$$ Variables and Patterns

The Parm class implements a pattern-matching system that allows for dynamic variable substitution and evaluation. The system is built around several specialized patterns that use $ and $$ prefixes to indicate different types of variable access and operations.

Pattern Types

The class defines four main pattern types:

1. Global Variables ($VARNAME):

   • Pattern: \$[A-Z]{2,}([-+*/%]\d+)?
   • Examples: $FOO, $BAR+4, $COUNT-2
   • Purpose: Access global variables stored in GlobalStore, with optional math operations

2. Loop Number ($$L):

   • Pattern: \$\$L
   • Purpose: Returns the current loop iteration number (0-indexed)

3. List Access ($$N and $$num):

   • Pattern: \$\$(?:N([-+*/%]\d+)?|(\d+))
   • Examples:
     • $$N - Access element at current loop index
     • $$N+2 - Access element at loop index + 2
     • $$N*3 - Access element at loop index * 3
     • $$1 - Access first element
     • $$5 - Access fifth element
   • Purpose: Retrieves items from an input list based on loop number or explicit index

4. Python Code (`code`):

   • Pattern: `([^`]+)`
   • Examples: `len("test")`, `math.sin(0.5)`
   • Purpose: Evaluates arbitrary Python expressions within a safe execution environment

How It Works

1. The system processes expressions in a specific order:

   • First: Global variables
   • Second: Loop numbers
   • Third: List access
   • Fourth: Python code evaluation

2. When eval() is called on a parameter, the system:

   • Checks if the parameter contains patterns
   • For each pattern match, it applies the corresponding transformation
   • Returns the fully resolved value

3. List access requires an input node that returns a list

   • The system calculates the appropriate index (with overflow protection)
   • Returns the corresponding list item as a string

4. The Python code evaluation runs in a sandboxed environment with limited built-ins and modules for security

This pattern system enables powerful template-based string manipulation and data access while maintaining a controlled execution environment.