Yaml Comment.

[PANDATEEN]

add a description attribute about property.
if used,when Serialize(),the result will have a yaml comment on that property.

[aaubry]

Thanks for this. I'll try to fix the CI so that the build passes. Could you add some tests to show that this works as intended ?

[PANDATEEN]

add a test about that.
please check if convenient.

[pawelpawlow]

@aaubry, is there any chance we have this merged and released soon?

[aaubry]

Sorry, I will check this later today and will release afterwards. 30 Mar 2021 12:50:14 PAWEŁ PAWŁOW ***@***.***>:

…

@aaubry[https://github.com/aaubry], is there any chance we have this merged and released soon? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub[#584 (comment)], or unsubscribe[https://github.com/notifications/unsubscribe-auth/AAEZ5X2XAP3QBVZP2YGK4GTTGG3HLANCNFSM4X5ZXATA]. [data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAEQAAABECAYAAAA4E5OyAAAAAXNSR0IArs4c6QAAAARzQklUCAgICHwIZIgAAAAoSURBVHic7cEBAQAAAIIg/69uSEABAAAAAAAAAAAAAAAAAAAAAACPBkiEAAHzeIrHAAAAAElFTkSuQmCC###24x24:true###][Tracking image][https://github.com/notifications/beacon/AAEZ5X4KSNCFKKZFDRSS3LDTGG3HLA5CNFSM4X5ZXATKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOGBFAHJQ.gif]

[aaubry]

I'm going to merge this, but I made some changes that you can check here:

1. I moved the functionality to a new class, as this has noting to do with handling default values, which is the responsibility of DefaultValuesObjectGraphVisitor.
2. I updated the test to actually perform assertions. Otherwise the test was not testing anything.

[PANDATEEN]

What dou you mean by "Otherwise the test was not testing anything"?
It at least test the great courage of mine.

[aaubry]

I'm sorry, I didn't mean to sound dismissive. What I mean is that the test was not making any assertion. This means that if a later change breaks the functionality, we won't find out unless we remember to check the output of the test.
I added assertions that check that the YAML at least contains the comments. This way the build will fail if someone breaks this feature.

[aaubry]

This feature has just been released, in YamlDotNet 10.1.0.

[ahannani-rms]

@aaubry - I just have a follow up question about this merge: is this change what may have caused a broken reference to be added to Yamldotnet's index.json in https://www.nuget.org?

We saw reference resolution issues during dotnet restore that stemmed from HTTP 404 on the following URI:
https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/page/8.0.0/10.0.1-pandateen-master-0006.json

[aaubry]

I'm not sure what you mean. What is index.json ?

[ahannani-rms]

The nuget manifest for your package in the public nuget repository: https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/index.json.

It now looks correct, but yesterday when we started seeing build failures shortly after this merge, it looked like the following:

{
	"@id": "https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/index.json",
	"@type": ["catalog:CatalogRoot", "PackageRegistration", "catalog:Permalink"],
	"commitId": "3ad1220b-820c-4fc6-a971-c28c12e43d70",
	"commitTimeStamp": "2021-03-30T12:32:22.6862928+00:00",
	"count": 4,
	"items": [{
		"@id": "https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/page/2.3.0-rc/3.9.0-pre239.json",
		"@type": "catalog:CatalogPage",
		"commitId": "a86247e0-dc5c-47a4-9c34-517c8086789a",
		"commitTimeStamp": "2021-02-02T12:14:33.9174985+00:00",
		"count": 64,
		"lower": "2.3.0-rc",
		"upper": "3.9.0-pre239"
	}, {
		"@id": "https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/page/3.9.0-pre240/4.1.1-pre0355.json",
		"@type": "catalog:CatalogPage",
		"commitId": "a86247e0-dc5c-47a4-9c34-517c8086789a",
		"commitTimeStamp": "2021-02-02T12:14:33.9174985+00:00",
		"count": 64,
		"lower": "3.9.0-pre240",
		"upper": "4.1.1-pre0355"
	}, {
		"@id": "https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/page/4.1.1-pre0356/8.0.0-pre0572.json",
		"@type": "catalog:CatalogPage",
		"commitId": "a86247e0-dc5c-47a4-9c34-517c8086789a",
		"commitTimeStamp": "2021-02-02T12:14:33.9174985+00:00",
		"count": 64,
		"lower": "4.1.1-pre0356",
		"upper": "8.0.0-pre0572"
	}, {
		"@id": "https://api.nuget.org/v3/registration5-gz-semver2/yamldotnet/page/8.0.0/10.0.1-pandateen-master-0006.json",
		"@type": "catalog:CatalogPage",
		"commitId": "3ad1220b-820c-4fc6-a971-c28c12e43d70",
		"commitTimeStamp": "2021-03-30T12:32:22.6862928+00:00",
		"count": 16,
		"lower": "8.0.0",
		"upper": "10.0.1-PANDATEEN-master-0006"
	}],
	"@context": {
		"@vocab": "http://schema.nuget.org/schema#",
		"catalog": "http://schema.nuget.org/catalog#",
		"xsd": "http://www.w3.org/2001/XMLSchema#",
		"items": {
			"@id": "catalog:item",
			"@container": "@set"
		},
		"commitTimeStamp": {
			"@id": "catalog:commitTimeStamp",
			"@type": "xsd:dateTime"
		},
		"commitId": {
			"@id": "catalog:commitId"
		},
		"count": {
			"@id": "catalog:count"
		},
		"parent": {
			"@id": "catalog:parent",
			"@type": "@id"
		},
		"tags": {
			"@id": "tag",
			"@container": "@set"
		},
		"reasons": {
			"@container": "@set"
		},
		"packageTargetFrameworks": {
			"@id": "packageTargetFramework",
			"@container": "@set"
		},
		"dependencyGroups": {
			"@id": "dependencyGroup",
			"@container": "@set"
		},
		"dependencies": {
			"@id": "dependency",
			"@container": "@set"
		},
		"packageContent": {
			"@type": "@id"
		},
		"published": {
			"@type": "xsd:dateTime"
		},
		"registration": {
			"@type": "@id"
		}
	}
}

Please note the 10.0.1-pandateen-master-0006 item.

[aaubry]

I don't know. For feature branches, the CI is configured to publish a pre-release package so that users can test the feature. This is what happened here, but I wouldn't expect this to cause problems.
Note that there seems to be a bug in nuget. If you happen to download the metadata while it is being updated, you get incomplete data in your cache and then you have to manually clear the cache.

[ahannani-rms]

Thanks @aaubry! Yeah, I actually ended up nuking the cache from our central artifactory as well as the local cache, to no avail. It kept getting refreshed with the same references from the public nuget repo, which leads me to believe the problem existed at the source (nuget.org).
However, this morning things got corrected.

[aaubry]

I'm glad that the problem went away. I will keep an eye on the feed to see if this happens again. Thanks

[ahannani-rms]

Much appreciated!