Triangulation: filter_iterator #8899
Conversation
| Finite_full_cell_const_iterator finite_full_cells_begin() const | ||
| { return Finite_full_cell_const_iterator(Finiteness_predicate(*this), full_cells_begin(), full_cells_end()); } | ||
| Finite_full_cell_const_iterator finite_full_cells_end() const | ||
| { return Finite_full_cell_const_iterator(Finiteness_predicate(*this), full_cells_end(), full_cells_end()); } | ||
| #else | ||
| /* |
There was a problem hiding this comment.
@sloriot When I uncomment the non-const version I get a compilation error. Can you explain why?
There was a problem hiding this comment.
Fixed. I don't get why I need to define extra operators. I would have expect that there would be an implicit conversion then.
|
See boostorg/iterator#92 for the source of the issue. |
|
@mglisse I only changed the filter for cells and not for vertices, which is enough for the testsuite. What is different for vertices? Is it maybe only not tested? |
|
It is the copy constructor (or conversion) of filter_iterator that is broken, in a way that mostly matters if you do ++ on a copy and the last element of the sequence does not satisfy the predicate but some earlier elements do (maybe also if you compare a copy to a non-copy, I didn't check). That's quite specific, so it is not surprising that tests don't always detect it. |
Summary of Changes
The testsuite of Triangulation has N's for some platforms. The test platforms have in common to use boost_1_88_0, but with VC++ or clang on Windows I can't reproduce the error with this version of boost.
In this PR I use the filter_iterator adaptor of CGAL, to see if the boost filter_iterator is the problem.
Release Management