decruft 0.1.2

Extract clean, readable content from web pages
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

<!-- {"url":"https://lean-lang.org/theorem_proving_in_lean4/dependent_type_theory.html"} -->
<!doctype html>
<html lang="en">
<head>
  <meta charset="utf-8" />
  <title>Dependent Type Theory</title>
</head>
<body>
  <main>
    <article>
      <h1>Dependent Type Theory</h1>
      <p>
        This fixture simulates a documentation page where highlighted code is rendered as a
        standalone <code>code.hl.block</code> element instead of a <code>pre</code> wrapper.
        Defuddle should preserve this as a fenced code block and should not expand internal links
        into markdown link syntax inside code. The surrounding prose is intentionally long enough
        to make main-content detection stable in tests and avoid scoring noise.
      </p>

      <code class="hl lean block" data-lean-context="examples">
        <span class="keyword token">def</span><span class="inter-text"> </span><span class="const token" id="m">m</span><span class="inter-text"> : </span><a href="https://lean-lang.org/doc/reference/4.26.0/Basic-Types/Natural-Numbers/#Nat___zero"><span class="const token">Nat</span></a><span class="inter-text"> := 1       -- m is a natural number</span>
      </code>

      <p>
        Additional text after the code block verifies that extraction continues normally and that
        the code is treated as block content rather than inline text. This also helps ensure the
        markdown output remains readable and that the issue does not regress in future updates.
      </p>
    </article>
  </main>
</body>
</html>