Add the ability to parse macros and include them in the interpretation

[pragunbhutani]

Currently, ragstar only reads model and source definitions and uses these to infer dbt compatible documentation for the models. Similarly, we should be able to read macros and include macro definitions in the context provided to the LLM to interpret models.

To do this one would have to:

1. Include parsing of macros in the DbtProject.parse() method
2. Scan model definitions for the names of macros that have been used
3. Add relevant macro definition to the context provided to the DocumentationGenerator