openschc/openschc

SCHC Implementation (for Python): see doc at

openschc's Issues

noAck FCN

if FCN size > 1 the noACK must number all the frames

CoAP unparsing error

There is a bug in CoAP unparse especially when UP and DW fields are present. It appears that the CoAP reconstruction was driven by the rule and not the header fields. ltn22/openschc contains in connector branch a new version of the unparse. What will be missing is to sort the option fields regarding their number and position, if the rule do not respect it.

coap parsing

Add an option in parse to indicate the CoAP port number. By default 5683

schc_send return code

Verify the code returned by schc_send. Should be None when no error occurs, False if no rules are found.

Problem in running examples/tcp

HI,
I just started to test examples/tcp and i got some errors. Could u please guide me?

after running Server on terminal 1 with the command "python3 ClientServerSimul.py --role server --compression false --rule ../configs/rule1.json", I got the following:

socket binded to 127.0.0.1 ip and port 12345
socket is listening
Connexion de 127.0.0.1 34362

then i ran client on another terminal: "python3 ClientServerSimul.py --role client --compression false --rule ../configs/rule1.json --time 20 --payload payload/testfile_small.txt"

Then for both terminal i got the same errors:

--------------------------- Iteration 1 --------------------------
---------- Client: Ip = 127.0.0.1 Port = 34362 ---------------
Traceback (most recent call last):
File "ClientServerSimul.py", line 100, in
schcConfig.start()
File "ClientServerSimul.py", line 70, in start
self.server.server()
File "/home/samira/openschc/examples/tcp/ServerConnection.py", line 48, in server
self.newThread = ClientThread.ClientThread(ipClient, portClient, self.clientSocket, self.configuration)
File "/home/samira/openschc/examples/tcp/ClientThread.py", line 28, in init
self.client_config()
File "/home/samira/openschc/examples/tcp/ClientThread.py", line 98, in client_config
self.clientConfigInServer.configSim()
File "/home/samira/openschc/examples/tcp/SchcConfig.py", line 68, in configSim
self.node0 = self.make_node(self.sim, self.rule_manager, self.devaddr)
File "/home/samira/openschc/examples/tcp/SchcConfig.py", line 85, in make_node
node = net_sim_core.SimulSCHCNode(sim, extra_config)
File "../../src/net_sim_core.py", line 89, in init
self.config, self, self.layer2, self.layer3, role, unique_peer)
File "../../src/protocol.py", line 124, in init
assert role in ["device", "core-server"]
AssertionError

test_compress has to fail

Make the test fails when the compression rule is not found.

pytestifying all tests

We need to have all tests written using the pytest so they can be run by GitHub Actions.

I see two locations for tests:

• the test/ directory. The README suggest one needs to run the scripts directly but test_frag.py suggests pytest is used
• different test_*.py script at the root, which I believe should be run directly.

My approach would to migrate all tests into the tests/ directory, using the pytest approach.

I can do the heavy-lifting, but @MarinoMtz, please indicate which tests we expect to run.

MIC error

The rule of the sender is:

{'RuleID': 1, 'RuleIDLength': 3, 'Fragmentation': {'FRDirection': 'DW', 'FRMode': 'ackOnError', 'FRModeProfile': {'FCNSize': 3, 'dtagSize': 2, 'MICALgorithm': 'crc32', 'WSize': 7, 'ackBehavior': 'afterAll1', 'tileSize': 17, 'maxRetry': 4, 'timeout': 600, 'L2WordSize': 8, 'lastTileInAll1': False, 'windowSize': 7}}}

Please look at the content of tile. The last two tiles the receiver received are different from the sender sent. So, MIC error happens at the receiver side.

It looks that the order is reversed. If you consider it, the content is still different.

Any idea ?

I am going to make a kit to reproduce this error.

Sender's tiles:

{'w-num': 0, 't-num': 6, 'tile': b'\x60\x05\x80'/17, 'sent': True
{'w-num': 0, 't-num': 5, 'tile': b'\x73\x64\x00'/17, 'sent': True
{'w-num': 0, 't-num': 4, 'tile': b'\x00\x40\x80'/17, 'sent': True
{'w-num': 0, 't-num': 3, 'tile': b'\xd2\x01\x00'/17, 'sent': True
{'w-num': 0, 't-num': 2, 'tile': b'\x00\x10\x00'/17, 'sent': True
{'w-num': 0, 't-num': 1, 'tile': b'\x84\x18\x00'/17, 'sent': True
{'w-num': 0, 't-num': 0, 'tile': b'\x37\x04\x00'/17, 'sent': True
{'w-num': 1, 't-num': 6, 'tile': b'\x01\x00\x00'/17, 'sent': True
{'w-num': 1, 't-num': 5, 'tile': b'\x00\x00\x00'/17, 'sent': True
{'w-num': 1, 't-num': 4, 'tile': b'\x00\x00\x00'/17, 'sent': True
{'w-num': 1, 't-num': 3, 'tile': b'\x00\x01\x80'/17, 'sent': True
{'w-num': 1, 't-num': 2, 'tile': b'\xc9\x20\x00'/17, 'sent': True
{'w-num': 1, 't-num': 1, 'tile': b'\x46\x80\x00'/17, 'sent': True
{'w-num': 1, 't-num': 0, 'tile': b'\x08\x00\x80'/17, 'sent': True
{'w-num': 2, 't-num': 6, 'tile': b'\x02\x06\x00'/17, 'sent': True
{'w-num': 2, 't-num': 5, 'tile': b'\x00\x00\x00'/17, 'sent': True
{'w-num': 2, 't-num': 4, 'tile': b'\x00\x00\x00'/17, 'sent': True
{'w-num': 2, 't-num': 3, 'tile': b'\x00\x01\x00'/17, 'sent': True
{'w-num': 2, 't-num': 2, 'tile': b'\xc8\xd2\x00'/17, 'sent': True
{'w-num': 2, 't-num': 1, 'tile': b'\x00\x07\x00'/17, 'sent': True
{'w-num': 2, 't-num': 0, 'tile': b'\x9d\xa4\x80'/17, 'sent': True
{'w-num': 3, 't-num': 6, 'tile': b'\x75\x20\x00'/17, 'sent': True
{'w-num': 3, 't-num': 5, 'tile': b'\x00\x17\x00'/17, 'sent': True
{'w-num': 3, 't-num': 4, 'tile': b'\x10\xf8\x80'/17, 'sent': True
{'w-num': 3, 't-num': 3, 'tile': b'\x56\x00\x00'/17, 'sent': True
{'w-num': 3, 't-num': 2, 'tile': b'\x18\x00'/9, 'sent': True
{'w-num': 3, 't-num': 1, 'tile': b'\x4c\x3c'/14, 'sent': True}]

Receiver's tiles:

{'w-num': 0, 't-num': 6, 'nb_tiles': 1, 'raw_tiles': b'\x60\x05\x80'/17}
{'w-num': 0, 't-num': 5, 'nb_tiles': 1, 'raw_tiles': b'\x73\x64\x00'/17}
{'w-num': 0, 't-num': 4, 'nb_tiles': 1, 'raw_tiles': b'\x00\x40\x80'/17}
{'w-num': 0, 't-num': 3, 'nb_tiles': 1, 'raw_tiles': b'\xd2\x01\x00'/17}
{'w-num': 0, 't-num': 2, 'nb_tiles': 1, 'raw_tiles': b'\x00\x10\x00'/17}
{'w-num': 0, 't-num': 1, 'nb_tiles': 1, 'raw_tiles': b'\x84\x18\x00'/17}
{'w-num': 0, 't-num': 0, 'nb_tiles': 1, 'raw_tiles': b'\x37\x04\x00'/17}
{'w-num': 1, 't-num': 6, 'nb_tiles': 1, 'raw_tiles': b'\x01\x00\x00'/17}
{'w-num': 1, 't-num': 5, 'nb_tiles': 1, 'raw_tiles': b'\x00\x00\x00'/17}
{'w-num': 1, 't-num': 4, 'nb_tiles': 1, 'raw_tiles': b'\x00\x00\x00'/17}
{'w-num': 1, 't-num': 3, 'nb_tiles': 1, 'raw_tiles': b'\x00\x01\x80'/17}
{'w-num': 1, 't-num': 2, 'nb_tiles': 1, 'raw_tiles': b'\xc9\x20\x00'/17}
{'w-num': 1, 't-num': 1, 'nb_tiles': 1, 'raw_tiles': b'\x46\x80\x00'/17}
{'w-num': 1, 't-num': 0, 'nb_tiles': 1, 'raw_tiles': b'\x08\x00\x80'/17}
{'w-num': 2, 't-num': 6, 'nb_tiles': 1, 'raw_tiles': b'\x02\x06\x00'/17}
{'w-num': 2, 't-num': 5, 'nb_tiles': 1, 'raw_tiles': b'\x00\x00\x00'/17}
{'w-num': 2, 't-num': 4, 'nb_tiles': 1, 'raw_tiles': b'\x00\x00\x00'/17}
{'w-num': 2, 't-num': 3, 'nb_tiles': 1, 'raw_tiles': b'\x00\x01\x00'/17}
{'w-num': 2, 't-num': 2, 'nb_tiles': 1, 'raw_tiles': b'\xc8\xd2\x00'/17}
{'w-num': 2, 't-num': 1, 'nb_tiles': 1, 'raw_tiles': b'\x00\x07\x00'/17}
{'w-num': 2, 't-num': 0, 'nb_tiles': 1, 'raw_tiles': b'\x9d\xa4\x80'/17}
{'w-num': 3, 't-num': 6, 'nb_tiles': 1, 'raw_tiles': b'\x75\x20\x00'/17}
{'w-num': 3, 't-num': 5, 'nb_tiles': 1, 'raw_tiles': b'\x00\x17\x00'/17}
{'w-num': 3, 't-num': 4, 'nb_tiles': 1, 'raw_tiles': b'\x10\xf8\x80'/17}
{'w-num': 3, 't-num': 3, 'nb_tiles': 1, 'raw_tiles': b'\x56\x00\x00'/17}
{'w-num': 3, 't-num': 1, 'nb_tiles': 1, 'raw_tiles': b'\x4c\x3c\x00'/17}
{'w-num': 3, 't-num': 7, 'nb_tiles': 1, 'raw_tiles': b'\x18\x00'/16}

Packet sender tried to send.

HEX: 6005b9b200103a4020010420c0dc100200000000000000792404680040040818000000000000b2348000e9da4ba900005c21f156000c130f

bytearray(b'`\x05\xb9\xb2\x00\x10:@ \x01\x04 \xc0\xdc\x10\x02\x00\x00\x00\x00\x00\x00\x00y$\x04h\x00@\x04\x08\x18\x00\x00\x00\x00\x00\x00\xb24\x80\x00\xe9\xdaK\xa9\x00\x00\!\xf1V\x00\x0c\x13\x0f')

Updating the API (synchronization of branch master and branch develop)

This issue is about the changes to synchronize the branch master with the branch develop. It stems from a discussion between openschc contributors.

The main question is related to the API changes, which impact "connectors" or "lower layers" (e.g. examples/udp, examples/gateway, examples/simulator, examples/scapy , etc. )

In order to make the connectors work with the branch develop some changes are necessary.
Some possibilities are listed in the last section, and feedback is welcome.

Architecture and API on branch master

For reference, the architecture and API of openschc in the master branch are "defined" (more or less accurately; the interface with the rule manager is incomplete, and the compression is not called in master) in :

• src/architecture.py
• docs/Internals/external-interface.svg - displayed on this doc page
• (and there exists a data flow docs/oschc-arch-00.svg although it is unclear if it is up-to-date with master).

Current changes on branch develop

These are some changes between the two branches:

1. there have been changes related to the identifier/address management, i.e. SCHCProtocol needs to have both the identifiers (addresses) of the core and the device (which itself is).

   • schc_recv and schc_send now take both addresses/identifiers as arguments

2. also there is a semantics change in schc_recv which returns a packet (when available) for Layer 3 (Upper Layer) instead of letting SCHCProtocol call the upper layer with the packet.

3. as a result of address/identifier management, and proper rule definition, internally, every used address is the in the format of the identifiers of the rule database, e.g. a string such as "lorawan:70b3d5499126b445":

   • there is the question of under which format the UpperLayer and LowerLayer send and receive those addresses, and how to convert to different formats.
     • in particular: the packet addresses are exchanged in SCHCProtocol.schc_send, SCHCProtocol.schc_recv, UpperLayer.recv_packet, and LowerLayer.send_packet (in design architecture.py, check in actual code).
   • In addition, for the UpperLayer there is the issue of knowing the address/identifier of the other end based on the IPv6 address.

4. there are changes in the scheduler to cancel all events related to one session src/net_sim_sched.py, e.g. addition of one method cancel_session

Proposed changes to synchronize both branches

1. for schc_recv, schc_send: keep both semantics from master and develop, considering the one from develop as more internal than the ones from branch master.

   • rationale(s):
     • the schc_recv semantics (in master) of letting SCHCProtocol pass the packet to the UpperLayer is consistent with the schc_send semantics of letting SCHCProtocol pass the packets to the LowerLayer ; and is probably expected by the connectors developers.
     • it makes sense to store one's own address/identifier in SCHCProtocol compared to passing it every time,
   • This can be achieved as follows: keep both pairs (of each branch), by renaming one pair, e.g.:
     • rename schc_recv, schc_send with branch master semantics as schc_recv_from_lower, schc_send_from_upper
     • or rename schc_recv, schc_send of develop to _schc_recv, _schc_send ;
     • and one version from the other (with one or two changes in the code)
   • Alternately: by renaming SCHCProtocol from branch develop as BaseSCHCProtocol and creating a class SCHCProtocol from branch master inheriting from BaseSCHCProtocol, and implementing the master semantics.

2. for the knowledge of own address/id in the SCHCProtocol (for the version of master of schc_recv/schc_send):

   • this can be implemented in LowerLayer.get_address
   • this can also be passed in the constructor of the SCHCProtocol. Rationale: the position is also passed to the constructor, so one should consider if this is consistent.

3. for the addresses/identifiers:

   • for the exchanges with the LowerLayer which do need MAC addresses
     • the internal representation of the addresses/identifiers of the rule manager can be consistently used in the openschc API, and helper functions can be provided to do conversions to any format.
     • (if both semantics of develop and master are needed it is possible to use inheritance+template methods, or to use object composition to have an optional address conversion object).
   • for the UpperLayer to retrieve the address/identifier of the other end, for some backends, some external step would be needed, as it is only available indirectly inside the rule set. Extra work is needed for this case, and for basic connector cases, a table can be constructed/provided (maybe in the most complex case, provided alongside with the rules).

4. for the cancel_session, some possibilities:

   • maybe review the code and implement cancel_session inside the SCHC and fragmentation code, not inside the scheduler. Rationale: the semantics of the SCHC sessions are unrelated to the semantics of the scheduler.
   • or rename cancel_session to cancel_event_group (introducing the equivalent concept of event group)
     • and maybe implement it as an adapter/wrapper to any scheduler. Rationale: there are at least three implementations of the Scheduler interface; one using 'asyncio' loops, one using Python 'sched' module, one "done from scratch", SimulScheduler; and implementing the semantics might require hacking their internals, or more reasonably implementing an adapter/wrapper anyway.
   • if the semantics of canceling session timers was already broken in master, this is not an urgent issue.

Feedback

Your help in selecting the changes would greatly welcomed. Thanks!

contextlib and micropython

It is not a case in general. But, my micropython doesn't have contextlib.
Is it acceptable on your micropython ? If not, the core SCHC codes which is under the src folder should have only features supported by micropython.

% micropython
MicroPython v1.12-35-g10709846f on 2020-08-03; darwin version
Use Ctrl-D to exit, Ctrl-E for paste mode

import sys
sys.version_info
(3, 4, 0)

import contextlib
Traceback (most recent call last):
File "", line 1, in
ImportError: no module named 'contextlib'

uncompressed packet is not processed

Hi,

I've added the no compressed rule and it works fine, the receivers receives fragments and do reassembly, then call the decompression which remove the nocompression rule ID

but then, the packet is lost. In the code of frag_recv.py, we have line 205

if not self.protocol.config.get("debug-fragment"):
# XXX
# XXX in hack105, we have separate databases for C/D and F/R.
# XXX need to merge them into one. Then, here searching database will
# XXX be moved into somewhere.
# XXX
#rule = self.protocol.rule_manager.FindRuleFromSCHCpacket(schc=schc_packet)
#dprint("debug: no-ack FindRuleFromSCHCpacket", rule)
uncompressed = self.protocol.process_decompress(schc_packet, self.sender_L2addr, "UP")
self.state = 'DONE_NO_ACK'
print ("return uncompress packet")
dprint(self.state)
return

originaly uncompressed variable didn't exist. I don't know how to give it to the upper layer. Help!!!

what is `stats/`?

@MarinoMtz I have a hard time understanding what stats/ is. Given that there is an __init__.py, I assume it's package which can be imported. I'm suprised to see log.txt which appears to be an artifact. I do see that it is imported a bit everywhere. I would recommend to, at least:

• rename, as "stats" is super super generic
• add a README.md to explain what it is and how it is used

What /interop is for?

We are cleaning the repo with @ltn22 and we want to know what this folder is used for, We see there are some python script to generate a 'golden environment', maybe @adjih can clarify this?

No Compression

There is no No Compression handling in schc_recv in protocol.py. A patch is currently on ltn22/openschc in connector branch

Demonstration with DLMS/COSEM

Someone wants to show DLMS over SCHC over LoRaWAN.
The DLMS-COMSEM implementation is here.

https://github.com/epri-dev/DLMS-COSEM

Could we make it ?

Pythonify tests/

Make the code in tests/ more Pythonic

Proper session management (was: "MAC address in protocol.py")

Why is there a test in protocol.py to test the device? For me we are independant of the device at this level.

  if dev_L2addr == b"\xaa\xbb\xcc\xee":

            if frag_rule[T_FRAG][T_FRAG_PROF][T_FRAG_DTAG] > 0:
                dtag = packet_bbuf.get_bits(frag_rule[T_FRAG][T_FRAG_PROF][T_FRAG_DTAG],
                                            position=frag_rule[T_RULEIDLENGTH])
            else:
                dtag = None
....

Pypi integration

kamene != scapy

Hi,

I'm a Scapy maintainer, and have bumped into this project because it was linked in https://github.com/phaethon/kamene/network/dependents

I've noticed that you were using kamene. I'm assuming you were using scapy-python3 before it was renamed. Please note that kamene is NOT an official Scapy version. Quoting https://scapy.net/

An independent fork of Scapy was created from v2.2.0 in 2015, aimed at supporting only Python3 (scapy3k). The fork diverged, did not follow evolutions and fixes, and has had its own life without contributions back to Scapy. Unfortunately, it has been packaged as python3-scapy in some distributions, and as scapy-python3 on PyPI leading to confusion amongst users. It should not be the case anymore soon. Scapy supports Python3 in addition to Python2 since 2.4.0. Scapy v2.4.0 should be favored as the official Scapy code base. The fork has been renamed as kamene.

Why this is a problem

There are two issues:

• for you: you are using a version of Scapy that clearly is buggy and not maintained. Development is stopped (or sometimes a few merged PRs now and then) and it's missing many improvements and bug fixes. Moreover we had issues with the owner of this fork trying to mislead users by using a confusing name.
• for us: we get bug reports that we fixed ages ago because some people still use this fork. This wastes some of our time

I'd advise switching to scapy = the original Scapy.

Thanks for your time !

PS: sorry if this feels copy/pasted. It is a bit :(

CoAP - Delta encoding in unparse code is not working

Because there is no full type value

rule value in bytearray

modify the add rule to store values in bytearray and add a display function for each field regarding its type.

Current state of OpenSCHC (sept 2022) and branch to use

This issue is to help newcomers (or people like myself who did not follow the evolutions in the past 2 years) select the right branch to use. This is only my understanding:

• OpenSCHC has a historical branch master where developments had been historically merged, but there have been no real developments since mid-2020.
• Currently @MarinoMtz and @ltn22 have been accelerating the development and debugging of OpenSCHC; such efforts have been on the branch develop. @twatteyne is also making the project into good shape (many tests, PyPI integration, CI, and a lot more).

If one wants to use OpenSCHC, it is suggested:

• the branch develop should be tried by newcomers. The example examples/scapy should be running (but requires some setup) and many tests (properly automated) are under quick updates to provide some running examples.
• the branch master has still several useful examples, such as examples/udp or examples/simulator or src/test_newschc.py but does not integrate compression (only fragmentation, and there are some bugs). Many of them will not run on develop due to changes in API. Discussion to fix most of it is in #117 and is ongoing work.

Create a gh-pages branch

To push the documentation generated by sphinx, I would need an existing gh-pages branch on th main repository

remove `schctest/`

@MarinoMtz what is schctest/? Looks like lots and lots of code, apparently duplicating some of the core OpenSCHC code?!? If this is some dead code, I strongly suggest we remove the folder.

AttributeError: `RuleManager` object has no attribute `find_context_bydevL2addr`

I just cloned the repository and followed the instruction to make the test:

_________________________________________________________________________________________ test_ruleman_01 __________________________________________________________________________________________

def test_ruleman_01():
    # XXX actually, it is not test code right now.
    RM = RuleManager()
    RM.add_context(context1, rule1,rule2,rule3)
    RM.add_context(context2, rule1)
    print(RM._db)
    #RM.add_rules(context1, [conflictingRule0])
    #RM.add_rules(context1, [bogusRule0])

  print(RM.find_context_bydevL2addr("AABBCCDD"))

E AttributeError: 'RuleManager' object has no attribute 'find_context_bydevL2addr'

tests/test_ruleman.py:72: AttributeError
--------------------------------------------------------------------------------------- Captured stdout call ---------------------------------------------------------------------------------------
[{'devL2Addr': 'AABBCCDD', 'dstIID': '2001:0db8:85a3::beef', 'comp': {"RuleID": 4, "RuleIDLength": 5, "Compression": []}, 'fragSender': {"RuleID": 4, "RuleIDLength": 3, "profile": {"MICAlgorithm": "RCS_RFC8724", "MICWordSize": 8, "L2WordSize": 8}, "Fragmentation": {"dir": "out", "FRMode": "AckOnError", "FRModeProfile": {"dtagSize": 2, "FCNSize": 3, "ackBehavior": "afterAll1", "WSize": 5, "windowSize": 7, "tileSize": 64}}}, 'fragReceiver': {"RuleID": 7, "RuleIDLength": 3, "Profile": {"MICAlgorithm": "RCS_RFC8724", "MICWordSize": 8, "L2WordSize": 8}, "Fragmentation": {"dir": "in", "FRMode": "NoAck", "FRModeProfile": {"dtagSize": 0, "WSize": 0, "FCNSize": 1, "windowSize": 1}}}}, {'devL2Addr': '', 'dstIID': '', 'comp': {"RuleID": 4, "RuleIDLength": 5, "Compression": []}}]
===================================================================================== short test summary info ======================================================================================
FAILED tests/test_frag.py::test_frag_ack_on_error_no_loss - assert "msg_type_queue -> ['SCHC_ACK_OK']" in "> Queue running event -> 0, callback -> _notify_start\n> 1 [schc] @0 recv-from-l3 None...
FAILED tests/test_frag.py::test_frag_ack_on_error_loss - assert "msg_type_queue -> ['SCHC_ACK_OK']" in "> Queue running event -> 0, callback -> _notify_start\n> 1 [schc] @2 recv-from-l3 None 20...
FAILED tests/test_frag.py::test_frag_no_ack_no_loss - assert 'SUCCESS: MIC matched' in "> Queue running event -> 0, callback -> _notify_start\n> 1 [schc] @4 recv-from-l3 None 2001:0db8:85a3:000...
FAILED tests/test_frag.py::test_frag_no_ack_loss - assert 'ERROR: MIC mismatched' in "> Queue running event -> 0, callback -> _notify_start\n> 1 [schc] @6 recv-from-l3 None 2001:0db8:85a3:0000:...
FAILED tests/test_ruleman.py::test_ruleman_01 - AttributeError: 'RuleManager' object has no attribute 'find_context_bydevL2addr'
=================================================================================== 5 failed, 8 passed in 0.21s ====================================================================================

move develop to Master

For many months Master as been quite and all the changes appears in the develop branch, with the following modifications:

• introduction of the coreconf format in the rule_manager
• add link to LNS through a proxy
• correct bugs in coap compression
• ICMPv6 ping message process

I propose you, for clarification to merge the develop to Master next week.

First documentation example

Hi,

in docs/General/UG_examples I save files related to examples on the user guide
-> I changed the name of udp_example to udp_schc which is sexier
-> I will add soon the documentation in the UG

I would like the program to finish when the file is sent. right now the device is blocked the core-server does not clear the buffer.

type of prefix and iid

The type of prefix is in integer by Parser.parse(), but in bytes by the rule manager. There is a part to convert into a type for comparison in the source code internally. I think that It's better to keep it as a single type without converting the type.

Is is okey to adopt "bytes" type of prefix and IID in Parser.parse() ?

get_address()

get_address() in the l2 layer is not needed anymore, correct ?

How to use this module

Hi,
I've a problem to use this module, there is no docs about how to use it, and for example, the udp example don't work:

Is there somewhere something that help to implement openschc ?

rule format

The rule format may have to change to allow several rules for fragmentation and compression for a specific device.

For me the constraints are:

• the rule must be the same in the device and in the core C/D
• in the core C/D we must easily look up for the rule of a specific device by its LPWAN @
• in the core C/D rules can be stored in a file (or in MongoDB)
• in the device the rule is easily associated to the device

In rulemanager.py, the rule is defined this way:

    db = [
        {
            "devL2Addr": ..,
            "dstIID": ..,
            "comp": {
                    "ruleID": ..,
                    "ruleLength": ..,
                    "compression": { ...}
                },
            "fragSender": {
                    "ruleID": ..,
                    "ruleLength": ..,
                    "fragmentation": { ...}
                }
            "fragReceiver": {
                    "ruleID": ..,
                    "ruleLength": ..,
                    "fragmentation": { ...}
                }
        }, ...
    ]

In my view the structure should be more:

    db = [
        {
            "devL2Addr": ..,
            "devIP": "   ",  # mandatory
             "rules": {
                [{
                      ruleID: xxx,
                      ruleLength: xxx, 
                      compression: {
                            [["fieldID", ..., ....],
                             ["ffffff", ...        ]
                     },
                    {
                      ruleID: xxx,
                      ruleLength: xxx, 
                      fragmentation: {
                           "direction": "up",
                           parameters....
                     }
                  }...
    
        }, 
        {.....}

 
    ]

no-compression RuleID is mandatory

The source code and the documentation must be improved to mandate that a "no-compression" RuleID MUST be present in the Rule set.
Define an attribute for this special "compression" rule.

includes not allowed in README.rst on GitHub

I used an include in the README.rst in examples/udp, because it made easier to build the Sphynx documentation this way.
GitHub does not process the include, therefore the README is incomplete as you browse the repo.
A workaround is suggested in https://gist.github.com/shaypal5/70044eeda587fae17b180e723496b057

remove type for TV

OpensSCHC may use the type store in TV to define an operation. The goal is to align openSCHC with the YANG Data Model and TV stores only bytearray. Regarding the field ID the data can be displayed differently.

remove obsolete files (such as the ones linked to the old rulemanager)

remove files such as test_basic.py that uses the add_rule function (old rule manager)

testfile.txt and testfile_large.txt in master-hack105 are not needed.

Extraneous files seem to have made their way into the master-hack105 branch.
They are testfile.txt and testfile_large.txt, in the src directory.
When the dust of the current code merge has settled, remember to delete these files.

GitHub Actions

add support for GitHub Actions to run all tests all the time

FindFragmentationRule() of RuleManager is to be implemented.

currently returns the first fragmentation rule, irrespective of the arguments.

client_server_simulation.txt

Does anyone know what client_server_simulation.txt is ?
frag_send.py contains below code. Why and what ?

f = open("client_server_simulation.txt", "w+")

proof-reading documentation

test_frag_new is broken

Trying to run the instructions in https://github.com/openschc/openschc/blob/master/src/README.md, I get the following error message:
$ python3 test_frag_new.py
<<<...>>>
Traceback (most recent call last):
File "test_frag_new.py", line 131, in
sim.run()
File "/Users/dominique/Desktop/Hackathon108/openschc/src/net_sim_core.py", line 337, in run
self.scheduler.run()
File "/Users/dominique/Desktop/Hackathon108/openschc/src/net_sim_sched.py", line 40, in run
callback(*args)
File "/Users/dominique/Desktop/Hackathon108/openschc/src/net_sim_core.py", line 240, in deliver_packet
count += self.send_packet_on_link(link, packet)
File "/Users/dominique/Desktop/Hackathon108/openschc/src/net_sim_core.py", line 307, in send_packet_on_link
node_to.event_receive(link.from_id, packet)
File "/Users/dominique/Desktop/Hackathon108/openschc/src/net_sim_core.py", line 96, in event_receive
self.layer2.event_receive_packet(sender_id, packet)
File "/Users/dominique/Desktop/Hackathon108/openschc/src/net_sim_layer2.py", line 80, in event_receive_packet
self.protocol.schc_recv(self.devaddr, packet)
File "/Users/dominique/Desktop/Hackathon108/openschc/src/protocol.py", line 243, in schc_recv
dtag_length = frag_rule[T_FRAG][T_FRAG_PROF][T_FRAG_DTAG]
TypeError: 'NoneType' object is not subscriptable

naming conventions in `README`

Per #119 (comment)

cleaning the repo

Hello,

We would like to make some cleaning in the repo. There is a lot of stale banches and currently only scapy is active. So we propose to delete all the unactive branches to keep only scapy as main. @dominique is the gh-page needed to generate the documation ?

Thanks.

Laurent

duplicate gen_rulemanager.py

There are two copies of the same file:

• gen_rulemanager.py
• src/gen_rulemanager.py

Doing a diff there are little differences sprinkled around.

Looking at the log:

• src/gen_rulemanager.py seems to be much more up to date, which changes from @ltn22 and @MarinoMtz dated back 25-May-2022
• gen_rulemanager.py seems to have been copied out of src/ for an IETF hackathon and forgotten

@MarinoMtz, should we remove gen_rulemanager.py?

uplink and downlink fragmentation

in the current fragmentation rule, there is no indication if the rule msut be used uplink or downlink. A new tag must be added to tell the direction