Does anyone know of a pod (or other tool/code that does the same) that extracts top-level comments and signatures from a file of code, into a hierarchical structure? Example: Given a project that has a structure like this:

src/models/routine.ts

The file has a

class Routine

with comments above it. The class has

function start(offsetSeconds: number)

with comments above it. It generates:

models.routine.md

with

## class Routine

that has

### function start

, with the comments as the content within each header. This would open for many things, one of the most exciting ones being: - Predictable "dendron-note-to-generated-note" linking - Guarding against comment rot through dendron discovering missing references after pod export run - Guarding against inverse documentation rot through dendron discovering missing references to notes from inside the generated markdown For comment rot: Given that text in

r.routines.lifecycle.md

referred to the code with

[[Routine model|models.routines#routine]]

, the missing link detection would identify if

src/models/routines.ts

or its (top level) signatures are renamed or removed. For "inverse documentation rot": Given that a comment in

src/models/routine.ts

said something like

The full routine lifecycle is visualized in [[r.routines.lifecycle#visualized]]

, the missing link detection would identify if

r.routines.lifecycle.md

or the

visualized

header, are renamed or removed.

don't know any but that would be awesome. you might be able to add ts-doc to do this

That's an interesting approach! I wonder about the application for other languages if ts-doc would be part of it. If the pod could be given a language aware extractor of some kind, it could become reusable across languages.

I know a lot of parsing libraries using https://tree-sitter.github.io/tree-sitter/ as a base