Add emission due to CX with neutral hydrogen to the TotalRadiatedPower model

[vsnever]

Cherab atomic data repository support three types of total radiated power coefficients:

• line power due to electron impact excitation, LineRadiationPower,
• continuum and line power due to recombination and Bremsstrahlung, ContinuumPower,
• line power due to charge exchange (CX) with thermal neutral hydrogen and its isotopes, CXRadiationPower.

However, only the first two coefficients are used in the TotalRadiatedPower model. Adding a third source of power loss does not seem difficult. Taking into account CX, the emissions calculated by this model will be as follows:

$$\epsilon_{\mathrm{total}} = \frac{1}{4 \pi \Delta\lambda} \left( n_{Z_\mathrm{i}} n_\mathrm{e} C_{\mathrm{excit}}(n_\mathrm{e}, T_\mathrm{e}) + n_{Z_\mathrm{i} + 1} n_\mathrm{e} C_{\mathrm{recomb}}(n_\mathrm{e}, T_\mathrm{e}) + n_{Z_\mathrm{i} + 1} n_\mathrm{hyd} C_{\mathrm{cx}}(n_\mathrm{e}, T_\mathrm{e}) \right)$$

where n_hyd is the total density of all hydrogen isotopes in the plasma.

This PR adds CX radiated power for hydrogen donors to the TotalRadiatedPower model, thus fixing #370.

Also, the evaluate() method of the _RadiatedPower core class was defined as cdef instead of cpdef. A changed it to cpdef just like in other atomic rate classes, because with cdef definition I could not create MockAtomicData class for the unit test.

[Mateasek]

Hi @vsnever , thanks for adding the CX radiated power. There are two points I would like you to look into and consider changes. Otherwise this is good to go.

[Mateasek]

Thanks fof the improvement @vsnever , this can be merged

[jacklovell]

Looks good: the documentation and test coverage is particularly welcome! Just a couple of small comments before merging.