issue in ‘data_generation/pretrain_data_gen.py’

[fankaisheng]

In line 161:

packet_direction = [] 
    feature_result = extract(label_pcap)
    for key in feature_result.keys():
        value = feature_result[key]
        packet_direction = [x // abs(x) for x in value.ip_lengths]

    if len(packet_direction) == len(packets):
        .....

It seems packet_direction is supposed to be a list that stores the direction sequence for every session in the input pcap.

However, inside the for loop over sessions the anonymous function packet_direction = [x // abs(x) for x in value.ip_lengths] overwrites packet_direction each time, so only the last session’s direction sequence is kept.

This makes the data processing impossible to work correctly if len(packet_direction) == len(packets) — can this code really reproduce the results correctly?