#3002 save() doesn't handle structs correctly

[Gunter Königsmann]

If a struct is defined and saved into a file:

defstruct(struct(a,b,c));
s:
[
    new(struct([1,2,3],2,3)),
    new(struct([4,5,5],5,6))
];
save("Test.lisp",s);
s[1]@a;

it can be loaded again within the same maxima session:

kill(s);
load("Test.lisp");
s[1]@a;

If maxima is restarted this doesn't work anymore, though:

load("Test.lisp");
s[1]@a;

in this case returns:

struct([1,2,3],2,3)@a

It seems like the struct isn't defined as a struct via defstruct() and new() but rather treated as a function with the struct elements as argument.

[Gunter Königsmann]

Found out why this did happen for me and documented it with commit #d0b2a1eacb1c2a47ed9bdef86c58bf9bdfc99851.

[Robert Dodier]

I'm reopening this bug report -- I just bumped into this problem. This is a bug in save -- it should save the structure definition, as it also saves arrays, functions, properties, etc.

When this bug is fixed, we can remove the text introduced in commit d0b2a1ea.

[Gunter Königsmann]

Thanks a lot! Don't know if there is a way to make maxima forget a struct's definition: That would feel like risking to delete the definition of an already-existing struct. If there is one I would be willing to write a few test cases.

[Robert Dodier]

Fixed by commit a5d1d3e. Closing this ticket.