Develop Reference

Obfuscating or removing string literal from all calls to Log-function in Android

We are building an Android application where we use Timber for Log-output. We have defined our own .e, .d, .v etc functions and use if (BuildConfig.DEBUG) to see if we should output the log. This takes care of the issue that we don't want to output Debug-logs in our releases but all the string literals used in our functions calls are still present in the compiled source code. We furthermore use ProGuard for obfuscation. To exemplify, in a class we can have:
somObj.normalFunction(variable)
Log.d("This secret class achieved its secret mission!");
In our release, this will not be seen in the app logs but if you reverse-engineer the APK you will see something like:
q.b(m)
z.a("This secret class achieved its secret mission!");
which can give a hint to the hackers about what class they are looking at.
So what we're looking for is to either be able to completely REMOVE all the Log function calls at compile time (using some pre-processing, annotation or something, but hopefully without having to add something before EVERY function call) OR to obfuscate all the String literal parameters to those function calls. So, two ideal solutions would be if the source, just before compilation, instead looks like:
q.b(m);
or
q.b(m);
z.a("jgasoisamgp23mmwaföfm,ak,ä")
Just by thinking I can see two bad ways to achieve this. Either we surround ALL calls to Log.d with if(BuildConfig.DEBUG) which will make the compiler remove them before compilation. But this is very tideous. OR, we make sure that every time you want to add a log-printout you need to do
Log.d(LogClass.getLog(1234))
and you then define ALL those logs inside LogClass and then remove them with if(BuildConfig.DEBUG) and return null in getLog if that's the case. But that makes it more tideous every time you want to add a log.
So finally, is there any GOOD solution to this?

DISCLAIMER: I work for PreEmptive, the company that makes PreEmptive Protection - DashO.
DashO is capable of removing calls to specific methods (e.g., Log methods). While this doesn't remove the instructions to load the string literal, just the call itself, DashO also offers String Encryption, which would offer some protection to those string literals.
As an example, I ran this class through DashO:
public class App {
public static void main(String[] args) {
Log.d("Secret message here");
}
}
After removing calls to Log.d with String Encryption on, the decompiled output looks like this:
public class App
{
public static void main(String[] paramArrayOfString)
{
a.replace("Bwpfpb7u|ih}z{?($0&", -12 - -61);
}
}
DashO offers other protections (e.g., Control Flow Obfuscation) that tend to break decompilers; I've turned those off for this demonstration.

What I would do is one or some of the following:
Use Timber (so you don't need to bother removing things or adding if statements). You simply do this once in your Application#onCreate(); if you are in DEBUG, then you plant a DebugTree that prints to the console. Else, you plant an "empty tree" that does nothing.
Simulate Timber but create your own "YourLogger" class and do the same (if you don't want to include a "third-party" library even though it's just one class). So you'd have YourLogger.v("tag", "string") and inside you'd do: if (debug) { Log.v(tag, string); } and so on and so forth for all other log types.
Use Proguard to strip the logging, and what not.
1 and 2 imply you go through your app and replace all Log. lines with either Timber. or YourLogger.
Option 3 wouldn't need that, since the code would be removed during obfuscation and the calls would do nothing, but this is mode complicated (and I haven't done it in years, don't even remember how it's done, probably easy to look it up).
I'd go for 1.
Update
Since apparently I don't know how to read, I think that in order to achieve this, your only bet is to have some sort of lookup mechanism for the actual text you want to emit/hide.
Strings.xml is the easiest, since it's included with Android and you can even localize the error messages (if that were needed, of course). Yes, there's a lookup time penalty, but I'd say unless you're iterating over thousands of items and logging different strings every time or something, the penalty wont' be noticeable (I don't have data for this, you'd have to benchmark).
Alternatively, instead of relying on resources, you could just use a file, read it, and load the strings in memory... a trade off; do you use more memory at the cost of simplicity and time to code the solution, or do you use the built-in mechanism and pay the CPU time?

Proguard deletes method parameters

Recently I found very strange thing with ProGuard. I have this code snippet
As you can see, method showTipHoodLock takes 2 parameters, fragmentManager and top (some offset)
but after I compile the app with minifyEnabled true
I got this on click callback
and this is Utils.showTipHoodLock method
As you can see, proguard deleted 2nd parameter (named top) from method signature and replaced its occurances with 0 literal.
Is it a bug, or a feature, or did I miss something?
P.S. If I change values in line
int coords[] = {0, 0}
to any other numbers, then everything works perfect, and nothing is deleted. Moreover The same snippet of code (which is copy-pasted) in different part of application (in other fragment), starts to work.

Is it a bug, or a feature, or did I miss something?
This is NOT a bug, this is a feature to optimise your code.
According to your piece of code, the second parameter is referenced by following logic as READ ONLY and its value is FIXED to be 0.
Proguard will remove (a kind of Proguard optimisation: Remove unused parameters or Propagate constant arguments) this parameter with this KNOWN FIXED value to simplify the invocation flow.

Is there a way to check for unnecessary #SuppressWarnings annotations?

I've recently begun using annotions to help reduce my lint warning count to filter out valid use-cases of warnings/errors.
As an example, let's assume I am 100% correct in having a variable that is never assigned a value (it's being done via GSON serialization but never x = 2 type assignment). Lint reports this as unused variable. I can add the #SuppressWarnings("unused") annotation and lint skips this.
Sometime in the future, would I be able to do an assignment to the x variable, and then have lint notify me that the #SuppressWarnings("unused") flag is no longer necessary? If not, I'd potentially have unneeded code in my application which I would like to avoid.
Any thoughts?

Keep sections are being removed on generation

I am using GreenDAO v1.2.0 to generate entity classes for an Android app.
I have set enableKeepSectionsByDefault() on my schema, and am getting these comment lines in my generated classes:
// KEEP METHODS - put your custom methods here
// KEEP METHODS END
I have put some custom methods between the comments, but the custom methods are removed every time I regenerate.
Is there something else I need to do to get this to work?

There is a parsing bug in greenDAO that will cause some or all of your KEEP METHODS to get wiped out by running the DaoGenerator.generateAll() method (even with schema.enableKeepSectionsByDefault() enabled)
The bug is that your Entity class (the one you have added custom imports, fields or methods to) REQUIRES a blank line at the end of the file.
for example
...
//KEEP_METHODS_END
}
... end of file
will fail, while
...
//KEEP_METHODS_END
}
... end of file
will work. Nasty little bug. Until it's fixed (if ever), ensure that your entity classes have a blank line at the end.
Hope this saves someone the torment I have experienced.
Ensure that the last line of the generated entity class is blank.

You need to call the schema.enableKeepSectionsByDefault();.

better way to do Debug only assert code

I am writing my first Android application and I am liberally using asserts() from junit.framework.Assert
I would like to find a way to ensure that the asserts are only compiled into the debug build, not in the release build.
I know how to query the android:debuggable attribute from the manifest so I could create a variable and accomplish this in the following fashon:
static final boolean mDebug = ...
if (mDebug)
Assert.assertNotNull(view);
Is there a better way to do this? i.e. I would prefer not to use an if() with each assert.
thanks

I think the Java language's assert keyword is likely what you want. Under the covers, the assert keyword essentially compiles into Dalvik byte code that does two things:
Checks whether the static variable assertionsDisabled (set in the class' static constructor via a call to java.lang.Class.desiredAssertionStatus()) is != 0 and if so, does nothing
If it is 0, then it checks the assertion expression and throws a java.lang.AssertionError if the expression resolves to false, effectively terminating your application.
The Dalvik runtime by default has assertions turned off, and therefore desiredAssertionStatus always returns 1 (or more precisely, some non-zero value). This is akin to running in "retail" mode. In order to turn on "debug" mode, you can run the following command against the emulator or the device:
adb shell setprop debug.assert 1
and this should do the trick (should work on the emulator or any rooted debugging-ready device).
Note however that the aforementioned Dalvik code that checks the value of assertionsDisabled and throws an AssertionError if the expression is false is always included in your byte code and liberal sprinkling of asserts in your code may lead to byte code bloat.
Please see this for a bit more detail: Can I use assert on Android devices?

If you're concerned about shipping code with the JUnit asserts in (or any other class path), you can use the ProGuard config option 'assumenosideeffects', which will strip out a class path on the assumption that removing it does nothing to the code.
Eg.
-assumenosideeffects class junit.framework.Assert {
*;
}
I have a common debug library I put all my testing methods in, and then use this option to strip it from my released apps.
This also removes the hard to spot problem of strings being manipulated that are never used in release code. For example if you write a debug log method, and in that method you check for debug mode before logging the string, you are still constructing the string, allocating memory, calling the method, but then opting to do nothing. Stripping the class out then removes the calls entirely, meaning as long as your string is constructed inside the method call, it goes away as well.
Make sure it is genuinely safe to just strip the lines out however, as it is done with no checking on ProGuard's part. Removing any void returning method will be fine, however if you are taking any return values from whatever you are removing, make sure you aren't using them for actual operational logic.

I mean if you were using a language feature, like assert(), the compiler should be able to strip that out. But this is an actual class and if a class is referenced by executable code it will be included or assumed included in the final product by the compiler.
However there is nothing stopping you from creating a script that removes all the references to the Assert class in all of your code base before compilation.
Another approach would be to make a test project that targets your application and within JUnit tests actually calls the Assert on the areas which you want to make sure work. I personally like this approach because it is a nice and clean separation of test and application.
If you are just worried about the having an if-statement everywhere, then just wrap Assert with your own class, DebuggableAssert which does that check before each call to Assert.X. It will be sort of less performant because of the method entry/exit and the conditionals but if you can maintain your code better then it might be worth it.