To help you get started, we've collected the most common ways that QtPy is being used within popular public projects.

mantidproject / mantid / qt / python / mantidqt / utils / qt / testing / nosy_qapplication.py View on Github

def str_from_button(self, enum_val):
        if enum_val == Qt.LeftButton:
            return "LeftButton"
        else:
            return ""

View more ways to use QtPy

QtPy: Abstraction layer for PyQt5/PySide2/PyQt6/PySide6

Copyright © 2009– The Spyder Development Team

Description

QtPy is a small abstraction layer that lets you write applications using a single API call to either PyQt or PySide.

It provides support for PyQt5, PySide2, PyQt6 and PySide6 using the Qt5 layout (where the QtGui module has been split into QtGui and QtWidgets).

Basically, you can write your code as if you were using PyQt or PySide directly, but import Qt modules from qtpy instead of PyQt5, PySide2, PyQt6 or PySide6.

Accordingly, when porting code between different Qt bindings (PyQt vs PySide) or Qt versions (Qt5 vs Qt6), QtPy makes this much more painless, and allows you to easily and incrementally transition between them. QtPy handles incompatibilities and differences between bindings or Qt versions for you while keeping your project running, so you can focus more on your own code and less on keeping track of supporting every Qt version and binding. Furthermore, when you do want to upgrade or support new bindings, it allows you to update your project module by module rather than all at once. You can check out examples of this approach in projects using QtPy, like git-cola.

Attribution and acknowledgments

This project is based on the pyqode.qt project and the spyderlib.qt module from the Spyder project, and also includes contributions adapted from qt-helpers, developed as part of the glue project.

Unlike pyqode.qt this is not a namespace package, so it is not tied to a particular project or namespace.

License

This project is released under the MIT license.

Requirements

You need PyQt5, PySide2, PyQt6 or PySide6 installed in your system to make use of QtPy. If several of these packages are found, PyQt5 is used by default unless you set the QT_API environment variable.

QT_API can take the following values:

• pyqt5 (to use PyQt5).
• pyside2 (to use PySide2).
• pyqt6 (to use PyQt6).
• pyside6 (to use PySide6).

Module aliases and constants

• QtCore.pyqtSignal, QtCore.pyqtSlot and QtCore.pyqtProperty (available on PyQt5/6) are instead exposed as QtCore.Signal, QtCore.Slot and QtCore.Property, respectively, following the Qt5 module layout.

• The Qt version being used can be checked with QtCore.__version__ (instead of QtCore.QT_VERSION_STR) as well as from qtpy.QT_VERSION.

• For PyQt6 enums, unscoped enum access was added by promoting the enums of the QtCore, QtGui, QtTest and QtWidgets modules.

• Compatibility is added between the QtGui and QtOpenGL modules for the QOpenGL* classes.

• To check the current binding version, you can use qtpy.PYSIDE_VERSION for PySide2/6 and qtpy.PYQT_VERSION for PyQt5/6. If the respective binding is not being used, the value of its attribute will be None.

• To check the current selected binding, you can use qtpy.API_NAME

• There are boolean values to check if Qt5/6, PyQt5/6 or PySide2/6 are being used: qtpy.QT5, qtpy.QT6, qtpy.PYQT5, qtpy.PYQT6, qtpy.PYSIDE2 and qtpy.PYSIDE6. True if currently being used, False otherwise.

Compat module

In the qtpy.compat module, you can find wrappers for QFileDialog static methods and SIP/Shiboken functions, such as:

• QFileDialog.getExistingDirectory wrapped with qtpy.compat.getexistingdirectory

• QFileDialog.getOpenFileName wrapped with qtpy.compat.getopenfilename

• QFileDialog.getOpenFileNames wrapped with qtpy.compat.getopenfilenames

• QFileDialog.getSaveFileName wrapped with qtpy.compat.getsavefilename

• sip.isdeleted and shiboken.isValid wrapped with qtpy.compat.isalive

Installation

pip install qtpy

or

conda install qtpy

Type checker integration

Type checkers have no knowledge of installed packages, so these tools require additional configuration.

A Command Line Interface (CLI) is offered to help with usage of QtPy (to get MyPy and Pyright/Pylance args/configurations).

Mypy

The mypy-args command helps you to generate command line arguments for Mypy that will enable it to process the QtPy source files with the same API as QtPy itself would have selected.

If you run

qtpy mypy-args

QtPy will output a string of Mypy CLI args that will reflect the currently selected Qt API. For example, in an environment where PyQt5 is installed and selected (or the default fallback, if no binding can be found in the environment), this would output the following:

--always-true=PYQT5 --always-false=PYSIDE2 --always-false=PYQT6 --always-false=PYSIDE6

Using Bash or a similar shell, this can be injected into the Mypy command line invocation as follows:

mypy --package mypackage $(qtpy mypy-args)

Pyright/Pylance

In the case of Pyright, instead of runtime arguments, it is required to create a config file for the project, called pyrightconfig.json or a pyright section in pyproject.toml. See here for reference. In order to set this configuration, QtPy offers the pyright-config command for guidance.

If you run

qtpy pyright-config

you will get the necessary configs to be included in your project files. If you don't have them, it is recommended to create the latter. For example, in an environment where PyQt5 is installed and selected (or the default fallback, if no binding can be found in the environment), this would output the following:

pyrightconfig.json:
{"defineConstant": {"PYQT5": true, "PYSIDE2": false, "PYQT6": false, "PYSIDE6": false}}

pyproject.toml:
[tool.pyright.defineConstant]
PYQT5 = true
PYSIDE2 = false
PYQT6 = false
PYSIDE6 = false

Note: These configurations are necessary for the correct usage of the default VSCode's type checking feature while using QtPy in your source code.

Testing matrix

Currently, QtPy runs tests for different bindings on Linux, Windows and macOS, using Python 3.7 and 3.11, and installing those bindings with conda and pip. For the PyQt bindings, we also check the installation of extra packages via pip.

Following this, the current test matrix looks something like this:

Note: The mentioned extra packages for the PyQt bindings are the following:

• PyQt3D and PyQt6-3D
• PyQtChart and PyQt6-Charts
• PyQtDataVisualization and PyQt6-DataVisualization
• PyQtNetworkAuth and PyQt6-NetworkAuth
• PyQtPurchasing
• PyQtWebEngine and PyQt6-WebEngine
• QScintilla and PyQt6-QScintilla

Contributing

Everyone is welcome to contribute! See our Contributing guide for more details.

Sponsors

QtPy is funded thanks to the generous support of

and the donations we have received from our users around the world through Open Collective: