toolComponent change #1277

michaelcfanning · 2019-02-11T18:16:04Z

@lgolding can you sanity check the schema change for the new toolComponent object?

ghost · 2019-02-11T20:54:03Z

src/Sarif/Schemata/sarif-schema.json

+        "extensionIndex": {
+          "description": "The index within the run.tool.extensions array of the tool component object which describes the plug-in or tool extension that produced the result.",
+          "type": "integer",
+          "minimum": -1


minimum [](start = 11, length = 7)

and: "default": -1 #Resolved

Yes, did some clean up for minimum/default, please take a look at update.

In reply to: 255688316 [](ancestors = 255688316)

ghost · 2019-02-11T20:55:05Z

src/Sarif/Schemata/sarif-schema.json

+        },
+
+        "extensions": {
+          "description": "Tool extensions that contributed to or reconfigured the analysis tool run.",


run [](start = 96, length = 3)

"run" => "that was run" (for parallelism with description of "driver", and also I think it's clearer). #Closed

ghost · 2019-02-11T20:59:14Z

src/Sarif/Schemata/sarif-schema.json

      "properties": {

        "name": {
-          "description": "The name of the tool.",
+          "description": "The name of the tool driver or extension.",


driver or extension [](start = 47, length = 19)

How about "component" since that's what you use in the rest of the properties below. #Resolved

ghost · 2019-02-11T21:01:02Z

src/Sarif/Schemata/sarif-schema.json

+        }
+      },
+
+      "required": [ "driver" ]


"required": [ "driver" ] [](start = 6, length = 24)

Yes. #Closed

ghost · 2019-02-11T21:01:18Z

src/Sarif/Schemata/sarif-schema.json

+          }
+        },
+
+


Extra blank line. #Closed

ghost

ghost · 2019-02-11T21:02:25Z

Looks fine. I agree you should add result.toolComponentIndex, default -1 (meaning it came from the driver, not an additional component). #Closed

ghost · 2019-02-11T21:05:41Z

Michael, didn't we talk about linking from a toolComponent to a file object, so we could have information like file hashes? (So: toolComponent.fileIndex). Less important, but I still would like it: an additional role value of toolComponent. #Closed

michaelcfanning · 2019-02-11T21:11:10Z

i put this in resultProvenance.extensionIndex. You don't populate this at all in cases where it came from the driver. Seems simpler for happy path production. ?

In reply to: 462492896 [](ancestors = 462492896)

michaelcfanning · 2019-02-11T21:11:32Z

yes! on it. ok, role, yes.

In reply to: 462493928 [](ancestors = 462493928)

ghost · 2019-02-11T21:16:00Z

Re: The default value: This is one of those cases where we need the default to be -1 for the sake of the serializer, because 0 is a valid value. If 0 were the default, the serializer would not emit 0, and it would like like it was absent. The serializer will refrain from emitting -1, so it will indeed be absent.
Re: location of extensionIndex: I think it belongs in result. Otherwise, if your tool has rule plug-ins, you have to instantiate a provenance object just to hold the extension index -- even if you would otherwise not have had to bother.

In reply to: 462495693 [](ancestors = 462495693,462492896)

ghost · 2019-02-11T21:17:18Z

src/Sarif/Schemata/sarif-schema.json

@@ -518,7 +520,7 @@
        "uriBaseId": {
          "description": "A string which indirectly specifies the absolute URI with respect to which a relative URI in the \"uri\" property is interpreted.",
          "type": "string"
-        },
+        },`


` [](start = 10, length = 1)

typo #Closed

michaelcfanning · 2019-02-11T21:18:27Z

the default is -1, do you see a problem i need to fix?
yes, you need to instantiate the provenance object, just as you have to if you specify invocationIndex. Can you explain why extensionIndex would be on result but invocationIndex would not? in general, we introduced the provenance object to provide option, unusual origin data for results. associating a specific result with a custom plug-in qualifies, in my opinion.

do you have a different set of standards for what belongs in provenance?

In reply to: 462497387 [](ancestors = 462497387,462495693,462492896)

ghost · 2019-02-11T21:21:10Z

No problem on the default. I misunderstood you.

In reply to: 462498159 [](ancestors = 462498159,462497387,462495693,462492896)

ghost · 2019-02-12T19:19:39Z

src/Sarif/Core/Tool.cs

        {
-            return this.NotificationsMetadata.HasAtLeastOneNonNullValue();
+            var toolComponent = new ToolComponent();


ToolComponent [](start = 36, length = 13)

Can you create a property bag directly? Do you need a toolComponent to hold it? #Closed

It is cumbersome to instantiate IDictionary<string, SerializedPropertyInfo> directly (I do it elsewhere in this change). This is the easier mechanism because you can use the SetProperty helper. Do you have other advice on creation? This specific changes retains the previous code pattern...

In reply to: 256107576 [](ancestors = 256107576)

ghost · 2019-02-12T19:21:15Z

src/Sarif/Visitors/InsertOptionalDataVisitor.cs

@@ -177,7 +177,7 @@ public override Message VisitMessage(Message node)
                }

                if (node.Text == null &&
-                    _run.Tool?.GlobalMessageStrings?.TryGetValue(node.MessageId, out formatString) == true)
+                    _run.Tool.Driver.GlobalMessageStrings?.TryGetValue(node.MessageId, out formatString) == true)


. [](start = 29, length = 1)

Why did you remove the null conditional operator? #Closed

check line 168, where it didn't exist. Tool.Driver is a required object. we could put an arg check at the top of the method

In reply to: 256108174 [](ancestors = 256108174)

ghost · 2019-02-12T19:26:25Z

src/Sarif.Converters/SemmleQlConverter.cs

@@ -36,6 +36,8 @@ private enum FieldIndex
        private CsvParser _parser;
        private List<Notification> _toolNotifications;

+        public override string ToolName { get { return "Semmle QL"; } }
+


In all the other converters you just wrote

public override string ToolName => "Semmle QL"; ``` #Closed

ghost · 2019-02-12T19:28:06Z

src/Sarif.Converters/CppCheckConverter.cs

@@ -20,6 +20,8 @@ public CppCheckConverter()
            _strings = new CppCheckStrings(_nameTable);
        }

+        public override string ToolName => ToolFormat.CppCheck;


CppCheck [](start = 54, length = 8)

I know we discussed, and I think we agreed, not to assume that the ToolFormat would also make a good tool name. In most of the other converters, you're just hard-coding a string here ("CppCheck"). #WontFix

in fact, most converters still use ToolFormat. I'm not going to invest more here, I spent hours on this change DRYing out test code, removing embedded literals. We can revisit this in the future if need be.

In reply to: 256111009 [](ancestors = 256111009)

ghost · 2019-02-12T19:29:37Z

src/Sarif.Converters.UnitTests/ConverterTestsBase.cs

@@ -43,5 +44,36 @@ public SarifLog RunTestCase(string inputData, string expectedResult, bool pretty

            return log;
        }
+
+        private static SarifLog BuildToolSpecificEmptyLog()


Nice. #Closed

ghost · 2019-02-12T19:31:47Z

src/Sarif.FunctionalTests/PropertyBagConverterTests.cs

+            int intPropertyValue = 42;
+
+            string stringPropertyName = nameof(stringPropertyName);
+            string stringPropetyValue = "'\"\\'";


stringPropetyValue [](start = 19, length = 18)

Propety => Property #Closed

ghost · 2019-02-12T19:44:00Z

src/Sarif.UnitTests/Readers/IsSuppressedInSourceConverterTests.cs

+        {            
+            string expected = CreateCurrentV2SarifLogText(
+                resultCount: 1,
+                (log) => {


log [](start = 17, length = 3)

I absolutely love this pattern. #Closed

ghost · 2019-02-12T19:45:03Z

...estData/PrereleaseCompatibilityTransformer/ExpectedOutputs/ComprehensiveToolProperties.sarif

+          "properties": {
+            "tags": [
+  "singleTag"
+],


formatting. #Closed

ghost · 2019-02-12T19:47:20Z

...itTests/TestData/PrereleaseCompatibilityTransformer/Inputs/ComprehensiveToolProperties.sarif

+          "tags": [ "singleTag" ],
+          "propertyKey": "propertyValue"
+        },
+        "globalMessageStrings": {


globalMessageStrings [](start = 9, length = 20)

In 11-28, these three properties didn't yet live in tool, right? Global message strings and rules metadata lived in run.resources, and notification metadata didn't yet exist. #Resolved

ghost · 2019-02-12T19:55:04Z

src/Sarif.UnitTests/Writers/ResultLogJsonWriterTests.cs

+                    {
+                        new Invocation
+                        {
+                            ConfigurationNotifications = new List<Notification>(s_notifications)


s_notifications [](start = 80, length = 15)

Hah. Really nice. #Closed

All the pieces were here. :)

In reply to: 256121795 [](ancestors = 256121795)

ghost · 2019-02-12T21:08:11Z

        string version = forceUpdate ? "2.0.0" : (string)sarifLog["version"];

Why is this no longer needed? If it's not needed, the comment is obsolete. #Pending

Refers to: src/Sarif/Writers/PrereleaseCompatibilityTransformer.cs:42 in 2dc4f02. [](commit_id = 2dc4f02, deletion_comment = True)

michaelcfanning · 2019-02-12T21:16:47Z

        string version = forceUpdate ? "2.0.0" : (string)sarifLog["version"];

This force argument always existed because some tests lie about their version: i.e., the profess to have a format that they don't. I cleaned that up and removed this arg. I'll remove the dead comment in the next PR. also, please note: the modified log arg can go away now. at this point, any transformation of a prerelease format entails a change (because a required property, tool, has been updated).

In reply to: 462935363 [](ancestors = 462935363)

Refers to: src/Sarif/Writers/PrereleaseCompatibilityTransformer.cs:42 in 2dc4f02. [](commit_id = 2dc4f02, deletion_comment = True)

mikefan and others added 2 commits February 11, 2019 10:13

Preliminary schema change.

fe6860a

Schema and code gen hint updates

6214334

ghost reviewed Feb 11, 2019

View reviewed changes

More schema updates

ddd7bd3

ghost reviewed Feb 11, 2019

View reviewed changes

src/Sarif/Schemata/sarif-schema.json Outdated

}

},

Copy link

ghost Feb 11, 2019 •

edited by ghost

Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Extra blank line. #Closed

ghost approved these changes Feb 11, 2019

View reviewed changes

PR feedback on schema

18a2d50

ghost reviewed Feb 11, 2019

View reviewed changes

michaelcfanning and others added 4 commits February 11, 2019 13:38

More schema PR feedback

110acb1

WIP. Code now builds. Ready to author prerelease transform.

4738f8e

Progress restoring unit tests

7cd8e46

Restore all tests to working order.

2dc4f02

michaelcfanning changed the title ~~Preliminary schema change for toolComponent change~~ toolComponent change Feb 12, 2019

ghost reviewed Feb 12, 2019

View reviewed changes

Respond to PR feedback.

b739595

michaelcfanning merged commit 1b04ab5 into master Feb 12, 2019

michaelcfanning deleted the tool-components branch February 12, 2019 21:11

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

toolComponent change #1277

toolComponent change #1277

michaelcfanning commented Feb 11, 2019

ghost Feb 11, 2019 •

edited by michaelcfanning

Loading

michaelcfanning Feb 11, 2019

ghost Feb 11, 2019 •

edited by ghost

Loading

ghost Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost Feb 11, 2019 •

edited by ghost

Loading

ghost Feb 11, 2019 •

edited by ghost

Loading

ghost left a comment

ghost commented Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost commented Feb 11, 2019 •

edited by michaelcfanning

Loading

michaelcfanning commented Feb 11, 2019

michaelcfanning commented Feb 11, 2019

ghost commented Feb 11, 2019

ghost Feb 11, 2019 •

edited by michaelcfanning

Loading

michaelcfanning commented Feb 11, 2019

ghost commented Feb 11, 2019

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

michaelcfanning Feb 12, 2019

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

michaelcfanning Feb 12, 2019

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

michaelcfanning Feb 12, 2019

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by ghost

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by ghost

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by ghost

Loading

michaelcfanning Feb 12, 2019

ghost commented Feb 12, 2019 •

edited by michaelcfanning

Loading

michaelcfanning commented Feb 12, 2019

toolComponent change #1277

toolComponent change #1277

Conversation

michaelcfanning commented Feb 11, 2019

ghost Feb 11, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

michaelcfanning Feb 11, 2019

Choose a reason for hiding this comment

ghost Feb 11, 2019 • edited by ghost Loading

Choose a reason for hiding this comment

ghost Feb 11, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

ghost Feb 11, 2019 • edited by ghost Loading

Choose a reason for hiding this comment

ghost Feb 11, 2019 • edited by ghost Loading

Choose a reason for hiding this comment

ghost left a comment

Choose a reason for hiding this comment

ghost commented Feb 11, 2019 • edited by michaelcfanning Loading

ghost commented Feb 11, 2019 • edited by michaelcfanning Loading

michaelcfanning commented Feb 11, 2019

michaelcfanning commented Feb 11, 2019

ghost commented Feb 11, 2019

ghost Feb 11, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

michaelcfanning commented Feb 11, 2019

ghost commented Feb 11, 2019

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

michaelcfanning Feb 12, 2019

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

michaelcfanning Feb 12, 2019

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

michaelcfanning Feb 12, 2019

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by ghost Loading

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by ghost Loading

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by michaelcfanning Loading

Choose a reason for hiding this comment

ghost Feb 12, 2019 • edited by ghost Loading

Choose a reason for hiding this comment

michaelcfanning Feb 12, 2019

Choose a reason for hiding this comment

ghost commented Feb 12, 2019 • edited by michaelcfanning Loading

michaelcfanning commented Feb 12, 2019

ghost Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost Feb 11, 2019 •

edited by ghost

Loading

ghost Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost Feb 11, 2019 •

edited by ghost

Loading

ghost Feb 11, 2019 •

edited by ghost

Loading

ghost commented Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost commented Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost Feb 11, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by ghost

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by ghost

Loading

ghost Feb 12, 2019 •

edited by michaelcfanning

Loading

ghost Feb 12, 2019 •

edited by ghost

Loading

ghost commented Feb 12, 2019 •

edited by michaelcfanning

Loading