Meh. The author confuses DFT with STFT. They refer to the STFT as "DFT", but then introduce the usual IDFT as the inverse transform, which is not the inverse of the STFT (the inverse STFT can be computed with the overlap-add method).
https://home.engineering.iastate.edu/~julied/classes/ee524/L...
I'll give you a hint: Cmd+F "window". See any results? Because the DFT doesn't have a window function. The STFT does have a window function.
Second hint: look at the image in the article (https://substackcdn.com/image/fetch/w_1456,c_limit,f_webp,q_...). Do you think this shows the time domain and frequency domain representations (i.e. the DFT) of a signal? No, it does not. It shows the time domain representation, and the STFT of the signal, aka the spectrogram.
Do you have any questions?