On Tue, 21 Aug 2007 13:39:36 +0200 Eike Rathke <[EMAIL PROTECTED]> wrote:
> Hi Takashi, > > On Tuesday, 2007-08-21 13:34:18 +0900, Takashi Nakamoto wrote: > > > > > Switching routine that choose one class from the four > > > > classes for creating an instance of a cell would be complex. > > > > > > I don't think so. You need to distinguish anyway, and subclassing may > > > actually ease things because switching can be done internally using > > > factory patterns and RTTI. > > > > I've considered this point. However, I couldn't find any factory patterns > > in ongoing source codes related to cells. Instead, I found many > > statements like "new ScStringCell ()". > > > This means we need to revise > > many lines related to the instantiation of Sc*Cell if we add > > ScRubyStringCell. Such revise results in complex code. > > No, it results in complex code analysis. You will have to find the > places anyway where you want to add ruby to string cells. Once changed > to a factory pattern the resulting code will be not more complex than > changing calls of > > ScStringCell* pCell = new ScStringCell(someText); > > to calls of > > ScStringCell* pCell = ScStringCell::createInstance(someText,maybeRuby); > > that returns a new'd ScRubyStringCell if maybeRuby is not empty, and > ScStringCell otherwise. A ScRubyStringCell should still have the > eCellType member variable set to CELLTYPE_STRING so you wouldn't need to > adapt the zillion places where that is used, but should be > distinguishable by means of RTTI where needed. > > > > If Sc*Cell class group was designed to be added a new cell type to > > from the start, subclassing would be the most attractive and cleanest > > solution for this problem. > > All cell types are derived from ScBaseCell so that shouldn't be a big > problem. We have CellType that so far was sufficient and efficient. > A real RTTI interface isn't available yet because most compilers didn't > know about RTTI back those days, which actually is the reason why the > ugly tools/rtti.hxx macros came to existence. We should strive for the > cleanest possible solution anyway, and if something is missing add the > missing pieces instead of fearing complex changes. I just expected a factory pattern and RTTI are already used. I agree about that shifting to a factory pattern and using RTTI interface is the cleanest solution. -- Japanese Native-Language project Good-Day Inc. Takashi Nakamoto --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
