[seqfan] transforms_pari.txt opv_mul_egf()

[Kevin Ryde]

In pari/gp code http://oeis.org/transforms_pari.txt there seems a small
gremlin in opv_mul_egf(),

    read("transforms_pari.txt");
    opv_mul_egf([1,2,3], [4,5,6])
    =>
    ***   in function opv_mul_egf: ...,binomial(i,j)*A[j+1])*B[i-j+1]);C
    ***                                                        ^---------
    ***   incorrect type in gtos [integer expected] (t_POL).

Would that B[i-j+1] go inside sum(j...)?  Doing so gets the right
results in code of A136629 (where maybe min would be max).

Also minor bit for transforms_pari.txt, I notice trv_chain_j() has B2 as
a global.  Would that be local, it's not a special second output is it?