Rollup merge of rust-lang#123769 - dtolnay:literal, r=fee1-dead

Improve escaping of byte, byte str, and c str proc-macro literals

This PR changes the behavior of `proc_macro::Literal::byte_character` (rust-lang#115268), `byte_string`, and `c_string` (rust-lang#119750) to improve their choice of escape sequences. 3 categories of changes are made:

1. Never use `\x00`. Always prefer `\0`, which is supported in all the same places.

2. Never escape `\'` inside double quotes and `\"` inside single quotes.

3. Never use `\x` for valid UTF-8 in literals that permit `\u`.

The second commit adds tests covering these cases, asserting the **old** behavior.

The third commit implements the behavior change and simultaneously updates the tests to assert the **new** behavior.

Author: matthiaskrgr

proc_macro/src/escape.rs

@@ -0,0 +1,57 @@
+#[derive(Copy, Clone)]
+pub(crate) struct EscapeOptions {
+    /// Produce \'.
+    pub escape_single_quote: bool,
+    /// Produce \".
+    pub escape_double_quote: bool,
+    /// Produce \x escapes for non-ASCII, and use \x rather than \u for ASCII
+    /// control characters.
+    pub escape_nonascii: bool,
+}
+
+pub(crate) fn escape_bytes(bytes: &[u8], opt: EscapeOptions) -> String {
+    let mut repr = String::new();
+
+    if opt.escape_nonascii {
+        for &byte in bytes {
+            escape_single_byte(byte, opt, &mut repr);
+        }
+    } else {
+        let mut chunks = bytes.utf8_chunks();
+        while let Some(chunk) = chunks.next() {
+            for ch in chunk.valid().chars() {
+                escape_single_char(ch, opt, &mut repr);
+            }
+            for &byte in chunk.invalid() {
+                escape_single_byte(byte, opt, &mut repr);
+            }
+        }
+    }
+
+    repr
+}
+
+fn escape_single_byte(byte: u8, opt: EscapeOptions, repr: &mut String) {
+    if byte == b'\0' {
+        repr.push_str("\\0");
+    } else if (byte == b'\'' && !opt.escape_single_quote)
+        || (byte == b'"' && !opt.escape_double_quote)
+    {
+        repr.push(byte as char);
+    } else {
+        // Escapes \t, \r, \n, \\, \', \", and uses \x## for non-ASCII and
+        // for ASCII control characters.
+        repr.extend(byte.escape_ascii().map(char::from));
+    }
+}
+
+fn escape_single_char(ch: char, opt: EscapeOptions, repr: &mut String) {
+    if (ch == '\'' && !opt.escape_single_quote) || (ch == '"' && !opt.escape_double_quote) {
+        repr.push(ch);
+    } else {
+        // Escapes \0, \t, \r, \n, \\, \', \", and uses \u{...} for
+        // non-printable characters and for Grapheme_Extend characters, which
+        // includes things like U+0300 "Combining Grave Accent".
+        repr.extend(ch.escape_debug());
+    }
+}

proc_macro/src/lib.rs

@@ -43,10 +43,12 @@
 pub mod bridge;
 
 mod diagnostic;
+mod escape;
 
 #[unstable(feature = "proc_macro_diagnostic", issue = "54140")]
 pub use diagnostic::{Diagnostic, Level, MultiSpan};
 
+use crate::escape::{escape_bytes, EscapeOptions};
 use std::ffi::CStr;
 use std::ops::{Range, RangeBounds};
 use std::path::PathBuf;
@@ -1356,40 +1358,61 @@ impl Literal {
     /// String literal.
     #[stable(feature = "proc_macro_lib2", since = "1.29.0")]
     pub fn string(string: &str) -> Literal {
-        let quoted = format!("{:?}", string);
-        assert!(quoted.starts_with('"') && quoted.ends_with('"'));
-        let symbol = &quoted[1..quoted.len() - 1];
-        Literal::new(bridge::LitKind::Str, symbol, None)
+        let escape = EscapeOptions {
+            escape_single_quote: false,
+            escape_double_quote: true,
+            escape_nonascii: false,
+        };
+        let repr = escape_bytes(string.as_bytes(), escape);
+        Literal::new(bridge::LitKind::Str, &repr, None)
     }
 
     /// Character literal.
     #[stable(feature = "proc_macro_lib2", since = "1.29.0")]
     pub fn character(ch: char) -> Literal {
-        let quoted = format!("{:?}", ch);
-        assert!(quoted.starts_with('\'') && quoted.ends_with('\''));
-        let symbol = &quoted[1..quoted.len() - 1];
-        Literal::new(bridge::LitKind::Char, symbol, None)
+        let escape = EscapeOptions {
+            escape_single_quote: true,
+            escape_double_quote: false,
+            escape_nonascii: false,
+        };
+        let repr = escape_bytes(ch.encode_utf8(&mut [0u8; 4]).as_bytes(), escape);
+        Literal::new(bridge::LitKind::Char, &repr, None)
     }
 
     /// Byte character literal.
     #[stable(feature = "proc_macro_byte_character", since = "1.79.0")]
     pub fn byte_character(byte: u8) -> Literal {
-        let string = [byte].escape_ascii().to_string();
-        Literal::new(bridge::LitKind::Byte, &string, None)
+        let escape = EscapeOptions {
+            escape_single_quote: true,
+            escape_double_quote: false,
+            escape_nonascii: true,
+        };
+        let repr = escape_bytes(&[byte], escape);
+        Literal::new(bridge::LitKind::Byte, &repr, None)
     }
 
     /// Byte string literal.
     #[stable(feature = "proc_macro_lib2", since = "1.29.0")]
     pub fn byte_string(bytes: &[u8]) -> Literal {
-        let string = bytes.escape_ascii().to_string();
-        Literal::new(bridge::LitKind::ByteStr, &string, None)
+        let escape = EscapeOptions {
+            escape_single_quote: false,
+            escape_double_quote: true,
+            escape_nonascii: true,
+        };
+        let repr = escape_bytes(bytes, escape);
+        Literal::new(bridge::LitKind::ByteStr, &repr, None)
     }
 
     /// C string literal.
     #[stable(feature = "proc_macro_c_str_literals", since = "1.79.0")]
     pub fn c_string(string: &CStr) -> Literal {
-        let string = string.to_bytes().escape_ascii().to_string();
-        Literal::new(bridge::LitKind::CStr, &string, None)
+        let escape = EscapeOptions {
+            escape_single_quote: false,
+            escape_double_quote: true,
+            escape_nonascii: false,
+        };
+        let repr = escape_bytes(string.to_bytes(), escape);
+        Literal::new(bridge::LitKind::CStr, &repr, None)
     }
 
     /// Returns the span encompassing this literal.