Python yield与实现方法代码分析

yield的功能类似于return，但是不同之处在于它返回的是生成器。

生成器

生成器是通过一个或多个yield表达式构成的函数，每一个生成器都是一个迭代器（但是迭代器不一定是生成器）。

如果一个函数包含yield关键字，这个函数就会变为一个生成器。

生成器并不会一次返回所有结果，而是每次遇到yield关键字后返回相应结果，并保留函数当前的运行状态，等待下一次的调用。

由于生成器也是一个迭代器，那么它就应该支持next方法来获取下一个值。

基本操作

# 通过`yield`来创建生成器
def func():
for i in xrange(10);
 yield i

# 通过列表来创建生成器
[i for i in xrange(10)]
# 通过`yield`来创建生成器
def func():
for i in xrange(10);
 yield i
# 通过列表来创建生成器
[i for i in xrange(10)]
Python
# 调用如下
>>> f = func()
>>> f # 此时生成器还没有运行
<generator object func at 0x7fe01a853820>
>>> f.next() # 当i=0时，遇到yield关键字，直接返回
>>> f.next() # 继续上一次执行的位置，进入下一层循环
...
>>> f.next()
>>> f.next() # 当执行完最后一次循环后，结束yield语句，生成StopIteration异常
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
StopIteration
>>>
# 调用如下
>>> f = func()
>>> f # 此时生成器还没有运行
<generator object func at 0x7fe01a853820>
>>> f.next() # 当i=0时，遇到yield关键字，直接返回
>>> f.next() # 继续上一次执行的位置，进入下一层循环
...
>>> f.next()
>>> f.next() # 当执行完最后一次循环后，结束yield语句，生成StopIteration异常
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
StopIteration
>>>

除了next函数，生成器还支持send函数。该函数可以向生成器传递参数。

>>> def func():
...  n = 0
...  while 1:
...   n = yield n #可以通过send函数向n赋值
...
>>> f = func()
>>> f.next() # 默认情况下n为0
>>> f.send(1) #n赋值1
>>> f.send(2)
>>>
>>> def func():
...  n = 0
...  while 1:
...   n = yield n #可以通过send函数向n赋值
...
>>> f = func()
>>> f.next() # 默认情况下n为0
>>> f.send(1) #n赋值1
>>> f.send(2)
>>>

应用

最经典的例子，生成无限序列。

常规的解决方法是，生成一个满足要求的很大的列表，这个列表需要保存在内存中，很明显内存限制了这个问题。

def get_primes(start):
for element in magical_infinite_range(start):
 if is_prime(element):
  return element
def get_primes(start):
for element in magical_infinite_range(start):
 if is_prime(element):
  return element

如果使用生成器就不需要返回整个列表，每次都只是返回一个数据，避免了内存的限制问题。

def get_primes(number):
while True:
 if is_prime(number):
  yield number
 number += 1
def get_primes(number):
while True:
 if is_prime(number):
  yield number
 number += 1

生成器源码分析

生成器的源码在Objects/genobject.c。

调用栈

在解释生成器之前，需要讲解一下Python虚拟机的调用原理。

Python虚拟机有一个栈帧的调用栈，其中栈帧的是PyFrameObject，位于Include/frameobject.h。

typedef struct _frame {
PyObject_VAR_HEAD
struct _frame *f_back; /* previous frame, or NULL */
PyCodeObject *f_code; /* code segment */
PyObject *f_builtins; /* builtin symbol table (PyDictObject) */
PyObject *f_globals; /* global symbol table (PyDictObject) */
PyObject *f_locals;  /* local symbol table (any mapping) */
PyObject **f_valuestack; /* points after the last local */
/* Next free slot in f_valuestack. Frame creation sets to f_valuestack.
 Frame evaluation usually NULLs it, but a frame that yields sets it
 to the current stack top. */
PyObject **f_stacktop;
PyObject *f_trace;  /* Trace function */

/* If an exception is raised in this frame, the next three are used to
 * record the exception info (if any) originally in the thread state. See
 * comments before set_exc_info() -- it's not obvious.
 * Invariant: if _type is NULL, then so are _value and _traceback.
 * Desired invariant: all three are NULL, or all three are non-NULL. That
 * one isn't currently true, but "should be".
 */
PyObject *f_exc_type, *f_exc_value, *f_exc_traceback;

PyThreadState *f_tstate;
int f_lasti;  /* Last instruction if called */
/* Call PyFrame_GetLineNumber() instead of reading this field
 directly. As of 2.3 f_lineno is only valid when tracing is
 active (i.e. when f_trace is set). At other times we use
 PyCode_Addr2Line to calculate the line from the current
 bytecode index. */
int f_lineno;  /* Current line number */
int f_iblock;  /* index in f_blockstack */
PyTryBlock f_blockstack[CO_MAXBLOCKS]; /* for try and loop blocks */
PyObject *f_localsplus[1]; /* locals+stack, dynamically sized */
} PyFrameObject;
typedef struct _frame {
PyObject_VAR_HEAD
struct _frame *f_back; /* previous frame, or NULL */
PyCodeObject *f_code; /* code segment */
PyObject *f_builtins; /* builtin symbol table (PyDictObject) */
PyObject *f_globals; /* global symbol table (PyDictObject) */
PyObject *f_locals;  /* local symbol table (any mapping) */
PyObject **f_valuestack; /* points after the last local */
/* Next free slot in f_valuestack. Frame creation sets to f_valuestack.
 Frame evaluation usually NULLs it, but a frame that yields sets it
 to the current stack top. */
PyObject **f_stacktop;
PyObject *f_trace;  /* Trace function */
/* If an exception is raised in this frame, the next three are used to
 * record the exception info (if any) originally in the thread state. See
 * comments before set_exc_info() -- it's not obvious.
 * Invariant: if _type is NULL, then so are _value and _traceback.
 * Desired invariant: all three are NULL, or all three are non-NULL. That
 * one isn't currently true, but "should be".
 */
PyObject *f_exc_type, *f_exc_value, *f_exc_traceback;

PyThreadState *f_tstate;
int f_lasti;  /* Last instruction if called */
/* Call PyFrame_GetLineNumber() instead of reading this field
 directly. As of 2.3 f_lineno is only valid when tracing is
 active (i.e. when f_trace is set). At other times we use
 PyCode_Addr2Line to calculate the line from the current
 bytecode index. */
int f_lineno;  /* Current line number */
int f_iblock;  /* index in f_blockstack */
PyTryBlock f_blockstack[CO_MAXBLOCKS]; /* for try and loop blocks */
PyObject *f_localsplus[1]; /* locals+stack, dynamically sized */
} PyFrameObject;

栈帧保存了给出代码的的信息和上下文，其中包含最后执行的指令，全局和局部命名空间，异常状态等信息。f_valueblock保存了数据，b_blockstack保存了异常和循环控制方法。

举一个例子来说明，

def foo():
x = 1
def bar(y):
 z = y + 2 #
def foo():
x = 1
def bar(y):
 z = y + 2 #

那么，相应的调用栈如下，一个py文件，一个类，一个函数都是一个代码块，对应者一个Frame，保存着上下文环境以及字节码指令。

c ---------------------------
a | bar Frame     | -> block stack: []
l |  (newest)    | -> data stack: [1, 2]
l ---------------------------
| foo Frame     | -> block stack: []
s |       | -> data stack: [.bar at 0x10d389680>, 1]
t ---------------------------
a | main (module) Frame  | -> block stack: []
c |  (oldest)   | -> data stack: []
k ---------------------------

c ---------------------------
a | bar Frame     | -> block stack: []
l |  (newest)    | -> data stack: [1, 2]
l ---------------------------
| foo Frame     | -> block stack: []
s |       | -> data stack: [.bar at 0x10d389680>, 1]
t ---------------------------
a | main (module) Frame  | -> block stack: []
c |  (oldest)   | -> data stack: []
k ---------------------------

每一个栈帧都拥有自己的数据栈和block栈，独立的数据栈和block栈使得解释器可以中断和恢复栈帧（生成器正式利用这点）。

Python代码首先被编译为字节码，再由Python虚拟机来执行。一般来说，一条Python语句对应着多条字节码（由于每条字节码对应着一条C语句，而不是一个机器指令，所以不能按照字节码的数量来判断代码性能）。

调用dis模块可以分析字节码，

from dis import dis
dis(foo)
   0 LOAD_CONST    1 (1) # 加载常量1
   3 STORE_FAST    0 (x) # x赋值为1
  6 LOAD_CONST    2 (<code>) # 加载常量2
   9 MAKE_FUNCTION   0 # 创建函数
   12 STORE_FAST    1 (bar)
  15 LOAD_FAST    1 (bar)
   18 LOAD_FAST    0 (x)
   21 CALL_FUNCTION   1 # 调用函数
   24 RETURN_VALUE  </code>

from dis import dis
dis(foo)
   0 LOAD_CONST    1 (1) # 加载常量1
   3 STORE_FAST    0 (x) # x赋值为1
  6 LOAD_CONST    2 (<code>) # 加载常量2
   9 MAKE_FUNCTION   0 # 创建函数
   12 STORE_FAST    1 (bar)
  15 LOAD_FAST    1 (bar)
   18 LOAD_FAST    0 (x)
   21 CALL_FUNCTION   1 # 调用函数
   24 RETURN_VALUE  </code>

其中，

第一行为代码行号；
第二行为偏移地址；
第三行为字节码指令；
第四行为指令参数；
第五行为参数解释。

第一行为代码行号；
第二行为偏移地址；
第三行为字节码指令；
第四行为指令参数；
第五行为参数解释。

生成器源码分析

由了上面对于调用栈的理解，就可以很容易的明白生成器的具体实现。

生成器的源码位于object/genobject.c。

生成器的创建

PyObject *
PyGen_New(PyFrameObject *f)
{
PyGenObject *gen = PyObject_GC_New(PyGenObject, &PyGen_Type); # 创建生成器对象
if (gen == NULL) {
 Py_DECREF(f);
 return NULL;
}
gen->gi_frame = f; # 赋予代码块
Py_INCREF(f->f_code); # 引用计数+1
gen->gi_code = (PyObject *)(f->f_code);
gen->gi_running = 0; # 0表示为执行，也就是生成器的初始状态
gen->gi_weakreflist = NULL;
_PyObject_GC_TRACK(gen); # GC跟踪
return (PyObject *)gen;
}

PyObject *
PyGen_New(PyFrameObject *f)
{
PyGenObject *gen = PyObject_GC_New(PyGenObject, &PyGen_Type); # 创建生成器对象
if (gen == NULL) {
 Py_DECREF(f);
 return NULL;
}
gen->gi_frame = f; # 赋予代码块
Py_INCREF(f->f_code); # 引用计数+1
gen->gi_code = (PyObject *)(f->f_code);
gen->gi_running = 0; # 0表示为执行，也就是生成器的初始状态
gen->gi_weakreflist = NULL;
_PyObject_GC_TRACK(gen); # GC跟踪
return (PyObject *)gen;
}

send与next

next与send函数，如下

static PyObject *
gen_iternext(PyGenObject *gen)
{
return gen_send_ex(gen, NULL, 0);
}
static PyObject *
gen_send(PyGenObject *gen, PyObject *arg)
{
return gen_send_ex(gen, arg, 0);
}

static PyObject *
gen_iternext(PyGenObject *gen)
{
return gen_send_ex(gen, NULL, 0);
}
static PyObject *
gen_send(PyGenObject *gen, PyObject *arg)
{
return gen_send_ex(gen, arg, 0);
}

从上面的代码中可以看到，send和next都是调用的同一函数gen_send_ex，区别在于是否带有参数。

static PyObject *
gen_send_ex(PyGenObject *gen, PyObject *arg, int exc)
{
PyThreadState *tstate = PyThreadState_GET();
PyFrameObject *f = gen->gi_frame;
PyObject *result;
if (gen->gi_running) { # 判断生成器是否已经运行
 PyErr_SetString(PyExc_ValueError,
     "generator already executing");
 return NULL;
}
if (f==NULL || f->f_stacktop == NULL) { # 如果代码块为空或调用栈为空，则抛出StopIteration异常
 /* Only set exception if called from send() */
 if (arg && !exc)
  PyErr_SetNone(PyExc_StopIteration);
 return NULL;
}
if (f->f_lasti == -1) { # f_lasti=1 代表首次执行
 if (arg && arg != Py_None) { # 首次执行不允许带有参数
  PyErr_SetString(PyExc_TypeError,
      "can't send non-None value to a "
      "just-started generator");
  return NULL;
 }
} else {
 /* Push arg onto the frame's value stack */
 result = arg ? arg : Py_None;
 Py_INCREF(result); # 该参数引用计数+1
 *(f->f_stacktop++) = result; # 参数压栈
}
/* Generators always return to their most recent caller, not
 * necessarily their creator. */
f->f_tstate = tstate;
Py_XINCREF(tstate->frame);
assert(f->f_back == NULL);
f->f_back = tstate->frame;
gen->gi_running = 1; # 修改生成器执行状态
result = PyEval_EvalFrameEx(f, exc); # 执行字节码
gen->gi_running = 0; # 恢复为未执行状态
/* Don't keep the reference to f_back any longer than necessary. It
 * may keep a chain of frames alive or it could create a reference
 * cycle. */
assert(f->f_back == tstate->frame);
Py_CLEAR(f->f_back);
/* Clear the borrowed reference to the thread state */
f->f_tstate = NULL;
/* If the generator just returned (as opposed to yielding), signal
 * that the generator is exhausted. */
if (result == Py_None && f->f_stacktop == NULL) {
 Py_DECREF(result);
 result = NULL;
 /* Set exception if not called by gen_iternext() */
 if (arg)
  PyErr_SetNone(PyExc_StopIteration);
}
if (!result || f->f_stacktop == NULL) {
 /* generator can't be rerun, so release the frame */
 Py_DECREF(f);
 gen->gi_frame = NULL;
}
return result;
}

static PyObject *
gen_send_ex(PyGenObject *gen, PyObject *arg, int exc)
{
PyThreadState *tstate = PyThreadState_GET();
PyFrameObject *f = gen->gi_frame;
PyObject *result;
if (gen->gi_running) { # 判断生成器是否已经运行
 PyErr_SetString(PyExc_ValueError,
     "generator already executing");
 return NULL;
}
if (f==NULL || f->f_stacktop == NULL) { # 如果代码块为空或调用栈为空，则抛出StopIteration异常
 /* Only set exception if called from send() */
 if (arg && !exc)
  PyErr_SetNone(PyExc_StopIteration);
 return NULL;
}
if (f->f_lasti == -1) { # f_lasti=1 代表首次执行
 if (arg && arg != Py_None) { # 首次执行不允许带有参数
  PyErr_SetString(PyExc_TypeError,
      "can't send non-None value to a "
      "just-started generator");
  return NULL;
 }
} else {
 /* Push arg onto the frame's value stack */
 result = arg ? arg : Py_None;
 Py_INCREF(result); # 该参数引用计数+1
 *(f->f_stacktop++) = result; # 参数压栈
}
/* Generators always return to their most recent caller, not
 * necessarily their creator. */
f->f_tstate = tstate;
Py_XINCREF(tstate->frame);
assert(f->f_back == NULL);
f->f_back = tstate->frame;
gen->gi_running = 1; # 修改生成器执行状态
result = PyEval_EvalFrameEx(f, exc); # 执行字节码
gen->gi_running = 0; # 恢复为未执行状态
/* Don't keep the reference to f_back any longer than necessary. It
 * may keep a chain of frames alive or it could create a reference
 * cycle. */
assert(f->f_back == tstate->frame);
Py_CLEAR(f->f_back);
/* Clear the borrowed reference to the thread state */
f->f_tstate = NULL;
/* If the generator just returned (as opposed to yielding), signal
 * that the generator is exhausted. */
if (result == Py_None && f->f_stacktop == NULL) {
 Py_DECREF(result);
 result = NULL;
 /* Set exception if not called by gen_iternext() */
 if (arg)
  PyErr_SetNone(PyExc_StopIteration);
}
if (!result || f->f_stacktop == NULL) {
 /* generator can't be rerun, so release the frame */
 Py_DECREF(f);
 gen->gi_frame = NULL;
}
return result;
}

字节码的执行

PyEval_EvalFrameEx函数的功能为执行字节码并返回结果。

# 主要流程如下，
for (;;) {
switch(opcode) { # opcode为操作码，对应着各种操作
 case NOP:
  goto fast_next_opcode;
 ...
 ...
 case YIELD_VALUE: # 如果操作码是yield
  retval = POP();
  f->f_stacktop = stack_pointer;
  why = WHY_YIELD;
  goto fast_yield; # 利用goto跳出循环
}
}
fast_yield:
...
return vetval; # 返回结果
# 主要流程如下，
for (;;) {
switch(opcode) { # opcode为操作码，对应着各种操作
 case NOP:
  goto fast_next_opcode;
 ...
 ...
 case YIELD_VALUE: # 如果操作码是yield
  retval = POP();
  f->f_stacktop = stack_pointer;
  why = WHY_YIELD;
  goto fast_yield; # 利用goto跳出循环
}
}
fast_yield:
...
return vetval; # 返回结果

举一个例子，f_back上一个Frame，f_lasti上一次执行的指令的偏移量，

import sys
from dis import dis
def func():
f = sys._getframe(0)
print f.f_lasti
print f.f_back
yield 1
print f.f_lasti
print f.f_back
yield 2
a = func()
dis(func)
a.next()
a.next()
import sys
from dis import dis
def func():
f = sys._getframe(0)
print f.f_lasti
print f.f_back
yield 1
print f.f_lasti
print f.f_back
yield 2
a = func()
dis(func)
a.next()
a.next()

结果如下，其中第三行的英文为操作码，对应着上面的opcode，每次switch都是在不同的opcode之间进行选择。

Python
  0 LOAD_GLOBAL    0 (sys)
   3 LOAD_ATTR    1 (_getframe)
   6 LOAD_CONST    1 (0)
   9 CALL_FUNCTION   1
   12 STORE_FAST    0 (f)
  15 LOAD_FAST    0 (f)
   18 LOAD_ATTR    2 (f_lasti)
   21 PRINT_ITEM  
   22 PRINT_NEWLINE  
  23 LOAD_FAST    0 (f)
   26 LOAD_ATTR    3 (f_back)
   29 PRINT_ITEM  
   30 PRINT_NEWLINE  
 31 LOAD_CONST    2 (1)
   34 YIELD_VALUE  # 此时操作码为YIELD_VALUE，直接跳转上述goto语句，此时f_lasti为当前指令，f_back为当前frame
   35 POP_TOP    
 36 LOAD_FAST    0 (f)
   39 LOAD_ATTR    2 (f_lasti)
   42 PRINT_ITEM  
   43 PRINT_NEWLINE  
  44 LOAD_FAST    0 (f)
   47 LOAD_ATTR    3 (f_back)
   50 PRINT_ITEM  
   51 PRINT_NEWLINE  
  52 LOAD_CONST    3 (2)
   55 YIELD_VALUE  
   56 POP_TOP    
   57 LOAD_CONST    0 (None)
   60 RETURN_VALUE  
<frame object at 0x7fa75fcebc20> #和下面的frame相同，属于同一个frame，也就是说在同一个函数（命名空间）内，frame是同一个。
<frame object at 0x7fa75fcebc20>
  0 LOAD_GLOBAL    0 (sys)
   3 LOAD_ATTR    1 (_getframe)
   6 LOAD_CONST    1 (0)
   9 CALL_FUNCTION   1
   12 STORE_FAST    0 (f)
  15 LOAD_FAST    0 (f)
   18 LOAD_ATTR    2 (f_lasti)
   21 PRINT_ITEM  
   22 PRINT_NEWLINE  
  23 LOAD_FAST    0 (f)
   26 LOAD_ATTR    3 (f_back)
   29 PRINT_ITEM  
   30 PRINT_NEWLINE  
  31 LOAD_CONST    2 (1)
   34 YIELD_VALUE  # 此时操作码为YIELD_VALUE，直接跳转上述goto语句，此时f_lasti为当前指令，f_back为当前frame
   35 POP_TOP    
  36 LOAD_FAST    0 (f)
   39 LOAD_ATTR    2 (f_lasti)
   42 PRINT_ITEM  
   43 PRINT_NEWLINE  
  44 LOAD_FAST    0 (f)
   47 LOAD_ATTR    3 (f_back)
   50 PRINT_ITEM  
   51 PRINT_NEWLINE  
  52 LOAD_CONST    3 (2)
   55 YIELD_VALUE  
   56 POP_TOP    
   57 LOAD_CONST    0 (None)
   60 RETURN_VALUE  
<frame object at 0x7fa75fcebc20> #和下面的frame相同，属于同一个frame，也就是说在同一个函数（命名空间）内，frame是同一个。
<frame object at 0x7fa75fcebc20>

总结

以上所述是小编给大家介绍的Python yield与实现方法代码分析网站的支持！