Pocket crawler

pocket.groovy

#!/usr/bin/env groovy

@Grab(group='org.seleniumhq.selenium', module='selenium-server', version='3.6.0')
@Grab(group='org.seleniumhq.selenium', module='selenium-java', version='3.6.0')
@Grab(group='org.seleniumhq.selenium', module='selenium-chrome-driver', version='3.6.0')

import org.openqa.selenium.By
import org.openqa.selenium.JavascriptExecutor
import org.openqa.selenium.WebDriver
import org.openqa.selenium.WebElement
import org.openqa.selenium.firefox.FirefoxDriver
import org.openqa.selenium.chrome.*

final String USERNAME = System.getenv("POCKET_USERNAME")
final String PASSWORD = System.getenv("POCKET_PASSWORD")
WebDriver driver = new ChromeDriver()

driver.get('http://www.getpocket.com/login')

driver.findElement(By.id('feed_id')).sendKeys(USERNAME)
driver.findElement(By.id('login_password')).sendKeys(PASSWORD)
driver.findElement(By.className('login-btn-email')).click()

sleep(3000)

int articleCount = driver.findElements(By.className('item')).size()

while (true) {
  ((JavascriptExecutor) driver).executeScript("window.scrollTo(0, document.body.scrollHeight)")
  sleep(3000)
  int newArticleCount = driver.findElements(By.className('item')).size()

  if (newArticleCount == articleCount) {
    break
  }

  articleCount = newArticleCount
}

def items = driver.findElements(By.className('item'))

items.each {
  println it.getAttribute('id')
  println it.findElement(By.className('title')).text
  String url = it.findElement(By.className('original_url')).getAttribute('href')
  println getUrlParams(url).url

}
println items.size()

Map<String, String> getUrlParams(String url) {
  def paramMap = new URL(url).getQuery().split('&').collectEntries {
     int idx = it.indexOf("=")
     [URLDecoder.decode(it[0..idx - 1], 'UTF-8'), URLDecoder.decode(it[idx + 1.. -1], 'UTF-8')]
  }

  return paramMap
}