KVSearch

Overview

This lib is providing a way to filter an array of any kind of object. You can provide an array of field you would like to "index", or you can generate your own query.

Installation

npm install @nexucis/kvsearch

Usage

Filter a list of object using a list of index

import { KVSearch } from '@nexucis/kvsearch';

const list = [
    {
        labels: { instance: 'demo.org', job: 'demo' },
        scrapePool: 'scrapePool demo'
    },
    {
        labels: { instance: 'k8s.org', job: 'constellation' },
        scrapePool: 'galaxy'
    }
]
const search = new KVSearch({
    indexedKeys: [
        'labels',
        'scrapePool',
    ],
});
search.filter('demo', list) // will match the first object
search.filter('constellation', list) // won't match any object present in the list, since the attribute `labels.jop` is not indexed

Here the indexed list says:

Since labels value is an object, then check if the pattern is matching a key in the object return by labels
Same thing for scrapePool, excepting it returns a string, so the code won't loop other a list of key, it will just check if the value of scrapePool is matching the pattern

Note that the matching is using the lib @nexucis/fuzzy, so it's not an exact match used.

Filter a list of object using a list of index, with some regexp

import { KVSearch } from '@nexucis/kvsearch';

const list = [
    {
        labels: { instance: 'demo.org', job: 'demo' },
        scrapePool: 'scrapePool demo'
    },
    {
        labels: { instance: 'k8s.org', job: 'constellation' },
        scrapePool: 'galaxy'
    }
]
const search = new KVSearch({
    indexedKeys: [
        'labels',
        ['labels', /.*/],
        'scrapePool',
    ],
});
search.filter('constel', list) // will match only the 2nd object

The difference here is we indexed the attributes of the labels object. In this example, by using the Regexp /.*/ we indexed every attribute of the object labels. That's why the pattern constellation is matching the second object. But we could also just index the field job of the labels like that ['labels', 'job']. It would have worked as well.

Filter a list of object using a specific query.

Using a list of index is simple, but it always used a fuzzy match. Probably sometimes you would like to do an exact match or a negative match depending on the context.

You can do it by creating your own query like that:

import { KVSearch } from '@nexucis/kvsearch';

const list = [
    {
        labels: { instance: 'demo.org', job: 'demo' },
        scrapePool: 'scrapePool demo'
    },
    {
        labels: { instance: 'k8s.org', job: 'constellation' },
        scrapePool: 'galaxy'
    }
]
const search = new KVSearch();
search.filterWithQuery({ keyPath: ['labels', /.*/], match: 'exact', pattern: 'constellation' })

Filter a list of object using a complex query.

It's possible to combine query together, so you can write multiple conditions.

import { KVSearch } from '@nexucis/kvsearch';

const list = [
    {
        labels: { instance: 'demo.org', job: 'demo' },
        scrapePool: 'scrapePool demo'
    },
    {
        labels: { instance: 'k8s.org', job: 'constellation' },
        scrapePool: 'galaxy'
    },
    {
        labels: { instance: 'awx.com', job: 'constellation' },
        scrapePool: 'galaxy',
    }
]
const search = new KVSearch();
search.filterWithQuery({
    operator: 'and',
    left: {
        keyPath: ['scrapePool'],
        match: 'fuzzy',
        pattern: 'gal'
    },
    right: {
        keyPath: ['labels', 'instance'],
        match: 'exact',
        pattern: 'awx.com'
    }
}) // this query is matching the last element of the list.

Note as it can be painful to write the query himself, a support to write it with a string in the codemirror editor is available. Check it here

Demo

A demo is available if you want to run it.

you need node 16 and npm 8
Clone the project
at the root of the project, run npm install
then run npm start

the demo should start with webpack and you can open your browser with the URL provided by webpack.

License