Add yaml config utils:

add yaml config utils

2 files changed, 285 insertions, 0 deletions

diff --git a/library/cpp/langs/langs.h b/library/cpp/langs/langs.h
new file mode 100644
index 0000000000..360ab6a832
--- /dev/null
+++ b/library/cpp/langs/langs.h
@@ -0,0 +1,229 @@
+#pragma once
+
+#include "scripts.h"
+
+#include <util/generic/strbuf.h>
+#include <util/system/defaults.h>
+
+#if defined(_win_)
+// LANG_LAO is #define in WinNT.h
+#undef LANG_LAO
+#endif
+
+// Language names are given according to ISO 639-2/B
+// Some languages are not present in ISO 639-2/B. Then ISO 639-3 is used.
+// http://en.wikipedia.org/wiki/List_of_ISO_639-1_codes
+enum ELanguage {
+    LANG_UNK = 0,           // Unknown
+    LANG_RUS = 1,           // Russian
+    LANG_ENG = 2,           // English
+    LANG_POL = 3,           // Polish
+    LANG_HUN = 4,           // Hungarian
+    LANG_UKR = 5,           // Ukrainian
+    LANG_GER = 6,           // German
+    LANG_FRE = 7,           // French
+    LANG_TAT = 8,           // Tatar
+    LANG_BEL = 9,           // Belarusian
+    LANG_KAZ = 10,          // Kazakh
+    LANG_ALB = 11,          // Albanian
+    LANG_SPA = 12,          // Spanish
+    LANG_ITA = 13,          // Italian
+    LANG_ARM = 14,          // Armenian
+    LANG_DAN = 15,          // Danish
+    LANG_POR = 16,          // Portuguese
+    LANG_ICE = 17,          // Icelandic
+    LANG_SLO = 18,          // Slovak
+    LANG_SLV = 19,          // Slovene
+    LANG_DUT = 20,          // Dutch (Netherlandish language)
+    LANG_BUL = 21,          // Bulgarian
+    LANG_CAT = 22,          // Catalan
+    LANG_HRV = 23,          // Croatian
+    LANG_CZE = 24,          // Czech
+    LANG_GRE = 25,          // Greek
+    LANG_HEB = 26,          // Hebrew
+    LANG_NOR = 27,          // Norwegian
+    LANG_MAC = 28,          // Macedonian
+    LANG_SWE = 29,          // Swedish
+    LANG_KOR = 30,          // Korean
+    LANG_LAT = 31,          // Latin
+    LANG_BASIC_RUS = 32,    // Simplified version of Russian (used at lemmer only)
+    LANG_BOS = 33,          // Bosnian
+    LANG_MLT = 34,          // Maltese
+    LANG_EMPTY = 35,        // Indicate that document is empty
+    LANG_UNK_LAT = 36,      // Any unrecognized latin language
+    LANG_UNK_CYR = 37,      // Any unrecognized cyrillic language
+    LANG_UNK_ALPHA = 38,    // Any unrecognized alphabetic language not fit into previous categories
+    LANG_FIN = 39,          // Finnish
+    LANG_EST = 40,          // Estonian
+    LANG_LAV = 41,          // Latvian
+    LANG_LIT = 42,          // Lithuanian
+    LANG_BAK = 43,          // Bashkir
+    LANG_TUR = 44,          // Turkish
+    LANG_RUM = 45,          // Romanian (also Moldavian)
+    LANG_MON = 46,          // Mongolian
+    LANG_UZB = 47,          // Uzbek
+    LANG_KIR = 48,          // Kirghiz
+    LANG_TGK = 49,          // Tajik
+    LANG_TUK = 50,          // Turkmen
+    LANG_SRP = 51,          // Serbian
+    LANG_AZE = 52,          // Azerbaijani
+    LANG_BASIC_ENG = 53,    // Simplified version of English (used at lemmer only)
+    LANG_GEO = 54,          // Georgian
+    LANG_ARA = 55,          // Arabic
+    LANG_PER = 56,          // Persian
+    LANG_CHU = 57,          // Church Slavonic
+    LANG_CHI = 58,          // Chinese
+    LANG_JPN = 59,          // Japanese
+    LANG_IND = 60,          // Indonesian
+    LANG_MAY = 61,          // Malay
+    LANG_THA = 62,          // Thai
+    LANG_VIE = 63,          // Vietnamese
+    LANG_GLE = 64,          // Irish (Gaelic)
+    LANG_TGL = 65,          // Tagalog (Filipino)
+    LANG_HIN = 66,          // Hindi
+    LANG_AFR = 67,          // Afrikaans
+    LANG_URD = 68,          // Urdu
+    LANG_MYA = 69,          // Burmese
+    LANG_KHM = 70,          // Khmer
+    LANG_LAO = 71,          // Lao
+    LANG_TAM = 72,          // Tamil
+    LANG_BEN = 73,          // Bengali
+    LANG_GUJ = 74,          // Gujarati
+    LANG_KAN = 75,          // Kannada
+    LANG_PAN = 76,          // Punjabi
+    LANG_SIN = 77,          // Sinhalese
+    LANG_SWA = 78,          // Swahili
+    LANG_BAQ = 79,          // Basque
+    LANG_WEL = 80,          // Welsh
+    LANG_GLG = 81,          // Galician
+    LANG_HAT = 82,          // Haitian Creole
+    LANG_MLG = 83,          // Malagasy
+    LANG_CHV = 84,          // Chuvash
+    LANG_UDM = 85,          // Udmurt
+    LANG_KPV = 86,          // Komi-Zyrian
+    LANG_MHR = 87,          // Meadow Mari (Eastern Mari)
+    LANG_SJN = 88,          // Sindarin
+    LANG_MRJ = 89,          // Hill Mari (Western Mari)
+    LANG_KOI = 90,          // Komi-Permyak
+    LANG_LTZ = 91,          // Luxembourgish
+    LANG_GLA = 92,          // Scottish Gaelic
+    LANG_CEB = 93,          // Cebuano
+    LANG_PUS = 94,          // Pashto
+    LANG_KMR = 95,          // Kurmanji
+    LANG_AMH = 96,          // Amharic
+    LANG_ZUL = 97,          // Zulu
+    LANG_IBO = 98,          // Igbo
+    LANG_YOR = 99,          // Yoruba
+    LANG_COS = 100,         // Corsican
+    LANG_XHO = 101,         // Xhosa
+    LANG_JAV = 102,         // Javanese
+    LANG_NEP = 103,         // Nepali
+    LANG_SND = 104,         // Sindhi
+    LANG_SOM = 105,         // Somali
+    LANG_EPO = 106,         // Esperanto
+    LANG_TEL = 107,         // Telugu
+    LANG_MAR = 108,         // Marathi
+    LANG_HAU = 109,         // Hausa
+    LANG_YID = 110,         // Yiddish
+    LANG_MAL = 111,         // Malayalam
+    LANG_MAO = 112,         // Maori
+    LANG_SUN = 113,         // Sundanese
+    LANG_PAP = 114,         // Papiamento
+    LANG_UZB_CYR = 115,     // Cyrillic Uzbek
+    LANG_TRANSCR_IPA = 116, // International Phonetic Alphabet Transcription
+    LANG_EMJ = 117,         // Emoji
+    LANG_UYG = 118,         // Uyghur
+    LANG_BRE = 119,         // Breton
+    LANG_SAH = 120,         // Yakut
+    LANG_KAZ_LAT = 121,     // Latin Kazakh
+    LANG_MAX
+};
+
+/**
+ * Converts string to corresponding enum. Will try to extract the primary language code from
+ * constructions like "en-cockney" or "zh_Hant". In case of failure will return `LANG_UNK`.
+ *
+ * @param name              Language name
+ * @return                  Language enum
+ */
+ELanguage LanguageByName(const TStringBuf& name);
+
+/**
+ * Same as `LanguageByName`, but in case of failure will return `LANG_MAX`.
+ *
+ * @see LanguageByName
+ */
+ELanguage LanguageByNameStrict(const TStringBuf& name);
+
+/**
+ * Converts language enum to corresponding ISO 639-2/B alpha-3 code. For languages missing in ISO
+ * standard convertions are:
+ *   - LANG_UNK:          "unk"
+ *   - LANG_BASIC_RUS:    "basic-rus"
+ *   - LANG_EMPTY:        "empty"
+ *   - LANG_UNK_LAT:      "unklat"
+ *   - LANG_UNK_CYR:      "unkcyr"
+ *   - LANG_UNK_ALPHA:    "unkalpha"
+ *   - LANG_BASIC_ENG:    "basic-eng"
+ *   - LANG_TRANSCR_IPA   "transcr-ipa"
+ * If language is missing in `ELanguage` or if it is a `LANG_MAX` then return value will be
+ * `nullptr`.
+ *
+ * @param language         Language enum
+ * @return                 Language ISO 639-2/B alpha-3 code
+ */
+const char* NameByLanguage(ELanguage language);
+
+/**
+ * Converts language enum to corresponding ISO 639-1 alpha-2 code. For languages missing in ISO
+ * standard convertions are:
+ *   - LANG_UNK:          "mis"
+ *   - LANG_BASIC_RUS:    "bas-ru"
+ *   - LANG_EMPTY:        ""
+ *   - LANG_UNK_LAT:      ""
+ *   - LANG_UNK_CYR:      ""
+ *   - LANG_UNK_ALPHA:    ""
+ *   - LANG_BASIC_ENG:    "bas-en"
+ *   - LANG_TRANSCR_IPA   "tr-ipa"
+ * If language is missing in `ELanguage` or if it is a `LANG_MAX` then return value will be
+ * `nullptr`.
+ *
+ * @param language         Language enum
+ * @return                 Language ISO 639-1 alpha-2 code
+ */
+const char* IsoNameByLanguage(ELanguage language);
+
+/**
+ * Converts language enum to corresponding human-readable language name. E.g. "Russian" for
+ * `LANG_RUS` or "Basic Russian" for `LANG_BASIC_RUS`. If language is missing in `ELanguage` or if
+ * it is a `LANG_MAX` then return value will be `nullptr`.
+ *
+ * @param language         Language enum
+ */
+const char* FullNameByLanguage(ELanguage language);
+
+/**
+ * Same as `LanguageByNameStrict` but in case of failure will throw `yexception`.
+ *
+ * @see LanguageByNameStrict
+ */
+ELanguage LanguageByNameOrDie(const TStringBuf& name);
+
+constexpr bool UnknownLanguage(const ELanguage language) noexcept {
+    return language == LANG_UNK || language == LANG_UNK_LAT || language == LANG_UNK_CYR || language == LANG_UNK_ALPHA || language == LANG_EMPTY;
+}
+
+EScript ScriptByLanguage(ELanguage language);
+EScript ScriptByGlyph(wchar32 glyph);
+
+namespace NCharsetInternal {
+    void InitScriptData(ui8 data[], size_t len);
+}
+
+inline bool LatinScript(ELanguage language) {
+    return ScriptByLanguage(language) == SCRIPT_LATIN;
+}
+
+inline bool CyrillicScript(ELanguage language) {
+    return ScriptByLanguage(language) == SCRIPT_CYRILLIC;
+}
diff --git a/library/cpp/langs/scripts.h b/library/cpp/langs/scripts.h
new file mode 100644
index 0000000000..4c47a33d2c
--- /dev/null
+++ b/library/cpp/langs/scripts.h
@@ -0,0 +1,56 @@
+#pragma once
+
+#include <util/generic/strbuf.h>
+
+// Writing systems, a.k.a. scripts
+//
+enum EScript {
+    SCRIPT_UNKNOWN = 0,
+    SCRIPT_LATIN,
+    SCRIPT_CYRILLIC,
+
+    SCRIPT_GREEK,
+    SCRIPT_ARABIC,
+    SCRIPT_HEBREW,
+    SCRIPT_ARMENIAN,
+    SCRIPT_GEORGIAN,
+
+    SCRIPT_HAN,
+    SCRIPT_KATAKANA,
+    SCRIPT_HIRAGANA,
+    SCRIPT_HANGUL,
+
+    SCRIPT_DEVANAGARI,
+    SCRIPT_BENGALI,
+    SCRIPT_GUJARATI,
+    SCRIPT_GURMUKHI,
+    SCRIPT_KANNADA,
+    SCRIPT_MALAYALAM,
+    SCRIPT_ORIYA,
+    SCRIPT_TAMIL,
+    SCRIPT_TELUGU,
+    SCRIPT_THAANA,
+    SCRIPT_SINHALA,
+
+    SCRIPT_MYANMAR,
+    SCRIPT_THAI,
+    SCRIPT_LAO,
+    SCRIPT_KHMER,
+    SCRIPT_TIBETAN,
+    SCRIPT_MONGOLIAN,
+
+    SCRIPT_ETHIOPIC,
+    SCRIPT_RUNIC,
+    SCRIPT_COPTIC,
+    SCRIPT_SYRIAC,
+
+    SCRIPT_OTHER,
+    SCRIPT_MAX
+};
+
+// According to ISO 15924 codes. See https://en.wikipedia.org/wiki/ISO_15924
+//
+EScript ScriptByName(const TStringBuf& name);
+EScript ScriptByNameOrDie(const TStringBuf& name);
+const char* IsoNameByScript(EScript script);
+const char* FullNameByScript(EScript script);