🏗 Merge C++ into python codebase (#504)

## Description:

Move esphome-core codebase into esphome (and a bunch of other refactors). See https://github.com/esphome/feature-requests/issues/97

Yes this is a shit ton of work and no there's no way to automate it :( But it will be worth it 👍

Progress:
- Core support (file copy etc): 80%
- Base Abstractions (light, switch): ~50%
- Integrations: ~10%
- Working? Yes, (but only with ported components).

Other refactors:
- Moves all codegen related stuff into a single class: `esphome.codegen` (imported as `cg`)
- Rework coroutine syntax
- Move from `component/platform.py` to `domain/component.py` structure as with HA
- Move all defaults out of C++ and into config validation.
- Remove `make_...` helpers from Application class. Reason: Merge conflicts with every single new integration.
- Pointer Variables are stored globally instead of locally in setup(). Reason: stack size limit.

Future work:
- Rework const.py - Move all `CONF_...` into a conf class (usage `conf.UPDATE_INTERVAL` vs `CONF_UPDATE_INTERVAL`). Reason: Less convoluted import block
- Enable loading from `custom_components` folder.

**Related issue (if applicable):** https://github.com/esphome/feature-requests/issues/97

**Pull request in [esphome-docs](https://github.com/esphome/esphome-docs) with documentation (if applicable):** esphome/esphome-docs#<esphome-docs PR number goes here>

## Checklist:
  - [ ] The code change is tested and works locally.
  - [ ] Tests have been added to verify that the new code works (under `tests/` folder).

If user exposed functionality or configuration variables are added/changed:
  - [ ] Documentation added/updated in [esphomedocs](https://github.com/OttoWinter/esphomedocs).

esphome/py_compat.py

@@ -74,8 +74,8 @@ def indexbytes(buf, i):
 if IS_PY2:
     def decode_text(data, encoding='utf-8', errors='strict'):
         # type: (str, str, str) -> unicode
-        return unicode(data, encoding='utf-8', errors=errors)
+        return unicode(data, encoding=encoding, errors=errors)
 else:
     def decode_text(data, encoding='utf-8', errors='strict'):
         # type: (bytes, str, str) -> str
-        return data.decode(encoding='utf-8', errors=errors)
+        return data.decode(encoding=encoding, errors=errors)