Function first_invalid_escape

crates/vm/src/vm/compile.rs:92–170

Scan `source[start..end]` for the first invalid escape sequence. Returns `Some((invalid_char, byte_offset_in_source))` for the first invalid escape found, or `None` if all escapes are valid. When `is_bytes` is true, `\u`, `\U`, and `\N` are treated as invalid (bytes literals only support byte-oriented escapes). Only reports the **first** invalid escape per string literal, matching `_PyUnicode_De

(
        source: &str,
        start: usize,
        end: usize,
        is_bytes: bool,
    )

    /// `_PyUnicode_DecodeUnicodeEscapeInternal2` which stores only the first
    /// `first_invalid_escape_char`.
    fn first_invalid_escape(
        source: &str,
        start: usize,
        end: usize,
        is_bytes: bool,
    ) -> Option<(char, usize)> {
        let raw = &source[start..end];
        let mut chars = raw.char_indices().peekable();
        while let Some((i, ch)) = chars.next() {
            if ch != '\\' {
                continue;
            }
            let Some((_, next)) = chars.next() else {
                break;
            };
            let valid = match next {
                '\\' | '\'' | '"' | 'a' | 'b' | 'f' | 'n' | 'r' | 't' | 'v' => true,
                '\n' => true,
                '\r' => {
                    if matches!(chars.peek(), Some(&(_, '\n'))) {
                        chars.next();
                    }
                    true
                }
                '0'..='7' => {
                    for _ in 0..2 {
                        if matches!(chars.peek(), Some(&(_, '0'..='7'))) {
                            chars.next();
                        } else {
                            break;
                        }
                    }
                    true
                }
                'x' | 'u' | 'U' => {
                    // \u and \U are only valid in string literals, not bytes
                    if is_bytes && next != 'x' {
                        false
                    } else {
                        let count = match next {
                            'x' => 2,
                            'u' => 4,
                            'U' => 8,
                            _ => unreachable!(),
                        };
                        for _ in 0..count {
                            if chars.peek().is_some_and(|&(_, c)| c.is_ascii_hexdigit()) {
                                chars.next();
                            } else {
                                break;
                            }
                        }
                        true
                    }
                }
                'N' => {
                    // \N{name} is only valid in string literals, not bytes
                    if is_bytes {