>>>>> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
Juergen> "Changing the citation key does not update references to it" Juergen> http://bugzilla.lyx.org/show_bug.cgi?id=2744 Juergen> The attached patch against 1.4svn fixes the bug by extending Juergen> changeRefsIfUnique. It works very nicely. OK for trunk and Juergen> branch? (I have another fix for bug 1684 in the pipe). Some comments: -void Buffer::changeRefsIfUnique(string const & from, string const & to) +void Buffer::changeRefsIfUnique(string const & from, string const & to, InsetBase::Code code) I do not like passing an inset code if it is not really generic (the codes are hardcoded later). If we follow this route, we should at least assert on code's value. + if (code == InsetBase::CITE_CODE) { + vector<pair<string, string> > keys; + fillWithBibKeys(keys); + vector<pair<string, string> >::const_iterator bit = keys.begin(); + vector<pair<string, string> >::const_iterator bend = keys.end(); + + for (; bit != bend; ++bit) + labels.push_back(bit->first); I am sure there is a nice way of doing the counting without building additional lists, but we would need Lars' help for that :) + } else + getLabelList(labels); + if (lyx::count(labels.begin(), labels.end(), from) > 1) return; If we were able to make the code below truly generic, the code above could be done by two separate helper functions. ParIterator it = par_iterator_begin(); ParIterator end = par_iterator_end(); Why not use an InsetIterator there? I know it was in the existing code, but it may be a good time to improve it. + if (contents == from) { inset->setContents(to); //inset->setButtonLabel(); changed_inset = true; + } else if (code == InsetBase::CITE_CODE + && tokenPos(contents, ',', from) != -1) { + vector<string> items = getVectorFromString(contents); + replace(items.begin(), items.end(), from, to); + inset->setContents(getStringFromVector(items)); I'd rather have a InsetCommand::replaceContents(from, to) that is specialized in InsetCite. This way the code in the present method would be truly generic. Finally, there is a changed_inset variable that does not seem to be used (not your doing). JMarc
