@math-blocks/parser

parserFactory can be used to generate TDOP/Pratt parsers. The reason for using this style of parsing is that it's much easier to change the precedence and add new operators that it would be other parser types.

function parserFactory<T extends {readonly type: string}, N, O>(
    getPrefixParselet: (token: T) => PrefixParselet<T, N, O>,
    getInfixParselet: (token: T) => InfixParselet<T, N, O> | null,
    getOpPrecedence: (arg0: O) => number,
    EOL: T,
): {parse: (arg0: Array<T>) => N}

The three generic properties are:

• T: token object type produced by a lexer
• N: node object type that the parser will produce
• O: operator type (usually an enum of union of strings)

getPrefixParselet and getInfixParselet are functions that return the parselet that should be used for the given token. Parselets are objects that contain a parse method and possible additional data. Infix parselets also keep track of the current operator.

In the future, support for post-fix operators such as ! and ' will be added.

Examples

• @math-blocks/editor-parser/src/editor-parser.ts
• @math-blocks/testing/src/text-parser.ts