Closed hcsch closed 1 year ago
Sorry for the spam, I just noticed that I was still missing something after setting up the changed code to test with the DBC that I have. I'll mark this PR ready for review again once I've actually tested the code :)
Hi, sorry for the late response! This looks good, but I don't have a test file to check it. I'll merge it anyway :)
If you're still interested, could you add a simple test as a new PR to make sure this continues to work?
I could add an example to testing/dbc-examples/example.dbc
, if that's fine by you, though I don't currently have a use for this anymore as I ended up wanting to parse databases at runtime without recompilation.
VECTOR__INDEPENDENT_SIG_MSG
messages kinda look like this, based on what I've seen:
BO_ 3221225472 VECTOR__INDEPENDENT_SIG_MSG: 0 Vector__XXX
SG_ Unassigned0 : 0|32@1- (1.2345,0) [0|20000] "" Vector__XXX
SG_ Unassigned3 : 0|16@1- (1,23450) [0|0] "" Vector__XXX
SG_ Unassigned1 : 0|37@1- (-1,0) [-3000|0] "" Vector__XXX
SG_ RandomSignal : 0|5@1- (1,0) [0|0] "" Vector__XXX
Basically just signals plucked out of other messages for later use in a real message again, but unused at the moment.
VECTOR__INDEPENDENT_SIG_MSG
is zero-size with non-zero-size signals, causing an error during codegen.As the signals in this message are unassigned to any real message, just ignoring it is probably the way to go.
See also https://github.com/cantools/cantools/issues/62