Interpret Z

This is an interpreter for Sailthru's Zephyr scripting language. I'd gotten fed up about posting every template to Sailthru in order to check if my code rendered a context correctly, so I build this. A heavy debt of gratitude and inspiration is owed to Ruslan Spivak, whose article series 'Let's build a simple interpreter' provided the basis for this program. You can read the series here, if interested.

Usage

Interpret Z exposes one function, render, which takes .txt or .html content and a context dict, and renders that dict into the given content, as follows:

html_content = interpret_z.render(some_html_template_content, context_dict)

where some_html_template_context is an (unrendered) string containing sailthru vars, and context_dict is a dictionary of vars to be rendered into the given HTML.

Improvements

I've only implemented the bare minimum of Sailthru functions necessary to get this project up and running. For the rest, we don't even raise a NotImplemented exception at the moment, so you'll be warned that a var is referenced before initialization. I welcome any and all pull requests to add more Zephyr functions.

Known Issues

1. Python and Zephyr treat the following expressions differently:

# 3 && 4 && 0
# Zephyr => false
# Python => 0

# 3 && 4 && 2
# Zephyr => true
# Python => 2

If you are going to use this interpreter, just avoid these sorts of statements. Given the limited nature of the programming people tend to do in Zephyr, it shouldn't be hard.

For more examples of this bug, see test_boolean_statements in tests/test_interpreter.py.

2. Limited support for Zephyr functions.

Mostly because of the time involved, I've only supported the Zephyr functions needed to render the templates I need to test for work. Feel free to make a PR to add support for more functions (though I imagine supporting something like Personalize is nearly impossible).

3. No support for binary operators

I haven't needed this, so I didn't implement it yet. Again, feel free to make a PR to add it.

4. No support for making your own dicts.

You can pass in dicts through context, obviously. But you can't initialize them in Zephyr code. I've never needed to do that, and I honestly don't even know if Zephyr supports it, but if they do this is something I should add at some point.

5. No support for keying into a dict.

Interpret Z support dot notation only. Again, I've never attempted to use dict['key'] in Zephyr, and I'm not sure why anyone would when you can just do dict.key. Still, it should be included for completeness' sake.