PEP 380 – Syntax for Delegating to a Subgenerator
- Gregory Ewing <greg.ewing at canterbury.ac.nz>
- Standards Track
- Python-Dev message
Table of Contents
A syntax is proposed for a generator to delegate part of its operations to another generator. This allows a section of code containing ‘yield’ to be factored out and placed in another generator. Additionally, the subgenerator is allowed to return with a value, and the value is made available to the delegating generator.
The new syntax also opens up some opportunities for optimisation when one generator re-yields values produced by another.
Guido officially accepted the PEP on 26th June, 2011.
A Python generator is a form of coroutine, but has the limitation that
it can only yield to its immediate caller. This means that a piece of
code containing a
yield cannot be factored out and put into a
separate function in the same way as other code. Performing such a
factoring causes the called function to itself become a generator, and
it is necessary to explicitly iterate over this second generator and
re-yield any values that it produces.
If yielding of values is the only concern, this can be performed without much difficulty using a loop such as
for v in g: yield v
However, if the subgenerator is to interact properly with the caller
in the case of calls to
things become considerably more difficult. As will be seen later, the
necessary code is very complicated, and it is tricky to handle all the
corner cases correctly.
A new syntax will be proposed to address this issue. In the simplest use cases, it will be equivalent to the above for-loop, but it will also handle the full range of generator behaviour, and allow generator code to be refactored in a simple and straightforward way.
The following new expression syntax will be allowed in the body of a generator:
yield from <expr>
where <expr> is an expression evaluating to an iterable, from which an
iterator is extracted. The iterator is run to exhaustion, during which
time it yields and receives values directly to or from the caller of
the generator containing the
yield from expression (the
Furthermore, when the iterator is another generator, the subgenerator
is allowed to execute a
return statement with a value, and that
value becomes the value of the
yield from expression.
The full semantics of the
yield from expression can be described
in terms of the generator protocol as follows:
- Any values that the iterator yields are passed directly to the caller.
- Any values sent to the delegating generator using
send()are passed directly to the iterator. If the sent value is None, the iterator’s
__next__()method is called. If the sent value is not None, the iterator’s
send()method is called. If the call raises StopIteration, the delegating generator is resumed. Any other exception is propagated to the delegating generator.
- Exceptions other than GeneratorExit thrown into the delegating
generator are passed to the
throw()method of the iterator. If the call raises StopIteration, the delegating generator is resumed. Any other exception is propagated to the delegating generator.
- If a GeneratorExit exception is thrown into the delegating
generator, or the
close()method of the delegating generator is called, then the
close()method of the iterator is called if it has one. If this call results in an exception, it is propagated to the delegating generator. Otherwise, GeneratorExit is raised in the delegating generator.
- The value of the
yield fromexpression is the first argument to the
StopIterationexception raised by the iterator when it terminates.
return exprin a generator causes
StopIteration(expr)to be raised upon exit from the generator.
Enhancements to StopIteration
For convenience, the
StopIteration exception will be given a
value attribute that holds its first argument, or None if there
are no arguments.
Python 3 syntax is used in this section.
- The statement
RESULT = yield from EXPR
is semantically equivalent to
_i = iter(EXPR) try: _y = next(_i) except StopIteration as _e: _r = _e.value else: while 1: try: _s = yield _y except GeneratorExit as _e: try: _m = _i.close except AttributeError: pass else: _m() raise _e except BaseException as _e: _x = sys.exc_info() try: _m = _i.throw except AttributeError: raise _e else: try: _y = _m(*_x) except StopIteration as _e: _r = _e.value break else: try: if _s is None: _y = next(_i) else: _y = _i.send(_s) except StopIteration as _e: _r = _e.value break RESULT = _r
- In a generator, the statement
is semantically equivalent to
except that, as currently, the exception cannot be caught by
exceptclauses within the returning generator.
- The StopIteration exception behaves as though defined thusly:
class StopIteration(Exception): def __init__(self, *args): if len(args) > 0: self.value = args else: self.value = None Exception.__init__(self, *args)
The Refactoring Principle
The rationale behind most of the semantics presented above stems from
the desire to be able to refactor generator code. It should be
possible to take a section of code containing one or more
expressions, move it into a separate function (using the usual
techniques to deal with references to variables in the surrounding
scope, etc.), and call the new function using a
The behaviour of the resulting compound generator should be, as far as
reasonably practicable, the same as the original unfactored generator
in all situations, including calls to
The semantics in cases of subiterators other than generators has been chosen as a reasonable generalization of the generator case.
The proposed semantics have the following limitations with regard to refactoring:
- A block of code that catches GeneratorExit without subsequently re-raising it cannot be factored out while retaining exactly the same behaviour.
- Factored code may not behave the same way as unfactored code if a StopIteration exception is thrown into the delegating generator.
With use cases for these being rare to non-existent, it was not considered worth the extra complexity required to support them.
There was some debate as to whether explicitly finalizing the
delegating generator by calling its
close() method while it is
suspended at a
yield from should also finalize the subiterator.
An argument against doing so is that it would result in premature
finalization of the subiterator if references to it exist elsewhere.
Consideration of non-refcounting Python implementations led to the decision that this explicit finalization should be performed, so that explicitly closing a factored generator has the same effect as doing so to an unfactored one in all Python implementations.
The assumption made is that, in the majority of use cases, the
subiterator will not be shared. The rare case of a shared subiterator
can be accommodated by means of a wrapper that blocks
close() calls, or by using a means other than
yield from to
call the subiterator.
Generators as Threads
A motivation for generators being able to return values concerns the use of generators to implement lightweight threads. When using generators in that way, it is reasonable to want to spread the computation performed by the lightweight thread over many functions. One would like to be able to call a subgenerator as though it were an ordinary function, passing it parameters and receiving a returned value.
Using the proposed syntax, a statement such as
y = f(x)
where f is an ordinary function, can be transformed into a delegation call
y = yield from g(x)
where g is a generator. One can reason about the behaviour of the
resulting code by thinking of g as an ordinary function that can be
suspended using a
When using generators as threads in this way, typically one is not
interested in the values being passed in or out of the yields.
However, there are use cases for this as well, where the thread is
seen as a producer or consumer of items. The
expression allows the logic of the thread to be spread over as many
functions as desired, with the production or consumption of items
occurring in any subfunction, and the items are automatically routed to
or from their ultimate source or destination.
close(), it is reasonable to expect
that if an exception is thrown into the thread from outside, it should
first be raised in the innermost generator where the thread is
suspended, and propagate outwards from there; and that if the thread
is terminated from outside by calling
close(), the chain of active
generators should be finalised from the innermost outwards.
The particular syntax proposed has been chosen as suggestive of its
meaning, while not introducing any new keywords and clearly standing
out as being different from a plain
Using a specialised syntax opens up possibilities for optimisation
when there is a long chain of generators. Such chains can arise, for
instance, when recursively traversing a tree structure. The overhead
__next__() calls and yielded values down and up the
chain can cause what ought to be an O(n) operation to become, in the
worst case, O(n**2).
A possible strategy is to add a slot to generator objects to hold a
generator being delegated to. When a
call is made on the generator, this slot is checked first, and if it
is nonempty, the generator that it references is resumed instead. If
it raises StopIteration, the slot is cleared and the main generator is
This would reduce the delegation overhead to a chain of C function calls involving no Python code execution. A possible enhancement would be to traverse the whole chain of generators in a loop and directly resume the one at the end, although the handling of StopIteration is more complicated then.
Use of StopIteration to return values
There are a variety of ways that the return value from the generator
could be passed back. Some alternatives include storing it as an
attribute of the generator-iterator object, or returning it as the
value of the
close() call to the subgenerator. However, the
proposed mechanism is attractive for a couple of reasons:
- Using a generalization of the StopIteration exception makes it easy for other kinds of iterators to participate in the protocol without having to grow an extra attribute or a close() method.
- It simplifies the implementation, because the point at which the return value from the subgenerator becomes available is the same point at which the exception is raised. Delaying until any later time would require storing the return value somewhere.
Some ideas were discussed but rejected.
Suggestion: There should be some way to prevent the initial call to __next__(), or substitute it with a send() call with a specified value, the intention being to support the use of generators wrapped so that the initial __next__() is performed automatically.
Resolution: Outside the scope of the proposal. Such generators should
not be used with
Suggestion: If closing a subiterator raises StopIteration with a
value, return that value from the
close() call to the delegating
The motivation for this feature is so that the end of a stream of values being sent to a generator can be signalled by closing the generator. The generator would catch GeneratorExit, finish its computation and return a result, which would then become the return value of the close() call.
Resolution: This usage of close() and GeneratorExit would be incompatible with their current role as a bail-out and clean-up mechanism. It would require that when closing a delegating generator, after the subgenerator is closed, the delegating generator be resumed instead of re-raising GeneratorExit. But this is not acceptable, because it would fail to ensure that the delegating generator is finalised properly in the case where close() is being called for cleanup purposes.
Signalling the end of values to a consumer is better addressed by other means, such as sending in a sentinel value or throwing in an exception agreed upon by the producer and consumer. The consumer can then detect the sentinel or exception and respond by finishing its computation and returning normally. Such a scheme behaves correctly in the presence of delegation.
close() is not to return a value, then raise an
exception if StopIteration with a non-None value occurs.
Resolution: No clear reason to do so. Ignoring a return value is not considered an error anywhere else in Python.
Under this proposal, the value of a
yield from expression would be
derived in a very different way from that of an ordinary
expression. This suggests that some other syntax not containing the
yield might be more appropriate, but no acceptable
alternative has so far been proposed. Rejected alternatives include
It has been suggested that some mechanism other than
return in the
subgenerator should be used to establish the value returned by the
yield from expression. However, this would interfere with the
goal of being able to think of the subgenerator as a suspendable
function, since it would not be able to return values in the same way
as other functions.
The use of an exception to pass the return value has been criticised as an “abuse of exceptions”, without any concrete justification of this claim. In any case, this is only one suggested implementation; another mechanism could be used without losing any essential features of the proposal.
It has been suggested that a different exception, such as
GeneratorReturn, should be used instead of StopIteration to return a
value. However, no convincing practical reason for this has been put
forward, and the addition of a
value attribute to StopIteration
mitigates any difficulties in extracting a return value from a
StopIteration exception that may or may not have one. Also, using a
different exception would mean that, unlike ordinary functions,
‘return’ without a value in a generator would not be equivalent to
Proposals along similar lines have been made before, some using the
yield * instead of
yield from. While
yield * is
more concise, it could be argued that it looks too similar to an
yield and the difference might be overlooked when reading
To the author’s knowledge, previous proposals have focused only on yielding values, and thereby suffered from the criticism that the two-line for-loop they replace is not sufficiently tiresome to write to justify a new syntax. By dealing with the full generator protocol, this proposal provides considerably more benefit.
Some examples of the use of the proposed syntax are available, and also a prototype implementation based on the first optimisation outlined above.
Examples and Implementation
A version of the implementation updated for Python 3.3 is available from tracker issue #11682
This document has been placed in the public domain.
Last modified: 2017-11-11 19:28:55+00:00 GMT