Iterating over marshal

[Tim Lesher]

I'm using the marshal library to unmarshal a file containing one or
more objects. The canonical way seems to be:

objs = []
while 1:
try:
objs.append(marshal.load(fobj))
except EOFError:
break

Maybe it's just me, but it seems as if this should be iterable. I can
get the behavior I want by writing:

def itermarshal(fobj):
while 1:
try:
yield marshal.load(fobj)
except EOFError:
raise StopIteration

objs = [obj for obj in itermarshal(fobj)]

But it seems that this should be built-in somewhere. Given that the
marshal library has been around since roughly forever, is it just that
no one's bothered to add iteration support to it, or am I missing
something?

Thanks.

 

[Fredrik Lundh]

I'm using the marshal library to unmarshal a file containing one or
more objects. The canonical way seems to be:

objs = []
while 1:
try:
objs.append(marshal.load(fobj))
except EOFError:
break

the canonical way to do this is to put the objects in a sequence
container *before* marshalling them.

</F>

 

[Tim Lesher]

I'm using the marshal library to unmarshal a file containing one or
more objects. The canonical way seems to be:

objs = []
while 1:
try:
objs.append(marshal.load(fobj))
except EOFError:
breakthe canonical way to do this is to put the objects in a sequence

container *before* marshalling them.

That makes sense; unfortunately, I'm unmarshalling from an external
data source I don't control (the Perforce version control client).