Mathlib-cernlib functions (clebsch,sixj) in root

I was looking for some clebsch, 3j, 6j coeficient in root. After looking around,
I tried to employ the functions from example,
Number 3 - compile and link with gcc finaly worked (no +a1 +z switches and no Xpm X11 dld libraries) and I was able to run rootCERN and make
float x[5]={1,1,1,1,1}
denlan_(x) with the result (Float_t)1.45206638158731027e-01

Now I tried to add to rootCERN.cxx lines
int i[6]={1,2,3,4,5,6};
to dummy block; then to CERNfunctions.h
extern “C” Float_t rwig3j_(Int_t *x);
then to CERNLinkDef.h
#pragma link C function rwig3j_;

to be as much as possible similar to denlan_ declaration.

What I am able to get is
Error: Symbol rwig3j_ is not defined in current scope
rwig3j_ is not there, denlan is.
Could you help me please? The same for dclebg_, dwig3j_ and even dislan_
Does anybody uses clebsch, sixj,threej, ninej under root??

clebsch: … 1/top.html
denlan: … 0/top.html

thank you,

SORRY, now I found some error I did - dont answer please
for next hour

the post on the topic? I have found the solution via the 3rd way from the reference above, with creating shared (and thus loadable) library finaly.
It would worth to have all the mathlib compiled somewhere, wouldnt it?