# Parameterized derived types **URL:** \ **Category:** Help\ **Created:** [April 18, 2022, 7:48pm UTC](https://fortran-lang.discourse.group/t/parameterized-derived-types/3263 "2022-04-18T19:48:46Z")\ **Posts on this page:** 5\ **Page:** 1
**Author:** ![james3](https://avatars.discourse-cdn.com/v4/letter/j/b5e925/32.png) [@james3](https://fortran-lang.discourse.group/u/james3)\ **Post date:** [April 18, 2022, 7:48pm UTC](https://fortran-lang.discourse.group/t/parameterized-derived-types/3263/1 "2022-04-18T19:48:46Z")
I was attempting something like the following, where I am initializing a parameterized derived type ```auto module test implicit none type x_t(n) integer, len :: n integer :: v(n) character(len = 8), dimension(n) :: str end type x_t type y_t type(x_t(1)) :: a type(x_t(2)) :: b end type y_t contains ! ... end module test program main use test implicit none type(x_t(1)), parameter :: a = x_t(1)([1], [character(len = 8) :: "a1"]) type(x_t(2)), parameter :: b = x_t(2)([1, 2], [character(len = 8) :: "b1", "b2"]) type(y_t), parameter :: y = y_t(a, b) print *, a print *, b print *, y end program main ``` Should this work? In `ifort` I get an internal compiler error ```auto Internal error store_pdtlen_data_space: Cannot find record problem.f90: catastrophic error: **Internal compiler error: internal abort** Please report this error along with the circumstances in which it occurred in a Software Problem Report. Note: File and line given may not be explicit cause of this error. ``` In `gfortran` I get errors beginning with ```auto 25 | type(x_t(1)), parameter :: a = x_t(1)([1], [character(len = 8) :: "a1"]) | 1 Error: No initializer for component ‘v’ given in the structure constructor at (1) ``` Is there a problem with my syntax, or are these compiler issues? ---
**Author:** ![ivanpribec](https://yyz2.discourse-cdn.com/free1/user_avatar/fortran-lang.discourse.group/ivanpribec/32/3290_2.png) [@ivanpribec](https://fortran-lang.discourse.group/u/ivanpribec)\ **Post date:** [April 18, 2022, 8:55pm UTC](https://fortran-lang.discourse.group/t/parameterized-derived-types/3263/2 "2022-04-18T20:55:39Z")
The following runs with `ifort` ```fortran module test implicit none public type x_t(n) integer, len :: n integer :: v(n) character(len=8) :: str(n) end type x_t type y_t type(x_t(n=1)) :: a type(x_t(n=2)) :: b end type y_t end module test program main use test implicit none type(x_t(n=1)) :: a type(x_t(n=2)) :: b type(y_t) :: y a = x_t(n=1)([1], [character(len=8) :: "a1"]) b = x_t(n=2)([1, 2], [character(len=8) :: "b1", "b2"]) y = y_t(a, b) print *, a print *, b print *, y end program main ``` Here’s the output: ```nohighlight ivan@maxwell:~/fortran/20220418_discourse_3263_pdt$ ifort -warn all main.f90 ivan@maxwell:~/fortran/20220418_discourse_3263_pdt$ ./a.out 1 a1 1 2 b1 a1 1 a1 1 2 b1 a1 ``` I have no clue if it should also work in the parameter case. I was surprised to see the structure constructor of the form `x_t(n=1)( ... )` work to begin with. One workaround if you really need the `parameter` attribute could be to abandon the PDT, and simply use parameter arrays of `integer` and `character` string type. Interestingly, the new Intel Fortran compiler `ifx` recognizes it as a not yet supported language feature: ```nohighlight $ ifx -warn all main.f90 /opt/intel/oneapi/compiler/2022.0.2/linux/bin-llvm/xfortcom[0xde27a9] main.f90(28): error #5533: Feature found on this line is not yet supported in ifx a = x_t(n=1)([1], [character(len=8) :: "a1"]) --------^ compilation aborted for main.f90 (code 3) ``` Hopefully, full support will be added soon. Concerning `gfortran`, support of parameterized derived types is it’s Achilles’ heel currently. Hopefully it will be improved in the future. ---
**Author:** ![nncarlson](https://yyz2.discourse-cdn.com/free1/user_avatar/fortran-lang.discourse.group/nncarlson/32/1175_2.png) [@nncarlson](https://fortran-lang.discourse.group/u/nncarlson)\ **Post date:** [April 18, 2022, 9:09pm UTC](https://fortran-lang.discourse.group/t/parameterized-derived-types/3263/3 "2022-04-18T21:09:38Z")
> [@james3](#): > > In `gfortran` I get errors beginning with > > ```auto > 25 | type(x_t(1)), parameter :: a = x_t(1)([1], [character(len = 8) :: "a1"]) > | 1 > Error: No initializer for component ‘v’ given in the structure constructor at (1) > > ``` > > Is there a problem with my syntax, or are these compiler issues? See [84120 – Syntax for used for PDT constructors is incorrect](https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84120) ---
**Author:** ![james3](https://avatars.discourse-cdn.com/v4/letter/j/b5e925/32.png) [@james3](https://fortran-lang.discourse.group/u/james3)\ **Post date:** [April 19, 2022, 12:29am UTC](https://fortran-lang.discourse.group/t/parameterized-derived-types/3263/4 "2022-04-19T00:29:57Z")
Thanks. I was able to test my example using `nagfor` and that works properly, but I think it will be wise for me to avoid using parameterized derived types for now. ---
**Author:** ![Harper](https://avatars.discourse-cdn.com/v4/letter/h/b5ac83/32.png) [@Harper](https://fortran-lang.discourse.group/u/Harper)\ **Post date:** [April 19, 2022, 5:54am UTC](https://fortran-lang.discourse.group/t/parameterized-derived-types/3263/5 "2022-04-19T05:54:47Z")
Internal compiler errors should always be reported to the compiler vendor. If your program was valid it should have been compiled. If not, the compiler should have told you about its bugs. (I have reported various internal compiler errors over the years to various vendors. They appreciate it if you can send them a very short program that produces the error. )