[serpent_skis]

The reason for using a match statement, even though there is only “one” case, is because it’s matching against a complex and (as you mentioned) deeply nested pattern. Lines 43-56 match the current AST node against the pattern (shown here as pseudo-python):

  with open(_) as f:
    _ = f.read()

(Where _ can be anything, and f can be any name). The traditional alternative to lines 43-56 would be dozens of lines that check the top level of the AST, and if that matches against what the top level of the pattern is expecting, diving one layer deeper and checking that level against the next level of the pattern, etc, until you reach the leaves of your pattern. Once your familiar with pattern matching syntax, this is much easier to quickly read/grok/audit.

[UncleEntity]

So it can do multi-level tree matching?

Very interesting…

One of the projects I occasionally poke at is a burs (bottom-up rewrite system) generator used to do cost-based tree rewriting. Think this is a concept that might make sense in that context, just need to find the time to play…