An easy way to deal with this is to resample at a uniform rate first.

OK - you can assess the amplitude without resampling: find min and max - half the difference is the amplitude.

The frequency is only available after resampling.

And: which phase ? Phase with respect to what ?

Ford, C.; Etter, D.M., "Wavelet basis reconstruction of nonuniformly sampled data," in Circuits and Systems II: Analog and Digital Signal Processing, IEEE Transactions on , vol.45, no.8, pp.1165-1168, Aug 1998

doi: 10.1109/82.718832

keywords: {sampled data systems;signal reconstruction;signal resolution;signal sampling;wavelet transforms;global information;multiresolutional setting;nonuniformly sampled data;signal processing technique;wavelet basis reconstruction;Discrete wavelet transforms;Multiresolution analysis;Nonuniform sampling;Sampling methods;Signal analysis;Signal processing;Signal reconstruction;Signal resolution;Signal sampling;Wavelet analysis},

URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=718832&isnumber=15534

Thanks you all for your valuable answers!

There is no need to re-sample.

The discrete Fourier transform (DFT) may be derived from linear regression, where a complete set of N complex sinusoids are the basis functions to be fitted to N points, using 

b = inv(X'*X)*X'*Y,                    (1)

where Y is a column vector of length N containing the sampled signal; X is an NxN matrix, containing the basis functions along its columns, evaluated at the sample times; and b is your complex spectrum - a column vector of length N .   

For a uniform sampling rate and a uniform error-weighting function, the basis set is orthonormal, so X'*X = I (the identity matrix). This tends to hide this fundamental fact. In this case, we simply get

b = X'*Y.                                     (2)

So for uniform sampling and a uniform error weight, we use (2), perhaps without realizing that we are actually doing linear regression.    

For non uniform sampling (and a uniform weight), you will need to use (1), because the basis set will not be orthogonal.