LwM2mNodeSenMLDecoder for SenML-JSON with unit tests. #681

cavenaghi9 · 2019-04-12T05:56:57Z

LwM2mNodeSenMLDecoder for SenML-JSON with unit tests, inspired by LwM2MNodeJsonDecoder.

Signed-off-by: Boya Zhang zhangboya@gmail.com

…2MNodeJsonDecoder. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

sbernard31 · 2019-04-15T09:36:45Z

...n-core/src/main/java/org/eclipse/leshan/core/node/codec/senml/LwM2mNodeSenMLJsonDecoder.java

+        // Extract baseName
+        LwM2mPath baseName = extractAndValidateBaseName(pack, path);
+        if (baseName == null)
+            baseName = path; // if no base name, use request path as base name


About "request path as basename, if basename is absent", I read quickly the specification §7.4.4 JSON, I didn't find anything about that.

It seems this part has changed with the new JSON format. I guess this is to respect more the SenML rfc. (If this is the case, I approve this change)

Anyway, I maybe missed it so if you find anything let me know.

Read-Composite is introduced in LwM2M 1.1. Table: 7.4.4.-6 JSON payload in the server request to read manufacturer name, battery level and registration lifetime shows an example of Read-Composite request like below:

[{"n":"/3/0/0"},
{"n":"/3/0/9"},
{"n":"/1/0/1"}]

In response to the above Read-Composite request the client will return a JSON payload as below:
[{"n":"/3/0/0", "vs":"Open Mobile Alliance"},
{"n" :"/3/0/9", "v":95},
{"n":"/1/0/1", "v":86400}]

In this case, the base name is absent, because of resources belong to different object instances, and object instances belong to different object. Do you think SenML JSON/CBOR encoder/decoder shall consider Read-Composite/Write-Composite/Observe-Composite now? Will appreciate if you share you opinion.

I suspect our current API is not really compatible with this.
So I would go to let this aside for now : one step at a time.

(when I say "let this aside", I mean support of Read-Composite/Write-Composite/Observe-Composite)

(when I say "let this aside", I mean support of Read-Composite/Write-Composite/Observe-Composite)

Understood! Is there any plan or on-going contributing activity related with Composite operation now?

There is no plan about that for now.
About on-going contribution, I just saw this #563 (comment)

After extract base name from record array, I am trying to update value of name field of each of record to full path. Therefore, rest of logic can ignore base name anymore. Do you have any comments on that?

I'm not sure to understand. I think we should remove extractAndValidateBaseName(pack, path) because we could face cases where there are several basenames ?

look at this https://tools.ietf.org/html/rfc8428#section-5.1.6

[ {"bn":"2001:db8::2/","bt":1.320078429e+09, "n":"temperature","u":"Cel","v":25.2}, {"n":"humidity","u":"%RH","v":30}, {"bn":"2001:db8::1/","n":"temperature","u":"Cel","v":12.3}, {"n":"humidity","u":"%RH","v":67} ]

if we want a more LWM2M one we could imagine something like this :
Reading object /2 (ACL) containing 2 instances

[ {"bn":"/2/0/","n":"0","v":3}, {"n":"1","v":1}, {"n":"3","v":123} {"bn":"/2/1/","n":"0","v":4}, {"n":"1","v":1}, {"n":"3","v":124} ]

(we could add unit tests like those)

So basename would be now handle in groupRecordsByInstanceId, maybe like this :

private static Map<Integer, Collection<SenMLRecord>> groupRecordsByInstanceId(Collection<SenMLRecord> records) throws CodecException { Map<Integer, Collection<SenMLRecord>> result = new HashMap<>(); String basename = ""; for (SenMLRecord record : records) { // update basename if present if (record.getBaseName() != null) basename = record.getBaseName(); // Build resource path String path = record.getName() == null ? basename : basename + record.getName(); LwM2mPath nodePath = new LwM2mPath(path); // Validate path if (!nodePath.isResourceInstance() && !nodePath.isResource()) { throw new CodecException( "Invalid path [%s] for resource, it should be a resource or a resource instance path", nodePath); } // Get SenML records for this instance Collection<SenMLRecord> recordForInstance = result.get(nodePath.getObjectInstanceId()); if (recordForInstance == null) { recordForInstance = new ArrayList<>(); result.put(nodePath.getObjectInstanceId(), recordForInstance); } // Add it to the list recordForInstance.add(record); } return result; }

sbernard31 · 2019-04-15T09:41:08Z

...n-core/src/main/java/org/eclipse/leshan/core/node/codec/senml/LwM2mNodeSenMLJsonDecoder.java

+    private static Map<Integer, Collection<SenMLRecord>> groupRecordsByInstanceId(Collection<SenMLRecord> records,
+            LwM2mPath baseName) throws CodecException {
+        Map<Integer, Collection<SenMLRecord>> result = new HashMap<>();
+


Following comment above, I feel that we don't need baseName parameter anymore.
And we should just use baseName from SenMLRecord.
Look at this part of the rfc to implement it correctly : https://tools.ietf.org/html/rfc8428#section-5.1.6

Same as above comments, I am trying to update value of name field of each of record to full path, then base name can be ignore, any comments for this change?

sbernard31 · 2019-04-15T09:48:10Z

...n-core/src/main/java/org/eclipse/leshan/core/node/codec/senml/LwM2mNodeSenMLJsonDecoder.java

+        // Create an entry for an empty instance if possible
+        if (result.isEmpty() && baseName.getObjectInstanceId() != null) {
+            result.put(baseName.getObjectInstanceId(), new ArrayList<SenMLRecord>());
+        }


If there is no more baseName parameter, we will probably remove this part of the code.
And so we should handle the case where result if empy in calling method.

sbernard31 · 2019-04-15T09:56:47Z

...n-core/src/main/java/org/eclipse/leshan/core/node/codec/senml/LwM2mNodeSenMLJsonDecoder.java

+        if (baseName == null)
+            baseName = path; // if no base name, use request path as base name
+
+        // Group JSON entry by instance


Just a details, but several time in comments you are using JSON or JSON entry, I feel this would be more appropriate to use SenML or SenML record.

Sorry for that, I tried to replace or remove irrelevant comments, but still missed something.

Signed-off-by: Boya Zhang <zhangboya@gmail.com>

sbernard31 · 2019-04-19T09:39:40Z

(@cavenaghi9, just to let you know I will not be available next 2,5 weeks, so don't worry if I don't reply)

cavenaghi9 · 2019-04-29T03:54:46Z

(@cavenaghi9, just to let you know I will not be available next 2,5 weeks, so don't worry if I don't reply)

2.5 weeks or 2~5 weeks? having a nice vacation.

sbernard31 · 2019-05-10T14:05:22Z

2.5 weeks or 2~5 weeks? having a nice vacation.

That was 2.5 week 😁

cavenaghi9 · 2019-05-21T02:58:10Z

2.5 weeks or 2~5 weeks? having a nice vacation.

That was 2.5 week 😁

Are you available now, not sure if your vacation is finished? I have to go on contribution.

sbernard31 · 2019-05-21T09:05:27Z

I have been back since May 9, but next week I will be unavailable too.

But anyway do not hesitate to contribute at any time. I just share my availability to avoid you to worry if I don't answer in time.

- update value of name field of each of record to full path, using name field directly when creating LwM2mPath object. - remove used codes, and update description. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

This reverts commit d279478. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

- update value of name field of each of record to full path, using name field directly when creating LwM2mPath object. - remove used codes, and update description. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

Signed-off-by: Boya Zhang <zhangboya@gmail.com>

sbernard31 · 2019-06-28T12:51:07Z

I'm a bit confuse with the current state of this branch.
Could you please squash all the commit and rebase this PR on branch 2.0.x ?

cavenaghi9 · 2019-07-01T09:30:27Z

squash

I'm a bit confuse with the current state of this branch.
Could you please squash all the commit and rebase this PR on branch 2.0.x ?

I forgot to sign-off, then I changed commit history.
I changed unit test in #682 , therefore I merge changes of 2.0.x to current branch.

any comments or question please let me know.

sbernard31 · 2019-07-19T15:24:11Z

I will not be available the next 2 weeks but I look at this as soon as I'm back.

Could you please squash all your commit and rebase it over branch 2.0.x ?
(do no hesitate to tell me if this is an issue for you)

sbernard31 · 2019-08-30T09:33:27Z

@cavenaghi9 do you plan to continue to work on this PR ? (it's just to know, no pressure :))

sbernard31 · 2020-10-22T09:57:43Z

I finished and integrated this PR in #911.

@cavenaghi9 thx again for all your contribution. 🙏

We didin't get any news from you for long time ago now. It seems you are no more active on github. I hope all is fine for you.

LwM2mNodeSenMLDecoder for SenML-JSON with unit tests, inspired by LwM…

59ced0d

…2MNodeJsonDecoder. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

sbernard31 reviewed Apr 15, 2019

View reviewed changes

update notes

8394849

Signed-off-by: Boya Zhang <zhangboya@gmail.com>

cavenaghi9 added 4 commits June 26, 2019 09:52

Fixing review comments:

53bf6a9

- update value of name field of each of record to full path, using name field directly when creating LwM2mPath object. - remove used codes, and update description. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

Revert "Fixing review comments:"

122552e

This reverts commit d279478. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

Fixing review comments:

4feb9bb

- update value of name field of each of record to full path, using name field directly when creating LwM2mPath object. - remove used codes, and update description. Signed-off-by: Boya Zhang <zhangboya@gmail.com>

eclipse-leshan#682: Add SenML-CBOR encoder in leshan-core + unit tests.

85eeb21

Signed-off-by: Boya Zhang <zhangboya@gmail.com>

cavenaghi9 force-pushed the lwm2mNode-senML-json-decoder branch from 9ed4b34 to 85eeb21 Compare June 26, 2019 01:54

merge with leshan-2.0.x branch.

0611407

Signed-off-by: Boya Zhang <zhangboya@gmail.com>

sbernard31 added this to the 2.0.0 milestone Aug 14, 2019

sbernard31 added the orphan PR with no more activity label Nov 22, 2019

sbernard31 closed this Jul 16, 2020

sbernard31 reopened this Jul 16, 2020

sbernard31 changed the base branch from 2.0.x to deprecated_2.0.x July 16, 2020 16:58

sbernard31 mentioned this pull request Oct 8, 2020

NIDD device support in leshan. #904

Closed

sbernard31 closed this Oct 22, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

LwM2mNodeSenMLDecoder for SenML-JSON with unit tests. #681

LwM2mNodeSenMLDecoder for SenML-JSON with unit tests. #681

cavenaghi9 commented Apr 12, 2019

sbernard31 Apr 15, 2019

cavenaghi9 Apr 16, 2019

sbernard31 Apr 16, 2019

sbernard31 Apr 17, 2019

cavenaghi9 Apr 18, 2019

sbernard31 Apr 18, 2019

cavenaghi9 May 1, 2019

sbernard31 May 10, 2019

sbernard31 Apr 15, 2019

cavenaghi9 May 1, 2019

sbernard31 Apr 15, 2019

sbernard31 Apr 15, 2019

cavenaghi9 Apr 16, 2019

sbernard31 commented Apr 19, 2019

cavenaghi9 commented Apr 29, 2019

sbernard31 commented May 10, 2019

cavenaghi9 commented May 21, 2019

sbernard31 commented May 21, 2019

sbernard31 commented Jun 28, 2019

cavenaghi9 commented Jul 1, 2019

sbernard31 commented Jul 19, 2019

sbernard31 commented Aug 30, 2019

sbernard31 commented Oct 22, 2020

LwM2mNodeSenMLDecoder for SenML-JSON with unit tests. #681

LwM2mNodeSenMLDecoder for SenML-JSON with unit tests. #681

Conversation

cavenaghi9 commented Apr 12, 2019

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sbernard31 commented Apr 19, 2019

cavenaghi9 commented Apr 29, 2019

sbernard31 commented May 10, 2019

cavenaghi9 commented May 21, 2019

sbernard31 commented May 21, 2019

sbernard31 commented Jun 28, 2019

cavenaghi9 commented Jul 1, 2019

sbernard31 commented Jul 19, 2019

sbernard31 commented Aug 30, 2019

sbernard31 commented Oct 22, 2020