{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Python 生成器与迭代器协议 (深入) 教程\n",
    "\n",
    "欢迎来到 Python 生成器与迭代器协议的深入教程！迭代是 Python 中一个非常核心的概念，理解其背后的迭代器协议以及强大的生成器机制，可以帮助你编写出更高效、内存友好且富有表现力的代码。\n",
    "\n",
    "**为什么深入学习迭代和生成器？**\n",
    "\n",
    "1.  **内存效率**：生成器允许按需生成值，而不是一次性在内存中创建整个序列，这对于处理大型数据集或无限序列至关重要。\n",
    "2.  **惰性求值 (Lazy Evaluation)**：值仅在需要时才被计算，可以节省计算资源。\n",
    "3.  **代码简洁性**：生成器提供了一种简洁的方式来创建迭代器。\n",
    "4.  **构建数据处理管道**：可以轻松地将多个生成器链接起来，形成高效的数据处理流。\n",
    "5.  **理解 Python 核心**：迭代协议是 `for` 循环、列表推导式、`map()`, `filter()` 等许多 Python 特性的基础。\n",
    "\n",
    "**本教程将涵盖：**\n",
    "\n",
    "1.  **迭代协议 (Iterator Protocol)**：`__iter__` 和 `__next__`。\n",
    "2.  **可迭代对象 (Iterable) vs 迭代器 (Iterator)**。\n",
    "3.  **生成器函数 (Generator Functions)**：使用 `yield` 关键字。\n",
    "4.  **生成器表达式 (Generator Expressions)**。\n",
    "5.  **`itertools` 模块**：强大的迭代工具。\n",
    "6.  **`yield from` 语句** (Python 3.3+)。\n",
    "7.  **生成器的高级特性**：`send()`, `throw()`, `close()` 方法 (传统协程基础)。\n",
    "8.  **应用场景与最佳实践**。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. 迭代协议 (Iterator Protocol)\n",
    "\n",
    "Python 的迭代协议定义了对象如何支持迭代。它依赖于两个核心的魔术方法：\n",
    "\n",
    "*   **`__iter__(self)`**：\n",
    "    *   当一个对象被传递给 `iter()` 内置函数时，或者当 `for` 循环开始时，会调用该对象的 `__iter__` 方法。\n",
    "    *   它必须返回一个**迭代器对象**。\n",
    "\n",
    "*   **`__next__(self)`**：\n",
    "    *   迭代器对象必须实现这个方法。\n",
    "    *   当调用 `next(iterator)` 内置函数时（`for` 循环在每次迭代时隐式调用它），会调用迭代器的 `__next__` 方法。\n",
    "    *   它应该返回序列中的下一个值。\n",
    "    *   当没有更多值可以返回时，它必须引发 `StopIteration` 异常。`for` 循环会自动捕获这个异常并终止循环。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. 可迭代对象 (Iterable) vs 迭代器 (Iterator)\n",
    "\n",
    "*   **可迭代对象 (Iterable)**：\n",
    "    *   任何实现了 `__iter__` 方法（返回一个迭代器）的对象都是可迭代的。\n",
    "    *   或者，如果一个对象实现了 `__getitem__` 方法并且可以从索引 0 开始接受整数参数（如序列），它也是可迭代的 (Python 会自动创建一个迭代器来遍历它)。\n",
    "    *   例子：列表 (`list`)、元组 (`tuple`)、字符串 (`str`)、字典 (`dict`)、集合 (`set`)、文件对象、自定义类（实现了 `__iter__` 或 `__getitem__`）。\n",
    "    *   你可以对一个可迭代对象多次调用 `iter()` 来获取新的迭代器，每个迭代器独立地遍历数据。\n",
    "\n",
    "*   **迭代器 (Iterator)**：\n",
    "    *   任何实现了 `__iter__` 方法和 `__next__` 方法的对象都是迭代器。\n",
    "    *   `__iter__` 方法对于迭代器来说，通常只需要返回 `self` (因为迭代器本身就是自己的迭代器)。\n",
    "    *   迭代器是有状态的：它们记住在迭代过程中的当前位置。\n",
    "    *   迭代器通常只能遍历一次。一旦 `__next__` 引发 `StopIteration`，它将继续引发该异常。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 示例：自定义一个可迭代对象和迭代器\n",
    "class MyRangeIterable:\n",
    "    \"\"\"一个简单的可迭代对象，类似于 range()\"\"\"\n",
    "    def __init__(self, start, end):\n",
    "        self.start = start\n",
    "        self.end = end\n",
    "        print(f\"MyRangeIterable initialized ({self.start} to {self.end})\")\n",
    "\n",
    "    def __iter__(self):\n",
    "        print(\"MyRangeIterable.__iter__ called, returning MyRangeIterator\")\n",
    "        # 返回一个新的迭代器实例\n",
    "        return MyRangeIterator(self.start, self.end)\n",
    "\n",
    "class MyRangeIterator:\n",
    "    \"\"\"一个迭代器，用于 MyRangeIterable\"\"\"\n",
    "    def __init__(self, start, end):\n",
    "        self.current = start\n",
    "        self.end = end\n",
    "        print(f\"MyRangeIterator initialized (current={self.current}, end={self.end})\")\n",
    "\n",
    "    def __iter__(self):\n",
    "        # 迭代器自身的 __iter__ 方法应该返回 self\n",
    "        print(\"MyRangeIterator.__iter__ called, returning self\")\n",
    "        return self\n",
    "\n",
    "    def __next__(self):\n",
    "        print(f\"MyRangeIterator.__next__ called (current={self.current})\")\n",
    "        if self.current < self.end:\n",
    "            value = self.current\n",
    "            self.current += 1\n",
    "            return value\n",
    "        else:\n",
    "            print(\"MyRangeIterator: Raising StopIteration\")\n",
    "            raise StopIteration\n",
    "\n",
    "print(\"--- Testing MyRangeIterable ---\")\n",
    "my_range_obj = MyRangeIterable(1, 4) # 可迭代对象\n",
    "\n",
    "print(\"\\nFirst iteration using for loop:\")\n",
    "for num in my_range_obj: # 隐式调用 iter(my_range_obj) 然后 next()\n",
    "    print(f\"  For loop got: {num}\")\n",
    "\n",
    "print(\"\\nSecond iteration using for loop (gets a new iterator):\")\n",
    "for num in my_range_obj:\n",
    "    print(f\"  For loop got: {num}\")\n",
    "\n",
    "print(\"\\nManual iteration:\")\n",
    "iterator1 = iter(my_range_obj) # 获取一个迭代器\n",
    "print(f\"Type of iterator1: {type(iterator1)}\")\n",
    "print(f\"next(iterator1): {next(iterator1)}\")\n",
    "print(f\"next(iterator1): {next(iterator1)}\")\n",
    "\n",
    "iterator2 = iter(my_range_obj) # 获取另一个独立的迭代器\n",
    "print(f\"next(iterator2): {next(iterator2)}\") # 从头开始\n",
    "\n",
    "print(f\"Continuing iterator1: {next(iterator1)}\")\n",
    "try:\n",
    "    print(f\"Continuing iterator1 (expect StopIteration): {next(iterator1)}\")\n",
    "except StopIteration as e:\n",
    "    print(f\"  Caught StopIteration as expected: {e}\")\n",
    "\n",
    "# 验证迭代器也是可迭代的\n",
    "iter_from_iter = iter(iterator2)\n",
    "print(f\"iterator2 is iter_from_iter: {iterator2 is iter_from_iter}\") # True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. 生成器函数 (Generator Functions)\n",
    "\n",
    "生成器函数是一种特殊的函数，它不使用 `return` 返回一个值，而是使用 `yield` 关键字“产生”一系列值。\n",
    "\n",
    "*   当调用一个生成器函数时，它**不会立即执行函数体**，而是返回一个**生成器对象 (generator object)**。\n",
    "*   生成器对象是一种特殊的迭代器：它自动实现了 `__iter__` 和 `__next__` 方法。\n",
    "*   每次在生成器对象上调用 `next()` 时，函数会从上次 `yield` 语句离开的地方继续执行，直到遇到下一个 `yield` 语句。\n",
    "*   `yield` 语句会“产生”一个值给调用者，并暂停函数的执行状态（包括局部变量）。\n",
    "*   当函数执行完毕（没有更多 `yield` 或遇到 `return` 语句，或正常退出）时，会自动引发 `StopIteration`。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def simple_generator_func(n):\n",
    "    print(\"Generator function: simple_generator_func called\")\n",
    "    i = 0\n",
    "    while i < n:\n",
    "        print(f\"Generator: yielding {i}\")\n",
    "        yield i # 产生值并暂停\n",
    "        i += 1\n",
    "        print(f\"Generator: resumed, i is now {i}\")\n",
    "    print(\"Generator: finished\")\n",
    "    # 隐式 StopIteration\n",
    "\n",
    "print(\"--- Testing simple_generator_func ---\")\n",
    "gen_obj = simple_generator_func(3) # 调用生成器函数，返回生成器对象\n",
    "print(f\"Type of gen_obj: {type(gen_obj)}\") # <class 'generator'>\n",
    "\n",
    "print(f\"\\nFirst next(gen_obj): {next(gen_obj)}\") # 开始执行，直到第一个yield\n",
    "print(f\"Second next(gen_obj): {next(gen_obj)}\")\n",
    "print(f\"Third next(gen_obj): {next(gen_obj)}\")\n",
    "try:\n",
    "    print(f\"Fourth next(gen_obj) (expect StopIteration): {next(gen_obj)}\")\n",
    "except StopIteration:\n",
    "    print(\"  Caught StopIteration as expected.\")\n",
    "\n",
    "print(\"\\nIterating with a for loop (uses a new generator object):\")\n",
    "for val in simple_generator_func(2):\n",
    "    print(f\"  For loop got: {val}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**生成器的优点：**\n",
    "*   **代码简洁**：创建迭代器的逻辑（状态管理、`StopIteration`）由 Python 自动处理。\n",
    "*   **内存高效**：值是按需生成的，适合处理大数据集或无限序列。\n",
    "\n",
    "**无限序列示例：**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fibonacci_generator():\n",
    "    \"\"\"生成一个无限的斐波那契数列。\"\"\"\n",
    "    a, b = 0, 1\n",
    "    while True:\n",
    "        yield a\n",
    "        a, b = b, a + b\n",
    "\n",
    "print(\"--- Fibonacci Generator ---\")\n",
    "fib_gen = fibonacci_generator()\n",
    "print(\"First 10 Fibonacci numbers:\")\n",
    "for _ in range(10):\n",
    "    print(next(fib_gen), end=\" \")\n",
    "print(\"\\n\")\n",
    "\n",
    "# 如果你想从头开始，需要重新创建生成器对象\n",
    "fib_gen2 = fibonacci_generator()\n",
    "print(f\"Next from fib_gen2: {next(fib_gen2)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. 生成器表达式 (Generator Expressions)\n",
    "\n",
    "生成器表达式提供了一种更简洁的方式来创建简单的生成器对象，其语法类似于列表推导式，但使用圆括号 `()` 而不是方括号 `[]`。\n",
    "\n",
    "`(expression for item in iterable if condition)`\n",
    "\n",
    "*   生成器表达式也返回一个生成器对象。\n",
    "*   它们也是惰性求值的，按需生成值。\n",
    "*   非常适合作为函数参数传递，尤其是当你不希望立即创建整个列表时。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "squares_list_comp = [x*x for x in range(5)] # 列表推导式，立即创建列表\n",
    "squares_gen_expr = (x*x for x in range(5))  # 生成器表达式，返回生成器对象\n",
    "\n",
    "print(f\"List comprehension: {squares_list_comp}, type: {type(squares_list_comp)}\")\n",
    "print(f\"Generator expression: {squares_gen_expr}, type: {type(squares_gen_expr)}\")\n",
    "\n",
    "print(\"\\nIterating over generator expression:\")\n",
    "for sq in squares_gen_expr:\n",
    "    print(sq, end=\" \")\n",
    "print(\"\\n\")\n",
    "\n",
    "# 再次迭代会发现它已经耗尽 (因为生成器是一次性的)\n",
    "print(\"Trying to iterate again (should be empty):\")\n",
    "for sq in squares_gen_expr: \n",
    "    print(sq, end=\" \") # 不会有输出\n",
    "print(\"\\n\")\n",
    "\n",
    "# 作为函数参数\n",
    "data = [1, 2, 3, 4, 5, 6]\n",
    "sum_of_even_squares = sum(x*x for x in data if x % 2 == 0)\n",
    "# 上面的 sum() 直接消耗了生成器表达式产生的值，没有创建中间列表\n",
    "print(f\"Sum of even squares: {sum_of_even_squares}\")\n",
    "\n",
    "# 如果生成器表达式是函数调用的唯一参数，可以省略外层圆括号\n",
    "sum_of_cubes = sum(x**3 for x in range(1, 4))\n",
    "print(f\"Sum of cubes (1,2,3): {sum_of_cubes}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. `itertools` 模块\n",
    "\n",
    "`itertools` 模块包含一系列用于创建高效迭代器的函数。这些函数受到 APL, Haskell, SML 等函数式编程语言中类似构造的启发。\n",
    "\n",
    "**一些常用的 `itertools` 函数：**\n",
    "\n",
    "*   **无限迭代器：**\n",
    "    *   `count(start=0, step=1)`: 从 `start` 开始，以 `step` 递增的无限序列。\n",
    "    *   `cycle(iterable)`: 无限重复 `iterable` 中的元素。\n",
    "    *   `repeat(object[, times])`: 重复 `object`，可以指定次数，否则无限重复。\n",
    "\n",
    "*   **处理有限序列的迭代器：**\n",
    "    *   `accumulate(iterable[, func, *, initial=None])`: 返回累积的总和（或其他二元函数的结果）。\n",
    "    *   `chain(*iterables)`: 将多个可迭代对象连接成一个序列。\n",
    "    *   `compress(data, selectors)`: 根据 `selectors` 中的真值过滤 `data` 中的元素。\n",
    "    *   `dropwhile(predicate, iterable)`: 当 `predicate` 为真时，跳过 `iterable` 中的元素，然后返回剩余所有元素。\n",
    "    *   `filterfalse(predicate, iterable)`: 返回 `iterable` 中 `predicate` 为假的元素。\n",
    "    *   `groupby(iterable, key=None)`: 将连续的具有相同键值（由 `key` 函数确定）的元素分组。\n",
    "    *   `islice(iterable, stop)` 或 `islice(iterable, start, stop[, step])`: 返回 `iterable` 的一个切片，类似于列表切片，但返回迭代器。\n",
    "    *   `starmap(function, iterable)`: 类似于 `map`，但 `iterable` 中的每个元素是一个元组，会解包作为 `function` 的参数。\n",
    "    *   `takewhile(predicate, iterable)`: 只要 `predicate` 为真，就从 `iterable` 中返回元素。\n",
    "    *   `tee(iterable, n=2)`: 返回 `n` 个独立的迭代器，它们都从同一个原始 `iterable` 中获取元素。\n",
    "    *   `zip_longest(*iterables, fillvalue=None)`: 类似于 `zip`，但会用 `fillvalue` 填充最短的迭代器，直到所有迭代器耗尽。\n",
    "\n",
    "*   **组合生成器：**\n",
    "    *   `product(*iterables, repeat=1)`: 笛卡尔积。\n",
    "    *   `permutations(iterable, r=None)`: 排列。\n",
    "    *   `combinations(iterable, r)`: 组合。\n",
    "    *   `combinations_with_replacement(iterable, r)`: 可重复组合。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import itertools\n",
    "\n",
    "print(\"--- itertools.count --- \")\n",
    "counter = itertools.count(10, 2)\n",
    "for _ in range(5):\n",
    "    print(next(counter), end=\" \") # 10 12 14 16 18\n",
    "print(\"\\n\")\n",
    "\n",
    "print(\"--- itertools.cycle --- \")\n",
    "cycler = itertools.cycle(\"ABC\")\n",
    "for _ in range(7):\n",
    "    print(next(cycler), end=\" \") # A B C A B C A\n",
    "print(\"\\n\")\n",
    "\n",
    "print(\"--- itertools.chain --- \")\n",
    "chained = itertools.chain([1, 2], \"XY\", (3, 4))\n",
    "print(list(chained)) # [1, 2, 'X', 'Y', 3, 4]\n",
    "\n",
    "print(\"--- itertools.islice --- \")\n",
    "sliced = itertools.islice(range(10), 2, 8, 2) # 从索引2到8 (不含)，步长2\n",
    "print(list(sliced)) # [2, 4, 6]\n",
    "\n",
    "print(\"--- itertools.groupby --- \")\n",
    "data = \"AAABBCDAA\"\n",
    "for key, group in itertools.groupby(data):\n",
    "    print(f\"Key: {key}, Group: {list(group)}\")\n",
    "# Key: A, Group: ['A', 'A', 'A']\n",
    "# Key: B, Group: ['B', 'B']\n",
    "# Key: C, Group: ['C']\n",
    "# Key: D, Group: ['D']\n",
    "# Key: A, Group: ['A', 'A']\n",
    "\n",
    "print(\"--- itertools.combinations --- \")\n",
    "combs = itertools.combinations(\"ABC\", 2)\n",
    "print(list(combs)) # [('A', 'B'), ('A', 'C'), ('B', 'C')]\n",
    "\n",
    "print(\"--- itertools.product --- \")\n",
    "prod = itertools.product(\"AB\", \"12\")\n",
    "print(list(prod)) # [('A', '1'), ('A', '2'), ('B', '1'), ('B', '2')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. `yield from` 语句 (Python 3.3+)\n",
    "\n",
    "`yield from <iterable>` 语句允许一个生成器将其部分操作委托给另一个可迭代对象 (通常是另一个生成器)。\n",
    "\n",
    "它主要做了以下事情：\n",
    "1.  迭代 `<iterable>`。\n",
    "2.  将从 `<iterable>` 中产生的每个值直接传递给当前生成器的调用者。\n",
    "3.  如果 `<iterable>` 本身是一个生成器，`yield from` 还会处理子生成器可能通过 `send()`, `throw()`, `close()` 接收到的值或异常，并将它们传递给子生成器。\n",
    "\n",
    "**用途：**\n",
    "*   **简化生成器嵌套**：避免写很多 `for item in sub_generator: yield item` 这样的代码。\n",
    "*   **构建协程管道** (虽然现代异步编程更多使用 `async/await`)。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sub_generator(start, end):\n",
    "    print(f\"  sub_generator: called with {start}, {end}\")\n",
    "    for i in range(start, end):\n",
    "        print(f\"  sub_generator: yielding {i}\")\n",
    "        yield i\n",
    "    print(\"  sub_generator: finished\")\n",
    "\n",
    "def delegating_generator_manual(iterables_list):\n",
    "    print(\"delegating_generator_manual: called\")\n",
    "    for iterable in iterables_list:\n",
    "        for item in iterable: # 手动迭代子可迭代对象\n",
    "            yield item\n",
    "    print(\"delegating_generator_manual: finished\")\n",
    "\n",
    "def delegating_generator_yield_from(iterables_list):\n",
    "    print(\"delegating_generator_yield_from: called\")\n",
    "    for iterable in iterables_list:\n",
    "        # 使用 yield from 委托给子可迭代对象\n",
    "        # 如果 iterable 是一个生成器，yield from 会建立一个双向通道\n",
    "        yield from iterable \n",
    "    print(\"delegating_generator_yield_from: finished\")\n",
    "\n",
    "print(\"--- Testing yield from ---\")\n",
    "data_sources = [\n",
    "    sub_generator(1, 3),    # 一个生成器\n",
    "    \"XY\",                 # 一个字符串 (可迭代)\n",
    "    (10, 11)              # 一个元组 (可迭代)\n",
    "]\n",
    "\n",
    "print(\"\\nUsing manual delegation:\")\n",
    "for item in delegating_generator_manual(list(data_sources)): # list() to consume sub_generator once\n",
    "    print(f\"Got item: {item}\")\n",
    "\n",
    "# 重新创建 data_sources 因为生成器会被消耗\n",
    "data_sources_2 = [\n",
    "    sub_generator(1, 3),\n",
    "    \"XY\",\n",
    "    (10, 11)\n",
    "]\n",
    "print(\"\\nUsing yield from:\")\n",
    "for item in delegating_generator_yield_from(data_sources_2):\n",
    "    print(f\"Got item: {item}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. 生成器的高级特性：`send()`, `throw()`, `close()`\n",
    "\n",
    "除了通过 `next()` 从生成器获取值，还可以向生成器发送值或异常，或者关闭它。这些特性使得生成器可以用作简单的**协程 (coroutine)** (这是 `async/await` 出现之前的协程概念)。\n",
    "\n",
    "*   **`generator.send(value)`**：\n",
    "    *   向生成器发送一个值，这个值会成为当前 `yield` 表达式的结果。\n",
    "    *   生成器会从暂停处恢复执行，直到遇到下一个 `yield` (产生一个值) 或终止。\n",
    "    *   在首次启动生成器时（即在第一次 `yield` 之前），必须发送 `None` (或者直接调用 `next(generator)`)。\n",
    "\n",
    "*   **`generator.throw(type[, value[, traceback]])`**：\n",
    "    *   在生成器暂停的地方（`yield` 表达式处）引发一个异常。\n",
    "    *   如果生成器内部捕获了这个异常，它可以继续执行并 `yield` 一个值，或者正常退出（引发 `StopIteration`），或者引发另一个异常。\n",
    "    *   如果生成器未捕获该异常，异常会传播给调用者。\n",
    "\n",
    "*   **`generator.close()`**：\n",
    "    *   在生成器暂停的地方引发一个 `GeneratorExit` 异常。\n",
    "    *   生成器通常应该捕获 `GeneratorExit`，执行清理操作，然后要么重新引发 `GeneratorExit`，要么引发 `StopIteration`，要么正常退出。\n",
    "    *   调用 `close()` 后，如果生成器尝试 `yield` 一个值，会引发 `RuntimeError`。\n",
    "    *   `close()` 之后的 `next()` 或 `send()` 调用也会引发 `StopIteration`。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def simple_coroutine():\n",
    "    print(\"Coroutine started\")\n",
    "    received_value = None\n",
    "    try:\n",
    "        while True:\n",
    "            received_value = yield received_value # yield 表达式的值是 send() 过来的值\n",
    "            print(f\"Coroutine received: {received_value}\")\n",
    "            if received_value == \"exit\":\n",
    "                print(\"Coroutine exiting normally\")\n",
    "                break\n",
    "            received_value = f\"Processed: {received_value}\"\n",
    "    except GeneratorExit:\n",
    "        print(\"Coroutine: Caught GeneratorExit, cleaning up...\")\n",
    "        # 执行清理操作\n",
    "        print(\"Coroutine: Cleaned up and closing.\")\n",
    "        # 不应再 yield 值，可以重新引发 GeneratorExit 或 StopIteration，或直接返回\n",
    "    except ValueError as e:\n",
    "        print(f\"Coroutine: Caught ValueError: {e}\")\n",
    "        yield f\"Error handled: {e}\" # 可以选择 yield 一个错误处理结果\n",
    "    finally:\n",
    "        print(\"Coroutine finally block executed\")\n",
    "\n",
    "print(\"--- Testing Coroutine send() ---\")\n",
    "co = simple_coroutine()\n",
    "next(co) # 启动协程，执行到第一个 yield，此时 received_value 为 None\n",
    "print(f\"Sent 10, got back: {co.send(10)}\")      # 发送 10, yield 返回 'Processed: 10'\n",
    "print(f\"Sent 'hello', got back: {co.send('hello')}\") # 发送 'hello', yield 返回 'Processed: hello'\n",
    "\n",
    "print(\"\\n--- Testing Coroutine throw() ---\")\n",
    "co2 = simple_coroutine()\n",
    "next(co2)\n",
    "try:\n",
    "    print(f\"Throwing ValueError, got back: {co2.throw(ValueError, 'Test error')}\")\n",
    "except ValueError as e:\n",
    "    print(f\"Caller caught an unhandled error from coroutine: {e}\") # 如果协程不处理并重新抛出\n",
    "\n",
    "print(f\"Sending 'after error' to co2, got back: {co2.send('after error')}\") # 协程可能已处理异常并继续\n",
    "\n",
    "print(\"\\n--- Testing Coroutine close() ---\")\n",
    "co3 = simple_coroutine()\n",
    "next(co3)\n",
    "co3.send(\"data before close\")\n",
    "co3.close() # 关闭协程，会引发 GeneratorExit\n",
    "\n",
    "try:\n",
    "    next(co3) # 尝试再次从已关闭的协程获取值\n",
    "except StopIteration:\n",
    "    print(\"Caught StopIteration after close, as expected.\")\n",
    "\n",
    "print(\"\\n--- Testing Coroutine exit command ---\")\n",
    "co4 = simple_coroutine()\n",
    "next(co4)\n",
    "co4.send(\"some data\")\n",
    "try:\n",
    "    co4.send(\"exit\") # 协程内部处理 exit 并正常结束\n",
    "except StopIteration:\n",
    "    print(\"Caught StopIteration after coroutine exited via 'exit' command.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "虽然 `async/await` 是现代 Python 中进行异步编程和协程的首选方式，但理解传统生成器协程的这些机制有助于理解 Python 异步历史以及某些库的底层实现。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 8. 应用场景与最佳实践\n",
    "\n",
    "**何时使用迭代器/生成器？**\n",
    "\n",
    "*   **处理大型数据集**：当数据无法一次性装入内存时（例如，读取大文件、数据库查询结果）。\n",
    "*   **无限序列**：如计数器、斐波那契数列、随机数流。\n",
    "*   **数据处理管道**：将多个生成器链接起来，以流式方式处理数据，每一步都是惰性的。\n",
    "    ```python\n",
    "    # lines = (line for line in open('large_file.txt'))\n",
    "    # non_empty_lines = (line for line in lines if line.strip())\n",
    "    # processed_lines = (process(line) for line in non_empty_lines)\n",
    "    # for result in processed_lines:\n",
    "    #     # ...\n",
    "    ```\n",
    "*   **需要自定义迭代行为的类**。\n",
    "*   **替代简单的列表推导式以节省内存**，如果结果列表很大且不需要立即全部使用。\n",
    "\n",
    "**最佳实践：**\n",
    "\n",
    "1.  **优先使用生成器表达式**：对于简单的惰性序列生成，生成器表达式最简洁。\n",
    "2.  **使用生成器函数**：当迭代逻辑复杂，需要多个 `yield` 或内部状态时。\n",
    "3.  **利用 `itertools`**：在自己动手实现复杂迭代逻辑之前，先看看 `itertools` 是否有现成的解决方案。\n",
    "4.  **理解迭代器是一次性的**：如果需要多次迭代，要么重新创建迭代器/生成器，要么将结果存储在列表中（如果内存允许）。\n",
    "5.  **`yield from` 可以使代码更扁平**：当委托给其他可迭代对象时。\n",
    "6.  **谨慎使用生成器的高级方法 (`send`, `throw`, `close`)**：它们引入了更复杂的控制流，对于大多数迭代场景是不必要的。现代异步编程应优先考虑 `async/await`。\n",
    "\n",
    "## 总结\n",
    "\n",
    "迭代器和生成器是 Python 中非常强大且基础的特性。它们不仅是许多内置功能（如 `for` 循环）的核心，还提供了一种优雅、高效的方式来处理数据流和序列。\n",
    "\n",
    "通过深入理解迭代协议、生成器函数、生成器表达式以及 `itertools` 模块，你可以编写出更 Pythonic、更高效、内存更友好的代码。"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}