pythonj1939asammdf

Create MF4 to be decode by J1939 DBC - asammdf

I'm trying to create a MF4 file to be decoded by a 'J1939.dbc' as the same way of CSS electronics

My code to generate a example file is like this:

from asammdf import MDF, SUPPORTED_VERSIONS, Signal
import numpy as np

sigs = []
mdf = MDF()

samples = [
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([217056256,217056256,217056256,217056256,217056256], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.ones(5, dtype=np.dtype('(8,)u1'))*111,
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32)
]   

types = [('CAN_DataFrame.BusChannel', 'u1'),
        ('CAN_DataFrame.ID', '<u4'), 
        ('CAN_DataFrame.IDE', 'u1'), 
        ('CAN_DataFrame.DLC', 'u1'), 
        ('CAN_DataFrame.DataLength', 'u1'),
        ('CAN_DataFrame.DataBytes', 'u1', (8,)), 
        ('CAN_DataFrame.Dir', 'u1'),
        ('CAN_DataFrame.EDL', 'u1'), 
        ('CAN_DataFrame.BRS', 'u1')]

t = np.arange(5, dtype=np.float64)

sig = Signal(
    np.core.records.fromarrays(samples, dtype=np.dtype(types)),
    t+10,
    name='Channel_structure_composition',
    comment='Structure channel composition',
)
sigs.append(sig)

mdf.append(sigs, comment='arrays', common_timebase=True)

mdf.save('demo.mf4', overwrite=True)
print('save')

On the picture:

1) The file generated is this

2) CAN logging is disable - when I use a file from CSS electronics the 'CAN logging' is enable-

enter image description here

But if a try to generate the decode by hand, it returns a blank file

import numpy as np
from asammdf import MDF, SUPPORTED_VERSIONS, Signal


mdf = MDF(r'demo.mf4')
databases = {
     "CAN": ["j1939.dbc"]
    }
extracted = mdf.extract_bus_logging(database_files=databases)
extracted.save('sorted.mf4')

Probably is need to specify the ID and DataByte column to the decode, but I have tried many things but is not working.

What I can do ?

Solution

  • The meta-data for bus logging has to be added as well to the new channel group:

    from asammdf import MDF, SUPPORTED_VERSIONS, Signal
import numpy as np
from asammdf.blocks.source_utils import Source

sigs = []
mdf = MDF()

samples = [
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([217056256,217056256,217056256,217056256,217056256], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.ones(5, dtype=np.dtype('(8,)u1'))*111,
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32),
    np.array([1,1,1,1,1], dtype=np.uint32)
]

types = [('CAN_DataFrame.BusChannel', 'u1'),
        ('CAN_DataFrame.ID', '<u4'),
        ('CAN_DataFrame.IDE', 'u1'),
        ('CAN_DataFrame.DLC', 'u1'),
        ('CAN_DataFrame.DataLength', 'u1'),
        ('CAN_DataFrame.DataBytes', 'u1', (8,)),
        ('CAN_DataFrame.Dir', 'u1'),
        ('CAN_DataFrame.EDL', 'u1'),
        ('CAN_DataFrame.BRS', 'u1')]

t = np.arange(5, dtype=np.float64)

sig = Signal(
    np.core.records.fromarrays(samples, dtype=np.dtype(types)),
    t+10,
    name='Channel_structure_composition',
    comment='Structure channel composition',
    source=Source(
        source_type=Source.SOURCE_BUS,
        bus_type=Source.BUS_TYPE_CAN,
        name="CAN bus",
        path="CAN bus",
        comment="",
    )
)
sigs.append(sig)

mdf.append(sigs, comment='arrays', common_timebase=True)

mdf.save('demo.mf4', overwrite=True)
print('save')