An algorithm is designed to perform Run-Length Encoding (RLE) on an input string of characters.

The pseudocode contains several missing parts, represented by the labels L1 to L5.

data ← L1
index ← 0
run_length ← L2
WHILE index < LEN(data) - 1
    IF data[index] = data[index + 1] THEN
        run_length ← run_length + 1
    ELSE
        OUTPUT run_length
        OUTPUT data[index]
        run_length ← L4
    ENDIF
    index ← L5
ENDWHILE
OUTPUT run_length
OUTPUT data[index]

What value should be written at point L2 of the algorithm to ensure the sequence matches are counted correctly?