There are several cases to consider.
A. If $q<0$, then as $\log(x^{-1})\sim(x-1)$ as $x\rightarrow1$ the integral (denoted as $J_{p,q}$ from now on) would diverge if $q\leq -1$ since
$$
J_{p,q}:=\int^1_0 x^p\log^q(x^{-1})\,dx \geq C_p\int^1_{1/2}\log^q(1/x)\,dx \geq C'_p\int^1_0(1-x)^q\,dx
$$
This shows that $J_{p,q}$ diverges for $q\leq -1$ (regardless of $p$).
B. From (A) it follows that if $p>0$ and $-1<q<0$, $J_{p,q}$ converges since $J_{p,q}=\int^{1/2}_0+\int^1_{1/2} x^p\log^q(x^{-1})\,dx$. The second integral is bounded by $A_p\int^1_{1/2}(1-x)^p\,dx$ for some constant $A_p$. The first integral has not problem as the integrand can be defined as a continuous function by setting its value at $x=0$ as $0$.
C. If $p,q>0$, then $J_{p,q}$ converges since $\lim_{x\rightarrow0}x^\alpha\log(x)=0$
for all $\alpha>0$, and so $g(x)= x^{p/q}\log(1/x))^q$ for $0<x\leq1$, $g(0)=0$ ddefines a nice continuous function.
The case $p<0$ and $-1<q<0$ may be analyzed along similar lines.
$$J_{p,q}=\int^{1/2}_0 +\int^1_{1/2} x^p\log^q(x^{-1})\,dx = I^1_{p,q} + I^2_{p,q}$$
The second integral converges since $I^2_{p,q}\leq B_p\int^1_{1/2}(1-x)^q\,dx$. Thus, it all boils down to see what happens to $I^1_{p,q}$.
D. If $-1<p<0$ (and $-1<q<0$) then $J_{p,q}$ converges since $I^1_{p,q}\leq \log^q(1/2)\int^{1/2}_1 x^p\,dx<\infty$.
E. If $p<-1$ (and $-1<q<0$) then $J_{p,q}$ diverges, for $\lim_{x\rightarrow0}x^\alpha\log x=0$ for all $\alpha>0$ and so,
$$
I^1_{p,q}=\int^{1/2}_0 x^{-1}\big(x^{(p+1)/q}\log(x^{-1})\big)^q\,dx\geq D\int^{1/2}_1x^{-1}\,dx =\infty$$
F. For the case $p=1$ (and $-1<q<0$) $J_{p,q}$ diverges. One can use substitution to directly evaluate the integral.
$$\int^{1/2}_0x^{-1}(-\log(x))^p\,dx=\int^\infty_{\log2} u^q\,du=\infty$$
You can complete the remaining case ($p=0$, $-1<q\leq 0$)