feat: Add Disable pairip license check patch

[hoo-dles]

Just a simple, little patch to bypass the client-side validation of the Google Play Licensing response. This should work for any app that only does the client validation (if we want to move the code to a shared location.)

Fixes #4776

[Nuckyz]

I would also take a look at changing the daily limit patch to returnEarly(0) as the method returns an int, and calling returnEarly without a value seems kinda weird

[dtricks]

I would also take a look at changing the daily limit patch to returnEarly(0) as the method returns an int, and calling returnEarly without a value seems kinda weird

#4604 (comment)

I find returnEarly(0) more readable as well

[hoo-dles]

I would also take a look at changing the daily limit patch to returnEarly(0) as the method returns an int, and calling returnEarly without a value seems kinda weird

@oSumAtrIX We currently don't have a helper method for returnEarly() that accepts primitive types other than boolean. returnEarly(false) does not seem any more readable in this instance since the original method returns an integer. Do we want to extend the existing helpers, or just leave the code as is?

[oSumAtrIX]

What's a daily limit patch? Regarding the utility function, i guess the simplest is to pass the return operand as a string 👍

[hoo-dles]

What's a daily limit patch?

The daily limit patch is the existing ads patch. App checks to see how many measurements the user has taken today, and then blocks functionality until an ad is watched.

Regarding the utility function, i guess the simplest is to pass the return operand as a string 👍

The constant size is hardcoded to a 4-bit const/4, how do you want to handle that?

[LisoUseInAIKyrios]

Can add an overloaded function

fun MutableMethod.returnEarly(Boolean)

That calls into:
fun MutableMethod.returnEarly(String)

So the usage for boolean return methods is a bit more explicit

[LisoUseInAIKyrios]

The constant size is hardcoded to a 4-bit const/4, how do you want to handle that?

Why not replace the instruction with regular const? If the 4-bit limit is the issue you're speaking of.

[hoo-dles]

The constant size is hardcoded to a 4-bit const/4, how do you want to handle that?

Why not replace the instruction with regular const? If the 4-bit limit is the issue you're speaking of.

That's definitely a possibility. Although I don't think that would work for a string literal (where you'd use const-string), which is confusing if the returnEarly() function signature accepts a string. I figured I would ask the team considering this a widely used helper throughout the patches project.

[LisoUseInAIKyrios]

Currently returnEarly (and it's similar returnLate) don't work with strings but feel free to add it to the existing util methods.

Instead of just boolean and string overloads, the utils method could be overloaded with all primitive types plus String, as that gives better type safety since the parameter type would need to match the method being patched. Then for the implementation just modify the existing private util method and have all the overloadeds call onto it.

[hoo-dles]

Currently returnEarly (and it's similar returnLate) don't work with strings but feel free to add it to the existing util methods.

Instead of just boolean and string overloads, the utils method could be overloaded with all primitive types plus String, as that gives better type safety since the parameter type would need to match the method being patched. Then for the implementation just modify the existing private util method and have all the overloadeds call onto it.

Pushed new commits that move pairip patch to shared location, and added returnEarly() and returnLate() overloads for primitives.

I didn't do an overload for String because all the Smali code I've seen treats strings as the Java object. Wasn't sure if const-string accepts a variable size string literal (opcode list says string_id param), or if it plays nicely with a return type of Ljava/lang/String;.

[Brosssh]

Description should be improved in my opinion. Basically every pairip version is based on libpairipcore.so, and that cannot be disabled (not by this patch, at least). May be worth noting it in the patch description, this patch is only bypassing the smali license check, but if the app has native checks, it won't open anyway.

[hoo-dles]

I'll change it to whatever language someone suggests :P