main: Dispatch verbs to objects separately ...

The existing implementation of code stores objects as a dictionary
containing a type, implementation-specific data and a list of verbs.

Unfortunately this creates a circular dependency: Verbs need to supply
all possible verbs when creating an object as a return value.

To solve this, separate the manipulations of objects from the dispatch
of verbs attached to them.

parse: Fix 'Expect to' to be 'Expected To'

main: Add way to append text

main: Pass subject to verbs

main: Namespace verbs in objects

main: Add text objects and name resolution

parse: Make symbols values case sensitive

parse: Represent values as a literal or reference ...

Instead of just passing back the token, instead decide what the value is
now and let the interpreter do something with it later.

Right now it just does the same thing it did before, but hopefully it
will turn text literals in to references to objects.

parse: Replace 'tokenizer' with 'tokenize'

parse: Make tokens case-sensitive ...

This should reduce ambiguity, especially since they read differently to
both eyes and screen readers.

main: Move parsing code to parse.py

Implement just enough parser to do a hello world

main: Add initial parser ...

This is very messy and incomplete code, but it's a start.

main: Skip parsing shebang if detected

main: Lowercase tokens and add special keyword type

code.txt: Remove useless done

code.txt: Add copyright information

code.txt: Add NewLang statement ...

This will be used to identify the file as source code and support
forwards and backwards compatibility with future features.

main.py: Support running using a shebang

code.txt: Replace 'Done' with 'EndIf' and 'EndSet' ...

This is much less ambiguous to humans the parser.

code.txt: Switch from 'Is' syntax to 'Set To' ...

'Subject Is Command' is hard to distinguish from 'Subject Verb Arguments'.
Let's do 'Set Subject To Command Done' instead.

Initial WIP commit