Clean up some more gratuitous differences in footnote parsing

* This reverses the choice to diverge from GitHub on the
  issue of back-to-back `[shortcut links][^and footnotes]`,
  since other implementations of this feature also do it.
* This stops pulldown-cmark from trying to parse `[^collapsed][]`
  footnote references. Those aren't a thing.

Fixes #772

specs/footnotes.txt

@@ -394,12 +394,13 @@ Songs that simply loop are a popular way to annoy people. [^examples3]
 Test case for the relationship between link references and footnotes.
  <https://gist.github.com/notriddle/5625a3c1cb70b3067a01d7465f9d10f1>
 
-In this case, pulldown-cmark treats all four `[cmark-gfm]` uses as links.
-GitHub, however, writes
+GitHub writes
 `My [cmark-gfm]<sup class="footnote-reference"><a href="#examples">1</a></sup>.`
-in the first paragraph. This seems like a bug, and would require some gymnastics for
-pulldown-cmark to copy it, but this test case is kept here to keep track of the decision to
-diverge from GFM's behavior.
+in the first paragraph. This seems like a bug, but [markdown-it] and [commonmark-hs]
+both do it, too, so it's probably not worth it to be different here.
+
+[markdown-it]: https://markdown-it.github.io/#md3=%7B%22source%22%3A%22%5B%5Efoo%5D%5B%5D%5Cn%5Cn%5B%5Efoo%5D%5Bbaz%5D%5Cn%5Cn%5Bbaz%5D%5B%5Efoo%5D%5Cn%5Cn%5B%5Efoo%5D%3A%20bar%5Cn%5Cn%5Bbaz%5D%3A%20https%3A%2F%2Frust-lang.org%22%2C%22defaults%22%3A%7B%22html%22%3Afalse%2C%22xhtmlOut%22%3Afalse%2C%22breaks%22%3Afalse%2C%22langPrefix%22%3A%22language-%22%2C%22linkify%22%3Atrue%2C%22typographer%22%3Atrue%2C%22_highlight%22%3Atrue%2C%22_strict%22%3Afalse%2C%22_view%22%3A%22html%22%7D%7D
+[commonmark-hs]: https://pandoc.org/try/?params=%7B%22text%22%3A%22%5B%5Efoo%5D%5B%5D%5Cn%5Cn%5B%5Efoo%5D%5Bbaz%5D%5Cn%5Cn%5Bbaz%5D%5B%5Efoo%5D%5Cn%5Cn%5B%5Efoo%5D%3A+bar%5Cn%5Cn%5Bbaz%5D%3A+https%3A%2F%2Frust-lang.org%22%2C%22to%22%3A%22html5%22%2C%22from%22%3A%22commonmark_x%22%2C%22standalone%22%3Afalse%2C%22embed-resources%22%3Afalse%2C%22table-of-contents%22%3Afalse%2C%22number-sections%22%3Afalse%2C%22citeproc%22%3Afalse%2C%22html-math-method%22%3A%22plain%22%2C%22wrap%22%3A%22auto%22%2C%22highlight-style%22%3Anull%2C%22files%22%3A%7B%7D%2C%22template%22%3Anull%7D
 
 ```````````````````````````````` example
 My [cmark-gfm][^c].
@@ -422,7 +423,7 @@ My [otherlink[^c]].
 
 [otherlink[^c]]: https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702
 .
-<p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a><sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
+<p>My [cmark-gfm]<sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
 <p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a><sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
 <p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a><sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
 <p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a> <sup class="footnote-reference"><a href="#c">1</a></sup>.</p>

specs/regression.txt

@@ -1668,3 +1668,24 @@ Second try]: https://rust-lang.org
 <h2>[First try</h2>
 <p>Second try]: https://rust-lang.org</p>
 ````````````````````````````````
+
+ISSUE 772
+
+```````````````````````````````` example
+[^foo][]
+
+[^foo][baz]
+
+[baz][^foo]
+
+[^foo]: bar
+
+[baz]: https://rust-lang.org
+.
+<p><sup class="footnote-reference"><a href="#foo">1</a></sup>[]</p>
+<p><a href="https://rust-lang.org">^foo</a></p>
+<p>[baz]<sup class="footnote-reference"><a href="#foo">1</a></sup></p>
+<div class="footnote-definition" id="foo"><sup class="footnote-definition-label">1</sup>
+<p>bar</p>
+</div>
+````````````````````````````````

src/parse.rs

@@ -435,6 +435,7 @@ impl<'input, F: BrokenLinkCallback<'input>> Parser<'input, F> {
                                     }
                                     (next, LinkType::Shortcut)
                                 }
+                                RefScan::UnexpectedFootnote => continue,
                             };
 
                             // FIXME: references and labels are mixed in the naming of variables
@@ -458,6 +459,7 @@ impl<'input, F: BrokenLinkCallback<'input>> Parser<'input, F> {
                                     .map(|(ix, label)| (label, label_start + ix))
                                     .filter(|(_, end)| *end == label_end)
                                 }
+                                RefScan::UnexpectedFootnote => continue,
                             };
 
                             let id = match &label {
@@ -479,13 +481,16 @@ impl<'input, F: BrokenLinkCallback<'input>> Parser<'input, F> {
                                 if !self.options.has_gfm_footnotes()
                                     || self.allocs.footdefs.contains(&self.allocs.cows[footref.0])
                                 {
-                                    self.tree[tos.node].next = node_after_link;
+                                    // use `next` instead of `node_after_link` because
+                                    // node_after_link is calculated for a [collapsed][] link,
+                                    // which footnotes don't support.
+                                    self.tree[tos.node].next = next;
                                     self.tree[tos.node].child = None;
                                     self.tree[tos.node].item.body =
                                         ItemBody::FootnoteReference(footref);
                                     self.tree[tos.node].item.end = end;
                                     prev = Some(tos.node);
-                                    cur = node_after_link;
+                                    cur = next;
                                     self.link_stack.clear();
                                     continue;
                                 }
@@ -1169,6 +1174,7 @@ enum RefScan<'a> {
     LinkLabel(CowStr<'a>, usize),
     // contains next node index
     Collapsed(Option<TreeIndex>),
+    UnexpectedFootnote,
     Failed,
 }
 
@@ -1238,12 +1244,13 @@ fn scan_reference<'a, 'b>(
         // TODO: this unwrap is sus and should be looked at closer
         let closing_node = tree[cur_ix].next.unwrap();
         RefScan::Collapsed(tree[closing_node].next)
-    } else if let Some((ix, ReferenceLabel::Link(label))) =
-        scan_link_label(tree, &text[start..], allow_footnote_refs, gfm_footnotes)
-    {
-        RefScan::LinkLabel(label, start + ix)
     } else {
-        RefScan::Failed
+        let label = scan_link_label(tree, &text[start..], allow_footnote_refs, gfm_footnotes);
+        match label {
+            Some((ix, ReferenceLabel::Link(label))) => RefScan::LinkLabel(label, start + ix),
+            Some((_ix, ReferenceLabel::Footnote(_label))) => RefScan::UnexpectedFootnote,
+            None => RefScan::Failed,
+        }
     }
 }
 

tests/suite/footnotes.rs

@@ -418,7 +418,7 @@ My [otherlink[^c]].
 
 [otherlink[^c]]: https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702
 "##;
-    let expected = r##"<p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a><sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
+    let expected = r##"<p>My [cmark-gfm]<sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
 <p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a><sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
 <p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a><sup class="footnote-reference"><a href="#c">1</a></sup>.</p>
 <p>My <a href="https://github.com/github/cmark-gfm/blob/1e230827a584ebc9938c3eadc5059c55ef3c9abf/test/extensions.txt#L702">cmark-gfm</a> <sup class="footnote-reference"><a href="#c">1</a></sup>.</p>

tests/suite/regression.rs

@@ -1955,3 +1955,26 @@ Second try]: https://rust-lang.org
 
     test_markdown_html(original, expected, false, false, false);
 }
+
+#[test]
+fn regression_test_124() {
+    let original = r##"[^foo][]
+
+[^foo][baz]
+
+[baz][^foo]
+
+[^foo]: bar
+
+[baz]: https://rust-lang.org
+"##;
+    let expected = r##"<p><sup class="footnote-reference"><a href="#foo">1</a></sup>[]</p>
+<p><a href="https://rust-lang.org">^foo</a></p>
+<p>[baz]<sup class="footnote-reference"><a href="#foo">1</a></sup></p>
+<div class="footnote-definition" id="foo"><sup class="footnote-definition-label">1</sup>
+<p>bar</p>
+</div>
+"##;
+
+    test_markdown_html(original, expected, false, false, false);
+}