Maxim Kuzmin

@mouse07410 From now it passes all tests and has support of larger constraints.

@mouse07410 #99 is already applied to this PR, so no need to apply it again. Crash looks rather strange, because I've tested parsing of 3GPP asn1 before pulling to repo...

@mouse07410 I've tried full build cycle on different PC five minutes ago (never mind, why =) ) and everything is OK, if not to mention dozens of warnings on asn1...

@mouse07410 Could you get backtrace for your crash?

@mouse07410 Sorry, I've just forgot to scroll up :( As far as I can see, the reason of crash is a freeing of unallocated memory. But I suppose that the...

@mouse07410 OK, I'll write it.

@mouse07410 > the part marker.default_value is not clone-constructed, but instead just the pointer is copied along with the rest of the content of the original object Exactly my thoughts. So,...

It works correctly for #115. Could you make a PR to my repo?

Your master and #99 got same problem, as @ikarso's version - they fail to decode my example of APER. My repo (just a fork) - https://github.com/AuthenticEshkinKot/asn1c.

@mouse07410 Finally, I've found that @ikarso's main_test.c differs from mine and has its own special ANY_to_type_aper. So, I've tested your master with it and everything is OK, it works with...