotated_type = _get_annotated_type(annotation)
+    if annotated_type is None:
+        return data
+
+    # ignore the first argument as it is the actual type
+    annotations = get_args(annotated_type)[1:]
+    for annotation in annotations:
+        if isinstance(annotation, PropertyInfo) and annotation.format is not None:
+            return _format_data(data, annotation.format, annotation.format_template)
+
+    return data
+
+
+def _format_data(data: object, format_: PropertyFormat, format_template: str | None) -> object:
+    if isinstance(data, (date, datetime)):
+        if format_ == "iso8601":
+            return data.isoformat()
+
+        if format_ == "custom" and format_template is not None:
+            return data.strftime(format_template)
+
+    if format_ == "base64" and is_base64_file_input(data):
+        binary: str | bytes | None = None
+
+        if isinstance(data, pathlib.Path):
+            binary = data.read_bytes()
+        elif isinstance(data, io.IOBase):
+            binary = data.read()
+
+            if isinstance(binary, str):  # type: ignore[unreachable]
+                binary = binary.encode()
+
+        if not isinstance(binary, bytes):
+            raise RuntimeError(f"Could not read bytes from {data}; Received {type(binary)}")
+
+        return base64.b64encode(binary).decode("ascii")
+
+    return data
+
+
+def _transform_typeddict(
+    data: Mapping[str, object],
+    expected_type: type,
+) -> Mapping[str, object]:
+    result: dict[str, object] = {}
+    annotations = get_type_hints(expected_type, include_extras=True)
+    for key, value in data.items():
+        type_ = annotations.get(key)
+        if type_ is None:
+            # we do not have a type annotation for this field, leave it as is
+            result[key] = value
+        else:
+            result[_maybe_transform_key(key, type_)] = _transform_recursive(value, annotation=type_)
+    return result
+
+
+async def async_maybe_transform(
+    data: object,
+    expected_type: object,
+) -> Any | None:
+    """Wrapper over `async_transform()` that allows `None` to be passed.
+
+    See `async_transform()` for more details.
+    """
+    if data is None:
+        return None
+    return await async_transform(data, expected_type)
+
+
+async def async_transform(
+    data: _T,
+    expected_type: object,
+) -> _T:
+    """Transform dictionaries based off of type information from the given type, for example:
+
+    ```py
+    class Params(TypedDict, total=False):
+        card_id: Required[Annotated[str, PropertyInfo(alias="cardID")]]
+
+
+    transformed = transform({"card_id": "<my card ID>"}, Params)
+    # {'cardID': '<my card ID>'}
+    ```
+
+    Any keys / data that does not have type information given will be included as is.
+
+    It should be noted that the transformations that this function does are not represented in the type system.
+    """
+    transformed = await _async_transform_recursive(data, annotation=cast(type, expected_type))
+    return cast(_T, transformed)
+
+
+async def _async_transform_recursive(
+    data: object,
+    *,
+    annotation: type,
+    inner_type: type | None = None,
+) -> object:
+    """Transform the given data against the expected type.
+
+    Args:
+        annotation: The direct type annotation given to the particular piece of data.
+            This may or may not be wrapped in metadata types, e.g. `Required[T]`, `Annotated[T, ...]` etc
+
+        inner_type: If applicable, this is the "inside" type. This is useful in certain cases where the outside type
+            is a container type such as `List[T]`. In that case `inner_type` should be set to `T` so that each entry in
+            the list can be transformed using the metadata from the container type.
+
+            Defaults to the same value as the `annotation` argument.
+    """
+    if inner_type is None:
+        inner_type = annotation
+
+    stripped_type = strip_annotated_type(inner_type)
+    if is_typeddict(stripped_type) and is_mapping(data):
+        return await _async_transform_typeddict(data, stripped_type)
+
+    if (
+        # List[T]
+        (is_list_type(stripped_type) and is_list(data))
+        # Iterable[T]
+        or (is_iterable_type(stripped_type) and is_iterable(data) and not isinstance(data, str))
+    ):
+        inner_type = extract_type_arg(stripped_type, 0)
+        return [await _async_transform_recursive(d, annotation=annotation, inner_type=inner_type) for d in data]
+
+    if is_union_type(stripped_type):
+        # For union types we run the transformation against all subtypes to ensure that everything is transformed.
+        #
+        # TODO: there may be edge cases where the same normalized field name will transform to two different names
+        # in different subtypes.
+        for subtype in get_args(stripped_type):
+            data = await _async_transform_recursive(data, annotation=annotation, inner_type=subtype)
+        return data
+
+    if isinstance(data, pydantic.BaseModel):
+        return model_dump(data, exclude_unset=True)
+
+    annotated_type = _get_annotated_type(annotation)
+    if annotated_type is None:
+        return data
+
+    # ignore the first argument as it is the actual type
+    annotations = get_args(annotated_type)[1:]
+    for annotation in annotations:
+        if isinstance(annotation, PropertyInfo) and annotation.format is not None:
+            return await _async_format_data(data, annotation.format, annotation.format_template)
+
+    return data
+
+
+async def _async_format_data(data: object, format_: PropertyFormat, format_template: str | None) -> object:
+    if isinstance(data, (date, datetime)):
+        if format_ == "iso8601":
+            return data.isoformat()
+
+        if format_ == "custom" and format_template is not None:
+            return data.strftime(format_template)
+
+    if format_ == "base64" and is_base64_file_input(data):
+        binary: str | bytes | None = None
+
+        if isinstance(data, pathlib.Path):
+            binary = await anyio.Path(data).read_bytes()
+        elif isinstance(data, io.IOBase):
+            binary = data.read()
+
+            if isinstance(binary, str):  # type: ignore[unreachable]
+                binary = binary.encode()
+
+        if not isinstance(binary, bytes):
+            raise RuntimeError(f"Could not read bytes from {data}; Received {type(binary)}")
+
+        return base64.b64encode(binary).decode("ascii")
+
+    return data
+
+
+async def _async_transform_typeddict(
+    data: Mapping[str, object],
+    expected_type: type,
+) -> Mapping[str, object]:
+    result: dict[str, object] = {}
+    annotations = get_type_hints(expected_type, include_extras=True)
+    for key, value in data.items():
+        type_ = annotations.get(key)
+        if type_ is None:
+            # we do not have a type annotation for this field, leave it as is
+            result[key] = value
+        else:
+            result[_maybe_transform_key(key, type_)] = await _async_transform_recursive(value, annotation=type_)
+    return result

mixpeek/_utils/_typing.py

@@ -0,0 +1,120 @@
+from __future__ import annotations
+
+from typing import Any, TypeVar, Iterable, cast
+from collections import abc as _c_abc
+from typing_extensions import Required, Annotated, get_args, get_origin
+
+from .._types import InheritsGeneric
+from .._compat import is_union as _is_union
+
+
+def is_annotated_type(typ: type) -> bool:
+    return get_origin(typ) == Annotated
+
+
+def is_list_type(typ: type) -> bool:
+    return (get_origin(typ) or typ) == list
+
+
+def is_iterable_type(typ: type) -> bool:
+    """If the given type is `typing.Iterable[T]`"""
+    origin = get_origin(typ) or typ
+    return origin == Iterable or origin == _c_abc.Iterable
+
+
+def is_union_type(typ: type) -> bool:
+    return _is_union(get_origin(typ))
+
+
+def is_required_type(typ: type) -> bool:
+    return get_origin(typ) == Required
+
+
+def is_typevar(typ: type) -> bool:
+    # type ignore is required because type checkers
+    # think this expression will always return False
+    return type(typ) == TypeVar  # type: ignore
+
+
+# Extracts T from Annotated[T, ...] or from Required[Annotated[T, ...]]
+def strip_annotated_type(typ: type) -> type:
+    if is_required_type(typ) or is_annotated_type(typ):
+        return strip_annotated_type(cast(type, get_args(typ)[0]))
+
+    return typ
+
+
+def extract_type_arg(typ: type, index: int) -> type:
+    args = get_args(typ)
+    try:
+        return cast(type, args[index])
+    except IndexError as err:
+        raise RuntimeError(f"Expected type {typ} to have a type argument at index {index} but it did not") from err
+
+
+def extract_type_var_from_base(
+    typ: type,
+    *,
+    generic_bases: tuple[type, ...],
+    index: int,
+    failure_message: str | None = None,
+) -> type:
+    """Given a type like `Foo[T]`, returns the generic type variable `T`.
+
+    This also handles the case where a concrete subclass is given, e.g.
+    ```py
+    class MyResponse(Foo[bytes]):
+        ...
+
+    extract_type_var(MyResponse, bases=(Foo,), index=0) -> bytes
+    ```
+
+    And where a generic subclass is given:
+    ```py
+    _T = TypeVar('_T')
+    class MyResponse(Foo[_T]):
+        ...
+
+    extract_type_var(MyResponse[bytes], bases=(Foo,), index=0) -> bytes
+    ```
+    """
+    cls = cast(object, get_origin(typ) or typ)
+    if cls in generic_bases:
+        # we're given the class directly
+        return extract_type_arg(typ, index)
+
+    # if a subclass is given
+    # ---
+    # this is needed as __orig_bases__ is not present in the typeshed stubs
+    # because it is intended to be for internal use only, however there does
+    # not seem to be a way to resolve generic TypeVars for inherited subclasses
+    # without using it.
+    if isinstance(cls, InheritsGeneric):
+        target_base_class: Any | None = None
+        for base in cls.__orig_bases__:
+            if base.__origin__ in generic_bases:
+                target_base_class = base
+                break
+
+        if target_base_class is None:
+            raise RuntimeError(
+                "Could not find the generic base class;\n"
+                "This should never happen;\n"
+                f"Does {cls} inherit from one of {generic_bases} ?"
+            )
+
+        extracted = extract_type_arg(target_base_class, index)
+        if is_typevar(extracted):
+            # If the extracted type argument is itself a type variable
+            # then that means the subclass itself is generic, so we have
+            # to resolve the type argument from the class itself, not
+            # the base class.
+            #
+            # Note: if there is more than 1 type argument, the subclass could
+            # change the ordering of the type arguments, this is not currently
+            # supported.
+            return extract_type_arg(typ, index)
+
+        return extracted
+
+    raise RuntimeError(failure_message or f"Could not resolve inner type variable at index {index} for {typ}")