Rollup merge of #77971 - jyn514:broken-intra-doc-links, r=mark-simulacrum

Deny broken intra-doc links in linkchecker

Since rustdoc isn't warning about these links, check for them manually.

This also fixes the broken links that popped up from the lint.

5 files changed, 68 insertions, 0 deletions

diff --git a/src/doc/embedded-book b/src/doc/embedded-book
-Subproject 79ab7776929c66db83203397958fa7037d5d9a3
+Subproject ca8169e69b479f615855d0eece7e318138fcfc0
diff --git a/src/doc/unstable-book/src/library-features/default-free-fn.md b/src/doc/unstable-book/src/library-features/default-free-fn.md
index 5dff73a94dd..d40a27dddf3 100644
--- a/src/doc/unstable-book/src/library-features/default-free-fn.md
+++ b/src/doc/unstable-book/src/library-features/default-free-fn.md
@@ -10,6 +10,8 @@ Adds a free `default()` function to the `std::default` module.  This function
 just forwards to [`Default::default()`], but may remove repetition of the word
 "default" from the call site.
 
+[`Default::default()`]: https://doc.rust-lang.org/nightly/std/default/trait.Default.html#tymethod.default
+
 Here is an example:
 
 ```rust
diff --git a/src/tools/cargo b/src/tools/cargo
-Subproject 12db56cdedbc2c26a9aa18f994c0188cdcc67df
+Subproject 79b397d72c557eb6444a2ba0dc00a211a226a35
diff --git a/src/tools/linkchecker/Cargo.toml b/src/tools/linkchecker/Cargo.toml
index 0994cd20662..e5d870c039d 100644
--- a/src/tools/linkchecker/Cargo.toml
+++ b/src/tools/linkchecker/Cargo.toml
@@ -7,3 +7,7 @@ edition = "2018"
 [[bin]]
 name = "linkchecker"
 path = "main.rs"
+
+[dependencies]
+regex = "1"
+once_cell = "1"
diff --git a/src/tools/linkchecker/main.rs b/src/tools/linkchecker/main.rs
index 4fe493a850d..f213944e0ab 100644
--- a/src/tools/linkchecker/main.rs
+++ b/src/tools/linkchecker/main.rs
@@ -21,6 +21,9 @@ use std::fs;
 use std::path::{Component, Path, PathBuf};
 use std::rc::Rc;
 
+use once_cell::sync::Lazy;
+use regex::Regex;
+
 use crate::Redirect::*;
 
 // Add linkcheck exceptions here
@@ -50,6 +53,44 @@ const LINKCHECK_EXCEPTIONS: &[(&str, &[&str])] = &[
     ("alloc/collections/btree_set/struct.BTreeSet.html", &["#insert-and-complex-keys"]),
 ];
 
+#[rustfmt::skip]
+const INTRA_DOC_LINK_EXCEPTIONS: &[(&str, &[&str])] = &[
+    // This will never have links that are not in other pages.
+    // To avoid repeating the exceptions twice, an empty list means all broken links are allowed.
+    ("reference/print.html", &[]),
+    // All the reference 'links' are actually ENBF highlighted as code
+    ("reference/comments.html", &[
+         "/</code> <code>!",
+         "*</code> <code>!",
+    ]),
+    ("reference/identifiers.html", &[
+         "a</code>-<code>z</code> <code>A</code>-<code>Z",
+         "a</code>-<code>z</code> <code>A</code>-<code>Z</code> <code>0</code>-<code>9</code> <code>_",
+         "a</code>-<code>z</code> <code>A</code>-<code>Z</code>] [<code>a</code>-<code>z</code> <code>A</code>-<code>Z</code> <code>0</code>-<code>9</code> <code>_",
+    ]),
+    ("reference/tokens.html", &[
+         "0</code>-<code>1",
+         "0</code>-<code>7",
+         "0</code>-<code>9",
+         "0</code>-<code>9",
+         "0</code>-<code>9</code> <code>a</code>-<code>f</code> <code>A</code>-<code>F",
+    ]),
+    ("reference/notation.html", &[
+         "b</code> <code>B",
+         "a</code>-<code>z",
+    ]),
+    // This is being used in the sense of 'inclusive range', not a markdown link
+    ("core/ops/struct.RangeInclusive.html", &["begin</code>, <code>end"]),
+    ("std/ops/struct.RangeInclusive.html", &["begin</code>, <code>end"]),
+    ("core/slice/trait.SliceIndex.html", &["begin</code>, <code>end"]),
+    ("alloc/slice/trait.SliceIndex.html", &["begin</code>, <code>end"]),
+    ("std/slice/trait.SliceIndex.html", &["begin</code>, <code>end"]),
+
+];
+
+static BROKEN_INTRA_DOC_LINK: Lazy<Regex> =
+    Lazy::new(|| Regex::new(r#"\[<code>(.*)</code>\]"#).unwrap());
+
 macro_rules! t {
     ($e:expr) => {
         match $e {
@@ -138,6 +179,14 @@ fn walk(cache: &mut Cache, root: &Path, dir: &Path, errors: &mut bool) {
     }
 }
 
+fn is_intra_doc_exception(file: &Path, link: &str) -> bool {
+    if let Some(entry) = INTRA_DOC_LINK_EXCEPTIONS.iter().find(|&(f, _)| file.ends_with(f)) {
+        entry.1.is_empty() || entry.1.contains(&link)
+    } else {
+        false
+    }
+}
+
 fn is_exception(file: &Path, link: &str) -> bool {
     if let Some(entry) = LINKCHECK_EXCEPTIONS.iter().find(|&(f, _)| file.ends_with(f)) {
         entry.1.contains(&link)
@@ -292,6 +341,19 @@ fn check(cache: &mut Cache, root: &Path, file: &Path, errors: &mut bool) -> Opti
             }
         }
     });
+
+    // Search for intra-doc links that rustdoc didn't warn about
+    // FIXME(#77199, 77200) Rustdoc should just warn about these directly.
+    // NOTE: only looks at one line at a time; in practice this should find most links
+    for (i, line) in contents.lines().enumerate() {
+        for broken_link in BROKEN_INTRA_DOC_LINK.captures_iter(line) {
+            if !is_intra_doc_exception(file, &broken_link[1]) {
+                *errors = true;
+                print!("{}:{}: broken intra-doc link - ", pretty_file.display(), i + 1);
+                println!("{}", &broken_link[0]);
+            }
+        }
+    }
     Some(pretty_file)
 }