Refactor encoding, part 2 (diag coded types) #292
Conversation
Signed-off-by: Andreas Lauser <andreas.lauser@mbition.io> Signed-off-by: Gerrit Ecke <gerrit.ecke@mbition.io>
andlaus
force-pushed
the
refactor_encoding2
branch
from
8dae7bd
to
4d09733
Compare
| cursor_bit_position=0, | ||
| origin_byte_position=0) | ||
| encode_state.cursor_bit_position = encode_state.cursor_bit_position | ||
| self.diag_coded_type.encode_into_pdu(internal_trouble_code, encode_state=tmp_state) |
There was a problem hiding this comment.
make a copy/clone function in the encode state, instead of all this manual attributes copying
There was a problem hiding this comment.
note that this is not a copy of the original state: the cursor and origin byte positions are different and coded_message is empty. Under the new regime this is required to encode only the diag_coded_type, but it is one of the temporary kludges mentioned in the PR description (i.e., it will be gone after the next two PRs)
| byte_val = dct.convert_internal_to_bytes("Hi", state, bit_position=0) | ||
| self.assertEqual(byte_val, bytes([0x48, 0x69, 0x0])) | ||
| state = EncodeState( | ||
| coded_message=bytearray([0x00]), parameter_values={}, is_end_of_pdu=False) |
There was a problem hiding this comment.
why was the 0x12 replaced with zero?
There was a problem hiding this comment.
that was done to avoid those pesky "overlapping parameter" warnings: convert_internal_to_bytes() simply did not care what already was in the PDU, encode_into_pdu() does...
| state = EncodeState( | ||
| coded_message=bytearray([0xcc]), parameter_values={}, cursor_byte_position=2) | ||
| dct.encode_into_pdu(0x12345, state) | ||
| self.assertEqual(state.coded_message.hex(), "cc00012345") |
There was a problem hiding this comment.
why did the assertion changed here?
There was a problem hiding this comment.
it's been a while since I did this, so take this with a grain of salt: I suppose it is because this sets the cursor byte position to 2, and the initial value of the PDU to 0xcc. I don't remember why the old code produced two leading zero bytes...
| dct = LeadingLengthInfoType( | ||
| base_data_type=DataType.A_UNICODE2STRING, | ||
| bit_length=8, | ||
| base_type_encoding=None, | ||
| is_highlow_byte_order_raw=False, | ||
| ) | ||
| state = DecodeState(bytes([0x12, 0x4, 0x61, 0x00, 0x39, 0x00]), cursor_byte_position=1) | ||
| state = DecodeState( | ||
| bytes([0x12, 0x4, 0x61, 0x00, 0x39, 0x00, 0xff, 0x00]), cursor_byte_position=1) |
There was a problem hiding this comment.
this tests for little-endian UTF-16 encoding, so the bytes representing the two letters are byte swapped compared to what's on line 106. (the first and the trailing two bytes are just garbage to check if the code properly works with arbitrary positions.)
… the corresponding length key Signed-off-by: Andreas Lauser <andreas.lauser@mbition.io> Signed-off-by: Gerrit Ecke <gerrit.ecke@mbition.io>
This converts the diag coded types to the new and shiny encoding infrastructure. Note that doing this as a stand-alone change requires some kludges like creating temporary
EncodeStateobjects. These will be removed by later PRs in the series...Andreas Lauser <andreas.lauser@mercedes-benz.com>, on behalf of MBition GmbH.
Provider Information