robocorp-browser

robocorp-browser 包是针对 Playwright 项目的轻量级包装，具有诸如Playwright对象的生命周期管理自动化（旨在与 robocorp-tasks 一起使用）等生活品质改进。

用法

👉 确认您已在配置中添加了依赖项；此库不是robocorp捆绑包的一部分。

• 用于自动化任务包的 conda.yaml
• 用于自动化操作包的 package.yaml
• 其余部分 requirements.txt、pyproject.toml、setup.py|cfg 等

from robocorp import browser, vault
from robocorp.tasks import task

@task
def automate_browser():
    """Start a browser to login in the surfed page."""
    # The configuration is used to set the basic `robocorp.browser` settings.
    # It must be called before calling APIs that create Playwright objects.
    browser.configure(
        # NOTE: `screenshot="only-on-failure"` is the default.
        # If this function finishes with an exception, it will make a screenshot and
        #  embed it into the logs.
        screenshot="only-on-failure",

        # By default, `headless` is False, unless running in a Linux container
        #  without a DISPLAY/WAYLAND_DISPLAY environment variable, but it
        #  can also be manually overridden.
        headless=True,  # won't display the browser window

        # Interactions may be run in slow motion (given in milliseconds).
        slowmo=100,
    )

    # The `browser.goto()` call may be used as a shortcut to get the current page and
    #  surf some URL (it may create the browser if not created already).
    browser.goto("https://example.com>")

    _login()  # call the login instructions


def _login():
    # APIs in `robocorp.browser` return the same browser instance, which is
    #  automatically closed when the task finishes.
    page = browser.page()

    # `robocorp.vault` is recommended for managing secrets.
    account = vault.get_secret("default-account")

    # Use the Playwright Browser API as usual to interact with the web elements.
    page.fill('//input[@ng-reflect-name="password"]', account["password"])
    page.click("input:text('Submit')")

🚀 现在开始使用我们的模板！

指南

• 浏览器配置
• 持久上下文

API参考

探索我们的API，获取详细的文档。

变更日志

发布列表及其对应更改可在变更日志中找到。