Regular Bytecode Expression (regbex)

⚠️ This library should only be used in Kotlin - at least for Regbex builders.

⚠️ This repository is in experimental state! If you encounter any issue, please open a GitHub issue or pull request!

A regular expression engine that's made for Java bytecode and objectweb ASM. It's designed to be easy to use

Missing Features / Known Issues

Look-Behind & Look-After (Won't add for now) (Can be replaced with thenCheckWithoutMovingPointer, but negative look-around would be impossible. Will add that later)

THE REPLACE FEATURE IS ALSO BUGGED! DON'T USE IT

Main Features

Fast - Tested in Minecraft, runs smoothly, and it usually needs to load a lot of classes while initializing the game, but it's still as smooth as it should be. (BUT Keep in mind that at the end of the day, it's still regex, if you are using a complicated pattern, we can't promise the efficiency of that)
Easy to use - Just like regex, it's easy to use - probably easier, because you don't need to remember all those crazy symbols anymore
Replacing - You can replace instructions like what you could do in regex, and you could also do Bugged, and will be fixed later
Capturing - Capture groups, and process them programmatically.

Using

Getting Started

Use jitpack:

Gradle (Kotlin)

repositories {
    ...
    maven { url ("https://jitpack.io") }
}

dependencies {
    implementation("com.github.fan87:Regular-Bytecode-Expression:<version>")
}

Gradle (Groovy)

allprojects {
    repositories {
        ...
        maven { url 'https://jitpack.io' }
    }
}
dependencies {
    implementation 'com.github.fan87:Regular-Bytecode-Expression:<version>'
}

Maven

<repositories>
    <repository>
        <id>jitpack.io</id>
        <url>https://jitpack.io</url>
    </repository>
</repositories>
<dependency>
    <groupId>com.github.fan87</groupId>
    <artifactId>Regular-Bytecode-Expression</artifactId>
    <version><version></version>
</dependency>

You could always find all available versions in GitHub release tab

Use Case

Let's say you want to do string obfuscation. Normally you would have to have a main for loop, a few if checks. But not anymore! With this library, you can match every LdcString, and replace them to obfuscated string.

Not only that, let's say you want to match a System.out.println("" /*Ldc String*/"), normally you have to save the current instruction index, have a main for loop, a lot of if check, but now you could just match getstatic, ldcstring, and method call, capture the ldc string, and that's where and what the string is, no for loop, only a few lines of code

Usage

You can check the Example Test Class for code example and usage.

Debugging

Struggling to debug what's wrong with your regbex? You can enable debug mode by setting matcher.debug to true (Or in java, matcher.setDebug(true)). It will print out all regbex elements, current regbex index, all instructions, current instructions index, and most importantly, the fail reason.

Using it in Java

If your codebase is in Java and you don't want to re-code your entire project, you can have classes in Kotlin that pre-defines every regbex you'll be using.

Attempting to build regbex pattern in Java is still technically possible, but it will make your code un-readable, and you shouldn't do that.

Efficiency

Just like regular expression, it's not designed to be high efficient, but to make your life easier.

In theory, it should be as fast as regex engines, bcs what I'm doing is (probably) how most regex engines are implemented.

Confused?

Please read the KDoc (Like JavaDoc, but for Kotlin) if any function is confusing you. For example, using TypeExpression class to match a type using regular expression could be confusing because you must assume the input could have L in the front and ; in the end of the class name, so if you want to match java.lang, you have to do L?java/lang/.*;?

FAQ

How to ignore let's say all Label Node?

Not Supported Yet

What is lazy

Lazy in regex is ?, for example, .*?, it means that it will match as few as possible.

Is it thread-safe?

About the matcher, no, not at all. Please create multiple matcher instances to match instructions in a different thread, even if they are the same.

How do I replace matched result with a group without crafting them myself?

You could replace them to ReplacePlaceHolder("groupName"), and the engine will convert that to the group content.

For more information, please check KDoc.

Development / Code Quality

About Testing

I'm not aiming to have coverage 100%, it's absolute time-wasting to me to write that many tests when I have confident that the code will work. So you may find the test quality to be a little worse than others.

fan87 / Regular-Bytecode-Expression