Fix MultithreadedAnalyzeCommandBase artifacts generation and enforcing JSON properties ordering

[eddynaka]

Description

The previous version of the sarif-sdk had a bug that was generating an artifact for each file that was analyzed. This update will guarantee that we will only generate an artifact if we have a result.

The sarif-sdk submodule change is also enforcing a specific ordering of the name/helpuri in the reportingDescriptor. This is expected since we want the name to be near the id.

Test

Checked all rebaselined files and only found two unexpected changes related to the artifacts. The reason we have those is that we are not producing any results, so the artifacts are getting dropped from the SARIF log.

[eddynaka]

artifacts

we will only emit an artifact if we have a result. This was part of the sarif-sdk change to improve the file size generated when utilizing the MultithreadedAnalyzeCommandBase class:
microsoft/sarif-sdk#2433

[michaelcfanning]

There should be an artifact entry for every artifact that is explicitly referenced. There's not a good reason to arbitrarily restrict this to results. We've lost important information here.

[eddynaka]

  ],

we will only emit an artifact if we have a result. This was part of the sarif-sdk change to improve the file size generated when utilizing the MultithreadedAnalyzeCommandBase class:
microsoft/sarif-sdk#2433

---

Refers to: src/Test.FunctionalTests.BinSkim.Driver/BaselineTestsData/Expected/gcc.stack_protector.a.sarif:52 in 64f53be. [](commit_id = 64f53be, deletion_comment = True)

[eddynaka]

name

with the sarif-sdk change, we are enforcing a specific order where name will be close to id and helpuri will be near the end.

Reviewed all files and only two has unexpected changes that are commented as well.

[michaelcfanning]

MultithreadedAnalyzeCommandBase

I'd document this isn't a BinSkim fix, it's an SDK update.

Update SARIF-SDK in order to fix `blah blah

[michaelcfanning]

  "artifacts": [

this is a bug, it shouldn't have gone away. Instead, the change should be that we see the addition of the artifact index.

---

Refers to: src/Test.FunctionalTests.BinSkim.Driver/BaselineTestsData/Expected/clang.stack_protector.a.sarif:41 in 3fa6d69. [](commit_id = 3fa6d69, deletion_comment = True)

[michaelcfanning]