70. www.newocr.com [ocr]

/*
  nama   : ocr
  base   : https://www.newocr.com
  by     : wolep
  update : 1 Desember 10 PM
  node   : v24.11.1
  note   : idk, enjoy
  module : esm (tested) / cjs (gak tau coba test yah :v)
*/

import fs from 'fs'

const ocr = {
    baseUrl: 'https://www.newocr.com',
    baseHeaders: {
        'Accept-Encoding': 'gzip, deflate, br, zstd',
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36'
    },
    log: (message) => console.log('[ocr] ' + message),
    error: async (functionName, r) => `${functionName} ${r.status} ${r.statusText}\n${await r.text() || '(empty response)'}`,

    upload: async function (imageBuffer) {
        this.log('uploading')
        const headers = this.baseHeaders

        const body = new FormData();
        body.append('userfile', new Blob([imageBuffer]), 'image.jpeg');
        body.append('preview', '1');

        const r = await fetch(this.baseUrl, {
            headers, body,
            'method': 'post'
        })
        if (!r.ok) throw Error(await this.error('upload', r))

        const html = await r.text()
        const u = html.match(/name ="u" value="(.+?)"/)?.[1]
        if (!u) throw Error(`gagal mendapatkan upload id`)
        return { u }
    },

    extract: async function (uobj) {
        this.log('extract text')
        const { u } = uobj
        const headers = {
            'content-type': 'application/x-www-form-urlencoded',
            ...this.baseHeaders
        }

        const body = new URLSearchParams({
            "l3": "ind",
            "l2[]": "ind",
            "r": "0",
            "psm": "6",
            "u": u,
            //"x1": "0",
            //"x2": "533",
            //"y1": "0",
            //"y2": "700",
            "ocr": "1"
        })

        const r = await fetch(this.baseUrl, {
            headers, body,
            'method': 'post'
        })
        if (!r.ok) throw Error(await this.error('extract', r))
        const html = await r.text()
        const text = html.match(/<textarea id="ocr-result"(?:.+?)>(.+?)<\/textarea>\n\n\t\t<\/div>/ms)?.[1]
        return text
    },

    read: async function(imageBuffer) {
        const uObj = await this.upload(imageBuffer)
        const extractObject = await this.extract(uObj)
        this.log('selesai')
        return extractObject
    }
}

// cara pakai
const buffer = fs.readFileSync("./input.jpeg")
ocr.read(buffer)
.then(console.log)
.catch(console.log)

/* output
string bla bla bla bla atau undefined kalau gagal membaca string
*/