Auto merge of rust-lang#137274 - yotamofek:pr/rustdoc/lazy-escape, r=…

…<try>

Allow lazy HTML escaping

Inspired by [this comment](rust-lang#136828 (comment)) by `@aDotInTheVoid` .
Makes `Escape` and `EscapeBodyText` accept any `impl fmt::Display`, instead of a `&str`, which allows us to avoid a few interim `String` allocations.
This opens up room for more lazifying, but I'll keep those for a separate PR.

Unfortunately, I think there might be a hit to performance because of the double vtable-indirection caused by wrapping a `fmt::Formatter` in another one, but I think that I should be able to gain that perf back by doing more lazy printing (either the small things improvements I made in this PR, or in later ones)

Probably better to review each commit individually.

src/librustdoc/clean/cfg.rs

@@ -189,14 +189,16 @@ impl Cfg {
     }
 
     /// Renders the configuration for long display, as a long plain text description.
-    pub(crate) fn render_long_plain(&self) -> String {
+    pub(crate) fn render_long_plain(&self) -> impl fmt::Display + '_ {
         let on = if self.should_use_with_in_description() { "with" } else { "on" };
 
-        let mut msg = format!("Available {on} {}", Display(self, Format::LongPlain));
-        if self.should_append_only_to_description() {
-            msg.push_str(" only");
-        }
-        msg
+        fmt::from_fn(move |f| {
+            write!(f, "Available {on} {}", Display(self, Format::LongPlain))?;
+            if self.should_append_only_to_description() {
+                f.write_str(" only")?;
+            }
+            Ok(())
+        })
     }
 
     fn should_capitalize_first_letter(&self) -> bool {

src/librustdoc/html/escape.rs

@@ -7,37 +7,58 @@ use std::fmt;
 
 use unicode_segmentation::UnicodeSegmentation;
 
+#[inline]
+fn escape(s: &str, mut w: impl fmt::Write, escape_quotes: bool) -> fmt::Result {
+    // Because the internet is always right, turns out there's not that many
+    // characters to escape: http://stackoverflow.com/questions/7381974
+    let pile_o_bits = s;
+    let mut last = 0;
+    for (i, ch) in s.char_indices() {
+        let s = match ch {
+            '>' => ">",
+            '<' => "&lt;",
+            '&' => "&amp;",
+            '\'' if escape_quotes => "&#39;",
+            '"' if escape_quotes => "&quot;",
+            _ => continue,
+        };
+        w.write_str(&pile_o_bits[last..i])?;
+        w.write_str(s)?;
+        // NOTE: we only expect single byte characters here - which is fine as long as we
+        // only match single byte characters
+        last = i + 1;
+    }
+
+    if last < s.len() {
+        w.write_str(&pile_o_bits[last..])?;
+    }
+    Ok(())
+}
+
+struct WriteEscaped<W: fmt::Write> {
+    writer: W,
+    escape_quotes: bool,
+}
+
+impl<W: fmt::Write> fmt::Write for WriteEscaped<W> {
+    #[inline]
+    fn write_str(&mut self, s: &str) -> fmt::Result {
+        escape(s, &mut self.writer, self.escape_quotes)
+    }
+}
+
 /// Wrapper struct which will emit the HTML-escaped version of the contained
 /// string when passed to a format string.
-pub(crate) struct Escape<'a>(pub &'a str);
+pub(crate) struct Escape<T>(pub T);
 
-impl fmt::Display for Escape<'_> {
+impl<T: fmt::Display> fmt::Display for Escape<T> {
+    #[inline]
     fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
-        // Because the internet is always right, turns out there's not that many
-        // characters to escape: http://stackoverflow.com/questions/7381974
-        let Escape(s) = *self;
-        let pile_o_bits = s;
-        let mut last = 0;
-        for (i, ch) in s.char_indices() {
-            let s = match ch {
-                '>' => "&gt;",
-                '<' => "&lt;",
-                '&' => "&amp;",
-                '\'' => "&#39;",
-                '"' => "&quot;",
-                _ => continue,
-            };
-            fmt.write_str(&pile_o_bits[last..i])?;
-            fmt.write_str(s)?;
-            // NOTE: we only expect single byte characters here - which is fine as long as we
-            // only match single byte characters
-            last = i + 1;
-        }
-
-        if last < s.len() {
-            fmt.write_str(&pile_o_bits[last..])?;
-        }
-        Ok(())
+        self.0.fmt(
+            &mut fmt
+                .options()
+                .create_formatter(&mut WriteEscaped { writer: fmt, escape_quotes: true }),
+        )
     }
 }
 
@@ -47,33 +68,15 @@ impl fmt::Display for Escape<'_> {
 /// This is only safe to use for text nodes. If you need your output to be
 /// safely contained in an attribute, use [`Escape`]. If you don't know the
 /// difference, use [`Escape`].
-pub(crate) struct EscapeBodyText<'a>(pub &'a str);
+pub(crate) struct EscapeBodyText<T>(pub T);
 
-impl fmt::Display for EscapeBodyText<'_> {
+impl<T: fmt::Display> fmt::Display for EscapeBodyText<T> {
     fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
-        // Because the internet is always right, turns out there's not that many
-        // characters to escape: http://stackoverflow.com/questions/7381974
-        let EscapeBodyText(s) = *self;
-        let pile_o_bits = s;
-        let mut last = 0;
-        for (i, ch) in s.char_indices() {
-            let s = match ch {
-                '>' => "&gt;",
-                '<' => "&lt;",
-                '&' => "&amp;",
-                _ => continue,
-            };
-            fmt.write_str(&pile_o_bits[last..i])?;
-            fmt.write_str(s)?;
-            // NOTE: we only expect single byte characters here - which is fine as long as we
-            // only match single byte characters
-            last = i + 1;
-        }
-
-        if last < s.len() {
-            fmt.write_str(&pile_o_bits[last..])?;
-        }
-        Ok(())
+        self.0.fmt(
+            &mut fmt
+                .options()
+                .create_formatter(&mut WriteEscaped { writer: fmt, escape_quotes: false }),
+        )
     }
 }
 

src/librustdoc/html/escape/tests.rs

@@ -1,3 +1,11 @@
+use std::iter;
+
+#[test]
+fn escape() {
+    use super::Escape as E;
+    assert_eq!(format!("<Hello> {}", E("<World>")), "<Hello> &lt;World&gt;");
+}
+
 // basic examples
 #[test]
 fn escape_body_text_with_wbr() {
@@ -47,21 +55,8 @@ fn escape_body_text_with_wbr_makes_sense() {
     use itertools::Itertools as _;
 
     use super::EscapeBodyTextWithWbr as E;
-    const C: [u8; 3] = [b'a', b'A', b'_'];
-    for chars in [
-        C.into_iter(),
-        C.into_iter(),
-        C.into_iter(),
-        C.into_iter(),
-        C.into_iter(),
-        C.into_iter(),
-        C.into_iter(),
-        C.into_iter(),
-    ]
-    .into_iter()
-    .multi_cartesian_product()
-    {
-        let s = String::from_utf8(chars).unwrap();
+    for chars in iter::repeat("aA_").take(8).map(str::chars).multi_cartesian_product() {
+        let s = chars.into_iter().collect::<String>();
         assert_eq!(s.len(), 8);
         let esc = E(&s).to_string();
         assert!(!esc.contains("<wbr><wbr>"));

src/librustdoc/html/format.rs

@@ -853,7 +853,7 @@ pub(crate) fn anchor<'a: 'cx, 'cx>(
                 f,
                 r#"<a class="{short_ty}" href="{url}" title="{short_ty} {path}">{text}</a>"#,
                 path = join_with_double_colon(&fqp),
-                text = EscapeBodyText(text.as_str()),
+                text = EscapeBodyText(text),
             )
         } else {
             f.write_str(text.as_str())

src/librustdoc/html/render/mod.rs

@@ -720,7 +720,7 @@ fn short_item_info(
             }
             DeprecatedSince::Future => String::from("Deprecating in a future version"),
             DeprecatedSince::NonStandard(since) => {
-                format!("Deprecated since {}", Escape(since.as_str()))
+                format!("Deprecated since {}", Escape(since))
             }
             DeprecatedSince::Unspecified | DeprecatedSince::Err => String::from("Deprecated"),
         };
@@ -1519,7 +1519,7 @@ pub(crate) fn notable_traits_button<'a, 'tcx>(
             write!(
                 f,
                 " <a href=\"#\" class=\"tooltip\" data-notable-ty=\"{ty}\">ⓘ</a>",
-                ty = Escape(&format!("{:#}", ty.print(cx))),
+                ty = Escape(ty.print(cx)),
             )
         })
     })

src/librustdoc/html/render/print_item.rs

@@ -534,12 +534,16 @@ fn extra_info_tags<'a, 'tcx: 'a>(
     import_def_id: Option<DefId>,
 ) -> impl Display + 'a + Captures<'tcx> {
     fmt::from_fn(move |f| {
-        fn tag_html<'a>(class: &'a str, title: &'a str, contents: &'a str) -> impl Display + 'a {
+        fn tag_html<'a>(
+            class: impl fmt::Display + 'a,
+            title: impl fmt::Display + 'a,
+            contents: impl fmt::Display + 'a,
+        ) -> impl Display + 'a {
             fmt::from_fn(move |f| {
                 write!(
                     f,
                     r#"<wbr><span class="stab {class}" title="{title}">{contents}</span>"#,
-                    title = Escape(title),
+                    title = Escape(&title),
                 )
             })
         }

src/librustdoc/lib.rs

@@ -8,6 +8,7 @@
 #![feature(debug_closure_helpers)]
 #![feature(file_buffered)]
 #![feature(format_args_nl)]
+#![feature(formatting_options)]
 #![feature(if_let_guard)]
 #![feature(impl_trait_in_assoc_type)]
 #![feature(iter_intersperse)]