The answer to my 2nd question I was able to figure out. Yes, pos() is relevant for InsetMath-derived objects too. See listing for nextAtom below.... the listing for nextInset is provided just for context.
According to me pit() is not relevant for InsetMath-derived objects ============= Inset <http://wiki.lyx.org/sourcedoc/svn/classlyx_1_1Inset.html> * lyx::DocIterator::nextInset ( ) const 00107 { 00108 LASSERT(!empty(), return 0); 00109 if (pos() == lastpos()) 00110 return 0; 00111 if (pos() > lastpos()) { 00112 LYXERR0("Should not happen, but it does: pos() = " 00113 << pos() << ", lastpos() = " << lastpos()); 00114 return 0; 00115 } 00116 if (inMathed()) 00117 return *nextAtom()*.nucleus(); 00118 return paragraph().getInset(pos()); ============== MathAtom <http://wiki.lyx.org/sourcedoc/svn/classlyx_1_1MathAtom.html> & lyx::DocIterator::nextAtom ( ) const 00161 { 00162 LASSERT(!empty(), ); 00163 //lyxerr << "lastpos: " << lastpos() << " next atom:\n" << *this << endl; 00164 LASSERT(pos() < lastpos(), ); 00165 *return cell()[pos()]*;>>>*This tells you the relevance of pos() for an InsetMath* 00166 } =============== On Fri, Jun 18, 2010 at 4:07 AM, Richard Heck <rgh...@comcast.net> wrote: > On 06/17/2010 06:13 PM, Sidharth Kshatriya wrote: > >> >> Q. What is the concept of a StableDocIterator and what is the use of >> zeroing out all the inset_ pointers in each CursorSlice? How does a >> stableDocIterator do its job in the UndoElement? >> >> The inset_ pointers have to be zeroed out here, because the insets to > which they point are about to be deleted. So the pointers would be invalid. > > Richard > > > > -- Sidharth Kshatriya www.sidk.info
