krabka-java-regex 1.0.7

//! Which `\p{name}` and `\P{name}` names Java's `Pattern` takes.
//!
//! `fancy_regex` matches a property name loosely: in any case, with or
//! without spaces, underscores and hyphens, and for every Unicode property and
//! script it knows. `Pattern.family` takes fewer, each in one spelling, and
//! refuses the rest with `Unknown character property name`. [`java_knows`]
//! says which it takes, so that the rewrite refuses a name Java refuses
//! instead of matching it as `fancy_regex` reads it. A name Java takes and
//! `fancy_regex` does not, such as a Unicode block, still fails to compile.
//!
//! The tables are those of JDK 25's `CharPredicates` and
//! `Character.UnicodeScript`.

/// Whether Java's `Pattern.family` takes `name`, the text between the braces
/// of `\p{name}`, or the letter of `\pL`. `unicode_classes` is `(?U)`.
///
/// - `key=value`: `sc` or `script` with a script, `gc` or
///   `general_category` with a name `forProperty` takes, and `blk` or
///   `block` with a block. The key is in any case.
/// - `In` and a block.
/// - `Is` and a Unicode property or POSIX class in any case
///   (`forUnicodeProperty`), a name `forProperty` takes, or a script.
/// - Otherwise a POSIX class in any case under `(?U)`, or a name
///   `forProperty` takes.
///
/// Every block name is taken here: `fancy_regex` knows none, so a block
/// fails to compile whatever its name.
pub(super) fn java_knows(name: &str, unicode_classes: bool) -> bool {
    if let Some((key, value)) = name.split_once('=') {
        return match key.to_lowercase().as_str() {
            "sc" | "script" => is_script(value),
            "gc" | "general_category" => FOR_PROPERTY.contains(&value),
            "blk" | "block" => true,
            _ => false,
        };
    }
    if name.starts_with("In") {
        return true;
    }
    if let Some(short) = name.strip_prefix("Is") {
        let upper = short.to_uppercase();
        return UNICODE_PROPERTIES.contains(&upper.as_str())
            || POSIX_NAMES.contains(&upper.as_str())
            || FOR_PROPERTY.contains(&short)
            || is_script(short);
    }
    (unicode_classes && POSIX_NAMES.contains(&name.to_uppercase().as_str()))
        || FOR_PROPERTY.contains(&name)
}

/// `Character.UnicodeScript.forName`: the name of a script or its alias, in
/// any case.
fn is_script(name: &str) -> bool { SCRIPTS.binary_search(&name.to_uppercase().as_str()).is_ok() }

/// The names `CharPredicates.getUnicodePredicate` takes, upper-cased.
const UNICODE_PROPERTIES: &[&str] = &[
    "ALPHABETIC",
    "ASSIGNED",
    "CONTROL",
    "EMOJI",
    "EMOJI_PRESENTATION",
    "EMOJI_MODIFIER",
    "EMOJI_MODIFIER_BASE",
    "EMOJI_COMPONENT",
    "EXTENDED_PICTOGRAPHIC",
    "HEXDIGIT",
    "HEX_DIGIT",
    "IDEOGRAPHIC",
    "JOINCONTROL",
    "JOIN_CONTROL",
    "LETTER",
    "LOWERCASE",
    "NONCHARACTERCODEPOINT",
    "NONCHARACTER_CODE_POINT",
    "TITLECASE",
    "PUNCTUATION",
    "UPPERCASE",
    "WHITESPACE",
    "WHITE_SPACE",
    "WORD",
];

/// The names `CharPredicates.getPosixPredicate` takes, upper-cased.
const POSIX_NAMES: &[&str] = &[
    "ALPHA", "LOWER", "UPPER", "SPACE", "PUNCT", "XDIGIT", "ALNUM", "CNTRL", "DIGIT", "BLANK",
    "GRAPH", "PRINT",
];

/// The names `CharPredicates.forProperty` takes, each in its one spelling.
const FOR_PROPERTY: &[&str] = &[
    "Cn",
    "Lu",
    "Ll",
    "Lt",
    "Lm",
    "Lo",
    "Mn",
    "Me",
    "Mc",
    "Nd",
    "Nl",
    "No",
    "Zs",
    "Zl",
    "Zp",
    "Cc",
    "Cf",
    "Co",
    "Cs",
    "Pd",
    "Ps",
    "Pe",
    "Pc",
    "Po",
    "Sm",
    "Sc",
    "Sk",
    "So",
    "Pi",
    "Pf",
    "L",
    "M",
    "N",
    "Z",
    "C",
    "P",
    "S",
    "LC",
    "LD",
    "L1",
    "all",
    "ASCII",
    "Alnum",
    "Alpha",
    "Blank",
    "Cntrl",
    "Digit",
    "Graph",
    "Lower",
    "Print",
    "Punct",
    "Space",
    "Upper",
    "XDigit",
    "javaLowerCase",
    "javaUpperCase",
    "javaAlphabetic",
    "javaIdeographic",
    "javaTitleCase",
    "javaDigit",
    "javaDefined",
    "javaLetter",
    "javaLetterOrDigit",
    "javaJavaIdentifierStart",
    "javaJavaIdentifierPart",
    "javaUnicodeIdentifierStart",
    "javaUnicodeIdentifierPart",
    "javaIdentifierIgnorable",
    "javaSpaceChar",
    "javaWhitespace",
    "javaISOControl",
    "javaMirrored",
];

/// The names and aliases of `Character.UnicodeScript` in JDK 25, upper-cased
/// and sorted.
const SCRIPTS: &[&str] = &[
    "ADLAM",
    "ADLM",
    "AGHB",
    "AHOM",
    "ANATOLIAN_HIEROGLYPHS",
    "ARAB",
    "ARABIC",
    "ARMENIAN",
    "ARMI",
    "ARMN",
    "AVESTAN",
    "AVST",
    "BALI",
    "BALINESE",
    "BAMU",
    "BAMUM",
    "BASS",
    "BASSA_VAH",
    "BATAK",
    "BATK",
    "BENG",
    "BENGALI",
    "BHAIKSUKI",
    "BHKS",
    "BOPO",
    "BOPOMOFO",
    "BRAH",
    "BRAHMI",
    "BRAI",
    "BRAILLE",
    "BUGI",
    "BUGINESE",
    "BUHD",
    "BUHID",
    "CAKM",
    "CANADIAN_ABORIGINAL",
    "CANS",
    "CARI",
    "CARIAN",
    "CAUCASIAN_ALBANIAN",
    "CHAKMA",
    "CHAM",
    "CHER",
    "CHEROKEE",
    "CHORASMIAN",
    "CHRS",
    "COMMON",
    "COPT",
    "COPTIC",
    "CPMN",
    "CPRT",
    "CUNEIFORM",
    "CYPRIOT",
    "CYPRO_MINOAN",
    "CYRILLIC",
    "CYRL",
    "DESERET",
    "DEVA",
    "DEVANAGARI",
    "DIAK",
    "DIVES_AKURU",
    "DOGR",
    "DOGRA",
    "DSRT",
    "DUPL",
    "DUPLOYAN",
    "EGYP",
    "EGYPTIAN_HIEROGLYPHS",
    "ELBA",
    "ELBASAN",
    "ELYM",
    "ELYMAIC",
    "ETHI",
    "ETHIOPIC",
    "GARA",
    "GARAY",
    "GEOR",
    "GEORGIAN",
    "GLAG",
    "GLAGOLITIC",
    "GONG",
    "GONM",
    "GOTH",
    "GOTHIC",
    "GRAN",
    "GRANTHA",
    "GREEK",
    "GREK",
    "GUJARATI",
    "GUJR",
    "GUKH",
    "GUNJALA_GONDI",
    "GURMUKHI",
    "GURU",
    "GURUNG_KHEMA",
    "HAN",
    "HANG",
    "HANGUL",
    "HANI",
    "HANIFI_ROHINGYA",
    "HANO",
    "HANUNOO",
    "HATR",
    "HATRAN",
    "HEBR",
    "HEBREW",
    "HIRA",
    "HIRAGANA",
    "HLUW",
    "HMNG",
    "HMNP",
    "HUNG",
    "IMPERIAL_ARAMAIC",
    "INHERITED",
    "INSCRIPTIONAL_PAHLAVI",
    "INSCRIPTIONAL_PARTHIAN",
    "ITAL",
    "JAVA",
    "JAVANESE",
    "KAITHI",
    "KALI",
    "KANA",
    "KANNADA",
    "KATAKANA",
    "KAWI",
    "KAYAH_LI",
    "KHAR",
    "KHAROSHTHI",
    "KHITAN_SMALL_SCRIPT",
    "KHMER",
    "KHMR",
    "KHOJ",
    "KHOJKI",
    "KHUDAWADI",
    "KIRAT_RAI",
    "KITS",
    "KNDA",
    "KRAI",
    "KTHI",
    "LANA",
    "LAO",
    "LAOO",
    "LATIN",
    "LATN",
    "LEPC",
    "LEPCHA",
    "LIMB",
    "LIMBU",
    "LINA",
    "LINB",
    "LINEAR_A",
    "LINEAR_B",
    "LISU",
    "LYCI",
    "LYCIAN",
    "LYDI",
    "LYDIAN",
    "MAHAJANI",
    "MAHJ",
    "MAKA",
    "MAKASAR",
    "MALAYALAM",
    "MAND",
    "MANDAIC",
    "MANI",
    "MANICHAEAN",
    "MARC",
    "MARCHEN",
    "MASARAM_GONDI",
    "MEDEFAIDRIN",
    "MEDF",
    "MEETEI_MAYEK",
    "MEND",
    "MENDE_KIKAKUI",
    "MERC",
    "MERO",
    "MEROITIC_CURSIVE",
    "MEROITIC_HIEROGLYPHS",
    "MIAO",
    "MLYM",
    "MODI",
    "MONG",
    "MONGOLIAN",
    "MRO",
    "MROO",
    "MTEI",
    "MULT",
    "MULTANI",
    "MYANMAR",
    "MYMR",
    "NABATAEAN",
    "NAGM",
    "NAG_MUNDARI",
    "NAND",
    "NANDINAGARI",
    "NARB",
    "NBAT",
    "NEWA",
    "NEW_TAI_LUE",
    "NKO",
    "NKOO",
    "NSHU",
    "NUSHU",
    "NYIAKENG_PUACHUE_HMONG",
    "OGAM",
    "OGHAM",
    "OLCK",
    "OLD_HUNGARIAN",
    "OLD_ITALIC",
    "OLD_NORTH_ARABIAN",
    "OLD_PERMIC",
    "OLD_PERSIAN",
    "OLD_SOGDIAN",
    "OLD_SOUTH_ARABIAN",
    "OLD_TURKIC",
    "OLD_UYGHUR",
    "OL_CHIKI",
    "OL_ONAL",
    "ONAO",
    "ORIYA",
    "ORKH",
    "ORYA",
    "OSAGE",
    "OSGE",
    "OSMA",
    "OSMANYA",
    "OUGR",
    "PAHAWH_HMONG",
    "PALM",
    "PALMYRENE",
    "PAUC",
    "PAU_CIN_HAU",
    "PERM",
    "PHAG",
    "PHAGS_PA",
    "PHLI",
    "PHLP",
    "PHNX",
    "PHOENICIAN",
    "PLRD",
    "PRTI",
    "PSALTER_PAHLAVI",
    "REJANG",
    "RJNG",
    "ROHG",
    "RUNIC",
    "RUNR",
    "SAMARITAN",
    "SAMR",
    "SARB",
    "SAUR",
    "SAURASHTRA",
    "SGNW",
    "SHARADA",
    "SHAVIAN",
    "SHAW",
    "SHRD",
    "SIDD",
    "SIDDHAM",
    "SIGNWRITING",
    "SIND",
    "SINH",
    "SINHALA",
    "SOGD",
    "SOGDIAN",
    "SOGO",
    "SORA",
    "SORA_SOMPENG",
    "SOYO",
    "SOYOMBO",
    "SUND",
    "SUNDANESE",
    "SUNU",
    "SUNUWAR",
    "SYLO",
    "SYLOTI_NAGRI",
    "SYRC",
    "SYRIAC",
    "TAGALOG",
    "TAGB",
    "TAGBANWA",
    "TAI_LE",
    "TAI_THAM",
    "TAI_VIET",
    "TAKR",
    "TAKRI",
    "TALE",
    "TALU",
    "TAMIL",
    "TAML",
    "TANG",
    "TANGSA",
    "TANGUT",
    "TAVT",
    "TELU",
    "TELUGU",
    "TFNG",
    "TGLG",
    "THAA",
    "THAANA",
    "THAI",
    "TIBETAN",
    "TIBT",
    "TIFINAGH",
    "TIRH",
    "TIRHUTA",
    "TNSA",
    "TODHRI",
    "TODR",
    "TOTO",
    "TULU_TIGALARI",
    "TUTG",
    "UGAR",
    "UGARITIC",
    "UNKNOWN",
    "VAI",
    "VAII",
    "VITH",
    "VITHKUQI",
    "WANCHO",
    "WARA",
    "WARANG_CITI",
    "WCHO",
    "XPEO",
    "XSUX",
    "YEZI",
    "YEZIDI",
    "YI",
    "YIII",
    "ZANABAZAR_SQUARE",
    "ZANB",
    "ZINH",
    "ZYYY",
    "ZZZZ",
];

#[cfg(test)]
mod tests {
    use {super::*, assert2::check};

    #[test]
    fn the_scripts_are_sorted_for_the_binary_search() {
        check!(SCRIPTS.is_sorted());
    }
}