It is just that Mathematica uses definite integrals $\int_{1}^{r}$ and uses K[1]
and K[2]
for integration variables, it makes it little hard to see.
But you can see it is same solution as by hand.
ClearAll["Global`*"]
ode1=(g[r]+G1[r])*(1/r^2+1/r (Derivative[1][g][r]+Derivative[1][G1][r])/(g[r]+G1[r]))==g[r]*(1/r^2+1/r Derivative[1][g][r]/g[r])
ode2=(g[r]+G1[r])*(1/r^2+1/r (Derivative[1][f][r]+Derivative[1][F1][r])/(f[r]+F1[r]))==g[r]*(1/r^2+1/r Derivative[1][f][r]/f[r])
Solve the first ode for $G(r)$ and plug the solution into second ode
solG=DSolve[ode1,G1[r],r][[1,1]]
ode2/.solG
This is Mathematica's solution for the above
DSolve[%, F1[r], r] // Simplify
Let solve it by hand and compare
ode3=F1'[r]==Solve[%%,F1'[r]][[1,1,2]]
ode3=Collect[ode3,{F1'[r],F1[r]}]
ode4=SubtractSides[ode3,ode3[[2,2]]]
So this is linear first ode. The integration factor is
mu=Exp[Integrate[-( (-C[1] f[r]+r^2 g[r] f'[r])/(r f[r] (C[1]+r g[r]))),r]]
Multiplying both sides of ode4 by this $\mu$ and integrating both sides gives (and remember to add new constant $c_2$ gives
sol=F1[r]==1/mu*(Integrate[mu*ode4[[2]],r]+C[2])
You see, it is same as Mathematica' but using indefinite integrals which makes it easier to read.
All code in one block
ClearAll["Global`*"]
ode1=(g[r]+G1[r])*(1/r^2+1/r (Derivative[1][g][r]+Derivative[1][G1][r])/(g[r]+G1[r]))==g[r]*(1/r^2+1/r Derivative[1][g][r]/g[r])
ode2=(g[r]+G1[r])*(1/r^2+1/r (Derivative[1][f][r]+Derivative[1][F1][r])/(f[r]+F1[r]))==g[r]*(1/r^2+1/r Derivative[1][f][r]/f[r])
solG=DSolve[ode1,G1[r],r][[1,1]]
ode2/.solG
DSolve[%,F1[r],r]//Simplify
ode3=F1'[r]==Solve[%%,F1'[r]][[1,1,2]]
ode3=Collect[ode3,{F1'[r],F1[r]}]
ode4=SubtractSides[ode3,ode3[[2,2]]]
mu=Exp[Integrate[-( (-C[1] f[r]+r^2 g[r] f'[r])/(r f[r] (C[1]+r g[r]))),r]]
sol=F1[r]==1/mu*(Integrate[mu*ode4[[2]],r]+C[2])
K[1],K[2]
for integration variables. !Mathematica graphics $\endgroup$