json-five 0.3.1

pub(crate) fn get_line_col_char(doc: &str, byte_offset: usize) -> (usize, usize, usize) {
    let mut lineno: usize = 1;
    let mut colno: usize = 0;
    if byte_offset == 0 {
        return (1, 1, 0)
    }

    assert!(byte_offset <= doc.len(), "requested byteoffset {} is not less than or equal to doc length ({})", byte_offset, doc.len());

    if byte_offset == doc.len() {
        let last_char_pos = doc.char_indices().last().unwrap(); // (byte_off, char)
        let (lineno, mut colno, mut codepoint_off) = get_line_col_char(doc, last_char_pos.0);
        colno += 1;
        codepoint_off += 1;
        return (lineno, colno, codepoint_off);
    }

    for (codepoint_off, (byte_off, char)) in doc.char_indices().enumerate() {
        colno += 1;
        if char == '\n' {
            if byte_off == byte_offset {
                return (lineno, colno, codepoint_off)
            }
            lineno += 1;
            colno = 0;
        }
        if byte_off < byte_offset {
            continue
        }
        if byte_off == byte_offset {
            return (lineno, colno, codepoint_off)
        }
        if byte_off > byte_offset {
            unreachable!("Byteoffset lands in the middle of a character")
        }
    }
    unreachable!("Reached end of document")
}


pub (crate) fn escape_double_quoted(input: &str) -> String {
    // In the worst case (every char requires a backslash), the output could
    // be roughly twice the length of `input`.
    let mut escaped = String::with_capacity(input.len() * 2);

    for c in input.chars() {
        match c {
            '"'  => { escaped.push('\\'); escaped.push('"');  }
            '\\' => { escaped.push('\\'); escaped.push('\\'); }
            '\n' => { escaped.push('\\'); escaped.push('n');  }
            '\r' => { escaped.push('\\'); escaped.push('r');  }
            '\t' => { escaped.push('\\'); escaped.push('t');  }
            '/'  => { escaped.push('\\'); escaped.push('/');  }
            '\u{0008}' => { escaped.push('\\'); escaped.push('b'); }
            '\u{000c}' => { escaped.push('\\'); escaped.push('f'); }
            '\u{2028}' => { escaped.push_str("\\u2028"); }
            '\u{2029}' => { escaped.push_str("\\u2029"); }
            _ => escaped.push(c),
        }
    }

    escaped
}

#[allow(dead_code)]
pub (crate) fn escape_single_quoted(input: &str) -> String {
    let mut escaped = String::with_capacity(input.len() * 2);
    for c in input.chars() {
        match c {
            '\''  => { escaped.push('\\'); escaped.push('\'');  }
            '\\' => { escaped.push('\\'); escaped.push('\\'); }
            '\n' => { escaped.push('\\'); escaped.push('n');  }
            '\r' => { escaped.push('\\'); escaped.push('r');  }
            '\t' => { escaped.push('\\'); escaped.push('t');  }
            '/'  => { escaped.push('\\'); escaped.push('/');  }
            '\u{0008}' => { escaped.push('\\'); escaped.push('b'); }
            '\u{000c}' => { escaped.push('\\'); escaped.push('f'); }
            '\u{2028}' => { escaped.push_str("\\u2028"); }
            '\u{2029}' => { escaped.push_str("\\u2029"); }
            _ => escaped.push(c),
        }
    }
    escaped
}

/// Unescape a JSON5-style (ES5.1-like) string.
/// - `\uXXXX` => exactly 4 hex digits (no `\u{...}`).
/// - `\xNN`   => exactly 2 hex digits.
/// - Single-char escapes like `\n`, `\r`, `\t`, `\b`, `\f`, `\v`, `\0`, `\\`, `\"`, `\/`.
/// Returns an owned `String` with all escapes resolved.
pub fn unescape(input: &str) -> Result<String, String> {
    let mut output = String::with_capacity(input.len());
    let mut chars = input.chars().peekable();

    while let Some(ch) = chars.next() {
        if ch != '\\' {
            // Normal character
            output.push(ch);
        } else {
            // We have a backslash; look at the next char
            let esc = chars.next().ok_or_else(|| err("Incomplete escape at end of string"))?;
            match esc {
                'a' => output.push('\x07'),
                'n' =>  output.push('\n'),
                'r' =>  output.push('\r'),
                't' =>  output.push('\t'),
                'b' =>  output.push('\x08'), // backspace
                'f' =>  output.push('\x0C'), // form-feed
                'v' =>  output.push('\x0B'),
                '0' =>  output.push('\0'),
                '\\' => output.push('\\'),
                '\'' => output.push('\''),
                '"'  => output.push('\"'),
                '/'  => output.push('/'), // optional in JSON5
                '\n' | '\r' | '\u{2028}' | '\u{2029}' => {
                    output.push(esc);
                }
                'x' => {
                    // \xNN => exactly 2 hex digits
                    let val = read_hex_digits(&mut chars, 2, "\\x")?;
                    output.push(char_from_u32(val)?);
                }
                'u' => {
                    // \uXXXX => exactly 4 hex digits (ES5.1)
                    let val = read_hex_digits(&mut chars, 4, "\\u")?;
                    output.push(char_from_u32(val)?);
                }
                _ => {
                    // Unknown escape
                    return Err(format!("Unknown escape character: {esc}"));
                }
            }
        }
    }
    Ok(output)
}

/// Read exactly `count` hex digits from `chars`, returning the combined u32.
/// The `context` helps produce a clearer error message (like `"\u"` or `"\x"`).
pub (crate) fn read_hex_digits<I: Iterator<Item = char>>(
    chars: &mut std::iter::Peekable<I>,
    count: usize,
    context: &str
) -> Result<u32, String> {
    let mut val = 0u32;
    for _ in 0..count {
        let c = chars.next().ok_or_else(|| err(format!("Incomplete {context} escape")))?;
        let digit = c
            .to_digit(16)
            .ok_or_else(|| err(format!("Invalid hex digit '{c}' in {context} escape")))?;
        val = (val << 4) | digit;
    }
    Ok(val)
}

/// Convert a u32 to a `char`, checking for valid Unicode scalar value range.
fn char_from_u32(u: u32) -> Result<char, String> {
    // In ES5.1, `\uXXXX` covers 0..=0xFFFF. If you need to disallow >0xFFFF, insert a check:
    // if u > 0xFFFF { return Err(err(format!("Code point out of range: U+{:X}", u))); }
    std::char::from_u32(u).ok_or_else(|| err(format!("Invalid Unicode code point U+{u:X}")))
}

fn err<S: Into<String>>(message: S) -> String {
    message.into()
}


#[cfg(all(not(feature = "unlimited_depth"), not(target_os = "windows"), debug_assertions))]
pub (crate) const MAX_DEPTH: usize = 1000;

#[cfg(all(not(feature = "unlimited_depth"), not(target_os = "windows"), not(debug_assertions)))]
pub (crate) const MAX_DEPTH: usize = 3000;


#[cfg(all(not(feature = "unlimited_depth"), target_os = "windows", debug_assertions))]
pub (crate) const MAX_DEPTH: usize = 700;

#[cfg(all(not(feature = "unlimited_depth"), target_os = "windows", not(debug_assertions)))]
pub (crate) const MAX_DEPTH: usize = 2000;

#[cfg(feature = "unlimited_depth")]
pub (crate) const MAX_DEPTH: usize = usize::MAX;

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_escape_single_quote_bug() {
        let input = "Hello'World";
        let result = escape_single_quoted(input);
        let expected = "Hello\\'World";
        assert_eq!(result, expected, "Single quote should be escaped as \\' not \\\"");
    }

    #[test]
    fn test_escape_double_quoted_comprehensive() {
        let input = "Hello\"World\n\t\r\\";
        let result = escape_double_quoted(input);
        let expected = "Hello\\\"World\\n\\t\\r\\\\";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_escape_single_quoted_comprehensive() {
        let input = "Hello'World\n\t\r\\";
        let result = escape_single_quoted(input);
        let expected = "Hello\\'World\\n\\t\\r\\\\";
        assert_eq!(result, expected);
    }

    #[test] 
    fn test_unescape_basic_escapes() {
        let input = "Hello\\nWorld\\t\\r\\\\";
        let result = unescape(input).unwrap();
        let expected = "Hello\nWorld\t\r\\";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_unescape_quotes() {
        let input = "He said \\\"Hello\\\" and she said \\'Hi\\'";
        let result = unescape(input).unwrap();
        let expected = "He said \"Hello\" and she said 'Hi'";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_unescape_unicode_valid() {
        let input = "Unicode: \\u0041\\u0042\\u2764";
        let result = unescape(input).unwrap();
        let expected = "Unicode: AB❤";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_unescape_hex_valid() {
        let input = "Hex: \\x41\\x42\\x21";
        let result = unescape(input).unwrap();
        let expected = "Hex: AB!";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_unescape_invalid_unicode_short() {
        let input = "Invalid: \\u12G";  // Invalid hex digit
        let result = unescape(input);
        assert!(result.is_err(), "Should fail on invalid unicode escape");
    }

    #[test]
    fn test_unescape_invalid_unicode_incomplete() {
        let input = "Incomplete: \\u123";  // Too few digits
        let result = unescape(input);
        assert!(result.is_err(), "Should fail on incomplete unicode escape");
    }

    #[test]
    fn test_unescape_invalid_hex_char() {
        let input = "Invalid hex: \\xZZ";
        let result = unescape(input);
        assert!(result.is_err(), "Should fail on invalid hex escape");
    }

    #[test]
    fn test_unescape_invalid_hex_incomplete() {
        let input = "Incomplete hex: \\x1";
        let result = unescape(input);
        assert!(result.is_err(), "Should fail on incomplete hex escape");
    }

    #[test]
    fn test_unescape_unknown_escape() {
        let input = "Unknown: \\z";
        let result = unescape(input);
        assert!(result.is_err(), "Should fail on unknown escape sequence");
    }

    #[test]
    fn test_unescape_incomplete_escape_at_end() {
        let input = "Incomplete: \\";
        let result = unescape(input);
        assert!(result.is_err(), "Should fail on incomplete escape at end");
    }

    #[test]
    fn test_unescape_line_continuation() {
        let input = "Line\\ncontinuation";
        let result = unescape(input).unwrap();
        let expected = "Line\ncontinuation";
        assert_eq!(result, expected);
    }

    #[test] 
    fn test_unescape_all_special_chars() {
        let input = "\\a\\b\\f\\n\\r\\t\\v\\0\\\\\\'\\\"";
        let result = unescape(input).unwrap();
        let expected = "\x07\x08\x0C\n\r\t\x0B\0\\'\"";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_unescape_unicode_line_separators() {
        let input = "\\u2028\\u2029";  // Line separator, Paragraph separator
        let result = unescape(input).unwrap();
        let expected = "\u{2028}\u{2029}";
        assert_eq!(result, expected);
    }

    #[test]
    fn test_read_hex_digits_valid() {
        let mut chars = "ABCD".chars().peekable();
        let result = read_hex_digits(&mut chars, 4, "\\u").unwrap();
        assert_eq!(result, 0xABCD);
    }

    #[test]
    fn test_read_hex_digits_invalid_char() {
        let mut chars = "12G4".chars().peekable();
        let result = read_hex_digits(&mut chars, 4, "\\u");
        assert!(result.is_err(), "Should fail on invalid hex character");
    }

    #[test]
    fn test_read_hex_digits_incomplete() {
        let mut chars = "12".chars().peekable();
        let result = read_hex_digits(&mut chars, 4, "\\u");
        assert!(result.is_err(), "Should fail on incomplete hex sequence");
    }
}