hdr_encode.py

from ctypes import *
import base64
import gzip, zlib

class HDR_COMPRESSED(BigEndianStructure):
    _pack_ = 1
    _fields_ = [
        ("cookie", c_int),
        ("length", c_int)
    ]

class HDR_ENCODED(BigEndianStructure):
    _pack_ = 1
    _fields_ = [
        ("cookie", c_int),
        ("significant_figures", c_int),
        ("lowest_trackable_value", c_longlong),
        ("highest_trackable_value", c_longlong),
        ("total_count", c_longlong)
    ]

class COUNT(BigEndianStructure):
    _fields_ = [("value", c_longlong)]

def main():

    # Base64 decode the string and store the compressed binary value in a bytestring
    compressed = bytearray(base64.decodestring('HISTiQAAAGd42pNpmdzBwMDAxAAGfgoMDMxuBjsWQLgMTG+hDDYGVMCExmdGo1nQ1LGg0VxQmhNKi0NpISgtDKUtoLQflPaG0v1Quh1Kp6GZww+lWXG4n1jAyDAKRsNtFIzG82g8j4IRCABtRQfZ'))
    # Create a ctypes struct over the top of the bytearray
    hdr_compressed = HDR_COMPRESSED.from_buffer(compressed)

    # Get the head fields which is followed by the compressed data
    print("compressed.cookie: {:#x}".format(hdr_compressed.cookie))
    print("compressed.length: {:d}".format(hdr_compressed.length))

    # Decompress the remaining data.
    encoded = bytearray(zlib.decompress(bytes(compressed[8:hdr_compressed.length + 8])))

    # Create a ctypes flyweight over the uncompressed data
    hdr_encode = HDR_ENCODED.from_buffer(encoded)

    # Read out the header fields.
    # Note this is the current jhiccup format which uses the older V0 format.
    # To be correct the cookie should be checked.
    print("encoded.cookie: {:#x}".format(hdr_encode.cookie))
    print("encoded.significant_figures: {:d}".format(hdr_encode.significant_figures))
    print("encoded.lowest_trackable_value: {:#d}".format(hdr_encode.lowest_trackable_value))
    print("encoded.highest_trackable_value: {:#d}".format(hdr_encode.highest_trackable_value))
    print("encoded.total_count: {:d}".format(hdr_encode.total_count))

    uncompressed_header_len = 32;
    word_size = 8

    # This is a hack, the histogram should be created from the above fields and figure out
    # the required data length.
    data_len = (len(encoded) - uncompressed_header_len) / word_size

    # Create an array over the remaining data
    CountsArrayType = COUNT * data_len;
    counts_array = CountsArrayType.from_buffer(encoded[uncompressed_header_len:len(encoded) + uncompressed_header_len])
    for i in counts_array:
        print i.value

if __name__ == '__main__':
    main()