> So I gave it another try... see attached. This is essentially based upon
> the 'row signature' approach mentioned by Jean-Marc in a later post.
The logic seems reasonable to me, but you're far more expert at that than I
am. Some small niggles with the code:
You'll get complaints here on 64bit machines std::size_t is bigger than
unsigned int. Please use lyx::size_type.
+ // Row signature; has row changed since last paint?
+ unsigned row_sig = rit->endpos() - rit->pos() + y;
No need to waste cycles...
InsetList::const_iterator iend = par.insetlist.end();
for ( ; ii != iend; ++ii) {
if (ii->pos >= rit->pos() && ii->pos < rit->endpos()
&& ii->inset->editable() == InsetBase::HIGHLY_EDITABLE)
insetinrow = true;
+ break;
}
This assumes a 1-to-1 correspondence between rowno and the row_sig()
container, a vector. Is that justifiable? Maybe it would be safer/cheaper
to use
std::map<lyx::size_type, lyx::size_type> row_sig_;
...
par.row_sig()[rowno] = row_sig;
? Means you'll also be able to remove the tests on size()...
// Add to row signature cache
if (rowno >= par.row_sig().size())
par.row_sig().push_back(row_sig);
else
par.row_sig()[rowno] = row_sig;
Be const correct. const bool select...
+ bool select = bv.cursor().selection();
Is this legal?
Anyway, you don't need to zero-initialize a std::container.
Paragraph::Paragraph()
- : begin_of_body_(0), pimpl_(new Paragraph::Pimpl(this))
+ : row_sig_(0), begin_of_body_(0), pimpl_(new Paragraph::Pimpl(this))
Yuck! Why are you making a variable public? Also, it appears to be neither
initialised in the constructor above, nor does it appear to be used in
your patch!
public:
///
InsetList insetlist;
+ ///
+ bool breakFlag;
--
Angus
