Markdown-extraction half of PtcRunner.Lisp.SpecValidator.

Parses the PTC-Lisp specification markdown into structured data — testable examples ({code, expected, section} tuples), TODO and BUG markers, illustrative skipped examples (; => ...), and per-section content hashes.

Pure functions over an in-memory string. No file I/O, no execution of any code. The companion SpecValidator module loads the spec file and runs the extracted examples through PtcRunner.Lisp.run/2.

Markdown shapes recognised

Single-line:

(+ 1 2)  ; => 3

Multi-line, with the marker on the closing line:

(let [x 1
      y 2]
  (+ x y))  ; => 3

Multi-line, with the marker on its own comment line below the code:

(filter odd? [1 2 3 4 5])
;;  => [1 3 5]

Section context is the most recent ## N. Title header seen.

Return shape

%{
  examples: [{"(+ 1 2)", 3, "## 1. ..."}, ...],
  todos:    [{"(some-fn x)", "not implemented", "## ..."}, ...],
  bugs:     [{"(buggy-fn)", "known issue",      "## ..."}, ...],
  skipped:  2
}