Movesense decode SBEM data from LogBook

问题

I'm trying to get the LogBook data over BLE to my App. This works fine for JSON, the data seems accurate. But it takes along time due to the JSON encoding. Getting the SBEM data is way faster. But I can't find any documentation on the encoding. I found out that the "Content" string is Base64 encoded. It starts with SBEM which means, it is uncompressed as stated here:

https://bitbucket.org/suunto/movesense-device-lib/src/5bcf0b40644a17d48977cf011ebcf6191650c6f0/MovesenseCoreLib/resources/movesense-api/mem/logbook.yaml?fileviewer=file-view-default#lines-186

But I couldn't find anything else.

Has somebody further information on that or found out what the encoding is like?

Best regards Alex

回答1:

First some clarification: When requesting the JSON log from MDS/Logbook/ service the data itself is transferred from Movesense sensor in SBEM format and the conversion is performed on the phone. If you have specific examples where the said conversion is slow (there very well might be) it's a good idea to add a bitbucket issue to movesense-mobile-lib.

About the SBEM format. This is "Suunto Oy internal" binary format for presenting xml (and nowadays json) files. This means that the interpretation of it may change when the format evolves. With that warning aside, here's the format:

Data is encoded in chunks with ID(1-2 bytes), length(1-4 bytes) and content
consists of two separate sections: Descriptors & Data which can be in separate "files" (like in Logbook service)
Descriptors describe the format of the data in data chunks ("format string")
Data chunks contain the binary data in described format.

If you want to learn about the SBEM format that the DataLogger / Logbook services use, see the "generated/sbem-code" folder that is created during the build.

And finally, here is a simple python code for parsing SBEM format:

from __future__ import print_function

import sys
import re

import glob, os

data_path = sys.argv[0]
descriptor_path = sys.argv[1]

ReservedSbemId_e_Escape = b"\255"
ReservedSbemId_e_Descriptor = 0

#print("data_path:",data_path)
print("descriptor_path:",descriptor_path)

# reads sbem ID upto uint16 from file
def readId(f):
    byte1 = f.read(1)
    id = None
    if not byte1:
        print("EOF found")
    elif byte1 < ReservedSbemId_e_Escape:
        id = int.from_bytes(byte1, byteorder='little')
        #print("one byte id:", id)
    else:
        # read 2 following bytes
        id_bytes = f.read(2)
        id = int.from_bytes(id_bytes, byteorder='little')         
        #print("two byte id:",id)
    return id

# reads sbem length upto uint32 from file
def readLen(f):
    byte1 = f.read(1)
    if byte1 < ReservedSbemId_e_Escape:
        datasize = int.from_bytes(byte1, byteorder='little')
        #print("one byte len:", len)

    else:
        # read 4 following bytes
        id_bytes = f.read(4)
        datasize = int.from_bytes(id_bytes, byteorder='little')         
        #print("4 byte len:",len)
    return datasize

# read sbem chunkheader from file
def readChunkHeader(f):
    id = readId(f)
    if id is None:
        return (None,None)

    datasize = readLen(f)
    ret = (id, datasize)
    print("SBEM chunk header:", ret)
    print(" offset:", f.tell())
    return ret

def readHeader(f):
    # read header
    header_bytes = f.read(8)
    print("SBEM Header: ", header_bytes)

def parseDescriptorChunk(data_bytes):
    print("parseDescriptorChunk data:", chunk_bytes)
    return

def parseDataChunk(data_bytes):
    print("parseDataChunk data:", chunk_bytes)
    return

# read descriptors
with open(descriptor_path, 'rb') as f_desc:

    readHeader(f_desc)

    while True:
        (id, datasize) = readChunkHeader(f_desc)
        if id is None:
#            print("None id:",id)
            break;
        chunk_bytes = f_desc.read(datasize)
        if (len(chunk_bytes) != datasize):
            print("ERROR: too few bytes returned.")
            break

        if id == ReservedSbemId_e_Descriptor:
            parseDescriptorChunk(chunk_bytes)                      
        else:
            print("WARNING: data chunk in descriptor file!")
            parseDataChunk(chunk_bytes)



# read data
with open(data_path, 'rb') as f_data:

    readHeader(f_data)

    while True:
        (id, datasize) = readChunkHeader(f_data)
        if id is None:
#            print("None id:",id)
            break;
        chunk_bytes = f_data.read(datasize)
        if (len(chunk_bytes) != datasize):
            print("ERROR: too few bytes returned.")
            break

        if id == ReservedSbemId_e_Descriptor:
            parseDescriptorChunk(chunk_bytes)                      
        else:
            parseDataChunk(chunk_bytes)

Full Disclaimer: I work for the Movesense team

来源：https://stackoverflow.com/questions/52992615/movesense-decode-sbem-data-from-logbook

标签

movesense