Source code for MambuPy.rest.mambuproduct

# coding: utf-8
"""Mambu Products objects.

.. autosummary::
   :nosignatures:
   :toctree: _autosummary

MambuProduct holds a product.

MambuProducts holds a list of products.

Uses mambugeturl.getproducturl as default urlfunc
"""


from ..mambugeturl import getproductsurl
from .mambustruct import MambuStruct
from .mamburestutils import MambuStructIterator

mod_urlfunc = getproductsurl


[docs]class MambuProduct(MambuStruct): """A Product from Mambu. With the default urlfunc, entid argument must be the ID of the product you wish to retrieve. """
[docs] def __init__(self, urlfunc=mod_urlfunc, entid="", *args, **kwargs): """Tasks done here: Just initializes the MambuStruct. """ MambuStruct.__init__(self, urlfunc, entid, *args, **kwargs)
[docs]class MambuProducts(MambuStruct): """A list of Products from Mambu. With the default urlfunc, entid argument must be empty at instantiation time to retrieve all the products according to any other filter you send to the urlfunc. """
[docs] def __init__(self, urlfunc=mod_urlfunc, entid="", *args, **kwargs): """By default, entid argument is empty. That makes perfect sense: you want several products, not just one """ MambuStruct.__init__(self, urlfunc, entid, *args, **kwargs)
[docs] def __iter__(self): return MambuStructIterator(self.attrs)
[docs] def convert_dict_to_attrs(self, *args, **kwargs): """The trick for iterable Mambu Objects comes here: You iterate over each element of the responded List from Mambu, and create a Mambu Product object for each one, initializing them one at a time, and changing the attrs attribute (which just holds a list of plain dictionaries) with a MambuProduct just created. .. todo:: pass a valid (perhaps default) urlfunc, and its corresponding id to entid to each MambuProduct, telling MambuStruct not to connect() by default. It's desirable to connect at any other further moment to refresh some element in the list. """ for n, r in enumerate(self.attrs): # ok ok, I'm modifying elements of a list while iterating it. BAD PRACTICE! try: params = self.params except AttributeError: params = {} kwargs.update(params) try: self.mambuproductclass except AttributeError: self.mambuproductclass = MambuProduct product = self.mambuproductclass(urlfunc=None, entid=None, *args, **kwargs) product.init(r, *args, **kwargs) self.attrs[n] = product
[docs]class AllMambuProducts(MambuStruct): """Singleton that holds ALL the Mambu products. With the default urlfunc, entid argument must be empty at instantiation time to retrieve all the products according to any other filter you send to the urlfunc. Caching singleton. You may not wish to retrieve from Mambu all the products every time you need to use them, so you can use this, which requests from Mambu one time only, and holds them here forever during your python session. To use it, instead of instantiating a MambuProducts() object when needing all the products from Mambu, instantiate an AllMambuProducts() object and the caching will be used by default. .. todo:: is there a better way to implement cache directly on MambuProducts()? Why have a cache? because Mambu products are not likely to change once you have them configured in Mambu. So it's better to prevent a lot of requests to get the same information over and over again. You are free to implement caches such as this on any other Mambu objects if you find it useful (perhaps groups or branches or any other Mambu entity do not change a lot on your own business?). Please try to be consistent on names and functionality. """ __instance = None
[docs] def __new__(cls, *args, **kwargs): if not cls.__instance: # cls.__instance = super(AllMambuProducts, cls).__new__(cls, *args, **kwargs) cls.__instance = super(AllMambuProducts, cls).__new__(cls) else: cls.__instance.noinit = True return cls.__instance
[docs] def __init__(self, urlfunc=mod_urlfunc, entid="", *args, **kwargs): """If you have already retrieved products, you don't contact Mambu. If you haven't, you do. """ try: getattr(self, "noinit") except AttributeError: MambuStruct.__init__(self, urlfunc, entid, *args, **kwargs)
[docs] def __iter__(self): return MambuStructIterator(self.attrs)
[docs] def __getattribute__(self, name): """Object-like get attribute""" try: return object.__getattribute__(self, name) except AttributeError: # Iterable AllMambuProducts singleton also uses a special # noinit property that should raise AttributeError when # not set if name in ["params", "noinit", "mambuproductclass"]: raise AttributeError return self[name]
[docs] def convert_dict_to_attrs(self, *args, **kwargs): """The trick for iterable Mambu Objects comes here: You iterate over each element of the responded List from Mambu, and create a Mambu Product object for each one, initializing them one at a time, and changing the attrs attribute (which just holds a list of plain dictionaries) with a MambuProduct just created. .. todo:: pass a valid (perhaps default) urlfunc, and its corresponding id to entid to each MambuProduct, telling MambuStruct not to connect() by default. It's desirable to connect at any other further moment to refresh some element in the list. """ for n, r in enumerate(self.attrs): # ok ok, I'm modifying elements of a list while iterating it. BAD PRACTICE! try: params = self.params except AttributeError: params = {} kwargs.update(params) try: self.mambuproductclass except AttributeError: self.mambuproductclass = MambuProduct product = self.mambuproductclass(urlfunc=None, entid=None, *args, **kwargs) product.init(r, *args, **kwargs) product._MambuStruct__urlfunc = getproductsurl self.attrs[n] = product