Hi Jakek,
"Jacek Caban" jacek@codeweavers.com wrote in message news:496A47AF.50506@codeweavers.com...
Hi Alistair,
Alistair Leslie-Hughes wrote:
Hi,
Changelog: mshtml: Implement IHTMLStyle get/put borderRightStyle
+static BOOL IsValidBorderStyle(BSTR v)
- Please use the same name convention as other functions in this file
- There is no test proving that the style is validated
- You didn't include get_borderRightStyle, which makes
put_borderRightStyle almost useless
static HRESULT WINAPI HTMLStyle_get_borderLeftStyle(IHTMLStyle *iface, BSTR *p)
That's not what changelog says.
Thanks for the catch.
I added a test case
V_VT(&arg) = VT_BSTR; V_BSTR(&arg) = a2bstr("invalid"); hres = IHTMLStyle_Invoke(pStyle, dispid, &IID_NULL, LOCALE_SYSTEM_DEFAUT, DISPATCH_PROPERTYPUT, ¶ms, &ret, NULL, NULL); ok_(__FILE__,line) (hres == S_OK, "invalid. ret: %08x\n", hres); VariantClear(&arg);
The problem is hres is DISP_E_EXCEPTION, whereas I'm expecting E_INVALIDARG (Tested on windows). It appears in ITypeInfo_fnInvoke, if an error is returned from the function call, the result is changed to return DISP_E_EXCEPTION (~ line 6320).
Should ITypeInfo_fnInvoke propergate the returned HRESULT instead of changing it?
Best Regards Alistair Leslie-Hughes