fixed subscription table

.venv/lib/python3.12/site-packages/prompt_toolkit/eventloop/__init__.py

@@ -0,0 +1,31 @@
+from __future__ import annotations
+
+from .async_generator import aclosing, generator_to_async_generator
+from .inputhook import (
+    InputHook,
+    InputHookContext,
+    InputHookSelector,
+    new_eventloop_with_inputhook,
+    set_eventloop_with_inputhook,
+)
+from .utils import (
+    call_soon_threadsafe,
+    get_traceback_from_context,
+    run_in_executor_with_context,
+)
+
+__all__ = [
+    # Async generator
+    "generator_to_async_generator",
+    "aclosing",
+    # Utils.
+    "run_in_executor_with_context",
+    "call_soon_threadsafe",
+    "get_traceback_from_context",
+    # Inputhooks.
+    "InputHook",
+    "new_eventloop_with_inputhook",
+    "set_eventloop_with_inputhook",
+    "InputHookSelector",
+    "InputHookContext",
+]

.venv/lib/python3.12/site-packages/prompt_toolkit/eventloop/async_generator.py

@@ -0,0 +1,125 @@
+"""
+Implementation for async generators.
+"""
+
+from __future__ import annotations
+
+from asyncio import get_running_loop
+from contextlib import asynccontextmanager
+from queue import Empty, Full, Queue
+from typing import Any, AsyncGenerator, Callable, Iterable, TypeVar
+
+from .utils import run_in_executor_with_context
+
+__all__ = [
+    "aclosing",
+    "generator_to_async_generator",
+]
+
+_T_Generator = TypeVar("_T_Generator", bound=AsyncGenerator[Any, None])
+
+
+@asynccontextmanager
+async def aclosing(
+    thing: _T_Generator,
+) -> AsyncGenerator[_T_Generator, None]:
+    "Similar to `contextlib.aclosing`, in Python 3.10."
+    try:
+        yield thing
+    finally:
+        await thing.aclose()
+
+
+# By default, choose a buffer size that's a good balance between having enough
+# throughput, but not consuming too much memory. We use this to consume a sync
+# generator of completions as an async generator. If the queue size is very
+# small (like 1), consuming the completions goes really slow (when there are a
+# lot of items). If the queue size would be unlimited or too big, this can
+# cause overconsumption of memory, and cause CPU time spent producing items
+# that are no longer needed (if the consumption of the async generator stops at
+# some point). We need a fixed size in order to get some back pressure from the
+# async consumer to the sync producer. We choose 1000 by default here. If we
+# have around 50k completions, measurements show that 1000 is still
+# significantly faster than a buffer of 100.
+DEFAULT_BUFFER_SIZE: int = 1000
+
+_T = TypeVar("_T")
+
+
+class _Done:
+    pass
+
+
+async def generator_to_async_generator(
+    get_iterable: Callable[[], Iterable[_T]],
+    buffer_size: int = DEFAULT_BUFFER_SIZE,
+) -> AsyncGenerator[_T, None]:
+    """
+    Turn a generator or iterable into an async generator.
+
+    This works by running the generator in a background thread.
+
+    :param get_iterable: Function that returns a generator or iterable when
+        called.
+    :param buffer_size: Size of the queue between the async consumer and the
+        synchronous generator that produces items.
+    """
+    quitting = False
+    # NOTE: We are limiting the queue size in order to have back-pressure.
+    q: Queue[_T | _Done] = Queue(maxsize=buffer_size)
+    loop = get_running_loop()
+
+    def runner() -> None:
+        """
+        Consume the generator in background thread.
+        When items are received, they'll be pushed to the queue.
+        """
+        try:
+            for item in get_iterable():
+                # When this async generator was cancelled (closed), stop this
+                # thread.
+                if quitting:
+                    return
+
+                while True:
+                    try:
+                        q.put(item, timeout=1)
+                    except Full:
+                        if quitting:
+                            return
+                        continue
+                    else:
+                        break
+
+        finally:
+            while True:
+                try:
+                    q.put(_Done(), timeout=1)
+                except Full:
+                    if quitting:
+                        return
+                    continue
+                else:
+                    break
+
+    # Start background thread.
+    runner_f = run_in_executor_with_context(runner)
+
+    try:
+        while True:
+            try:
+                item = q.get_nowait()
+            except Empty:
+                item = await loop.run_in_executor(None, q.get)
+            if isinstance(item, _Done):
+                break
+            else:
+                yield item
+    finally:
+        # When this async generator is closed (GeneratorExit exception, stop
+        # the background thread as well. - we don't need that anymore.)
+        quitting = True
+
+        # Wait for the background thread to finish. (should happen right after
+        # the last item is yielded).
+        await runner_f

.venv/lib/python3.12/site-packages/prompt_toolkit/eventloop/inputhook.py

@@ -0,0 +1,191 @@
+"""
+Similar to `PyOS_InputHook` of the Python API, we can plug in an input hook in
+the asyncio event loop.
+
+The way this works is by using a custom 'selector' that runs the other event
+loop until the real selector is ready.
+
+It's the responsibility of this event hook to return when there is input ready.
+There are two ways to detect when input is ready:
+
+The inputhook itself is a callable that receives an `InputHookContext`. This
+callable should run the other event loop, and return when the main loop has
+stuff to do. There are two ways to detect when to return:
+
+- Call the `input_is_ready` method periodically. Quit when this returns `True`.
+
+- Add the `fileno` as a watch to the external eventloop. Quit when file descriptor
+  becomes readable. (But don't read from it.)
+
+  Note that this is not the same as checking for `sys.stdin.fileno()`. The
+  eventloop of prompt-toolkit allows thread-based executors, for example for
+  asynchronous autocompletion. When the completion for instance is ready, we
+  also want prompt-toolkit to gain control again in order to display that.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import os
+import select
+import selectors
+import sys
+import threading
+from asyncio import AbstractEventLoop, get_running_loop
+from selectors import BaseSelector, SelectorKey
+from typing import TYPE_CHECKING, Any, Callable, Mapping
+
+__all__ = [
+    "new_eventloop_with_inputhook",
+    "set_eventloop_with_inputhook",
+    "InputHookSelector",
+    "InputHookContext",
+    "InputHook",
+]
+
+if TYPE_CHECKING:
+    from _typeshed import FileDescriptorLike
+    from typing_extensions import TypeAlias
+
+    _EventMask = int
+
+
+class InputHookContext:
+    """
+    Given as a parameter to the inputhook.
+    """
+
+    def __init__(self, fileno: int, input_is_ready: Callable[[], bool]) -> None:
+        self._fileno = fileno
+        self.input_is_ready = input_is_ready
+
+    def fileno(self) -> int:
+        return self._fileno
+
+
+InputHook: TypeAlias = Callable[[InputHookContext], None]
+
+
+def new_eventloop_with_inputhook(
+    inputhook: Callable[[InputHookContext], None],
+) -> AbstractEventLoop:
+    """
+    Create a new event loop with the given inputhook.
+    """
+    selector = InputHookSelector(selectors.DefaultSelector(), inputhook)
+    loop = asyncio.SelectorEventLoop(selector)
+    return loop
+
+
+def set_eventloop_with_inputhook(
+    inputhook: Callable[[InputHookContext], None],
+) -> AbstractEventLoop:
+    """
+    Create a new event loop with the given inputhook, and activate it.
+    """
+    # Deprecated!
+
+    loop = new_eventloop_with_inputhook(inputhook)
+    asyncio.set_event_loop(loop)
+    return loop
+
+
+class InputHookSelector(BaseSelector):
+    """
+    Usage:
+
+        selector = selectors.SelectSelector()
+        loop = asyncio.SelectorEventLoop(InputHookSelector(selector, inputhook))
+        asyncio.set_event_loop(loop)
+    """
+
+    def __init__(
+        self, selector: BaseSelector, inputhook: Callable[[InputHookContext], None]
+    ) -> None:
+        self.selector = selector
+        self.inputhook = inputhook
+        self._r, self._w = os.pipe()
+
+    def register(
+        self, fileobj: FileDescriptorLike, events: _EventMask, data: Any = None
+    ) -> SelectorKey:
+        return self.selector.register(fileobj, events, data=data)
+
+    def unregister(self, fileobj: FileDescriptorLike) -> SelectorKey:
+        return self.selector.unregister(fileobj)
+
+    def modify(
+        self, fileobj: FileDescriptorLike, events: _EventMask, data: Any = None
+    ) -> SelectorKey:
+        return self.selector.modify(fileobj, events, data=None)
+
+    def select(
+        self, timeout: float | None = None
+    ) -> list[tuple[SelectorKey, _EventMask]]:
+        # If there are tasks in the current event loop,
+        # don't run the input hook.
+        if len(getattr(get_running_loop(), "_ready", [])) > 0:
+            return self.selector.select(timeout=timeout)
+
+        ready = False
+        result = None
+
+        # Run selector in other thread.
+        def run_selector() -> None:
+            nonlocal ready, result
+            result = self.selector.select(timeout=timeout)
+            os.write(self._w, b"x")
+            ready = True
+
+        th = threading.Thread(target=run_selector)
+        th.start()
+
+        def input_is_ready() -> bool:
+            return ready
+
+        # Call inputhook.
+        # The inputhook function is supposed to return when our selector
+        # becomes ready. The inputhook can do that by registering the fd in its
+        # own loop, or by checking the `input_is_ready` function regularly.
+        self.inputhook(InputHookContext(self._r, input_is_ready))
+
+        # Flush the read end of the pipe.
+        try:
+            # Before calling 'os.read', call select.select. This is required
+            # when the gevent monkey patch has been applied. 'os.read' is never
+            # monkey patched and won't be cooperative, so that would block all
+            # other select() calls otherwise.
+            # See: http://www.gevent.org/gevent.os.html
+
+            # Note: On Windows, this is apparently not an issue.
+            #       However, if we would ever want to add a select call, it
+            #       should use `windll.kernel32.WaitForMultipleObjects`,
+            #       because `select.select` can't wait for a pipe on Windows.
+            if sys.platform != "win32":
+                select.select([self._r], [], [], None)
+
+            os.read(self._r, 1024)
+        except OSError:
+            # This happens when the window resizes and a SIGWINCH was received.
+            # We get 'Error: [Errno 4] Interrupted system call'
+            # Just ignore.
+            pass
+
+        # Wait for the real selector to be done.
+        th.join()
+        assert result is not None
+        return result
+
+    def close(self) -> None:
+        """
+        Clean up resources.
+        """
+        if self._r:
+            os.close(self._r)
+            os.close(self._w)
+
+        self._r = self._w = -1
+        self.selector.close()
+
+    def get_map(self) -> Mapping[FileDescriptorLike, SelectorKey]:
+        return self.selector.get_map()

.venv/lib/python3.12/site-packages/prompt_toolkit/eventloop/utils.py

@@ -0,0 +1,101 @@
+from __future__ import annotations
+
+import asyncio
+import contextvars
+import sys
+import time
+from asyncio import get_running_loop
+from types import TracebackType
+from typing import Any, Awaitable, Callable, TypeVar, cast
+
+__all__ = [
+    "run_in_executor_with_context",
+    "call_soon_threadsafe",
+    "get_traceback_from_context",
+]
+
+_T = TypeVar("_T")
+
+
+def run_in_executor_with_context(
+    func: Callable[..., _T],
+    *args: Any,
+    loop: asyncio.AbstractEventLoop | None = None,
+) -> Awaitable[_T]:
+    """
+    Run a function in an executor, but make sure it uses the same contextvars.
+    This is required so that the function will see the right application.
+
+    See also: https://bugs.python.org/issue34014
+    """
+    loop = loop or get_running_loop()
+    ctx: contextvars.Context = contextvars.copy_context()
+
+    return loop.run_in_executor(None, ctx.run, func, *args)
+
+
+def call_soon_threadsafe(
+    func: Callable[[], None],
+    max_postpone_time: float | None = None,
+    loop: asyncio.AbstractEventLoop | None = None,
+) -> None:
+    """
+    Wrapper around asyncio's `call_soon_threadsafe`.
+
+    This takes a `max_postpone_time` which can be used to tune the urgency of
+    the method.
+
+    Asyncio runs tasks in first-in-first-out. However, this is not what we
+    want for the render function of the prompt_toolkit UI. Rendering is
+    expensive, but since the UI is invalidated very often, in some situations
+    we render the UI too often, so much that the rendering CPU usage slows down
+    the rest of the processing of the application.  (Pymux is an example where
+    we have to balance the CPU time spend on rendering the UI, and parsing
+    process output.)
+    However, we want to set a deadline value, for when the rendering should
+    happen. (The UI should stay responsive).
+    """
+    loop2 = loop or get_running_loop()
+
+    # If no `max_postpone_time` has been given, schedule right now.
+    if max_postpone_time is None:
+        loop2.call_soon_threadsafe(func)
+        return
+
+    max_postpone_until = time.time() + max_postpone_time
+
+    def schedule() -> None:
+        # When there are no other tasks scheduled in the event loop. Run it
+        # now.
+        # Notice: uvloop doesn't have this _ready attribute. In that case,
+        #         always call immediately.
+        if not getattr(loop2, "_ready", []):
+            func()
+            return
+
+        # If the timeout expired, run this now.
+        if time.time() > max_postpone_until:
+            func()
+            return
+
+        # Schedule again for later.
+        loop2.call_soon_threadsafe(schedule)
+
+    loop2.call_soon_threadsafe(schedule)
+
+
+def get_traceback_from_context(context: dict[str, Any]) -> TracebackType | None:
+    """
+    Get the traceback object from the context.
+    """
+    exception = context.get("exception")
+    if exception:
+        if hasattr(exception, "__traceback__"):
+            return cast(TracebackType, exception.__traceback__)
+        else:
+            # call_exception_handler() is usually called indirectly
+            # from an except block. If it's not the case, the traceback
+            # is undefined...
+            return sys.exc_info()[2]
+
+    return None

.venv/lib/python3.12/site-packages/prompt_toolkit/eventloop/win32.py

@@ -0,0 +1,72 @@
+from __future__ import annotations
+
+import sys
+
+assert sys.platform == "win32"
+
+from ctypes import pointer
+
+from ..utils import SPHINX_AUTODOC_RUNNING
+
+# Do not import win32-specific stuff when generating documentation.
+# Otherwise RTD would be unable to generate docs for this module.
+if not SPHINX_AUTODOC_RUNNING:
+    from ctypes import windll
+
+from ctypes.wintypes import BOOL, DWORD, HANDLE
+
+from prompt_toolkit.win32_types import SECURITY_ATTRIBUTES
+
+__all__ = ["wait_for_handles", "create_win32_event"]
+
+
+WAIT_TIMEOUT = 0x00000102
+INFINITE = -1
+
+
+def wait_for_handles(handles: list[HANDLE], timeout: int = INFINITE) -> HANDLE | None:
+    """
+    Waits for multiple handles. (Similar to 'select') Returns the handle which is ready.
+    Returns `None` on timeout.
+    http://msdn.microsoft.com/en-us/library/windows/desktop/ms687025(v=vs.85).aspx
+
+    Note that handles should be a list of `HANDLE` objects, not integers. See
+    this comment in the patch by @quark-zju for the reason why:
+
+        ''' Make sure HANDLE on Windows has a correct size
+
+        Previously, the type of various HANDLEs are native Python integer
+        types. The ctypes library will treat them as 4-byte integer when used
+        in function arguments. On 64-bit Windows, HANDLE is 8-byte and usually
+        a small integer. Depending on whether the extra 4 bytes are zero-ed out
+        or not, things can happen to work, or break. '''
+
+    This function returns either `None` or one of the given `HANDLE` objects.
+    (The return value can be tested with the `is` operator.)
+    """
+    arrtype = HANDLE * len(handles)
+    handle_array = arrtype(*handles)
+
+    ret: int = windll.kernel32.WaitForMultipleObjects(
+        len(handle_array), handle_array, BOOL(False), DWORD(timeout)
+    )
+
+    if ret == WAIT_TIMEOUT:
+        return None
+    else:
+        return handles[ret]
+
+
+def create_win32_event() -> HANDLE:
+    """
+    Creates a Win32 unnamed Event .
+    http://msdn.microsoft.com/en-us/library/windows/desktop/ms682396(v=vs.85).aspx
+    """
+    return HANDLE(
+        windll.kernel32.CreateEventA(
+            pointer(SECURITY_ATTRIBUTES()),
+            BOOL(True),  # Manual reset event.
+            BOOL(False),  # Initial state.
+            None,  # Unnamed event object.
+        )
+    )