huskium 1.0.3

UI Automation Page Objects design pattern.

huskium

---

Copyright

• Developer: Johnny Chou

---

Overview

• huskium is a Page Object framework built on Selenium and Appium.
• huskium leverages Python’s data descriptors to simplify and enhance UI automation.

---

Usage

• Build page objects simply and elegantly using the Page and Element(s) classes.
• Write test scripts simply and elegantly using the constructed Page objects.

---

Page Object Example Code

1. Construct a Page object in any Python file, each page recommended to be a separate Page class.

# my_page.py


from huskium import Page, Element, Elements
from huskium import By
from huskium import dynamic


class MyPage(Page):
    
    # Static element: The most common way to set up Page objects.
    # Element is a data descriptor of Page, allowing simple setup as shown below:
    search_field = Element(By.NAME, 'q', remark='Search input box')
    search_results = Elements(By.TAG_NAME, 'h3', remark='All search results')
    search_result1 = Element(By.XPATH, '(//h3)[1]', remark='First search result')
    
    # Dynamic element: Rarely used, typically determined during test case execution.
    # Must use @dynamic to enable the descriptor's functionality.
    @dynamic
    def search_result(self, order: int = 1):
        return Element(By.XPATH, f'(//h3)[{order}]', remark=f'Search result no.{order}')
    
    # For dynamic elements as properties, use the following format:
    @property
    @dynamic
    def keyword_results(self):
        return Elements(By.XPATH, f'//*[contains(text(), "{Keyword.text1}")]')
    
    # To record dynamic elements statically, use standard data descriptor dynamic assignment:
    # 1. Create a data descriptor object (e.g., static_element).
    static_element = Element()
    
    # 2. Define a function and call "dynamic" to assign a value to "static_element".
    # The logic for dynamic parameters is the same as in Element.
    # After calling "dynamic_element", you can also use "static_element" to operate it.
    def dynamic_element(self, par):
        return self.static_element.dynamic(By.XPATH, f'//*[contains(text(), "{par}")]')
    
    # 3. Use the standard method for a data descriptor.
    def dynamic_element(self, par):
        self.static_element = Element(By.XPATH, f'//*[contains(text(), "{par}")]')
        return self.static_element

2. After constructing the Page object, you can begin writing test cases.

# test_my_page.py


from selenium import webdriver
from my_page import MyPage


class TestMyPage:
    
    driver = webdriver.Chrome()

    # Set up a page object. All actions will be triggered from this object.
    my_page = MyPage(driver)

    # The Page object can also call driver-related methods.
    my_page.get("https://google.com")

    # Example of a wait usage:
    # Wait until an element is visible, then take a screenshot.
    my_page.search_field.wait_visible()
    my_page.save_screenshot("my/file/image1.png")

    # All actions automatically handle explicit waits.
    # No need to manually call wait methods unless required, 
    # e.g. Equivalent to: 
    # my_page.search_field.wait_clickable().send_keys(keyword).wait_clickable().submit()
    my_page.search_field.send_keys(keyword).submit()

    # Various wait states are available.
    my_page.loading_image.wait_absent()
    my_page.search_results.wait_all_visible()
    my_page.save_screenshot("my/file/image2.png")

    # Assertions can be made directly:
    search_keyword = 'dinner'
    assert my_page.keyword_results(search_keyword).quantity > 1
    assert search_keyword in my_page.search_result1.text

    # Reuse found elements through existing sessions:
    # Once an element (e.g., `my_page.search_result1`) is located, 
    # it will use the same session unless the element becomes stale.
    # No need to store the found element in a separate variable.
    # Just perform actions directly:
    my_page.search_result1.click()
    ...

    driver.close()

---

Timeout Global Settings

1. In addition to setting timeouts for individual elements and methods, a global timeout setting is also available. See the example below:

from huskium import Timeout


# Set the default timeout for all Elements.
# The huskium default is 30 seconds. You can change it as needed:
Timeout.DEFAULT = 60

# If you don’t want any waiting, you can also set it to 0:
Timeout.DEFAULT = 0

# Set the reraise behavior for timeouts on all Elements.
# The huskium default is True, with the following logic:
# True: Raise a TimeoutException if the element times out.
# False: Return False if the element times out, without raising a TimeoutException.
Timeout.RERAISE = False

2. The priority order for timeout values is as follows:

• P1: Method-Level:
  • page.element.wait_method(timeout=20)
• P2: Element-Level:
  • element = Element(..., timeout=10, ...)
• P3: Global-Level:
  • Timeout.DEFAULT = 60

3. The priority order for timeout reraise behavior is as follows:

• P1: Method-Level:
  • page.element.wait_method(reraise=True)
• P2: Global-Level:
  • Timeout.RERAISE = False

---

Cache Global Settings

1. Cache determines whether the Element class should reference a previously located WebElement for actions or locate the element again for each action.
2. The benefits of Cache are evident when the same Element is accessed multiple times, such as performing .text followed by .click().
3. Note that Elements does not support cache. For multiple elements, the state can be highly unstable, so each action must locate the elements again to ensure expected behavior.

from huskium import Cache


# Set the default cache for all Element.
# The default is True. You can change it as needed:
Cache.ELEMENT = False


# You can also configure the cache for an individual Element:
element = Element(..., cache=False)

4. The priority order for cache value is as follows:

• P1: Element-Level:
  • element = Element(..., cache=False)
• P2: Global-Level:
  • Cache.ELEMENT = False

---

Wait Actions

We offer a comprehensive set of wait methods, extending the official expected_conditions in ec_extension.py and encapsulating them into corresponding methods. Below are the extended methods for Element(s):

# Element
page.element.wait_present()
page.element.wait_absent()
page.element.wait_visible()
page.element.wait_invisible()
page.element.wait_clickable()
page.element.wait_unclickable()
page.element.wait_selected()
page.element.wait_unselected()

# Elements
page.elements.wait_all_present()
page.elements.wait_all_absent()
page.elements.wait_all_visible()
page.elements.wait_any_visible()

# You can set default timeout and reraise behavior for all wait functions.
page.element.wait_visible(timeout=10, reraise=True)
# Recommended to use default settings (timeout=30, reraise=True) for simplicity.
page.element.wait_visible()

# For reverse conditions like invisible and unclickable, 
# use the "present" parameter to define if existence is required.
# Element must be present and invisible (default).
page.element.wait_invisible(present=True)
# Element can be absent or [present and invisible].
page.element.wait_invisible(present=False)
# Element must be present and unclickable (default).
page.element.wait_unclickable(present=True)
# Element can be absent or [present and unclickable].
page.element.wait_unclickable(present=False)

# Selection states are tied to user actions, 
# so the element must be present; no "present" parameter is provided.
page.element.wait_selected()
page.element.wait_unselected()

---

Appium Extended Actions

We have extended Appium with highly convenient action features, including:

# Offset allows you to define swipe directions. 
# It supports eight standard directions: 
# UP, DOWN, LEFT, RIGHT, UPPER_LEFT, UPPER_RIGHT, LOWER_LEFT, LOWER_RIGHT.
# Area lets you define the swipeable range, 
# defaulting to the full device screen (FULL), or you can customize it.

from huskium import Offset, Area

# Page swiping. Refer to function docstrings for details.
page.swipe_by(Offset.UP, Area.FULL)
page.swipe_by(Offset.DOWN)
page.swipe_by(Offset.LEFT)
page.swipe_by(Offset.RIGHT)
page.swipe_by(Offset.UPPER_LEFT)
page.swipe_by(Offset.UPPER_RIGHT)
page.swipe_by(Offset.LOWER_LEFT)
page.swipe_by(Offset.LOWER_RIGHT)

# Page flicking. Refer to function docstrings for details.
# All Offset directions are supported.
page.flick_by(Offset.UP, Area.FULL)

# Element swiping until an element is visible.
# All Offset directions are supported.
page.element.swipe_by(Offset.UP, Area.FULL)

# Element flicking until an element is visible.
# All Offset directions are supported.
page.element.flick_by(Offset.UP, Area.FULL)

# Page draw lines.
# Define dots coordinates
dots = [{"x": x1, "y": y1}, {"x": x2, "y": y2}, {"x": x3, "y": y3}, ...]
# Alternatively, use element locations if available
dots = page.elements.locations
page.draw_lines(dots)

# Page draw gesture.
# 9-grid coordinates, or define your own
dots = page.elements.locations  
# 9-grid gesture string (1–9 represent grid positions). This example draws a reverse Z.
gesture = "9875321"  
page.draw_gesture(dots, gesture)

---

Other Actions

All Selenium and Appium features are included. Here are some examples:

# ActionChains
page.element.scroll_to_element().perform()
page.element.move_to_element().drag_and_drop().perform()

# Temporarily store ActionChains and execute later
page.element.move_to_element().drag_and_drop()
...  # Process other logic before executing perform()
page.element.perform()

# Select options
page.element.options
page.element.select_by_value()

---

Logstack

Using logstack to log specific frame information. The logstack module extends logging functionality, allowing you to capture information for specific frames, such as those starting with a designated prefix (e.g., test), without tracing all frames manually.

from huskium import logstack

# Configure logging using either logging.basicConfig() or logstack.config().
# logstack.config() simplifies the default settings. You can use it as shown below
# to output the log file to "./log.log".
logstack.config()

# Use logstack in your code to log specific frames
def some_func():
    ...
    # Logs information from the first frame with the prefix (default: test)
    logstack.info("Log from some function.", prefix="test")

def test_func():
    ...
    # Logs frame info for test_func, not some_func
    some_func()

# Example log output:
# 2025-01-04 18:20:48 | INFO | testing.py:32 | test_func | Log from some function.

Inheritance

You can also extend the Page and Element classes to include custom methods. There’s no need to manually handle descriptors, and the inheritance usage remains unchanged.

from huskium import Page as HuskyPage
from huskium import Element as HuskyElement


class Page(HuskyPage):

    def extended_func(self, par):
        ...


class Element(HuskyElement):

    def extended_func(self, par):
        ...

---

TODO

Keep tracking the Appium version.