• Site
  • Status do aplicativo
  • Documentação da API
  • Blog
Telegram Icon Comunidade
PT
Русский
Português
English
中文 (中国)
Tiếng Việt
Fazer login Teste grátis
  • Site
  • Status do aplicativo
  • Documentação da API
  • Blog
  • Comunidade








  • Portuguese (Brazil)









    Русский










    Português










    English










    中文 (中国)










    Tiếng Việt








Fazer login Teste grátis

Scripts Python personalizados

Crie fluxos de automação poderosos com Python e a API do Multilogin. Crie, gerencie e execute perfis de navegador com scripts totalmente personalizáveis.

search icon

Tente uma palavra-chave diferente ou verifique se há erros de digitação. Se ainda precisar de ajuda, entre em contato com o suporte.

  • Começando com a automação do Multilogin
  • Automação básica com CLI
  • Automação low-code com Postman
  • Executor de scripts e scripts pré-definidos
  • Puppeteer, Selenium e Playwright
  • Scripts Python personalizados
  • Soluções rápidas com as Ferramentas de Desenvolvedor
  • Ferramentas de automação externas
  • Início
  • breadcrumb separator bar
  • Multilogin
  • breadcrumb separator bar
  • Automação de tarefas com API
  • breadcrumb separator bar
  • Scripts Python personalizados
  • breadcrumb separator bar
  • Como automatizar postagens no Pinterest em telefones em nuvem com Appium

Como automatizar postagens no Pinterest em telefones em nuvem com Appium

Escrito por Anton L ( Atualizado em 20 de Julho de 2026 )

Atualizado em 20 de Julho de 2026

Telefones em nuvem Multilogin suportam ADB, então você pode conectá-los a ferramentas de automação Android como o Appium.

Neste exemplo, um script em Python lida com trabalho repetitivo no Pinterest em vários telefones em nuvem. Ele se conecta a cada telefone, faz login na sua conta do Pinterest, move as imagens designadas, executa algumas ações de feed de exemplo e publica as imagens como Pins.

➡️ Você não precisa abrir cada telefone em nuvem e repetir as mesmas ações manualmente. Adicione seus telefones e contas a devices.txt, inicie o script e deixe-o trabalhar na lista para você.

O script processa os telefones em nuvem um por vez. Ele completa o fluxo de trabalho completo em um telefone antes de passar para o próximo.

Este script é uma prova de conceito. Ele mostra o que você pode construir com Multilogin telefones em nuvem, ADB e Appium. A automação do Pinterest não está integrada ao Multilogin.

 
Your browser does not support HTML5 video.

O que o script faz

Uma vez que tudo esteja pronto, o script:

  • Conecta a um telefone em nuvem através do ADB
  • Instala o Pinterest a partir de um arquivo APK
  • Muda imagens do seu computador para o telefone em nuvem
  • Faz login na conta do Pinterest que você fornecer
  • Executa algumas interações de feed de exemplo para mostrar a navegação automatizada no aplicativo
  • Publica as imagens designadas
  • Passa para o próximo telefone em nuvem na lista

Cada telefone em nuvem pode usar sua própria conta do Pinterest e faixa de imagens. Isso facilita a organização de conteúdo para vários perfis sociais a partir de uma única configuração.

As interações de feed de exemplo também podem ser adaptadas para um fluxo de trabalho de aquecimento de conta. Confira Como aquecer um perfil no Multilogin para mais detalhes.

Lembre-se de que algumas interações automatizadas não garantem a confiança da conta ou previnem restrições da plataforma. Sempre siga as regras do Pinterest e automatize apenas contas que você está autorizado a gerenciar.

 

Antes de começar

Você precisará de:

  • Appium e UiAutomator2 instalados no seu computador: Documentação do Appium
  • Um telefone em nuvem Multilogin iniciado com ADB habilitado: Como usar ADB em telefones em nuvem
  • Uma conta de teste do Pinterest e seus detalhes de login
  • O arquivo APK do Pinterest
  • Uma pasta chamada images com as imagens que você deseja publicar
  • Um arquivo de texto chamado devices.txt

Mantenha o script, o arquivo APK, devices.txt, e a pasta images no mesmo diretório do projeto. Isso ajuda o script a encontrar tudo automaticamente.

 

Configurando devices.txt

Adicione um telefone em nuvem por linha usando este formato:

IP:PORT ADB_PASSWORD PINTEREST_EMAIL PINTEREST_PASSWORD IMAGE_RANGE

Exemplo:

199.190.44.226:25922 j3fYxT email@email.com password 1-3

Aqui está o que cada valor significa:

  • IP:PORT: o endereço ADB do telefone em nuvem
  • ADB_PASSWORD: a senha usada para conectar via ADB
  • PINTEREST_EMAIL: o email da conta do Pinterest
  • PINTEREST_PASSWORD: a senha da conta do Pinterest
  • IMAGE_RANGE: as imagens atribuídas àquela conta

Por exemplo, 1-3 informa ao script para usar as três primeiras imagens da pasta images ordenadas alfabeticamente.

Você pode dividir imagens entre vários telefones em nuvem:

199.190.44.226:25922 j3fYxT user1@email.com password1 1-3
199.190.44.227:25922 k9pQrS user2@email.com password2 4-5

O primeiro telefone recebe as imagens 1–3. O segundo telefone recebe as imagens 4–5.

Deixe a faixa de imagens de fora se você quiser que um telefone use todas as imagens na pasta.

Mantenha seus detalhes de login seguros. devices.txt armazena senhas e outras informações sensíveis como texto simples. Use contas de teste, mantenha o arquivo privado, nunca o faça upload para um repositório público e delete-o quando não precisar mais.

 

Executando o script

  1. Inicie cada telefone em nuvem incluído em devices.txt
  2. Ative o ADB em cada telefone em nuvem
  3. Inicie o servidor Appium no seu computador
  4. Verifique se o script, o arquivo APK, devices.txt, e a pasta images estão no mesmo local
  5. Salve o script abaixo como appium_script.py

appium_script.py

"""

How to use
-----
Run this script and include it in the folder with two files: devices.txt and images folder

devices.txt   -- one phone per line:  IP:PORT TOKEN USERNAME PASSWORD [IMAGE_RANGE]
                 IMAGE_RANGE is an amount of images which will be used for specific account, e.g. "1-3"
                 into the alphabetically-sorted list of images found in images/
                 Skip it to give that phone every image in the folder. Example, splitting 5 images across 2
                 phones (phone 1 gets images 1-3, phone 2 gets images 4-5):
                     199.190.44.226:25922 j3fYxT user1 pass1 1-3
                     199.190.44.227:25922 k9pQrS user2 pass2 4-5
images/       -- a folder of images (jpg/png), or a single image file.

Requires:  pip install Appium-Python-Client
           Appium server running locally (appium) and ANDROID adb on PATH.

The script installs the APK and images on the phone, opens Pinterest, enters the account credentials, runs a few sample feed interactions, and publishes the assigned images.
"""

import argparse
import random
import re
import subprocess
import sys
import time
from pathlib import Path
from urllib.parse import quote as urlquote

from appium import webdriver
from appium.options.android import UiAutomator2Options
from appium.webdriver.common.appiumby import AppiumBy

APPIUM_SERVER_URL = "http://localhost:4723"
PINTEREST_PACKAGE = "com.pinterest"

CONNECT_RETRIES = 6
CONNECT_RETRY_SLEEP = 2
INSTALL_RETRIES = 3

# Warm up limit
WARM_UP_MIN_ACTIONS = 3
WARM_UP_MAX_ACTIONS = 5

# Folder we push images into
REMOTE_IMAGE_DIR = "/sdcard/Pictures/bulk_publisher"



# adb connection function

def adb(*args, serial=None, timeout=60):
    cmd = ["adb"] + (["-s", serial] if serial else []) + list(args)
    return subprocess.run(cmd, capture_output=True, text=True, timeout=timeout)


def connect_and_login(serial: str, token: str) -> bool:
    """ADB connection, uses the phone name/address and glogin."""
    print(f"   [CONNECTING] {serial}: connecting via adb...")
    for attempt in range(1, CONNECT_RETRIES + 1):
        adb("connect", serial)
        state = adb("get-state", serial=serial).stdout.strip()
        if state == "device":
            break
        print(f"   [CONNECTING] {serial}: not online yet (attempt {attempt}/{CONNECT_RETRIES}, state={state or 'unknown'}), retrying...")
        adb("disconnect", serial)
        time.sleep(CONNECT_RETRY_SLEEP)
    else:
        print(f"   [CONNECT_FAIL] {serial}")
        return False

    print(f"   [CONNECTED] {serial}: online, logging in (glogin)...")
    result = adb("shell", "glogin", token, serial=serial)
    output = (result.stdout + result.stderr).lower()
    if "success" in output or "already logged" in output:
        print(f"   [OK] {serial}: logged in")
        return True
    print(f"   [GLOGIN_FAIL] {serial}: {result.stdout or result.stderr}")
    return False


def install_apk(serial: str, apk_path: str) -> bool:
    for attempt in range(INSTALL_RETRIES):
        result = adb("install", "-r", apk_path, serial=serial, timeout=180)
        if "success" in result.stdout.lower():
            return True
        time.sleep(3)
    print(f"   [INSTALL_FAIL] {serial}: {result.stdout or result.stderr}")
    return False


def grant_full_media_access(serial: str):
    """Grants full photo/video library access via ADB """
    adb("shell", "pm", "grant", PINTEREST_PACKAGE, "android.permission.READ_MEDIA_IMAGES", serial=serial)
    adb("shell", "pm", "grant", PINTEREST_PACKAGE, "android.permission.READ_MEDIA_VIDEO", serial=serial)


def sanitize_filename(name: str) -> str:
    """Name mapping (if the script is not able to find the image by default)"""
    stem, suffix = Path(name).stem, Path(name).suffix
    return re.sub(r"[^A-Za-z0-9._-]", "_", stem) + suffix


def push_images(serial: str, image_paths, remote_dir=REMOTE_IMAGE_DIR) -> list:
    """Push local images to the phone and trigger a media scan"""
    adb("shell", "mkdir", "-p", remote_dir, serial=serial)
    remote_paths = []
    used_names = set()
    total = len(image_paths)
    for i, local_path in enumerate(image_paths, start=1):
        safe_name = sanitize_filename(Path(local_path).name)
        candidate, n = safe_name, 1
        while candidate in used_names:  # two originals sanitized to the same name
            candidate = f"{Path(safe_name).stem}_{n}{Path(safe_name).suffix}"
            n += 1
        used_names.add(candidate)
        remote_path = f"{remote_dir}/{candidate}"
        print(f"   [UPLOADING] {serial}: pushing image {i}/{total}: {Path(local_path).name}")
        push_result = adb("push", local_path, remote_path, serial=serial)
        if push_result.returncode != 0 or "error" in push_result.stdout.lower():
            print(f"   [PUSH_FAIL] {serial}: {Path(local_path).name}: "
                  f"{push_result.stdout.strip() or push_result.stderr.strip()}")
            continue

        scan_uri = "file://" + urlquote(remote_path)
        scan_result = adb(
            "shell", "am", "broadcast", "-a",
            "android.intent.action.MEDIA_SCANNER_SCAN_FILE",
            "-d", scan_uri,
            serial=serial,
        )
        if scan_result.returncode != 0:
            print(f"   [SCAN_FAIL] {serial}: {Path(local_path).name}: "
                  f"{scan_result.stdout.strip() or scan_result.stderr.strip()}")
        remote_paths.append(remote_path)
    return remote_paths


def disconnect(serial: str):
    adb("disconnect", serial)



def build_driver(serial: str):
    """Start the Appium session then launch Pinterest. """
    capabilities = dict(
        platformName="Android",
        automationName="uiautomator2",
        udid=serial,
        newCommandTimeout=180,
        noReset=True,
        language="en",
        locale="US",
    )
    driver = webdriver.Remote(APPIUM_SERVER_URL, options=UiAutomator2Options().load_capabilities(capabilities))
    driver.activate_app(PINTEREST_PACKAGE)
    return driver


def find_any(driver, candidates, timeout=20, poll=0.5, require_enabled=False):
    """Try each (by, value) locator candidate in order until one resolves,
    polling until the timeout instead of failing on the first miss. """
    deadline = time.time() + timeout
    last_err = None
    while time.time() < deadline:
        for by, value in candidates:
            try:
                el = driver.find_element(by=by, value=value)
                if el.is_displayed() and (not require_enabled or el.get_attribute("enabled") == "true"):
                    return el
            except Exception as exc:  # noqa: BLE001 - trying multiple strategies on purpose
                last_err = exc
        time.sleep(poll)
    raise TimeoutError(f"No candidate locator matched within {timeout}s: {candidates}") from last_err


def tap_any(driver, candidates, timeout=20, require_enabled=False):
    find_any(driver, candidates, timeout=timeout, require_enabled=require_enabled).click()


def tap_if_present(driver, candidates, timeout=3) -> bool:
    """Like tap_any, but returns False instead of raising when nothing matches
    -- for popups that only sometimes appear."""
    try:
        tap_any(driver, candidates, timeout=timeout)
        return True
    except TimeoutError:
        return False

# Stage 1: First login screen email only.
EMAIL_ONLY_FIELD = [
    (AppiumBy.XPATH, "//*[@resource-id='com.pinterest:id/email_address']//android.widget.EditText"),
]
CONTINUE_BUTTON = [
    (AppiumBy.ID, "com.pinterest:id/continue_email_bt"),
]

# Stage 2: combined email + password screen, then "Log In".
LOGIN_EMAIL_FIELD = [
    (AppiumBy.XPATH, "//*[@resource-id='com.pinterest:id/email']//android.widget.EditText"),
]
LOGIN_PASSWORD_FIELD = [
    (AppiumBy.XPATH, "//*[@resource-id='com.pinterest:id/password']//android.widget.EditText"),
]
LOGIN_SUBMIT_BUTTON = [
    (AppiumBy.ID, "com.pinterest:id/login_bt"),
]

# Popups that can appear at unpredictable points after login.
EMAIL_UPDATE_DISMISS = [
    (AppiumBy.ID, "com.pinterest:id/actionPromptDismissButton"),  # "Is this your email?" -> close
]
NOTIFICATIONS_PERMISSION_ALLOW = [
    (AppiumBy.ID, "com.android.permissioncontroller:id/permission_allow_button"),
]
PHOTOS_PERMISSION_ALLOW_ALL = [
    (AppiumBy.ID, "com.android.permissioncontroller:id/permission_allow_all_button"),
]

# Opens the "Start creating now" bottom sheet. Scripts looking for the selectors depending on phone's resolution and size
CREATE_ICON_BUTTON = [
    (AppiumBy.XPATH,
     '//android.widget.Button[@content-desc="Create"]'
     '/android.widget.FrameLayout[@resource-id="com.pinterest:id/icon_button_container"]'
     '/android.widget.Button[@resource-id="com.pinterest:id/icon_button"]'),
    (AppiumBy.XPATH,
     '//android.widget.FrameLayout[@content-desc="Create"]'
     '/android.widget.LinearLayout[@resource-id="com.pinterest:id/tab_content"]'),
]

# Bottom nav "Home" tab 
HOME_TAB_BUTTON = [
    (AppiumBy.ACCESSIBILITY_ID, "Home"),
    (AppiumBy.ID, "com.pinterest:id/bottom_nav_home_icon"),
]

# "Start creating now" sheet that appears after tapping Create -- pick "Pin"
PIN_ACTION_BUTTON = [
    (AppiumBy.XPATH, '//android.widget.Button[@content-desc="Pin"]'),
]

# Media picker "Next" button (enabled once a photo is selected).
GALLERY_NEXT_BUTTON = [
    (AppiumBy.ID, "com.pinterest:id/end_container_text_button"),
]

# Final submit on the Create Pin details screen -- labelled "Create".
CREATE_PIN_SUBMIT_BUTTON = [
    (AppiumBy.ID, "com.pinterest:id/create_gestalt_button"),
]

# --- Warm-up locators (home feed browsing: open a pin, like/save/comment) ---

# Generic home-feed pin card. The script finds any pins on screen for that phone/account/session.
PIN_CARD = [
    # Finds an actually-clickable match. Fall back to any match if no clickable one is found, in case a
    (AppiumBy.XPATH, '//*[contains(@content-desc, "Pin from") and @clickable="true"]'),
    (AppiumBy.XPATH, '//*[contains(@content-desc, "Pin from")]'),
]
REACT_BUTTON = [
    (AppiumBy.XPATH, '//android.widget.ImageView[@content-desc="React to this Pin"]'),
]
SAVE_BUTTON = [
    (AppiumBy.ID, "com.pinterest:id/save_pinit_bt"),
    (AppiumBy.XPATH, '//android.widget.Button[@content-desc="Save"]'),
]
COMMENTS_MODULE = [
    (AppiumBy.ID, "com.pinterest:id/new_comments_module_container"),
]
CLOSEUP_BACK_BUTTON = [
    (AppiumBy.ID, "com.pinterest:id/closeup_back_button"),
    (AppiumBy.ACCESSIBILITY_ID, "Back"),
]


def dismiss_popups(driver):
    """Clears the 'Is this your email?' prompt and both Android permission
    dialogs (notifications, photo access) if any of them are on screen."""
    tap_if_present(driver, EMAIL_UPDATE_DISMISS, timeout=2)
    tap_if_present(driver, NOTIFICATIONS_PERMISSION_ALLOW, timeout=2)
    tap_if_present(driver, PHOTOS_PERMISSION_ALLOW_ALL, timeout=2)


def ensure_on_home_feed(driver, max_attempts=6, settle_seconds=1.5):
    """Confirms the Create icon (home feed) is visible, actively navigating
    back if not."""
    for _ in range(max_attempts):
        try:
            find_any(driver, CREATE_ICON_BUTTON, timeout=4)
            return
        except TimeoutError:
            pass
        if not tap_if_present(driver, HOME_TAB_BUTTON, timeout=3):
            try:
                driver.back()
            except Exception:  # noqa: BLE001 - best-effort recovery, keep retrying
                pass
        time.sleep(settle_seconds)

    # Sanity check if the app is minimized
    driver.activate_app(PINTEREST_PACKAGE)
    try:
        find_any(driver, CREATE_ICON_BUTTON, timeout=10)
        return
    except TimeoutError:
        pass
    raise TimeoutError("Could not get back to the Pinterest home feed (Create icon never reappeared)")


def login(driver, username: str, password: str):

    try:
        find_any(driver, EMAIL_ONLY_FIELD, timeout=15).send_keys(username)
        tap_any(driver, CONTINUE_BUTTON)
    except TimeoutError:
        pass  # already on the combined email+password screen

    email_field = find_any(driver, LOGIN_EMAIL_FIELD, timeout=20)
    email_field.clear()
    email_field.send_keys(username)
    find_any(driver, LOGIN_PASSWORD_FIELD).send_keys(password)
    tap_any(driver, LOGIN_SUBMIT_BUTTON)
    dismiss_popups(driver)
    ensure_on_home_feed(driver)
    dismiss_popups(driver)


def publish_pin(driver, remote_image_path: str):

    ensure_on_home_feed(driver)
    dismiss_popups(driver)
    tap_any(driver, CREATE_ICON_BUTTON, timeout=10)
    dismiss_popups(driver)
    # Some sessions may skip the "Start creating now" sheet and go straight
    # to the photo picker, so a missing sheet here isn't fatal.
    tap_if_present(driver, PIN_ACTION_BUTTON, timeout=6)
    dismiss_popups(driver)  # photo-access permission dialog typically appears here

    # Match on "<upload folder>/<filename>", not filename alone -- a phone that
    # already has a same-named file elsewhere (e.g. Download/).
    unique_suffix = f"{Path(REMOTE_IMAGE_DIR).name}/{Path(remote_image_path).name}"
    thumbnail = [(AppiumBy.XPATH, f"//*[contains(@content-desc, '{unique_suffix}')]")]
    tap_any(driver, thumbnail, timeout=25)  # gallery grid can take a while to populate
    # "Next" starts disabled until the selection registers -- wait for enabled=true,
    # not just present, or the click silently no-ops on a disabled button.
    tap_any(driver, GALLERY_NEXT_BUTTON, timeout=15, require_enabled=True)

    # 45s here because Pinterest has to finish uploading/processing
    # the image before this button renders.
    tap_any(driver, CREATE_PIN_SUBMIT_BUTTON, timeout=45)
    ensure_on_home_feed(driver)
    dismiss_popups(driver)


def _open_random_pin(driver) -> bool:
    """Taps a random visible pin card on the home feed."""
    for by, value in PIN_CARD:
        try:
            cards = [c for c in driver.find_elements(by=by, value=value) if c.is_displayed()]
        except Exception:  # noqa: BLE001 - try the next candidate
            continue
        if not cards:
            continue
        try:
            random.choice(cards).click()
            return True
        except Exception:  # noqa: BLE001 - a missed tap just means "try something else"
            continue
    return False


def _close_pin(driver):
    """Closes the pin closeup view."""
    try:
        if find_any(driver, CREATE_ICON_BUTTON, timeout=1).is_displayed():
            return
    except TimeoutError:
        pass
    if not tap_if_present(driver, CLOSEUP_BACK_BUTTON, timeout=3):
        try:
            driver.back()
        except Exception:  # noqa: BLE001 - best-effort recovery
            pass


def _like_current_pin(driver) -> bool:
    return tap_if_present(driver, REACT_BUTTON, timeout=4)


def _save_current_pin(driver) -> bool:
    return tap_if_present(driver, SAVE_BUTTON, timeout=4)


def _view_comments(driver) -> bool:
    if not tap_if_present(driver, COMMENTS_MODULE, timeout=4):
        return False
    time.sleep(random.uniform(1, 2))
    try:
        driver.back()
    except Exception:  # noqa: BLE001 - best-effort recovery
        pass
    return True


def _scroll_feed(driver):
    """Swipes the feed by a random amount, mimicking idle scrolling rather than
    only ever tapping the first pin that happens to be visible."""
    size = driver.get_window_size()
    x = int(size["width"] * 0.5)
    start_y = int(size["height"] * random.uniform(0.55, 0.7))
    end_y = int(size["height"] * random.uniform(0.2, 0.35))
    driver.swipe(x, start_y, x, end_y, random.randint(400, 900))


WARM_UP_PIN_ACTIONS = [_like_current_pin, _save_current_pin, _view_comments]


def warm_up(driver, min_actions=WARM_UP_MIN_ACTIONS, max_actions=WARM_UP_MAX_ACTIONS):
    """Runs a randomized handful of home-feed browsing actions (open a pin,
    then like/save/view its comments, then back) right after login and before
    any pin gets published"""
    ensure_on_home_feed(driver)
    dismiss_popups(driver)

    action_count = random.randint(min_actions, max_actions)
    print(f"   [WARMUP] running {action_count} warm-up action(s)")

    done = 0
    attempts = 0
    while done < action_count and attempts < action_count * 3:
        attempts += 1
        if random.random() < 0.3:
            _scroll_feed(driver)
            time.sleep(random.uniform(1, 2))
            continue
        if not _open_random_pin(driver):
            _scroll_feed(driver)  # nothing tappable yet -- load more of the feed
            time.sleep(1)
            continue
        time.sleep(random.uniform(1.5, 2.5))  # let the closeup view finish loading
        dismiss_popups(driver)
        random.choice(WARM_UP_PIN_ACTIONS)(driver)
        time.sleep(random.uniform(1, 2))
        _close_pin(driver)
        dismiss_popups(driver)
        # Re-check after every iteration.
        try:
            ensure_on_home_feed(driver)
        except TimeoutError as exc:
            print(f"   [WARMUP] lost the home feed mid warm-up, stopping early: {exc}")
            break
        time.sleep(random.uniform(0.5, 1.5))
        done += 1

    ensure_on_home_feed(driver)
    dismiss_popups(driver)

# Image upload

def parse_image_range(range_str: str, line_no: int):
    """Parses a range like '1-3' or a single index like '4'."""
    try:
        if "-" in range_str:
            start_str, end_str = range_str.split("-", 1)
            start, end = int(start_str), int(end_str)
        else:
            start = end = int(range_str)
    except ValueError:
        raise ValueError(f"devices.txt line {line_no}: invalid image range {range_str!r} (expected e.g. '1-3' or '4')")
    if start < 1 or end < start:
        raise ValueError(f"devices.txt line {line_no}: invalid image range {range_str!r} (expected e.g. '1-3' or '4')")
    return start, end


def parse_devices(path: str):
    devices = []
    for line_no, raw_line in enumerate(Path(path).read_text().splitlines(), start=1):
        line = raw_line.strip()
        if not line or line.startswith("#"):
            continue
        parts = line.split()
        if len(parts) == 4:
            serial, token, username, password = parts
            image_range = None  # -> gets every image in the folder, see main()
        elif len(parts) == 5:
            serial, token, username, password, range_str = parts
            image_range = parse_image_range(range_str, line_no)
        else:
            raise ValueError(
                f"devices.txt line {line_no}: expected 4 or 5 fields "
                f"(IP:PORT TOKEN USERNAME PASSWORD [IMAGE_RANGE]), got {len(parts)}: {raw_line!r}"
            )
        devices.append(dict(serial=serial, token=token, username=username, password=password, image_range=image_range))
    return devices


def resolve_images(images_arg: str):
    """Returns every image found, sorted alphabetically. Per-device IMAGE_RANGE
    fields in devices.txt index into this same sorted list."""
    p = Path(images_arg)
    if p.is_file():
        return [str(p)]
    return sorted([str(f) for f in p.iterdir() if f.suffix.lower() in (".jpg", ".jpeg", ".png")])

# Auto-discovery: if an argument is omitted, look for it next to the script

def autodiscover_devices_file(base: Path) -> Path:
    candidate = base / "devices.txt"
    if candidate.is_file():
        return candidate
    raise FileNotFoundError(f"No devices.txt found in {base}")


def autodiscover_apk(base: Path) -> Path:
    apks = sorted(base.glob("*.apk"))
    if not apks:
        raise FileNotFoundError(f"No .apk file found in {base}")
    if len(apks) > 1:
        print(f">> multiple .apk files found in {base}, using {apks[0].name}")
    return apks[0]


def autodiscover_images(base: Path) -> Path:
    images_dir = base / "images"
    if images_dir.is_dir():
        return images_dir
    loose = [f for f in base.iterdir() if f.suffix.lower() in (".jpg", ".jpeg", ".png")]
    if loose:
        return base
    raise FileNotFoundError(f"No images/ folder or .jpg/.png files found in {base}")


def run_device(device: dict, apk_path: str, image_paths: list):
    serial = device["serial"]
    print(f"\n=== {serial} ===")

    if not connect_and_login(serial, device["token"]):
        return
    if not install_apk(serial, apk_path):
        disconnect(serial)
        return

    grant_full_media_access(serial)
    remote_images = push_images(serial, image_paths)

    driver = None
    try:
        driver = build_driver(serial)
        login(driver, device["username"], device["password"])
        try:
            warm_up(driver)
        except Exception as exc:  # noqa: BLE001 - warm-up is best-effort, publishing still matters
            print(f"   [WARMUP_FAILED] {serial}: {exc}")
        published = 0
        total = len(remote_images)
        for i, remote_image in enumerate(remote_images, start=1):
            print(f"   [PUBLISHING] {serial}: pin {i}/{total}: {Path(remote_image).name}")
            try:
                publish_pin(driver, remote_image)
                published += 1
            except Exception as exc:  # noqa: BLE001 - one slow/failed pin shouldn't cost the rest
                print(f"   [PIN_FAILED] {serial}: {Path(remote_image).name}: {exc}")
            time.sleep(2)
        print(f"   [DONE] {serial}: published {published}/{len(remote_images)} pin(s)")
    except Exception as exc:  # noqa: BLE001 - keep going to the next phone either way
        print(f"   [ERROR] {serial}: {exc}")
    finally:
        if driver:
            driver.quit()
        disconnect(serial)


def main():
    script_dir = Path(__file__).resolve().parent

    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("devices_file", nargs="?", default=None,
                         help=f"IP:PORT TOKEN USERNAME PASSWORD [IMAGE_RANGE], one per line "
                              f"(default: devices.txt next to this script, i.e. {script_dir})")
    parser.add_argument("apk_path", nargs="?", default=None,
                         help="Path to the Pinterest APK to install (default: the .apk found next to this script)")
    parser.add_argument("images", nargs="?", default=None,
                         help="Folder of images, or a single image file "
                              "(default: an images/ folder, or loose .jpg/.png files, next to this script)")
    args = parser.parse_args()

    try:
        devices_file = Path(args.devices_file) if args.devices_file else autodiscover_devices_file(script_dir)
        apk_path = Path(args.apk_path) if args.apk_path else autodiscover_apk(script_dir)
        images_arg = Path(args.images) if args.images else autodiscover_images(script_dir)
    except FileNotFoundError as exc:
        sys.exit(f"{exc}\nPass devices_file / apk_path / images explicitly, or place them next to the script.")

    print(f">> devices file: {devices_file}")
    print(f">> apk: {apk_path}")
    print(f">> images: {images_arg}")

    try:
        devices = parse_devices(devices_file)
    except ValueError as exc:
        sys.exit(str(exc))

    all_images = resolve_images(images_arg)
    if not all_images:
        sys.exit(f"No images found at {images_arg}")

    print(f">> {len(devices)} phone(s) queued, {len(all_images)} image(s) available:")
    for i, image_path in enumerate(all_images, start=1):
        print(f"     {i}: {Path(image_path).name}")

    for device in devices:
        start, end = device["image_range"] or (1, len(all_images))
        if end > len(all_images):
            print(f"   [SKIP] {device['serial']}: image range {start}-{end} exceeds "
                  f"the {len(all_images)} image(s) available")
            continue
        image_subset = all_images[start - 1:end]
        run_device(device, str(apk_path), image_subset)


if __name__ == "__main__":
    main()
 
 
  1. Abra seu IDE ou terminal e execute:
    python appium_script.py

O script mostrará seu progresso no terminal. Ele se conectará ao primeiro telefone, completará o fluxo de trabalho do Pinterest, desconectará e, em seguida, continuará com o próximo telefone.

Quer tentar outro aplicativo Android? Este é um script de demonstração, não uma ferramenta pronta para todos os serviços. Você pode adaptá-lo, mas cada aplicativo usa botões e localizadores de elementos diferentes. Use Appium Inspector para encontrar os localizadores corretos e atualizar o script para o seu aplicativo.

 

Este artigo foi útil?

Dê seu feedback sobre este artigo

Neste artigo

  • O que o script faz
  • Antes de começar
  • Configurando devices.txt
  • Executando o script
  • appium_script.py

Comunidade Multilogin

Mantenha-se informado, compartilhe suas opiniões e envolva-se com os outros!

Telegram Icon Junte-se a nós no Telegram

Leia mais sobre o assunto

Blog Post Img

10 Best Datacenter Proxies for Web Scraping (2025 Edition)

Apr 2, 2025 5 min read
Google SERP Img

What is a Google SERP Proxy and Why Should You Care?

Apr 1, 2025 6 min read
UK Proxy Img

What Are Dedicated UK Proxies? Everything You Need to Know

Apr 1, 2025 6 min read
Related Article Title Icon

Artigos relacionados

  • Como automatizar o login em um site
  • Fazendo login no Multilogin automaticamente
  • Como recuperar tokens API com Python
  • Exportando as credenciais de proxy dos perfis

Soluções Mobile

  • Telefone na nuvem
  • Celular virtual
  • Celular remoto
  • Fazenda de celulares
  • Celular na nuvem
  • Emulador android na nuvem
  • Automação de ações rápidas de IA

Multi-contabilidade

  • Crie várias contas no Instagram
  • Criar várias contas no TikTok
  • Criar várias contas no Reddit
  • Criar várias contas no Telegram
  • Crie várias contas no Facebook
  • Criar várias contas no YouTube
  • Crie várias contas no LinkedIn

Comparação

  • Multilogin vs. MoreLogin
  • Multilogin vs. FlashID
  • Multilogin vs. DuoPlus
  • Multilogin vs. VMOS Cloud
  • Multilogin vs. Octo Browser
  • Multilogin vs. AdsPower
  • Multilogin vs. GoLogin

Proxies de plataforma

  • Proxy móvel
  • Proxy Reddit
  • Proxy do Facebook
  • Proxy SOCKS5
  • Proxy do Instagram
  • Proxy OnlyFans
  • Proxy do LinkedIn

Casos de Uso

  • Celular na nuvem para TikTok
  • Celular na nuvem para Instagram
  • Celular na nuvem para Reddit
  • Celular na nuvem para Facebook
  • Celular na nuvem para YouTube

Recursos

  • Central de conhecimentos
  • Documentação da API
  • Glossário
  • Blog
  • Status do aplicativo
  • Notas de versão

FERRAMENTAS GRATUITAS

  • YouTube views to money calculator
  • Instagram money calculator
  • Conversor online de URL para texto
  • Google local SERP checker
  • Random address generator

Entre em contato

  • Suporte 24/7
    support@multilogin.com
  • Programa de indicação
  • Programa de afiliados
  • Página de preços
  • Carreiras
GDPR Compliant

© 2026 Multilogin. All rights reserved.

  • Privacy policy
  • Terms of service
  • Cookie policy
Multilogin watermark
  • SOLUÇÕES MOBILE

    • Telefone na nuvem
    • Celular virtual
    • Celular remoto
    • Fazenda de celulares
    • Celular na nuvem
    • Emulador android na nuvem
    • Automação de ações rápidas de IA
  • MULTI-CONTABILIDADE

    • Crie várias contas no Instagram
    • Criar várias contas no TikTok
    • Criar várias contas no Reddit
    • Criar várias contas no Telegram
    • Crie várias contas no Facebook
    • Criar várias contas no YouTube
    • Crie várias contas no LinkedIn
  • COMPARAÇÃO

    • Multilogin vs. MoreLogin
    • Multilogin vs. FlashID
    • Multilogin vs. DuoPlus
    • Multilogin vs. VMOS Cloud
    • Multilogin vs. Octo Browser
    • Multilogin vs. AdsPower
    • Multilogin vs. GoLogin
  • PROXIES DE PLATAFORMA

    • Proxy móvel
    • Proxy Reddit
    • Proxy do Facebook
    • Proxy SOCKS5
    • Proxy do Instagram
    • Proxy OnlyFans
    • Proxy do LinkedIn
  • CASOS DE USO

    • Celular na nuvem para TikTok
    • Celular na nuvem para Instagram
    • Celular na nuvem para Reddit
    • Celular na nuvem para Facebook
    • Celular na nuvem para YouTube
  • RECURSOS

    • Central de conhecimentos
    • Documentação da API
    • Glossário
    • Blog
    • Status do aplicativo
    • Notas de versão
  • FERRAMENTAS GRATUITAS

    • YouTube views to money calculator
    • Instagram money calculator
    • Conversor online de URL para texto
    • Google local SERP checker
    • Random address generator
  • ENTRE EM CONTATO

    • Suporte 24/7
      support@multilogin.com
    • Programa de indicação
    • Programa de afiliados
    • Página de preços
    • Carreiras
GDPR Compliant
  • Privacy policy
  • Terms of service
  • Cookie policy

© 2026 Multilogin. All rights reserved.

Expand