Skip to content
/ manen Public

An implementation of the Page Object Model design pattern, and other utilities for web scraping and automation

License

Notifications You must be signed in to change notification settings

kodaho/manen

Repository files navigation

🌔 Manen


PyPI

An implementation of the Page Object Model design pattern, and other utilities for web scraping and automation.


PyPIDocumentationChangelogIssue tracker

Manen is a package built to enhance developer experience when using Selenium. Among the core features, you can find:

  • an implementation of the Page Object Model design pattern
  • a class which improves the operability of a Selenium WebDriver
  • a function to easily find and isolate DOM elements inside a Selenium page

This package aims to provide you the tools to write more concise, flexible and powerful code compared to what you would do by using only Selenium.

Note

For now, only Selenium Chrome WebDriver is supported. Other browsers will be supported in the future, as well as other automation tools such as Playwright or Scrapy.

📥 Installation

The package can be installed using the official Python package manager pip.

pip install manen

✨ Features

  • manen.finder.find allows to easily get element(s) in a HTML page. This function support several very different use cases, to help reduce your code complexity when fetching for elements (example: using default values, trying different selectors, iterating over several elements).
  • manen.browser defines an enhanced Selenium WebDriver called Browser
  • manen.page_object_model is an implementation of the Page Object Model design pattern. It will wrap a HTML page, component and DOM values inside Python classes and objects, providing a better way to interact with a web page.

🚀 Getting started

Manen features will be explored by a simple example: going to the PyPI page, searching for a specific package and extracting some information from the search results.

First thing to do is to initialize a WebDriver instance. It can be done using the usual way provided by Selenium, but an alternative is to use the Browser class provided by Manen. Note that both ways are equivalent, but Browser provides some additional features, that won't be explored here.

from manen.browser import ChromeBrowser

browser = ChromeBrowser.initialize()
browser.get("https://pypi.org")

PyPI home page

We are now on the home page of PyPI. What we are going to do now is building a class that will inherit from Page from the manen.page_object_model.component module. This Python class will be a reflect of the HTML page, allowing us to access DOM elements in the same way we access attributes. Note the whole page object model design pattern is implemented with type hints (a bit like in Pydantic model).

from manen.page_object_model.types import href, input_value
from manen.page_object_model.config import CSS, Attribute, DatetimeFormat, XPath
from manen.page_object_model.component import Page, Component


class HomePage(Page):
    query: Annotated[input_value, CSS("input[name='q']")]


class SearchResultPage(Page):
    class Result(Component):
        name: Annotated[str, CSS("h3 span.package-snippet__name")]
        version: Annotated[str, CSS("h3 span.package-snippet__version")]
        link: Annotated[href, CSS("a.package-snippet")]
        description: Annotated[str, CSS("p.package-snippet__description")]
        release_datetime: A[
            datetime,
            DatetimeFormat("%Y-%m-%dT%H:%M:%S%z"),
            Attribute("datetime"),
            CSS("span.package-snippet__created time"),
        ]

    nb_results: Annotated[
        int,
        XPath("//*[@id='content']//form/div[1]/div[1]/p/strong"),
    ]
    results: Annotated[
        list[Result],
        CSS("ul[aria-label='Search results'] li"),
    ]

The Page class encapsulates the whole current HTML page available through the driver. Each DOM value we want to extract is then represented by a class attribute, with a type (what to extract) and a selector (where to extract it). Depending on the type of the value, Manen will automatically execute the appropriate DOM content extraction on the HTML element (for example, it will extract the inner text for a str type, the HTML attribute href for a HRef , or the property innerHTML for InnerHTML).

A Component captures a sub-part of an HTML page. All the elements defined under this will be fetched inside the HTML element represented by the Component class.

Here the class HomePage defines an Input element, that will be linked to the search bar. Filling the search bar is done by assigning a value to the attribute query.

from selenium.webdriver.common.keys import Keys

page = HomePage(browser) # A Page object is initialized only with a WebDriver instance

page.query = "manen"
page.query += Keys.ENTER

After submitting the form, we are redirected to the search results page.

PyPI home page

The SearchResultPage will then be used to extract the results.

page = SearchResultPage(browser)

print(page.nb_results)
# 3

print(page.results[0])
# <__main__.SearchResultPage.Result at 0x1058e97c0>

Manen provides a model_dump method, quite similar to the one in Pydantic to easily get all the attributes of a component or a page.

print(page.results[0].model_dump())
# {'name': 'manen',
#  'version': '0.2.0',
#  'link': 'https://pypi.org/project/manen/',
#  'description': 'A package around Selenium with an implementation of the page object model, an enhanced WebDriver and a CLI.',
#  'release_datetime': datetime.datetime(2022, 2, 19, 12, 10, 31, tzinfo=datetime.timezone.utc)}

Tip

Other DOM elements are also implemented, such as ImageSrc, Input, Checkbox... Each one of them is used to target a specific attribute from a DOM value and can enable interaction with it, in a flawless Pythonic way. Check the documentation for the list of available elements.

Let's finally close the Selenium WebDriver to avoid any remaining running applications once we exit the Python program.

browser.quit()

🦾 Going further

The documentation provides an extensive overview of the possibilities offered by the package. It also contains several user guides to help you get started with the package.

A set of examples is also available in the examples/ directory of the source code repository. They are designed to show you how to use the package in a real-world context.

Don't hesitate to open an issue if you have any question or concern about this project!

Looking to contribute to fix or add new features? Just read this page, fork the repository and start doing the modifications you want.