Last active
December 7, 2020 13:55
-
-
Save cupcakearmy/242b54ee6b1a914896390c91846aa4d4 to your computer and use it in GitHub Desktop.
Lunr Test
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import { readFileSync, writeFileSync } from 'fs' | |
import { brotliCompressSync, gzipSync } from 'zlib' | |
import hf from 'human-format' | |
import lunr from 'lunr' | |
// https://gutenberg.org/ebooks/6130 | |
const file = readFileSync('./6130-0.txt', 'utf8') | |
const opts = { | |
scale: 'binary', | |
unit: 'B', | |
} | |
const data = [['Pieces', 'Size', 'Index', 'GZip', 'BR', 'HSize', 'HIndex', 'HGZip', 'HBr']] | |
for (const pieces of [1, 10, 20, 50, 100, 200, 500, 1000, 2000, 5000]) { | |
// for (const pieces of [1, 5, 10, 20, 50, 100, 200, 500, 2000, 3000, 5000]) { | |
// const size = Math.floor(file.length / pieces) | |
const size = 10000 | |
const hSize = hf(Buffer.from(file.slice(0, size)).length) | |
console.log('Pieces', pieces) | |
console.log('Piece Size', hSize) | |
const idx = lunr(function () { | |
this.ref('id') | |
this.field('text') | |
for (let i = 0; i < pieces; i++) { | |
const position = (i * size) % file.length | |
this.add({ | |
id: Math.random().toString(36), | |
text: file.slice(position, position + size), | |
}) | |
} | |
}) | |
const prebuilt = JSON.stringify(idx) | |
const bytes = Buffer.from(prebuilt) | |
const gz = gzipSync(bytes, { level: 9 }) | |
const br = brotliCompressSync(bytes) | |
const hIndex = hf(bytes.length, opts) | |
const hGz = hf(gz.length, opts) | |
const hBr = hf(br.length, opts) | |
console.log(`Normal:\t${hIndex}`) | |
console.log(`GZip:\t${hGz}`) | |
console.log(`Brotli:\t${hBr}`) | |
data.push([pieces, size, bytes.length, gz.length, br.length, hSize, hIndex, hGz, hBr]) | |
} | |
writeFileSync('./data.csv', data.map((row) => row.join(',')).join('\n')) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Pieces | Size | Index | GZip | BR | HSize | HIndex | HGZip | HBr | |
---|---|---|---|---|---|---|---|---|---|
1 | 10000 | 37811 | 6063 | 4248 | 10.04 k | 36.92 KiB | 5.92 KiB | 4.15 KiB | |
10 | 10000 | 257608 | 43311 | 34045 | 10.04 k | 251.57 KiB | 42.3 KiB | 33.25 KiB | |
20 | 10000 | 502006 | 86938 | 68083 | 10.04 k | 490.24 KiB | 84.9 KiB | 66.49 KiB | |
50 | 10000 | 1161161 | 202147 | 157147 | 10.04 k | 1.11 MiB | 197.41 KiB | 153.46 KiB | |
100 | 10000 | 2178638 | 392040 | 292482 | 10.04 k | 2.08 MiB | 382.85 KiB | 285.63 KiB | |
200 | 10000 | 4145798 | 765040 | 540214 | 10.04 k | 3.95 MiB | 747.11 KiB | 527.55 KiB | |
500 | 10000 | 9776558 | 1952672 | 1137839 | 10.04 k | 9.32 MiB | 1.86 MiB | 1.09 MiB | |
1000 | 10000 | 19225463 | 4548799 | 2109189 | 10.04 k | 18.33 MiB | 4.34 MiB | 2.01 MiB | |
2000 | 10000 | 38005044 | 11370691 | 4089003 | 10.04 k | 36.24 MiB | 10.84 MiB | 3.9 MiB | |
5000 | 10000 | 94455961 | 35521544 | 10314160 | 10.04 k | 90.08 MiB | 33.88 MiB | 9.84 MiB |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment