Filter Languages

There is a unique filter language for each filterable data type. The documentation-bRoKeN-lInK describes the various languages.

Grammars

malloy-filter use moo and nearley to generate parsers for each sub language. The grammars are in grammars.

After changing a .ne file use npm run build to update the generated parser.

API

Refer to the clause interfaces for details on the output of the parser, or look at the parser tests in src/test/ for more examples of parser output.

import {StringFilterExpression, StringClause} from '@malloydata/malloy-filter'

// or ...
// NumberFilterExpression, NumberClause
// BooleanFilterExpression, BooleanClause
// TemporalFilterExpression, TemporalClause

// Each has two entry points, "parse" and "unparse"

const stringParse = StringFilterExpression.parse('FOO%,-FOOD')
if (stringParse.parsed) {
    console.log(JSON.stringify(stringParse.parsed, null, 2));
    /*
     * Output will look something like this ...
     * { operator: ',',
     *   members: [
     *     {operator: 'starts', values: ['FOO']}
     *     {operator: '=', not: true, values: ['FOOD']}
     *   ],
     * }
     */
    console.log(StringFilterExpression.unparse(stringParse.parse));
    /*
     * There will be minor variations on unparse, for example in this
     * case there will be space after the comma:
     *
     * FOO%, -FOOD
     */
} else {
    console.log(stringParse.log);
}

Testing

The malloy-filter tests use Jest.

npm run build
npm run test