Using IDL union types in Blink is a bit tricky. Here are some tips to use union types correctly.
For each union type, the code generator creates a C++ class which is used like an “impl” class of a normal interface type. The name of a generated class is a type name of the union type. For example, the code generator will create
StringOrFloat class for
(DOMString or float).
The code generator puts generated classes into separate files. You need to include generated header files when you use union types in core/modules.
The file name for a generated class is basically the same as its class name, but we use some aliases to avoid too-long file names (See https://crbug.com/611437 why we need to avoid long file names).
The paths for generated classes depend on the places union types are used. If a union type is used only in IDL files under modules/, the include path is
bindings/modules/v8/foo_or_bar.h. Otherwise, the include path is
bindings/core/v8/foo_or_bar.h. For example, given following definitions:
// core/fileapi/file_reader.idl readonly attribute (DOMString or ArrayBuffer)? result; // dom/common_definitions.idl typedef (ArrayBuffer or ArrayBufferView) BufferSource; // modules/encoding/text_decoder.idl DOMString decode(optional BufferSource input, optional TextDecodeOptions options); // modules/fetch/request.idl typedef (Request or USVString) RequestInfo;
The include paths will be:
array_buffer_or_array_buffer_view.h is located under core/ even it is used by
request.idl which is located under modules/.
Special NOTE: If you are going to use a union type under core/ and the union type is currently used only under modules/, you will need to update the include path for the union type under modules/.
Due to the requirements of GN, we need to put generated file names in GN files. Please update