What is the rationale for the encoding procedure described in the DVB-S2 standard for LDPC codes?

Question

I have not found an explanation on why accumulating the given parity check addresses and then xoring the accumulated values gives the codeword of the LDPC code. The close it