OPM中提供属性分类
常规的使用如下方法
BEGIN_OPMPROP_MAP()
OPMPROP_ENTRY(0, DispID, PROPCAT_Data,0, 0, 0, “”, 0, 1, IID_NULL, IID_NULL, “”)
END_OPMPROP_MAP()其中DispIDP是R属性的 显示ID；OPCAT_为前缀的就是已经预定义的属性分类，其数值范围是-1~-25。
如果需要自定义分类，则需要在PROPCAT_的位置上设为自定义的数值，然后 在头文件中改写
STDMETHODIMP GetCategoryName(
/* [in] */ PROPCAT propcat,
/* [in] */ LCID lcid,
/* [out] */ BSTR* pbstrName)
{if (propcat == 1){*pbstrName = ::SysAllocString(L”SQSize”);return S_OK;}elsereturn S_FALSE;
}注意看原来例程的写法，按照那个写法程序不会调用这个函数。我也不知道为什么。
执行完这个函数后，GetElementStrings函数被调用。
这个函数的作用就是分解复杂的属性，如当你所描述的属性是一个点时，你需要将x,y,z的坐标同时显示在属性窗
口中。 这个函数就是你实现如何分解这些属性。当然也可以不实现它。
AsdkSquareWrapper例程里没写清楚地好像就是这些了。我的表达能力不好，希望大家能看明白。
OPM中属性的下拉列表的支持
扩展属性中实现 下拉列表
接口需要继承
public IOPMPropertyExtensionImpl实现其中的
//OPM calls this function for each property to obtain a list of strings and cookies if they are available.
//For our textstyle property we would like to display all the textstyles currently available in the database.
//This function is declared on the IPerPropertyBrowsing interface. Our IOPMPropertyExtensionImpl
//class implements this member by reading the values in the OPM property map. (You set this up in your
//head file when you use BEGIN_OPMPROP_MAP, OPMPROP_ENTRY, END_OPMPROP_MAP macros.)
//Since we need a dynamic list of entries in this drop down list and a static map cannot implement this,
//we need to override this function a provide dynamic list of text styles to OPM.STDMETHOD(GetPredefinedStrings)(/* [in] */ DISPID dispID,/* [out] */ CALPOLESTR *pCaStringsOut,/* [out] */ CADWORD *pCaCookiesOut);//OPM calls this function when the user selects an element from a drop down list. OPM provides
//the cookie that we associated with the element in the GetPredefinedStrings function. We are
//responsible for mapping this cookie back to a value that the properties corresponding put_ function
//can understand.
//In this particular case all we need to do is to provide the name of the text style as
//the put_TextStyle method needs that.STDMETHOD(GetPredefinedValue)(/* [in] */ DISPID dispID, /* [out] */ DWORD dwCookie, /* [out] */ VARIANT *pVarOut);具体实现的思路就是根据不同的 dispID 设置不同的列表
具体实现方法可以参考polysamp
注意：上述函数返回值是和get/put系列函数的形参严格对应，自动过滤掉不符合的项。
How to create a polyline-like vertex edit with a spin control in OPM?
http://adndevblog.typepad.com/autocad/2012/12/how-to-create-a-polyline-like-vertex-edit-with-a-spin-control-in-opm.html
By Gopinath Taget
The following snippets of code can be used to implement a polyline-like vertex edit in OPM (Object Property Manager) using IOPMPropertyExpander interface for a custom entity.
Lets assume the custom object (called AsDkRings) has two new variables to reflect an array of vertices:
AcGePoint3d m_polyline[5]; int m_numbervertices;For simplicity, lets say the custom entity has a maximum of five vertices.
The custom class also has two corresponding access functions. (Note the vertex number parameter):
Acad::ErrorStatus AsDkRings::polyline(AcGePoint3d& vertex, int vertexNumber)Acad::ErrorStatus AsDkRings::setPolyline(AcGePoint3d vertex,int vertexNumber)In AsDkRings:: subWoldDraw(), we draw a polyline connecting the five vertices.
The key to getting a spin control in OPM is to return a grouping number. This number will determine the number of elements to group together.
//IOPMPropertyExpanderSTDMETHODIMP CRings::GetElementGrouping(/* [in] */ DISPID dispID,/* [out] */ short *groupingNumber){..........................} else if (dispID == 5){*groupingNumber = 4;return S_OK;}return E_NOTIMPL;}A grouping number of 4 means there is one entry (called "Vertex") plus 3 entries (Vertex X, Y and Z) to group together into one property. What this would do is put a spin control into the first item (vertex in this case), so that this could be used to traverse an array of 3 remaining grouped items (which is Vertex X, Y and Z).
Next, we should specify the number of items that the property is going to display. Here this would be the number of vertices = 5. This would mean that the spin contol can go up to a maximum of 5 steps, for five vertices.
//IOPMPropertyExpanderSTDMETHODIMP CRings::GetGroupCount(/* [in] */ DISPID dispID,/* [out] */ long *nGroupCnt){...............................} else if (dispID == 5){*nGroupCnt = m_numbervertices; // Number of verticesreturn S_OK;}return E_NOTIMPL;}Next, we set the string to display in the OPM.
//IOPMPropertyExpanderSTDMETHODIMP CRings::GetElementStrings(/* [in] */ DISPID dispID,/* [out] */ OPMLPOLESTR __RPC_FAR *pCaStringsOut,/* [out] */ OPMDWORD __RPC_FAR *pCaCookiesOut){................................................................} else if (dispID == 5){// For the VerticespCaStringsOut->cElems = 4;pCaStringsOut->pElems = (LPOLESTR*)CoTaskMemAlloc(sizeof(LPOLESTR)*4);pCaStringsOut->pElems[0] = SysAllocString(L"Vertex");pCaStringsOut->pElems[1] = SysAllocString(L"Vertex X");pCaStringsOut->pElems[2] = SysAllocString(L"Vertex Y");pCaStringsOut->pElems[3] = SysAllocString(L"Vertex Z");pCaCookiesOut->cElems = 4;pCaCookiesOut->pElems = (DWORD*)CoTaskMemAlloc(sizeof(DWORD)*4);pCaCookiesOut->pElems[0] = 10;pCaCookiesOut->pElems[1] = 1;pCaCookiesOut->pElems[2] = 2;pCaCookiesOut->pElems[3] = 3;return S_OK;}return E_NOTIMPL;}The cookies count is the unique identifier for each property item that will be used to get and set values. The table below will give you an idea of cookie value for each vertex value. Please remember that vertex entry has a spin control that goes like 1, 2, 3, 4.........and so on.
| Property String | Cookie Value | |||||
|---|---|---|---|---|---|---|
| Vertex = 1 | Vertex = 2 | Vertex = 3 | Vertex = 4 | ....and so on | ||
| Vertex | 0 | 4 | 8 | 12 | ....and so on | |
| Vertex X | 1 | 5 | 9 | 13 | ....and so on | |
| Vertex Y | 2 | 6 | 10 | 14 | ....and so on | |
| Vertex Z | 3 | 7 | 11 | 15 | ....and so on |
So just by using the cookie values, you must get the vertex number and find out if it is an x, y, or z coordinate. The following code does that.
// Get the coordinate index (x=0, y= 1, z=2) from dwCookie
for (int i = 1; i < 4; i++) {vertex = (double(dwCookie) - i) / 4;if( vertex == (double(dwCookie) - i) / 4) {index = i -1;break;}}index will return 0 for vertex X, 1 for vertex Y and 2 for vertex Z for the corrosponding cookie values.
So this is all that is needed to implement a poly-line vertex edit.
