MsgPack exception - use_list=False

[ahopkins]

I know you removed the use_list option in the loads method when deserializing. See #370

However, I think it should be added back, even with an overwritable default. Without it, it will cause a difficult to diagnose exception.

Second, I see you added the encoding option. It looks like that is being depracated: https://github.com/msgpack/msgpack-python#deprecating-encoding-option

I would suggest going back to

    @classmethod
    def loads(cls, value, use_list=False):
        """
        Deserialize value using ``msgpack.loads``.

        :param value: bytes
        :returns: obj
        """
        if value is None:
            return None
        return msgpack.loads(value, use_list=use_list)

Furthermore, I am trying to use it with a simple string loaded serializer: aiocache.serializers.MsgPackSerializer. Which, to me means adding some sensible defaults.

{
    'cache': "aiocache.RedisCache",
    'serializer': {
        'class': "aiocache.serializers.MessagePackSerializer"
    },
    'plugins': [
        {'class': "aiocache.plugins.HitMissRatioPlugin"},
        {'class': "aiocache.plugins.TimingPlugin"}
    ],
}

[argaen]

Hi, I'm working on #379. This will allow you to pass the use_list param for that specific serializer. You will be able to do:

{
    'cache': "aiocache.RedisCache",
    'serializer': {
        'class': "aiocache.serializers.MessagePackSerializer",
        'use_list': False,
    },
    'plugins': [
        {'class': "aiocache.plugins.HitMissRatioPlugin"},
        {'class': "aiocache.plugins.TimingPlugin"}
    ],
}

Regarding encoding, thanks for the tip. But I want to avoid this:

>>> import msgpack
>>> msgpack.dumps('hola')
b'\xa4hola'
>>> msgpack.loads(b'\xa4hola')
b'hola'

If I encode a string, I want a string to be returned. Right now using encoding is the only way I've seen. In version 1.0 there is the raw one but not available in 0.5.1. Once 1.0 is released I'll update it

[ahopkins]

Thanks. I love the implementation. I guess I am used to getting the bytes string back having used msgpack in the past. I had to set encoding=None or else I was getting an error.

[argaen]

Glad its useful.

What do you mean by "getting an error"? encoding=utf-8 is still a valid pattern in 0.5.1

[ahopkins]

Sorry for being vague. I worked through the exception and do not have it in front of me. Effectively it was something like this: UnicodeDecodeError: codec can't decode byte 0xde in position 0

…

On Tue, Jan 23, 2018 at 9:39 PM Manuel Miranda ***@***.***> wrote: Glad its useful. What do you mean by "getting an error"? encoding=utf-8 is still a valid pattern in 0.5.1 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#380 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAKJfVohsbFYp9PMLVjuwK-iOlWbwix9ks5tNjVpgaJpZM4RoSN1> .

[argaen]

Closing until we get more data to investigate