Separate Multiple Frames With a Newline

[alonalbert]

When an obfuscated frame resolves to multiple clear frames, separate them with a newline.

See #432

[mrjameshamilton]

Thanks for the PR @alonalbert . It looks like the issue was introduced in #200: there were two places where println was previously used which were replaced by append; therefore a new line no longer gets added when building up the result string.

[mrjameshamilton]

Based on the change in #200: it looks like this should be like the following to revert to the previous behaviour:

Suggested change

	result.append(trimmedLine);
	if (retracedFrames.hasNext()) {
	result.append("\n");
	}
	result.append(trimmedLine);
	result.append(System.lineSeparator());

[mrjameshamilton]

Line 245 also had a similar change, which no longer adds a newline.

[alonalbert]

I like the use of System.lineSeparator() instead of "\n" but if I accept your suggesion, we would end up with a newline separating each batch of frames. For example:

09-10 17:57:52.303  9066  9066 D Foobar  : 	at S0.a.e(SourceFile:30)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at H.a.k(SourceFile:160)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at m.n.m(SourceFile:132)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at Z0.a.d(SourceFile:9)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at o1.v.n(SourceFile:81)

Will become:

09-10 17:57:52.303  9066  9066 D Foobar  : 	at com.example.myapplication.Foo.foo(Foo.java:7)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at com.example.myapplication.MainActivity.onClick(MainActivity.java:38)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at com.example.myapplication.MainActivity.Greeting$lambda$1$lambda$0(MainActivity.java:43)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at S0.MainActivity$$ExternalSyntheticLambda0.invoke(MainActivity.java:0)

09-10 17:57:52.303  9066  9066 D Foobar  : 	at androidx.compose.foundation.ClickablePointerInputNode$pointerInput$3.invoke-k-4lQ0M(ClickablePointerInputNode.java:987)
09-10 17:57:52.303  9066  9066 D Foobar  : 	at androidx.compose.foundation.ClickablePointerInputNode$pointerInput$3.invoke(ClickablePointerInputNode.java:981)

09-10 17:57:52.303  9066  9066 D Foobar  : 	at androidx.compose.foundation.gestures.TapGestureDetectorKt$detectTapAndPress$2$1.invokeSuspend(TapGestureDetectorKt.java:255)

09-10 17:57:52.303  9066  9066 D Foobar  : 	at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(BaseContinuationImpl.java:33)

09-10 17:57:52.303  9066  9066 D Foobar  : 	at kotlinx.coroutines.DispatchedTaskKt.resume(DispatchedTaskKt.java:179)

We need to either use an if (retracedFrames.hasNext()) as in my PR or maybe better, to change line 182 to use print instead of println:

            stackTraceWriter.print(deobf);

I'm not sure what you mean about line 25. It seems to work without changing it.

I'm updating the PR.

[mrjameshamilton]

I'm not sure what you mean about line 25. It seems to work without changing it.

I meant the line here:

            // Print out the original line.
            result.append(obfuscatedLine);

which used to add a newline before #200 :

                // Print out the original line.
                stackTraceWriter.println(obfuscatedLine);

With the other changes, I'm not sure this is required to be changed though.

[mrjameshamilton]

With your latest change, if I take the following simple stacktrace and mapping, everything is printed on one line, so your original suggestion of adding the new line only if (retracedFrames.hasNext()) seems to be correct.

Caused by: java.lang.RuntimeException: AAA
	at com.example.demo.sub.TestComponent2.a(Unknown Source)
	at com.example.demo.DemoApplication.main(Unknown Source)

com.example.demo.DemoApplication -> com.example.demo.DemoApplication:
    void <init>() -> <init>
    void main(java.lang.String[]) -> main
com.example.demo.TestComponent -> com.example.demo.TestComponent:
    void <init>() -> <init>
    int getAnswer() -> a
com.example.demo.sub.TestComponent2 -> com.example.demo.sub.TestComponent2:
    void <init>() -> <init>
    java.lang.String getMessage() -> a

[alonalbert]

Hmm, for me, it works fine with the latest version.

I'm on a Linux machine. What system are you on?

[mkrichevsky]

@mrjameshamilton @alonalbert does current version look good to be merged or do we need to make more changes for this issue?

[sonarcloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud

[alonalbert]

I re-pushed the original version which I think is safer.

[mrjameshamilton]

I re-pushed the original version which I think is safer.

Thanks! This version seems to work well on my sample as well.

[mkrichevsky]

@mrjameshamilton what is the expected timeline when the fix can be released? Just to plan around approximate schedule. Thanks a lot for your efforts reviewing this.

[mrjameshamilton]

@mrjameshamilton what is the expected timeline when the fix can be released? Just to plan around approximate schedule. Thanks a lot for your efforts reviewing this.

I'll make a release within the next week, we need a release for Java 23 support anyway.