Simplify NBT implementation

 * Remove the TYPE class variable to rely in the actual type of each tag
 * Change the from_object and to_object to use schemas describing the data instead of using integer ranges and choosing types arbitrarily

changes/257.feature.md

@@ -1,9 +1,12 @@
 - Added the `NBTag` to deal with NBT data:
   - The `NBTag` class is the base class for all NBT tags and provides the basic functionality to serialize and deserialize NBT data from and to a `Buffer` object.
   - The classes `EndNBT`, `ByteNBT`, `ShortNBT`, `IntNBT`, `LongNBT`, `FloatNBT`, `DoubleNBT`, `ByteArrayNBT`, `StringNBT`, `ListNBT`, `CompoundNBT`, `IntArrayNBT`and `LongArrayNBT` were added and correspond to the NBT types described in the [NBT specification](https://wiki.vg/NBT#Specification).
-  - NBT tags can be created using the `NBTag.from_object()` method, which automatically selects the correct tag type based on the object's type and works recursively for lists and dictionaries.
-  - The `NBTag.to_object()` method can be used to convert an NBT tag back to a Python object.
+  - NBT tags can be created using the `NBTag.from_object()` method and a schema that describes the NBT tag structure.
+    Compound tags are represented as dictionaries, list tags as lists, and primitive tags as their respective Python types.
+    The implementation allows to add custom classes to the schema to handle custom NBT tags if they inherit the `:class: NBTagConvertible` class.
+  - The `NBTag.to_object()` method can be used to convert an NBT tag back to a Python object. Use include_schema=True to include the schema in the output, and `include_name=True` to include the name of the tag in the output. In that case the output will be a dictionary with a single key that is the name of the tag and the value is the object representation of the tag.
   - The `NBTag.serialize()` can be used to serialize an NBT tag to a new `Buffer` object.
   - The `NBTag.deserialize(buffer)` can be used to deserialize an NBT tag from a `Buffer` object.
   - If the buffer already exists, the `NBTag.write_to(buffer, with_type=True, with_name=True)` method can be used to write the NBT tag to the buffer (and in that case with the type and name in the right format).
   - The `NBTag.read_from(buffer, with_type=True, with_name=True)` method can be used to read an NBT tag from the buffer (and in that case with the type and name in the right format).
+  - The `NBTag.value` property can be used to get the value of the NBT tag as a Python object.