fixup! feat: Add option to scan and register HTML anchors

docs/changelog.md

@@ -1,10 +1 @@
 --8<-- "CHANGELOG.md"
-
-[](#hello){#hello2}
-
-## Hello
-
-Hello.
-
-Link to [Hello 1][hello1].
-Link to [Hello 2][hello2].

docs/index.md

@@ -1,10 +1,8 @@
 --8<-- "README.md"
 
-[](#hello){#hello1}
+[](){#hello}
+## Hello world
 
-## Hello
+helllo.
 
-Hello.
-
-Link to [Hello 1][hello1].
-Link to [Hello 2][hello2].
+[hello][hello]

src/mkdocs_autorefs/references.py

@@ -3,8 +3,10 @@
 from __future__ import annotations
 
 import re
+import unicodedata
 from html import escape, unescape
-from typing import TYPE_CHECKING, Any, Callable, Match, Tuple
+from itertools import zip_longest
+from typing import TYPE_CHECKING, Any, Callable, ClassVar, Match, Tuple
 from urllib.parse import urlsplit
 from xml.etree.ElementTree import Element
 
@@ -204,6 +206,8 @@ def fix_refs(html: str, url_mapper: Callable[[str], str]) -> tuple[str, list[str
 class AnchorScannerTreeProcessor(Treeprocessor):
     """Tree processor to scan and register HTML anchors."""
 
+    _htags: ClassVar[set[str]] = {"h1", "h2", "h3", "h4", "h5", "h6"}
+
     def __init__(self, plugin: AutorefsPlugin, md: Markdown | None = None) -> None:
         """Initialize the tree processor.
 
@@ -217,12 +221,24 @@ def run(self, root: Element) -> None:  # noqa: D102
         if self.plugin.current_page is not None:
             self._scan_anchors(root)
 
-    def _scan_anchors(self, parent: Element) -> None:
-        for el in parent:
-            if el.tag == "a" and (hid := el.get("id")):
-                self.plugin.register_anchor(self.plugin.current_page, hid, el.get("href", "").lstrip("#"))  # type: ignore[arg-type]
+    @staticmethod
+    def _slugify(value: str, separator: str = "-") -> str:
+        value = unicodedata.normalize("NFKD", str(value)).encode("ascii", "ignore").decode("ascii")
+        value = re.sub(r"[^\w\s-]", "", value.lower())
+        return re.sub(r"[-_\s]+", separator, value).strip("-_")
+
+    def _scan_anchors(self, parent: Element) -> str | None:
+        hid = None
+        for el, next_el in zip_longest(parent, parent[1:], fillvalue=Element("/")):
+            if el.tag == "a":
+                hid = el.get("id")
+            elif el.tag == "p" and (hid := self._scan_anchors(el)):
+                href = (next_el.get("id") or self._slugify(next_el.text or "")) if next_el.tag in self._htags else ""
+                self.plugin.register_anchor(self.plugin.current_page, hid, href)  # type: ignore[arg-type]
+                hid = None
             else:
-                self._scan_anchors(el)
+                hid = self._scan_anchors(el)
+        return hid
 
 
 class AutorefsExtension(Extension):

tests/test_plugin.py

@@ -60,27 +60,3 @@ def test_dont_make_relative_urls_relative_again() -> None:
             plugin.get_item_url("hello", from_url="baz/bar/foo.html", fallback=lambda _: ("foo.bar.baz",))
             == "../../foo/bar/baz.html#foo.bar.baz"
         )
-
-
-def test_register_html_anchors() -> None:
-    """Check that HT?ML anchors are registered when enabled."""
-    plugin = AutorefsPlugin()
-    plugin.scan_toc = False
-    plugin.scan_anchors = True
-
-    class Page:
-        url = "/page/url"
-
-    plugin.on_page_content(
-        """
-        <a id="foo.bar">
-        <a href="#foo.baz">
-        <a id="foo.qux" href="#fooqux">
-        <a href="quxfoo" id="qux.foo">
-        """,
-        page=Page(),  # type: ignore[arg-type]
-    )
-    assert "foo.bar" in plugin._url_map
-    assert "foo.baz" not in plugin._url_map
-    assert "foo.qux" in plugin._url_map
-    assert "qux.foo" in plugin._url_map

tests/test_references.py

@@ -2,10 +2,14 @@
 
 from __future__ import annotations
 
+from functools import partial
+from textwrap import dedent
+
 import markdown
 import pytest
 
-from mkdocs_autorefs.references import AutorefsExtension, fix_refs, relative_url
+from mkdocs_autorefs.plugin import AutorefsPlugin
+from mkdocs_autorefs.references import AnchorScannerTreeProcessor, AutorefsExtension, fix_refs, relative_url
 
 
 @pytest.mark.parametrize(
@@ -224,3 +228,25 @@ def test_external_references() -> None:
     output, unmapped = fix_refs(source, url_map.__getitem__)
     assert output == '<a class="autorefs autorefs-external" href="https://example.com">example</a>'
     assert unmapped == []
+
+
+def test_register_html_anchors() -> None:
+    """Check that HTML anchors are registered when enabled."""
+    plugin = AutorefsPlugin()
+    md = markdown.Markdown(extensions=["attr_list", AutorefsExtension(partial(AnchorScannerTreeProcessor, plugin))])
+    plugin.current_page = ""
+    md.convert(
+        dedent(
+            """
+            [](){#foo}
+            ## Heading
+
+            Paragraph 1.
+
+            [](){#bar}
+            Paragraph 2.
+            """,
+        ),
+    )
+    assert plugin._url_map["foo"] == "#heading"
+    assert plugin._url_map["bar"] == "#bar"