Українською

Yar. Ukrainian Numbers Spellout

The library to convert the numbers into Ukrainian text. Example project is here https://hryvnia.yar.org.ua Source code is here https://gitlab.com/yar.ua/hryvnia/

Supported Features

• [x] Number spellout – converting numbers into text (numerals).
• [x] Numeral inflection - modifying grammatical form of numerals to match required grammatical characteristics.
• [x] Agreement - setting grammatical forms of compound numerals to produce grammatically and syntactically correct text.
• [x] Stress.
• [x] Cardinal numbers in range [0...1×1027)1, 2.
• [x] Ordinal numbers in range [0...1×1027)1.
• [x] Fractional numbers with whole, each element in range [0...1×1027).
• [x] Decimal numbers in range (0...1×1027) with precision up to 1×10-27 3.

1 Numbers outside of the range will be spelled out digit-wise and inflected only last digit

2 1×1027 is 1 000 000 000 000 000 000 000 000 000

3 1×10-27 is 0,000 000 000 000 000 000 000 000 001

Future Features

• [ ] Negative numbers.
• [ ] Contracted ordinal numbers (e.g. "1-й", "1000-на").
• [ ] Scientific notation of exponential numbers (e.g. "1e5").
• [ ] Roman numbers (e.g. "XIV").

Supported Grammatical Attributes

• cases: nominative, genitive, dative, accusative, instrumental, locative, vocative
• gender: masculine, feminine, neuter
• number: singular, plural
• animacy: inanimate, animate

Using Syntax Tree API

First you need to define the root word. Typically the word governs the depndent words, but numerals impose governing back at the parent word. However the desired form of the root word is still important and should be explicitly specified.

Here is an example of paradigm definition. It specifies every form and grammatical category for it. The last component is describing immutable grammatical categories, like gender and animacy for nouns.

const HRYVNYA = [
  [
    [{ case: "nominative", number: "singular" }, "гри́вня"],
    [{ case: "genitive", number: "singular" }, "гри́вні"],
    [{ case: "dative", number: "singular" }, "гри́вні"],
    [{ case: "accusative", number: "singular" }, "гри́вню"],
    [{ case: "instrumental", number: "singular" }, "гри́вні"],
    [{ case: "locative", number: "singular" }, "гри́вні"],
    [{ case: "vocative", number: "singular" }, "гри́вня"],
    [{ case: "nominative", number: "plural" }, "гри́вні"],
    [{ case: "genitive", number: "plural" }, "гри́вень"],
    [{ case: "dative", number: "plural" }, "гри́вням"],
    [{ case: "accusative", number: "plural" }, "гри́вні"],
    [{ case: "instrumental", number: "plural" }, "гри́внями"],
    [{ case: "locative", number: "plural" }, "гри́внях"],
    [{ case: "vocative", number: "plural" }, "гри́вні"],
  ],
  { gender: "feminine", animacy: "inanimate" },
]; // as paradigm.ParadigmDefinition

After the paradigm definition we can start building the syntax tree. Every node of the tree starts with a lexeme, so we create a lexeme based off the paradigm definition. And the specify the target form of the whole phrase.

const hryvnia = new paradigm.ParadigmLexeme("", ...HRYVNYA);
hryvnia.updateForm({ case: "nominative" });
const rootNode = new syntax.SyntaxNode(hryvnia);

The next 2 functions have the main logic of inflection. First we build the syntax tree on our root node. In this case we want a cardinal number but there are also available functions to build ordinal, decimal, and fractional trees. And then we agree the tree, this resolves the final form of each node.

cardinals.build_tree(rootNode, amount);
rootNode.agree();

At this point we can already access the correct text of the phrase or individual nodes.

rootNode.text();
rootNode.lexeme.text();