coreutils 0.11.0

// This file is part of the uutils coreutils package.
//
// For the full copyright and license information, please view the LICENSE
// file that was distributed with this source code.
// spell-checker:ignore αbcdef ; (people) kkos
// spell-checker:ignore aabcccd aabcd aabd abbb abbbd abbcabc abbcac abbcbbbd abbcbd
// spell-checker:ignore abbccd abcabc abcac acabc andand bigcmp bignum emptysub
// spell-checker:ignore orempty oror bcdef fedcb GHSA xjfw pqfw

use uutests::new_ucmd;

#[test]
fn test_no_arguments() {
    new_ucmd!()
        .fails_with_code(2)
        .usage_error("missing operand");
}

#[test]
fn test_simple_values() {
    // null or 0 => EXIT_VALUE == 1
    new_ucmd!().args(&[""]).fails_with_code(1).stdout_only("\n");
    new_ucmd!()
        .args(&["0"])
        .fails_with_code(1)
        .stdout_only("0\n");
    new_ucmd!()
        .args(&["00"])
        .fails_with_code(1)
        .stdout_only("00\n");
    new_ucmd!()
        .args(&["-0"])
        .fails_with_code(1)
        .stdout_only("-0\n");

    // non-null and non-0 => EXIT_VALUE = 0
    new_ucmd!().args(&["1"]).succeeds().stdout_only("1\n");
}

#[test]
fn test_simple_arithmetic() {
    new_ucmd!()
        .args(&["1", "+", "1"])
        .succeeds()
        .stdout_only("2\n");

    new_ucmd!()
        .args(&["1", "-", "1"])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["3", "*", "2"])
        .succeeds()
        .stdout_only("6\n");

    new_ucmd!()
        .args(&["4", "/", "2"])
        .succeeds()
        .stdout_only("2\n");

    new_ucmd!()
        .args(&["4", "=", "2"])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["4", "=", "4"])
        .succeeds()
        .stdout_only("1\n");
}

#[test]
fn test_complex_arithmetic() {
    new_ucmd!()
        .args(&["9223372036854775807", "+", "9223372036854775807"])
        .succeeds()
        .stdout_only("18446744073709551614\n");

    new_ucmd!()
        .args(&[
            "92233720368547758076549841651981984981498415651",
            "%",
            "922337203685",
        ])
        .succeeds()
        .stdout_only("533691697086\n");

    new_ucmd!()
        .args(&[
            "92233720368547758076549841651981984981498415651",
            "*",
            "922337203685",
        ])
        .succeeds()
        .stdout_only("85070591730190566808700855121818604965830915152801178873935\n");

    new_ucmd!()
        .args(&[
            "92233720368547758076549841651981984981498415651",
            "-",
            "922337203685",
        ])
        .succeeds()
        .stdout_only("92233720368547758076549841651981984059161211966\n");

    new_ucmd!()
        .args(&["9", "/", "0"])
        .fails()
        .stderr_only("expr: division by zero\n");
}

#[test]
fn test_parenthesis() {
    new_ucmd!()
        .args(&["(", "1", "+", "1", ")", "*", "2"])
        .succeeds()
        .stdout_only("4\n");

    new_ucmd!()
        .args(&["1", "(", ")"])
        .fails_with_code(2)
        .stderr_only("expr: syntax error: unexpected argument '('\n");
}

#[test]
fn test_or() {
    new_ucmd!()
        .args(&["0", "|", "foo"])
        .succeeds()
        .stdout_only("foo\n");

    new_ucmd!()
        .args(&["foo", "|", "bar"])
        .succeeds()
        .stdout_only("foo\n");

    new_ucmd!()
        .args(&["14", "|", "1"])
        .succeeds()
        .stdout_only("14\n");

    new_ucmd!()
        .args(&["-14", "|", "1"])
        .succeeds()
        .stdout_only("-14\n");

    new_ucmd!()
        .args(&["1", "|", "a", "/", "5"])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["foo", "|", "a", "/", "5"])
        .succeeds()
        .stdout_only("foo\n");

    new_ucmd!()
        .args(&["0", "|", "10", "/", "5"])
        .succeeds()
        .stdout_only("2\n");

    new_ucmd!()
        .args(&["12", "|", "9a", "+", "1"])
        .succeeds()
        .stdout_only("12\n");

    new_ucmd!().args(&["", "|", ""]).fails().stdout_only("0\n");

    new_ucmd!().args(&["", "|", "0"]).fails().stdout_only("0\n");

    new_ucmd!()
        .args(&["", "|", "00"])
        .fails()
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["", "|", "-0"])
        .fails()
        .stdout_only("0\n");
}

#[test]
fn test_and() {
    new_ucmd!()
        .args(&["foo", "&", "1"])
        .succeeds()
        .stdout_only("foo\n");

    new_ucmd!()
        .args(&["14", "&", "1"])
        .succeeds()
        .stdout_only("14\n");

    new_ucmd!()
        .args(&["-14", "&", "1"])
        .succeeds()
        .stdout_only("-14\n");

    new_ucmd!()
        .args(&["-1", "&", "10", "/", "5"])
        .succeeds()
        .stdout_only("-1\n");

    new_ucmd!()
        .args(&["0", "&", "a", "/", "5"])
        .fails()
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["", "&", "a", "/", "5"])
        .fails()
        .stdout_only("0\n");

    new_ucmd!().args(&["", "&", "1"]).fails().stdout_only("0\n");

    new_ucmd!().args(&["", "&", ""]).fails().stdout_only("0\n");
}

#[test]
fn test_parenthesized_short_circuit_dead_branches() {
    new_ucmd!()
        .args(&["1", "|", "(", "1", "/", "0", ")"])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["0", "&", "(", "1", "/", "0", ")"])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["1", "|", "(", "0", "&", "(", "1", "/", "0", ")", ")"])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["0", "&", "(", "1", "|", "(", "1", "/", "0", ")", ")"])
        .fails_with_code(1)
        .stdout_only("0\n");
}

#[test]
fn test_length_fail() {
    new_ucmd!().args(&["length", "αbcdef", "1"]).fails();
}

#[test]
fn test_length() {
    new_ucmd!()
        .args(&["length", "abcdef"])
        .succeeds()
        .stdout_only("6\n");

    new_ucmd!()
        .args(&["abcdef", "length"])
        .fails_with_code(2)
        .stderr_only("expr: syntax error: unexpected argument 'length'\n");
}

#[test]
fn test_regex_empty() {
    new_ucmd!().args(&["", ":", ""]).fails().stdout_only("0\n");
    new_ucmd!()
        .args(&["abc", ":", ""])
        .fails()
        .stdout_only("0\n");
}

#[test]
fn test_regex_trailing_backslash() {
    new_ucmd!()
        .args(&["\\", ":", "\\\\"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["\\", ":", "\\"])
        .fails()
        .stderr_only("expr: Trailing backslash\n");
    new_ucmd!()
        .args(&["abc\\", ":", "abc\\\\"])
        .succeeds()
        .stdout_only("4\n");
    new_ucmd!()
        .args(&["abc\\", ":", "abc\\"])
        .fails()
        .stderr_only("expr: Trailing backslash\n");
}

#[test]
// expr mismatches `\[^a]` against `[^a]` on Windows for arm64 only; it passes
// everywhere else, including x86_64 Windows.
#[cfg_attr(
    all(windows, target_arch = "aarch64"),
    ignore = "bracket expression mismatches on Windows arm64"
)]
fn test_regex_caret() {
    new_ucmd!()
        .args(&["a^b", ":", "a^b"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["a^b", ":", "a\\^b"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["abc", ":", "^abc"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["^abc", ":", "^^abc"])
        .succeeds()
        .stdout_only("4\n");
    new_ucmd!()
        .args(&["b", ":", "a\\|^b"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["ab", ":", "\\(^a\\)b"])
        .succeeds()
        .stdout_only("a\n");
    new_ucmd!()
        .args(&["^abc", ":", "^abc"])
        .fails()
        .stdout_only("0\n");
    new_ucmd!()
        .args(&["^^^^^^^^^", ":", "^^^"])
        .succeeds()
        .stdout_only("2\n");
    new_ucmd!()
        .args(&["ab[^c]", ":", "ab[^c]"])
        .succeeds()
        .stdout_only("3\n"); // Matches "ab["
    new_ucmd!()
        .args(&["ab[^c]", ":", "ab\\[^c]"])
        .succeeds()
        .stdout_only("6\n");
    new_ucmd!()
        .args(&["[^a]", ":", "\\[^a]"])
        .succeeds()
        .stdout_only("4\n");
    new_ucmd!()
        .args(&["\\a", ":", "\\\\[^^]"])
        .succeeds()
        .stdout_only("2\n");
    // Patterns are anchored to the beginning of the pattern "^bc"
    new_ucmd!()
        .args(&["abc", ":", "bc"])
        .fails()
        .stdout_only("0\n");
    new_ucmd!()
        .args(&["^a", ":", "^^[^^]"])
        .succeeds()
        .stdout_only("2\n");
    new_ucmd!()
        .args(&["abc", ":", "ab[^c]"])
        .fails()
        .stdout_only("0\n");
}

#[test]
fn test_regex_dollar() {
    new_ucmd!()
        .args(&["a$b", ":", "a\\$b"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["a", ":", "a$\\|b"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["ab", ":", "a\\(b$\\)"])
        .succeeds()
        .stdout_only("b\n");
    new_ucmd!()
        .args(&["a$c", ":", "a$\\c"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["$a", ":", "$a"])
        .succeeds()
        .stdout_only("2\n");
    new_ucmd!()
        .args(&["a", ":", "a$\\|b"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["-5", ":", "-\\{0,1\\}[0-9]*$"])
        .succeeds()
        .stdout_only("2\n");
    new_ucmd!()
        .args(&["$", ":", "$"])
        .fails()
        .stdout_only("0\n");
    new_ucmd!()
        .args(&["a$", ":", "a$\\|b"])
        .fails()
        .stdout_only("0\n");
}

#[test]
fn test_regex_range_quantifier() {
    new_ucmd!()
        .args(&["a", ":", "a\\{1\\}"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["aaaaaaaaaa", ":", "a\\{1,\\}"])
        .succeeds()
        .stdout_only("10\n");
    new_ucmd!()
        .args(&["aaa", ":", "a\\{,3\\}"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["aa", ":", "a\\{1,3\\}"])
        .succeeds()
        .stdout_only("2\n");
    new_ucmd!()
        .args(&["aaaa", ":", "a\\{,\\}"])
        .succeeds()
        .stdout_only("4\n");
    new_ucmd!()
        .args(&["a", ":", "ab\\{,3\\}"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["abbb", ":", "ab\\{,3\\}"])
        .succeeds()
        .stdout_only("4\n");
    new_ucmd!()
        .args(&["abcabc", ":", "\\(abc\\)\\{,\\}"])
        .succeeds()
        .stdout_only("abc\n");
    new_ucmd!()
        .args(&["a", ":", "a\\{,6\\}"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["{abc}", ":", "\\{abc\\}"])
        .succeeds()
        .stdout_only("5\n");
    new_ucmd!()
        .args(&["a{bc}", ":", "a\\(\\{bc\\}\\)"])
        .succeeds()
        .stdout_only("{bc}\n");
    new_ucmd!()
        .args(&["{b}", ":", "a\\|\\{b\\}"])
        .succeeds()
        .stdout_only("3\n");
    new_ucmd!()
        .args(&["{", ":", "a\\|\\{"])
        .succeeds()
        .stdout_only("1\n");
    new_ucmd!()
        .args(&["{}}}", ":", "\\{\\}\\}\\}"])
        .succeeds()
        .stdout_only("4\n");
    new_ucmd!()
        .args(&["a{}}}", ":", "a\\{\\}\\}\\}"])
        .fails()
        .stderr_only("expr: Invalid content of \\{\\}\n");
    new_ucmd!()
        .args(&["ab", ":", "ab\\{\\}"])
        .fails()
        .stderr_only("expr: Invalid content of \\{\\}\n");
    new_ucmd!()
        .args(&["_", ":", "a\\{12345678901234567890\\}"])
        .fails()
        .stderr_only("expr: Regular expression too big\n");
    new_ucmd!()
        .args(&["_", ":", "a\\{12345678901234567890,\\}"])
        .fails()
        .stderr_only("expr: Regular expression too big\n");
    new_ucmd!()
        .args(&["_", ":", "a\\{,12345678901234567890\\}"])
        .fails()
        .stderr_only("expr: Regular expression too big\n");
    new_ucmd!()
        .args(&["_", ":", "a\\{1,12345678901234567890\\}"])
        .fails()
        .stderr_only("expr: Regular expression too big\n");
    new_ucmd!()
        .args(&["_", ":", "a\\{1,1234567890abcdef\\}"])
        .fails()
        .stderr_only("expr: Invalid content of \\{\\}\n");
}

#[test]
fn test_regex_newline() {
    new_ucmd!()
        .args(&["line1\nline2\nline3 ", ":", ".*line2.*"])
        .succeeds()
        .stdout_only("18\n");
}

#[test]
fn test_regex_catastrophic_backtracking() {
    let input = "a".repeat(30) + "c";
    new_ucmd!()
        .args(&[input.as_str(), ":", "\\(a\\+a\\+\\)\\+b"])
        .fails_with_code(1)
        .stdout_only("\n");
}

#[test]
fn test_substr() {
    new_ucmd!()
        .args(&["substr", "abc", "1", "1"])
        .succeeds()
        .stdout_only("a\n");

    new_ucmd!()
        .args(&["abc", "substr", "1", "1"])
        .fails_with_code(2)
        .stderr_only("expr: syntax error: unexpected argument 'substr'\n");
}

#[test]
fn test_builtin_functions_precedence() {
    new_ucmd!()
        .args(&["substr", "ab cd", "3", "1", "!=", " "])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["substr", "ab cd", "3", "1", "=", " "])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["length", "abcd", "!=", "4"])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["length", "abcd", "=", "4"])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["index", "abcd", "c", "!=", "3"])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["index", "abcd", "c", "=", "3"])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["match", "abcd", "ab\\(.*\\)", "!=", "cd"])
        .fails_with_code(1)
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["match", "abcd", "ab\\(.*\\)", "=", "cd"])
        .succeeds()
        .stdout_only("1\n");
}

#[test]
fn test_invalid_substr() {
    new_ucmd!()
        .args(&["substr", "abc", "0", "1"])
        .fails_with_code(1)
        .stdout_only("\n");

    new_ucmd!()
        .args(&["substr", "abc", &(usize::MAX.to_string() + "0"), "1"])
        .fails_with_code(1)
        .stdout_only("\n");

    new_ucmd!()
        .args(&["substr", "abc", "0", &(usize::MAX.to_string() + "0")])
        .fails_with_code(1)
        .stdout_only("\n");
}

#[test]
fn test_escape() {
    new_ucmd!().args(&["+", "1"]).succeeds().stdout_only("1\n");

    new_ucmd!()
        .args(&["1", "+", "+", "1"])
        .succeeds()
        .stdout_only("2\n");

    new_ucmd!()
        .args(&["2", "*", "+", "3"])
        .succeeds()
        .stdout_only("6\n");

    new_ucmd!()
        .args(&["(", "1", ")", "+", "1"])
        .succeeds()
        .stdout_only("2\n");
}

#[test]
fn test_invalid_syntax() {
    let invalid_syntaxes = [["12", "12"], ["12", "|"], ["|", "12"]];

    for invalid_syntax in invalid_syntaxes {
        new_ucmd!()
            .args(&invalid_syntax)
            .fails_with_code(2)
            .stderr_contains("syntax error");
    }
}

#[test]
fn test_num_str_comparison() {
    new_ucmd!()
        .args(&["1a", "<", "1", "+", "1"])
        .succeeds()
        .stdout_is("1\n");
}

#[test]
fn test_missing_closing_parenthesis_reports_syntax_error() {
    new_ucmd!()
        .args(&["(", "1", "/", "0"])
        .fails()
        .stderr_contains("expecting ')' after '0'");
}

#[test]
fn test_long_input() {
    // Giving expr an arbitrary long expression should succeed rather than end with a segfault due to a stack overflow.
    #[cfg(all(not(windows), not(target_os = "openbsd")))]
    const MAX_NUMBER: usize = 40000;
    #[cfg(all(not(windows), not(target_os = "openbsd")))]
    const RESULT: &str = "800020000\n";

    // On OpenBSD, crash with default MAX_NUMBER
    #[cfg(target_os = "openbsd")]
    const MAX_NUMBER: usize = 10000;
    #[cfg(target_os = "openbsd")]
    const RESULT: &str = "50005000\n";

    // On windows there is 8192 characters input limit
    #[cfg(windows)]
    const MAX_NUMBER: usize = 1300; // 7993 characters (with spaces)
    #[cfg(windows)]
    const RESULT: &str = "845650\n";

    let mut args: Vec<String> = vec!["1".to_string()];

    for i in 2..=MAX_NUMBER {
        args.push('+'.to_string());
        args.push(i.to_string());
    }

    new_ucmd!().args(&args).succeeds().stdout_is(RESULT);
}

#[test]
fn test_deeply_nested_expression() {
    // A deeply nested expression should succeed rather than abort due to a
    // stack overflow in the parser (GHSA-gr9g-xjfw-pqfw).
    #[cfg(not(windows))]
    const DEPTH: usize = 10000;

    // On windows there is 8192 characters input limit
    #[cfg(windows)]
    const DEPTH: usize = 2000; // 8003 characters (with spaces)

    let mut args: Vec<&str> = vec!["("; DEPTH];
    args.push("1");
    args.extend(std::iter::repeat_n(")", DEPTH));
    new_ucmd!().args(&args).succeeds().stdout_is("1\n");
}

#[test]
fn test_deeply_nested_length() {
    // Keyword operands recurse in the parser just like parentheses do.
    #[cfg(not(windows))]
    const DEPTH: usize = 10000;

    // On windows there is 8192 characters input limit
    #[cfg(windows)]
    const DEPTH: usize = 1000; // 7002 characters (with spaces)

    let mut args: Vec<&str> = vec!["length"; DEPTH];
    args.push("1");
    new_ucmd!().args(&args).succeeds().stdout_is("1\n");
}

/// Test cases for expr arithmetic and string operations
mod expr_arithmetic {
    use uutests::new_ucmd;

    #[test]
    fn test_addition() {
        new_ucmd!()
            .args(&["8", "+", "5"])
            .succeeds()
            .stdout_only("13\n");
    }

    #[test]
    fn test_subtraction() {
        new_ucmd!()
            .args(&["9", "-", "4"])
            .succeeds()
            .stdout_only("5\n");
    }

    #[test]
    fn test_multiplication() {
        new_ucmd!()
            .args(&["7", "*", "6"])
            .succeeds()
            .stdout_only("42\n");
    }

    #[test]
    fn test_integer_division() {
        new_ucmd!()
            .args(&["120", "/", "8"])
            .succeeds()
            .stdout_only("15\n");
    }

    #[test]
    fn test_modulo_remainder() {
        new_ucmd!()
            .args(&["120", "%", "7"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_add_with_negative() {
        new_ucmd!()
            .args(&["6", "+", "-4"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_two_negatives_added() {
        new_ucmd!()
            .args(&["-3", "+", "-5"])
            .succeeds()
            .stdout_only("-8\n");
    }

    #[test]
    fn test_double_dash_neg_large() {
        new_ucmd!()
            .args(&["--", "-7", "+", "15"])
            .succeeds()
            .stdout_only("8\n");
    }

    #[test]
    fn test_neg_large_no_dash() {
        new_ucmd!()
            .args(&["-7", "+", "15"])
            .succeeds()
            .stdout_only("8\n");
    }

    #[test]
    fn test_double_dash_neg_small() {
        new_ucmd!()
            .args(&["--", "-2", "+", "9"])
            .succeeds()
            .stdout_only("7\n");
    }

    #[test]
    fn test_neg_small_no_dash() {
        new_ucmd!()
            .args(&["-2", "+", "9"])
            .succeeds()
            .stdout_only("7\n");
    }

    #[test]
    fn test_double_dash_positive() {
        new_ucmd!()
            .args(&["--", "3", "+", "6"])
            .succeeds()
            .stdout_only("9\n");
    }

    #[test]
    fn test_parens_simple_mod() {
        new_ucmd!()
            .args(&["(", "120", "%", "7", ")"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_parens_mod_minus() {
        new_ucmd!()
            .args(&["(", "120", "%", "7", ")", "-", "8"])
            .succeeds()
            .stdout_only("-7\n");
    }

    #[test]
    fn test_parens_div_mod_minus() {
        new_ucmd!()
            .args(&["18", "/", "(", "120", "%", "7", ")", "-", "6"])
            .succeeds()
            .stdout_only("12\n");
    }

    #[test]
    fn test_parens_div_nested_mod() {
        new_ucmd!()
            .args(&["24", "/", "(", "(", "120", "%", "7", ")", "+", "5", ")"])
            .succeeds()
            .stdout_only("4\n");
    }

    #[test]
    fn test_parens_add_mod() {
        new_ucmd!()
            .args(&["9", "+", "(", "120", "%", "7", ")"])
            .succeeds()
            .stdout_only("10\n");
    }

    #[test]
    fn test_0bang() {
        new_ucmd!()
            .args(&["00", "<", "0!"])
            .fails()
            .code_is(1)
            .stdout_only("0\n");
    }

    #[test]
    fn test_00() {
        new_ucmd!()
            .args(&["00"])
            .fails()
            .code_is(1)
            .stdout_only("00\n");
    }

    #[test]
    fn test_minus0() {
        new_ucmd!()
            .args(&["-0"])
            .fails()
            .code_is(1)
            .stdout_only("-0\n");
    }

    #[test]
    fn test_andand() {
        new_ucmd!()
            .args(&["0", "&", "1", "/", "0"])
            .fails()
            .code_is(1)
            .stdout_only("0\n");
    }

    #[test]
    fn test_oror() {
        new_ucmd!()
            .args(&["1", "|", "1", "/", "0"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_orempty() {
        new_ucmd!()
            .args(&["", "|", ""])
            .fails()
            .code_is(1)
            .stdout_only("0\n");
    }

    #[test]
    fn test_fail_a() {
        new_ucmd!()
            .args(&["3", "+", "-"])
            .fails()
            .code_is(2)
            .no_stdout()
            .stderr_contains("non-integer argument");
    }

    #[test]
    fn test_bigcmp() {
        new_ucmd!()
            .args(&[
                "--",
                "-2417851639229258349412352",
                "<",
                "2417851639229258349412352",
            ])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_anchor() {
        new_ucmd!()
            .args(&["a\nb", ":", "a$"])
            .fails()
            .code_is(1)
            .stdout_only("0\n");
    }

    #[test]
    fn test_emptysub() {
        new_ucmd!()
            .args(&["a", ":", "\\(b\\)*"])
            .fails()
            .code_is(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre1() {
        new_ucmd!()
            .args(&["abc", ":", "a\\(b\\)c"])
            .succeeds()
            .stdout_only("b\n");
    }

    #[test]
    fn test_bre2() {
        new_ucmd!()
            .args(&["a(", ":", "a("])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre3() {
        new_ucmd!()
            .args(&["_", ":", "a\\("])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Unmatched ( or \\(");
    }

    #[test]
    fn test_bre4() {
        new_ucmd!()
            .args(&["_", ":", "a\\(b"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Unmatched ( or \\(");
    }

    #[test]
    fn test_bre5() {
        new_ucmd!()
            .args(&["a(b", ":", "a(b"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre6() {
        new_ucmd!()
            .args(&["a)", ":", "a)"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre7() {
        new_ucmd!()
            .args(&["_", ":", "a\\)"])
            .fails_with_code(2)
            .stderr_contains("Unmatched ) or \\)");
    }

    #[test]
    fn test_bre8() {
        new_ucmd!()
            .args(&["_", ":", "\\)"])
            .fails_with_code(2)
            .stderr_contains("Unmatched ) or \\)");
    }

    #[test]
    fn test_bre9() {
        new_ucmd!()
            .args(&["ab", ":", "a\\(\\)b"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre10() {
        new_ucmd!()
            .args(&["a^b", ":", "a^b"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre11() {
        new_ucmd!()
            .args(&["a$b", ":", "a$b"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre12() {
        new_ucmd!()
            .args(&["", ":", "\\($\\)\\(^\\)"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre13() {
        new_ucmd!()
            .args(&["b", ":", "a*\\(b$\\)c*"])
            .succeeds()
            .stdout_only("b\n");
    }

    #[test]
    fn test_bre14() {
        new_ucmd!()
            .args(&["X|", ":", "X\\(|\\)", ":", "(", "X|", ":", "X\\(|\\)", ")"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre15() {
        new_ucmd!()
            .args(&["X*", ":", "X\\(*\\)", ":", "(", "X*", ":", "X\\(*\\)", ")"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre16() {
        new_ucmd!()
            .args(&["abc", ":", "\\(\\)"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre17() {
        new_ucmd!()
            .args(&["{1}a", ":", "\\(\\{1\\}a\\)"])
            .succeeds()
            .stdout_only("{1}a\n");
    }

    #[test]
    fn test_bre18() {
        new_ucmd!()
            .args(&["X*", ":", "X\\(*\\)", ":", "^*"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre19() {
        new_ucmd!()
            .args(&["{1}", ":", "\\{1\\}"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre20() {
        new_ucmd!()
            .args(&["{", ":", "{"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre21() {
        new_ucmd!()
            .args(&["abbcbd", ":", "a\\(b*\\)c\\1d"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre22() {
        new_ucmd!()
            .args(&["abbcbbbd", ":", "a\\(b*\\)c\\1d"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre23() {
        new_ucmd!()
            .args(&["abc", ":", "\\(.\\)\\1"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre24() {
        new_ucmd!()
            .args(&["abbccd", ":", "a\\(\\([bc]\\)\\2\\)*d"])
            .succeeds()
            .stdout_only("cc\n");
    }

    #[test]
    fn test_bre25() {
        new_ucmd!()
            .args(&["abbcbd", ":", "a\\(\\([bc]\\)\\2\\)*d"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre26() {
        new_ucmd!()
            .args(&["abbbd", ":", "a\\(\\(b\\)*\\2\\)*d"])
            .succeeds()
            .stdout_only("bbb\n");
    }

    #[test]
    fn test_bre27() {
        new_ucmd!()
            .args(&["aabcd", ":", "\\(a\\)\\1bcd"])
            .succeeds()
            .stdout_only("a\n");
    }

    #[test]
    fn test_bre28() {
        new_ucmd!()
            .args(&["aabcd", ":", "\\(a\\)\\1bc*d"])
            .succeeds()
            .stdout_only("a\n");
    }

    #[test]
    fn test_bre29() {
        new_ucmd!()
            .args(&["aabd", ":", "\\(a\\)\\1bc*d"])
            .succeeds()
            .stdout_only("a\n");
    }

    #[test]
    fn test_bre30() {
        new_ucmd!()
            .args(&["aabcccd", ":", "\\(a\\)\\1bc*d"])
            .succeeds()
            .stdout_only("a\n");
    }

    #[test]
    fn test_bre31() {
        new_ucmd!()
            .args(&["aabcccd", ":", "\\(a\\)\\1bc*[ce]d"])
            .succeeds()
            .stdout_only("a\n");
    }

    #[test]
    fn test_bre32() {
        new_ucmd!()
            .args(&["aabcccd", ":", "\\(a\\)\\1b\\(c\\)*cd"])
            .succeeds()
            .stdout_only("a\n");
    }

    #[test]
    fn test_bre33() {
        new_ucmd!()
            .args(&["a*b", ":", "a\\(*\\)b"])
            .succeeds()
            .stdout_only("*\n");
    }

    #[test]
    fn test_bre34() {
        new_ucmd!()
            .args(&["ab", ":", "a\\(**\\)b"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre35() {
        new_ucmd!()
            .args(&["ab", ":", "a\\(***\\)b"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre36() {
        new_ucmd!()
            .args(&["*a", ":", "*a"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre37() {
        new_ucmd!()
            .args(&["a", ":", "**a"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre38() {
        new_ucmd!()
            .args(&["a", ":", "***a"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre39() {
        new_ucmd!()
            .args(&["ab", ":", "a\\{1\\}b"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre40() {
        new_ucmd!()
            .args(&["ab", ":", "a\\{1,\\}b"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre41() {
        new_ucmd!()
            .args(&["aab", ":", "a\\{1,2\\}b"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre42() {
        new_ucmd!()
            .args(&["_", ":", "a\\{1"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Unmatched \\{");
    }

    #[test]
    fn test_bre43() {
        new_ucmd!()
            .args(&["_", ":", "a\\{1a"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Unmatched \\{");
    }

    #[test]
    fn test_bre44() {
        new_ucmd!()
            .args(&["_", ":", "a\\{1a\\}"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Invalid content of \\{\\}");
    }

    #[test]
    fn test_bre45() {
        new_ucmd!()
            .args(&["a", ":", "a\\{,2\\}"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre46() {
        new_ucmd!()
            .args(&["a", ":", "a\\{,\\}"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre47() {
        new_ucmd!()
            .args(&["_", ":", "a\\{1,x\\}"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Invalid content of \\{\\}");
    }

    #[test]
    fn test_bre48() {
        new_ucmd!()
            .args(&["_", ":", "a\\{1,x"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Unmatched \\{");
    }

    #[test]
    fn test_bre49() {
        new_ucmd!()
            .args(&["_", ":", "a\\{32768\\}"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Regular expression too big\n");
    }

    #[test]
    fn test_bre50() {
        new_ucmd!()
            .args(&["_", ":", "a\\{1,0\\}"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("Invalid content of \\{\\}");
    }

    #[test]
    fn test_bre51() {
        new_ucmd!()
            .args(&["acabc", ":", ".*ab\\{0,0\\}c"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre52() {
        new_ucmd!()
            .args(&["abcac", ":", "ab\\{0,1\\}c"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre53() {
        new_ucmd!()
            .args(&["abbcac", ":", "ab\\{0,3\\}c"])
            .succeeds()
            .stdout_only("4\n");
    }

    #[test]
    fn test_bre54() {
        new_ucmd!()
            .args(&["abcac", ":", ".*ab\\{1,1\\}c"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre55() {
        new_ucmd!()
            .args(&["abcac", ":", ".*ab\\{1,3\\}c"])
            .succeeds()
            .stdout_only("3\n");
    }

    #[test]
    fn test_bre56() {
        new_ucmd!()
            .args(&["abbcabc", ":", ".*ab\\{2,2\\}c"])
            .succeeds()
            .stdout_only("4\n");
    }

    #[test]
    fn test_bre57() {
        new_ucmd!()
            .args(&["abbcabc", ":", ".*ab\\{2,4\\}c"])
            .succeeds()
            .stdout_only("4\n");
    }

    #[test]
    fn test_bre58() {
        new_ucmd!()
            .args(&["aa", ":", "a\\{1\\}\\{1\\}"])
            .succeeds()
            .stdout_only("1\n");
    }

    #[test]
    fn test_bre59() {
        new_ucmd!()
            .args(&["aa", ":", "a*\\{1\\}"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre60() {
        new_ucmd!()
            .args(&["aa", ":", "a\\{1\\}*"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_bre61() {
        new_ucmd!()
            .args(&["acd", ":", "a\\(b\\)?c\\1d"])
            .fails_with_code(1)
            .stdout_only("\n");
    }

    #[test]
    fn test_bre62() {
        new_ucmd!()
            .args(&["--", "-5", ":", "-\\{0,1\\}[0-9]*$"])
            .succeeds()
            .stdout_only("2\n");
    }

    #[test]
    fn test_fail_c() {
        new_ucmd!()
            .args::<&str>(&[])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("missing operand")
            .stderr_contains("Try")
            .stderr_contains("for more information");
    }

    const BIG: &str = "98782897298723498732987928734";
    const BIG_P1: &str = "98782897298723498732987928735";
    const BIG_SUM: &str = "197565794597446997465975857469";
    const BIG_PROD: &str = "9758060798730154302876482828124348356960410232492450771490";

    #[test]
    fn test_bignum_add() {
        new_ucmd!()
            .args(&[BIG, "+", "1"])
            .succeeds()
            .stdout_only(format!("{BIG_P1}\n"));
    }

    #[test]
    fn test_bignum_add1() {
        new_ucmd!()
            .args(&[BIG, "+", BIG_P1])
            .succeeds()
            .stdout_only(format!("{BIG_SUM}\n"));
    }

    #[test]
    fn test_bignum_sub() {
        new_ucmd!()
            .args(&[BIG_P1, "-", "1"])
            .succeeds()
            .stdout_only(format!("{BIG}\n"));
    }

    #[test]
    fn test_bignum_sub1() {
        new_ucmd!()
            .args(&[BIG_SUM, "-", BIG])
            .succeeds()
            .stdout_only(format!("{BIG_P1}\n"));
    }

    #[test]
    fn test_bignum_mul() {
        new_ucmd!()
            .args(&[BIG_P1, "*", BIG])
            .succeeds()
            .stdout_only(format!("{BIG_PROD}\n"));
    }

    #[test]
    fn test_bignum_div() {
        new_ucmd!()
            .args(&[BIG_PROD, "/", BIG])
            .succeeds()
            .stdout_only(format!("{BIG_P1}\n"));
    }

    #[test]
    fn test_se0() {
        new_ucmd!()
            .args(&["9", "9"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: unexpected argument '9'");
    }

    #[test]
    fn test_se1() {
        new_ucmd!()
            .args(&["2", "a"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: unexpected argument 'a'");
    }

    #[test]
    fn test_se2() {
        new_ucmd!()
            .args(&["2", "+"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: missing argument after '+'");
    }

    #[test]
    fn test_se3() {
        new_ucmd!()
            .args(&["2", ":"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: missing argument after ':'");
    }

    #[test]
    fn test_se4() {
        new_ucmd!()
            .args(&["length"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: missing argument after 'length'");
    }

    #[test]
    fn test_se5() {
        new_ucmd!()
            .args(&["(", "2"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: expecting ')' after '2'");
    }

    #[test]
    fn test_se6() {
        new_ucmd!()
            .args(&["(", "2", "a"])
            .fails_with_code(2)
            .no_stdout()
            .stderr_contains("syntax error: expecting ')' instead of 'a'");
    }
}

/// Test that `expr` correctly detects and handles locales
mod locale_aware {
    use uutests::new_ucmd;

    #[test]
    #[cfg_attr(wasi_runner, ignore = "WASI: no locale data, every locale is C")]
    fn test_expr_collating() {
        for (loc, code, output) in [
            ("C", 0, "1\n"),
            ("fr_FR.UTF-8", 1, "0\n"),
            ("fr_FR.utf-8", 1, "0\n"),
            ("en_US", 1, "0\n"),
        ] {
            new_ucmd!()
                .args(&["50n", ">", "-51"])
                .env("LC_ALL", loc)
                .run()
                .code_is(code)
                .stdout_only(output);
        }
    }
}

/// Tests for multibyte character arithmetic in expr
#[cfg(target_os = "linux")]
mod expr_multibyte_arithmetic {
    use uutests::new_ucmd;

    use uucore::os_str_from_bytes;

    trait AsByteSlice<'a> {
        fn into_bytes(self) -> &'a [u8];
    }

    impl<'a> AsByteSlice<'a> for &'a str {
        fn into_bytes(self) -> &'a [u8] {
            self.as_bytes()
        }
    }

    impl<'a> AsByteSlice<'a> for &'a [u8] {
        fn into_bytes(self) -> &'a [u8] {
            self
        }
    }

    impl<'a, const N: usize> AsByteSlice<'a> for &'a [u8; N] {
        fn into_bytes(self) -> &'a [u8] {
            self
        }
    }

    const EXPRESSION: &[u8] = "\u{6C49}\u{5B57}\u{6D4B}\u{8BD5}".as_bytes(); // 汉字测试 – 4 chars, 12 bytes

    #[derive(Debug, Default, Clone, Copy)]
    struct TestCase {
        pub locale: &'static str,
        pub out: Option<&'static [u8]>,
        pub code: i32,
    }

    impl TestCase {
        const FR: Self = Self::new("fr_FR.UTF-8");
        const C: Self = Self::new("C");

        const fn new(locale: &'static str) -> Self {
            Self {
                locale,
                out: None,
                code: 0,
            }
        }

        fn out(mut self, out: impl AsByteSlice<'static>) -> Self {
            self.out = Some(out.into_bytes());
            self
        }

        fn code(mut self, code: i32) -> Self {
            self.code = code;
            self
        }
    }

    fn check_test_case(args: &[&[u8]], tc: &TestCase) {
        // The WASI build ships no locale data, so every locale behaves like
        // "C", and WASI can only carry valid UTF-8 through argv. Run just the
        // cases that don't depend on either.
        if std::env::var("UUTESTS_WASM_RUNNER").is_ok()
            && (tc.locale != "C" || args.iter().any(|arg| std::str::from_utf8(arg).is_err()))
        {
            return;
        }

        let args = args
            .iter()
            .map(|arg: &&[u8]| os_str_from_bytes(arg).unwrap())
            .collect::<Vec<_>>();

        let res = new_ucmd!().env("LC_ALL", tc.locale).args(&args).run();

        res.code_is(tc.code);

        if let Some(out) = tc.out {
            let mut out = out.to_owned();
            out.push(b'\n');
            res.stdout_is_bytes(&out);
        } else {
            res.no_stdout();
        }
    }

    // LENGTH EXPRESSIONS

    // sanity check
    #[test]
    fn test_mb_length_full() {
        let args: &[&[u8]] = &[b"length", b"abcdef"];

        let cases = &[TestCase::FR.out("6"), TestCase::C.out("6")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // A single multibyte character in the beginning of the string \xCE\xB1 is
    // UTF-8 for "U+03B1 (2-byte UTF-8 sequence, used as multibyte prefix test)"
    #[test]
    fn test_mb_length_ascii_prefix() {
        let args: &[&[u8]] = &[b"length", b"\xCE\xB1bcdef"];

        let cases = &[TestCase::FR.out("6"), TestCase::C.out("7")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // A single multibyte character in the middle of the string \xCE\xB4 is
    // UTF-8 for "U+03B4 (2-byte UTF-8 sequence, used as multibyte middle test)"
    #[test]
    fn test_mb_length_ascii_middle() {
        let args: &[&[u8]] = &[b"length", b"abc\xCE\xB4ef"];

        let cases = &[TestCase::FR.out("6"), TestCase::C.out("7")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // A single multibyte character in the end of the string
    #[test]
    fn test_mb_length_ascii_suffix() {
        let args: &[&[u8]] = &[b"length", b"fedcb\xCE\xB1"];

        let cases = &[TestCase::FR.out("6"), TestCase::C.out("7")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // A invalid multibyte sequence
    #[test]
    fn test_mb_length_invalid_seq() {
        let args: &[&[u8]] = &[b"length", b"\xB1aaa"];

        let cases = &[TestCase::FR.out("4"), TestCase::C.out("4")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // An incomplete multibyte sequence at the end of the string
    #[test]
    fn test_mb_length_incomplete_end() {
        let args: &[&[u8]] = &[b"length", b"aaa\xCE"];

        let cases = &[TestCase::FR.out("4"), TestCase::C.out("4")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // An incomplete multibyte sequence at the end of the string
    #[test]
    fn test_mb_length_expression() {
        let args: &[&[u8]] = &[b"length", EXPRESSION];

        let cases = &[TestCase::FR.out("4"), TestCase::C.out("12")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // INDEX EXPRESSIONS

    // sanity check
    #[test]
    fn test_i1() {
        let args: &[&[u8]] = &[b"index", b"abcdef", b"fb"];

        let cases = &[TestCase::FR.out("2"), TestCase::C.out("2")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Search for a single-octet
    #[test]
    fn test_i2() {
        let args: &[&[u8]] = &[b"index", b"\xCE\xB1bc\xCE\xB4ef", b"b"];

        let cases = &[TestCase::FR.out("2"), TestCase::C.out("3")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_i3() {
        let args: &[&[u8]] = &[b"index", b"\xCE\xB1bc\xCE\xB4ef", b"f"];

        let cases = &[TestCase::FR.out("6"), TestCase::C.out("8")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Search for multibyte character.
    // In the C locale, the search string is treated as two octets.
    // the first of them (\xCE) matches the first octet of the input string.
    #[test]
    fn test_i4() {
        let args: &[&[u8]] = &[b"index", b"\xCE\xB1bc\xCE\xB4ef", b"\xCE\xB4"];

        let cases = &[TestCase::FR.out("4"), TestCase::C.out("1")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Invalid multibyte sequence in the input string, treated as a single
    // octet.
    #[test]
    fn test_i5() {
        let args: &[&[u8]] = &[b"index", b"\xCEbc\xCE\xB4ef", b"\xCE\xB4"];

        let cases = &[TestCase::FR.out("4"), TestCase::C.out("1")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Invalid multibyte sequence in the search string, treated as a single
    // octet. In multibyte locale, there should be no match, expr returns and
    // prints zero, and terminates with exit-code 1 (as per POSIX).
    #[test]
    fn test_i6() {
        let args: &[&[u8]] = &[b"index", b"\xCE\xB1bc\xCE\xB4ef", b"\xB4"];

        let cases = &[TestCase::FR.out("0").code(1), TestCase::C.out("6")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Edge-case: invalid multibyte sequence BOTH in the input string and in
    // the search string: expr should find a match.
    #[test]
    fn test_i7() {
        let args: &[&[u8]] = &[b"index", b"\xCE\xB1bc\xB4ef", b"\xB4"];

        let cases = &[TestCase::FR.out("4")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // SUBSTR EXPRESSIONS

    // sanity check
    #[test]
    fn test_s1() {
        let args: &[&[u8]] = &[b"substr", b"abcdef", b"2", b"3"];

        let cases = &[TestCase::FR.out("bcd"), TestCase::C.out("bcd")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s2() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xCE\xB4ef", b"1", b"1"];

        let cases = &[TestCase::FR.out(b"\xCE\xB1"), TestCase::C.out(b"\xCE")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s3() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xCE\xB4ef", b"3", b"2"];

        let cases = &[TestCase::FR.out(b"c\xCE\xB4"), TestCase::C.out("bc")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s4() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xCE\xB4ef", b"4", b"1"];

        let cases = &[TestCase::FR.out(b"\xCE\xB4"), TestCase::C.out("c")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s5() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xCE\xB4ef", b"4", b"2"];

        let cases = &[TestCase::FR.out(b"\xCE\xB4e"), TestCase::C.out(b"c\xCE")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s6() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xCE\xB4ef", b"6", b"1"];

        let cases = &[TestCase::FR.out(b"f"), TestCase::C.out(b"\xB4")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s7() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xCE\xB4ef", b"7", b"1"];

        let cases = &[TestCase::FR.out("").code(1), TestCase::C.out(b"e")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    #[test]
    fn test_s8() {
        let args: &[&[u8]] = &[b"substr", b"\xCE\xB1bc\xB4ef", b"3", b"3"];

        let cases = &[TestCase::FR.out(b"c\xB4e"), TestCase::C.out(b"bc\xB4")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // MATCH EXPRESSIONS

    // sanity check
    #[test]
    fn test_m1() {
        let args: &[&[u8]] = &[b"match", b"abcdef", b"ab"];

        let cases = &[TestCase::FR.out("2"), TestCase::C.out("2")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }
    #[test]
    fn test_m2() {
        let args: &[&[u8]] = &[b"match", b"abcdef", b"\\(ab\\)"];

        let cases = &[TestCase::FR.out("ab"), TestCase::C.out("ab")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // The regex engine should match the '.' to the first multibyte character.
    #[test]
    fn test_m3() {
        let args: &[&[u8]] = &[b"match", b"\xCE\xB1bc\xCE\xB4ef", b".bc"];

        let cases = &[TestCase::FR.out("3"), TestCase::C.out("0").code(1)];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // The opposite of the previous test: two dots should only match the two
    // octets in single-byte locale.
    #[test]
    fn test_m4() {
        let args: &[&[u8]] = &[b"match", b"\xCE\xB1bc\xCE\xB4ef", b"..bc"];

        let cases = &[TestCase::FR.out("0").code(1), TestCase::C.out("4")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Match with grouping - a single dot should return the two octets
    #[test]
    fn test_m5() {
        let args: &[&[u8]] = &[b"match", b"\xCE\xB1bc\xCE\xB4ef", b"\\(.b\\)c"];

        let cases = &[TestCase::FR.out(b"\xCE\xB1b"), TestCase::C.out("").code(1)];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Invalid multibyte sequences - regex should not match in multibyte locale
    // (POSIX requirement)
    #[test]
    fn test_m6() {
        let args: &[&[u8]] = &[b"match", b"\xCEbc\xCE\xB4ef", b"\\(.\\)"];

        let cases = &[TestCase::FR.out("").code(1), TestCase::C.out(b"\xCE")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }

    // Character classes: in the multibyte case, the regex engine understands
    // there is a single multibyte character in the brackets.
    // In the single byte case, the regex engine sees two octets in the
    // character class ('\xCE' and '\xB1') - and it matches the first one.
    #[test]
    fn test_m7() {
        let args: &[&[u8]] = &[b"match", b"\xCE\xB1bc\xCE\xB4ef", b"\\(.\\)"];

        let cases = &[TestCase::FR.out(b"\xCE\xB1"), TestCase::C.out(b"\xCE")];

        for tc in cases {
            check_test_case(args, tc);
        }
    }
}

#[test]
#[cfg_attr(wasi_runner, ignore = "WASI: no locale data, every locale is C")]
fn test_emoji_operations() {
    new_ucmd!()
        .args(&["🚀", "=", "🚀"])
        .succeeds()
        .stdout_only("1\n");

    new_ucmd!()
        .args(&["🚀", "!=", "🚀"])
        .fails()
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["🚀", "=", "🧨"])
        .fails()
        .stdout_only("0\n");

    new_ucmd!()
        .args(&["length", "🦀🚀🎯"])
        .env("LC_ALL", "fr_FR.UTF-8")
        .succeeds()
        .stdout_only("3\n");

    new_ucmd!()
        .args(&["🌍", "!=", "🌎"])
        .succeeds()
        .stdout_only("1\n");
}

#[cfg(all(feature = "feat_diagnostics", not(wasi_runner)))]
mod diagnostics {
    use super::*;
    #[cfg(unix)]
    #[test]
    fn test_snippet_points_at_the_unexpected_argument() {
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["6", "+", "7", "spare"])
            .fails_with_code(2);

        // The whole report: the `expr: ` prefix of the plain form, the
        // expression echoed back, a caret on `spare`, and the quoting advice.
        assert_eq!(
            result.stderr_as_displayed(),
            "\
expr: syntax error: unexpected argument 'spare'
   ╭─[ expr:1:7 ]
   │
 1 │ 6 + 7 spare
   │       ──┬──
   │         ╰──── the expression was already complete here
   │
   │ Help: the shell may have expanded an operator; quote it as '*' or escape it as \\*
───╯"
        );
    }

    #[cfg(unix)]
    #[test]
    fn test_snippet_points_at_the_dangling_operator() {
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["6", "*"])
            .fails_with_code(2);

        assert_eq!(
            result.stderr_as_displayed(),
            "\
expr: syntax error: missing argument after '*'
   ╭─[ expr:1:3 ]
   │
 1 │ 6 *
   │   ─
   │
   │ Help: every operator needs a value on both sides
───╯"
        );
    }

    #[cfg(unix)]
    #[test]
    fn test_snippet_points_at_the_non_integer_operand() {
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["oops", "+", "4"])
            .fails_with_code(2);

        assert_eq!(
            result.stderr_as_displayed(),
            "\
expr: non-integer argument
   ╭─[ expr:1:1 ]
   │
 1 │ oops + 4
   │ ────
   │
   │ Help: arithmetic operators need integers; use = or != to compare strings instead
───╯"
        );
    }

    #[cfg(unix)]
    #[test]
    fn test_unclosed_parenthesis_points_at_the_last_operand() {
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["(", "6", "+", "7"])
            .fails_with_code(2);

        // No help line here: the label alone says what should follow.
        assert_eq!(
            result.stderr_as_displayed(),
            "\
expr: syntax error: expecting ')' after '7'
   ╭─[ expr:1:7 ]
   │
 1 │ ( 6 + 7
   │       ─
───╯"
        );
    }

    #[cfg(unix)]
    #[test]
    fn test_errors_without_a_position_stay_plain() {
        // Division by zero is raised once the expression is already parsed and
        // carries no operand, so there is nothing to point at. The terminal
        // turns `\n` into `\r\n`, which is undone before comparing.
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["6", "/", "0"])
            .fails_with_code(2);
        assert_eq!(
            result.stderr_str().replace("\r\n", "\n"),
            "expr: division by zero\n"
        );
        // An empty expression keeps its usage hint.
        new_ucmd!()
            .terminal_sim_stderr()
            .fails_with_code(2)
            .stderr_contains("missing operand")
            .stderr_contains("for more information");
    }

    #[cfg(unix)]
    #[test]
    fn test_snippet_points_at_the_operand_that_failed_not_its_twin() {
        // Both operands read `a`, but only the right-hand one of `+` failed
        // arithmetic; the caret must not drift to the first occurrence, a
        // perfectly valid string operand of `=`.
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["a", "=", "2", "+", "a"])
            .fails_with_code(2);

        assert_eq!(
            result.stderr_as_displayed(),
            "\
expr: non-integer argument
   ╭─[ expr:1:9 ]
   │
 1 │ a = 2 + a
   │         ─
   │
   │ Help: arithmetic operators need integers; use = or != to compare strings instead
───╯"
        );
    }

    #[cfg(unix)]
    #[test]
    fn test_computed_operand_stays_plain() {
        // The failing operand is the result of `substr`, not an argument, so
        // there is nothing to point at and the plain message is kept.
        let result = new_ucmd!()
            .terminal_sim_stderr()
            .args(&["substr", "abc", "1", "2", "+", "1"])
            .fails_with_code(2);
        assert_eq!(
            result.stderr_str().replace("\r\n", "\n"),
            "expr: non-integer argument\n"
        );
    }

    #[test]
    fn test_plain_message_is_the_default() {
        // The test harness pipes stderr, so the report must not appear.
        new_ucmd!()
            .args(&["6", "+", "7", "spare"])
            .fails_with_code(2)
            .stderr_is("expr: syntax error: unexpected argument 'spare'\n");
    }
}