FiwareObjectConverter (FOC)

This is a simple implementation to serialize Python2- and -3-Objects into a Fiware-Entity and back. The generated JSON-Strings can be POSTed to their API.

There is also the posibillity to ignore the Metadata while parsing back to the (specified) Python-Object.

For more Information about Fiwire-Orion visit the following website

Fiware Entity

Entity consits of
    - Id
    - Type
    - has n Attributes:
        - Name
        - Type
        - Value
        has n-Metadata:
            - Name
            - Type
            - Value

Usage

Let's create a class which contains a unicode-string in Python via:

class FooBar(object):
    def __init__(self):
        self.myStr = u'Hi!'

Object 2 FiwareEntity

This class can be simply serialized to Json with:

from objectFiwareConverter import ObjectFiwareConverter

json = ObjectFiwareConverter.obj2Fiware(FooBar(), ind=4)
# indent is set to 4 for readability

The str json contains the following

{
    "myStr": {
        "type": "string",
        "value": "Hi!",
        "metadata": {
            "python": {
                "type": "dataType",
                "value": "unicode"
            }
        }
    },
    "type": "FooBar",
    "id": "FooBarbc86c90d-6cca-41c3-878e-cbb58908056c"
}

This is a simple class for demonstration. The Data-Structure can be arbitrary complex.

The "type"- and "id"-values can be set manually. To do so, just add self.type = "YOUR_TYPE" and/or self.id = "YOUR_ID" to FooBar.

You also have the option to not set the id and value. Just do the following:

json = ObjectFiwareConverter.obj2Fiware(FooBar(), ind=4, showIdValue=False)

which would just create:

{
    "myStr": {
        "type": "string",
        "value": "Hi!",
        "metadata": {
            "python": {
                "type": "dataType",
                "value": "unicode"
            }
        }
    }
}

(E.g.: This json-string can be PATCHed to the v2-Api of the Context-Broker)

FiwareEntity 2 Object

A (Representation-) Class is needed to convert it back. Let's set a Class and then parse the JSON-Object into it:

class MyVeryOwnFooBar(object): 
    def __init__(self):
        self.myStr = u' ' # Set Unicode DataType

mvofb = MyVeryOwnFooBar()
ObjectFiwareConverter.fiware2Obj(json, mvofb) # the json from above
print mvofb.myStr # prints "Hi!"

Missing variables which are defined by the JSON-String are ignored if not set in the class. If variables are defined which are not specified in the JSON-String then those variables are not touched.

Further Information

Fiware Entity Information (Type/ID)

type = ClassName
id = ClassName + Universally Unique Identifier (UUID)

The id consist of the Class-Name + a random generated uuid by uuid4() and the type is simply the Class-Name. Also: All Objects, which are converted back from json may contain an id and type Attribute. They can be accessed with getattr, if

Ignoring MetaData/Additional MetaData and excluding MetaDAta

To ignore the metadata, do the following:

mvofb = MyVeryOwnFooBar()
ObjectFiwareConverter.fiware2Obj(json, mvofb, useMetaData=False) # the json from above

The conversion between a Python-Object and a JSON-String is not bidirectional by ignoring the metadata. By ignoring the metadata some Python-DataTypes are "converted into a simple type":

Complex, Tuple --> List
Unicode        --> String

NOTE: Python3's strings are already unicode by default, so here nothing is changed. Additionally long-types no longer exists. Those are then set as int

---

NOTE: The above example will throw an TypeError, because the class MyVeryOwnFooBar awaits an unicode but would be overwritten with a string, because the metadata is ignored. This behaviour can be turned off with the following:

mvofb = MyVeryOwnFooBar()
ObjectFiwareConverter.fiware2Obj(json, mvofb, useMetaData=False, ignoreWrongDataType=True) # the json from above

---

It is also possible to give a Data-Type-Object-Structure while converting to json. Simply create a dict containing the concrete Data-Type for Data and the additional Information will be added into the Metadata.

Example:

class AFooBar(object): 
    def __init__(self):
        self.x = 32.123
        self.y = 42.123

myData = dict(y='float32', x='float32')
json = ObjectFiwareConverter.obj2Fiware(AFooBar(), ind=4, dataTypeDict=myData)

wolud result to:

{
    "y": {
        "type": "number", 
        "value": 42.123, 
        "metadata": {
            "python": {
                "type": "dataType", 
                "value": "float"
            }, 
            "dataType": {
                "type": "dataType", 
                "value": "float32"
            }
        }
    }, 
    "x": {
        "type": "number", 
        "value": 32.123, 
        "metadata": {
            "python": {
                "type": "dataType", 
                "value": "float"
            }, 
            "dataType": {
                "type": "dataType", 
                "value": "float32"
            }
        }
    }, 
    "type": "AFooBar", 
    "id": "AFooBar703e581d-3068-4b62-bfa3-c713707e6929"
}

---

To exclude the python-metadata, while creating the json. You can use ignorePythonMetaData=True as here:

json = ObjectFiwareConverter.obj2Fiware(FooBar(), ind=4, ignorePythonMetaData=True) # json fron above

This creates simply the following json:

{
    "myStr": {
        "type": "string",
        "value": "Hi!"
        },
    "type": "FooBar",
    "id": "FooBarbc86c90d-6cca-41c3-878e-cbb58908056c"
}

---

At last, if you simply cannot create a class which contains the needed values (or everything is dynamically), just use the setAttr- Parameter.

class SomeTempObject(object):
    pass

sto = SomeTempObject()
ObjectFiwareConverter.fiware2Obj(json, sto, setAttr=True) # The json from above

The Values are added via setattr and can be accesed by getattr:

print getattr(sto, 'myStr') # --> would print Hi!