Doxyqml turns .qml into pseudo-C++ which Doxygen can then use to generate documentation.
Doxyqml uses the standard Python setup tools, so you can install it with pip:
pip install doxyqml
or manually with:
python setup.py install
Edit your Doxygen file: first you need to set the FILTER_PATTERNS
key to
associate .qml files with Doxyqml:
FILTER_PATTERNS = *.qml=doxyqml
Note: On Windows you may need to use the full path of the doxyqml.py
file
instead. For example if you installed Python 2.7 in C:\Python27
, use this:
FILTER_PATTERNS = *.qml=C:\Python27\Lib\site-packages\doxyqml\doxyqml.py
Then you must add .qml files to FILE_PATTERNS
:
FILE_PATTERNS = *.qml
QML is partly-typed: functions are untyped, properties and signals are. Doxyqml provides a way to define types when they are missing or not precise enough.
Functions in QML are untyped, but you can define types in the documentation like this:
/**
* Create a user
* @param type:string firstname User firstname
* @param type:string lastname User lastname
* @param type:int User age
* @return type:User The User object
*/
function createUser(firstname, lastname, age);
QML properties are typed, so Doxyqml uses them by default. You can nevertheless
overwrite the type using the same type:<name>
syntax. This is useful to
document property aliases:
/** type:string The user lastname */
property alias lastname: someObject.text
QML signals are typed, so there is no need to use the type:<name>
syntax to
document their parameters. Using type:<name>
syntax in signal documentation
will not work: Doxyqml won't strip it out and Doxygen will confuse it with the
parameter name.
/**
* User just logged in
* @param user The user which logged in
*/
signal loggedIn(User user)