appium-pytest-kit 0.1.0

Reusable Appium 2.x + pytest mobile test framework

appium-pytest-kit

appium-pytest-kit is a reusable Appium 2.x + pytest framework library for Python 3.11+.

• pip install appium-pytest-kit (or install from GitHub — see below)
• appium-pytest-kit-init to bootstrap configuration, or --framework to scaffold a full project
• Write tests immediately with built-in fixtures and zero boilerplate

Full documentation: DOCUMENTATION.md

---

Installation

From PyPI (once published)

pip install appium-pytest-kit

From GitHub

# latest main branch
pip install git+https://github.com/gianlucasoare/appium-pytest-kit.git

# specific branch
pip install git+https://github.com/gianlucasoare/appium-pytest-kit.git@main

# specific tag
pip install git+https://github.com/gianlucasoare/appium-pytest-kit.git@v0.1.0

Local clone (editable, for development)

git clone https://github.com/gianlucasoare/appium-pytest-kit.git
cd appium-pytest-kit
pip install -e ".[dev]"

---

Quickstart

python -m venv .venv
source .venv/bin/activate
pip install git+https://github.com/gianlucasoare/appium-pytest-kit.git
appium-pytest-kit-init          # creates .env with starter config
# or scaffold a full project:
appium-pytest-kit-init --framework --root my-project
pytest -q

Edit .env with your device and app details, then write tests.

---

Step-by-step: test a real app in 5 minutes

This example tests the Android Calculator on an emulator. See DOCUMENTATION.md for the full iOS walkthrough and all options.

1. Start Appium and an emulator

appium &
emulator -avd Pixel_7_API_33 &
adb devices    # confirm emulator-5554 is listed

2. Configure .env

APP_PLATFORM=android
APP_APPIUM_URL=http://127.0.0.1:4723
APP_DEVICE_NAME=emulator-5554
APP_PLATFORM_VERSION=13
APP_APP_PACKAGE=com.google.android.calculator
APP_APP_ACTIVITY=com.android.calculator2.Calculator
APP_NO_RESET=true

3. Write a test

# tests/test_calculator.py
import pytest
from appium.webdriver.common.appiumby import AppiumBy

BTN_2      = (AppiumBy.ACCESSIBILITY_ID, "2")
BTN_PLUS   = (AppiumBy.ACCESSIBILITY_ID, "plus")
BTN_3      = (AppiumBy.ACCESSIBILITY_ID, "3")
BTN_EQUALS = (AppiumBy.ACCESSIBILITY_ID, "equals")
RESULT     = (AppiumBy.RESOURCE_ID, "com.google.android.calculator:id/result_final")


@pytest.mark.integration
def test_addition(actions):
    actions.tap(BTN_2)
    actions.tap(BTN_PLUS)
    actions.tap(BTN_3)
    actions.tap(BTN_EQUALS)
    assert actions.text(RESULT) == "5"

4. Run it

pytest -m integration -v

---

Built-in fixtures

Fixture	Scope	Description
settings	session	Resolved AppiumPytestKitSettings — access any config field
appium_server	session	Server URL and whether it is framework-managed
driver	function	Live appium.webdriver.Remote, quit automatically after each test
waiter	function	Explicit waits with WaitTimeoutError on timeout
actions	function	High-level UI helpers: tap, type_text, text, exists, swipe, and more

---

Session modes

Control driver lifecycle per test or across the whole session:

APP_SESSION_MODE=clean          # default: fresh driver per test
APP_SESSION_MODE=clean-session  # shared driver, app reset between tests
APP_SESSION_MODE=debug          # shared driver, no reset (fast local debugging)

---

Device resolution (3-tier priority)

1. Explicit — APP_DEVICE_NAME / APP_UDID set in .env or CLI
2. Profile — APP_DEVICE_PROFILE=pixel7 resolved from data/devices.yaml
3. Auto-detect — adb devices (Android) or xcrun simctl / xctrace (iOS)

# Use a named profile from data/devices.yaml
pytest --app-device-profile pixel7

# Auto-detect (no device settings needed)
pytest

---

Failure diagnostics

On test failure the framework automatically captures:

• Screenshot → artifacts/screenshots/<test_id>.png
• Page source → artifacts/pagesource/<test_id>.xml
• Video (if policy allows) → artifacts/videos/<test_id>.mp4

APP_VIDEO_POLICY=never    # default
APP_VIDEO_POLICY=failed   # record and save only on failure
APP_VIDEO_POLICY=always   # record every test
APP_ARTIFACTS_DIR=artifacts

Allure attachments are added automatically when allure-pytest is installed.

---

Configuration

Settings are loaded from .env → environment variables → CLI flags (highest wins).

pytest --app-platform ios
pytest --app-device-name "Pixel 7" --app-platform-version 13
pytest --appium-url http://192.168.1.10:4723
pytest --app-app-package com.example.app --app-app-activity .MainActivity
pytest --app-session-mode clean-session
pytest --app-device-profile pixel7
pytest --app-video-policy failed
pytest --app-is-simulator
pytest --app-capabilities-json '{"autoGrantPermissions": true}'
pytest --app-manage-appium-server    # start Appium automatically
pytest --app-reporting-enabled       # write artifacts/appium-pytest-kit/summary.json

See DOCUMENTATION.md § Configuration for the full settings table.

---

Expanded waits

waiter.for_clickable(locator)                     # wait for element to be tappable
waiter.for_invisibility(locator)                  # wait for element to disappear
waiter.for_text_contains(locator, "partial text") # wait for text substring
waiter.for_text_equals(locator, "exact text")     # wait for exact text match
waiter.for_all_visible([loc1, loc2, loc3])        # wait for all elements to appear
waiter.for_any_visible([loc1, loc2])              # wait for first visible element

---

Expanded actions

actions.tap_if_present(locator)       # tap only if visible, returns bool
actions.clear(locator)                # clear a text field
actions.swipe(sx, sy, ex, ey)         # W3C Pointer swipe gesture
actions.scroll_down()                 # swipe up on screen center
actions.scroll_up()                   # swipe down on screen center

---

Extension hooks

Implement these in your conftest.py to customise behaviour without touching the framework:

# conftest.py

def pytest_appium_pytest_kit_capabilities(capabilities, settings):
    """Add extra capabilities before each driver session."""
    return {"autoGrantPermissions": True, "language": "en"}


def pytest_appium_pytest_kit_configure_settings(settings):
    """Replace settings at session start."""
    return settings.model_copy(update={"implicit_wait": 2.0})


def pytest_appium_pytest_kit_driver_created(driver, settings):
    """Run setup immediately after each driver is created."""
    driver.orientation = "PORTRAIT"

---

Project scaffolding

Generate a full project structure in one command:

appium-pytest-kit-init --framework --root my-project

Creates:

my-project/
├── data/devices.yaml          # device profiles
├── pages/
│   ├── base_page.py           # BasePage composition class
│   └── example_page.py        # starter page object
├── flows/                     # reusable multi-step flows
├── tests/
│   ├── android/test_smoke.py
│   └── ios/test_smoke.py
├── conftest.py
├── pytest.ini
└── .env.example

---

Public API

Top-level imports (stable):

from appium_pytest_kit import (
    AppiumPytestKitSettings,
    AppiumPytestKitError,
    ConfigurationError,
    DeviceResolutionError,
    LaunchValidationError,
    WaitTimeoutError,
    ActionError,
    DriverCreationError,
    DeviceInfo,
    DriverConfig,
    MobileActions,
    Waiter,
    build_driver_config,
    create_driver,
    load_settings,
    apply_cli_overrides,
)

Stable public modules (direct import):

• appium_pytest_kit.settings
• appium_pytest_kit.driver
• appium_pytest_kit.waits
• appium_pytest_kit.actions
• appium_pytest_kit.errors
• appium_pytest_kit.interfaces — CapabilitiesAdapter protocol for custom adapters

Private/internal modules (no compatibility guarantee):

• appium_pytest_kit._internal.*

---

Fixture lifecycle

flowchart TD
    A["pytest start"] --> B["load defaults + .env + env vars"]
    B --> C["apply --app-* CLI overrides"]
    C --> D["settings fixture (session)"]
    D --> E{"APP_MANAGE_APPIUM_SERVER"}
    E -->|"true"| F["start local Appium server"]
    E -->|"false"| G["use APP_APPIUM_URL"]
    F --> H["appium_server fixture"]
    G --> H
    H --> I{"session_mode"}
    I -->|"clean-session / debug"| J["_driver_shared (session)"]
    I -->|"clean"| K["driver per test"]
    J --> K
    K --> L["waiter / actions fixtures"]
    K --> M["test executes"]
    M --> N{"failed?"}
    N -->|"yes"| O["capture screenshot + page source"]
    N --> P["video stop (per policy)"]
    O --> P
    P --> Q["driver.quit() if clean mode"]
    Q --> R["optional report summary flush"]
    R --> S["optional server stop"]

---

Local development

pip install -e ".[dev]"
ruff check .
pytest -q
pytest --collect-only examples/basic/tests -q