bin2ml
bin2ml copied to clipboard
br0kej
A command line tool for extracting machine learning ready data from software binaries powered by Radare2
Is there a way to add a decompiling feature to the tool so that it leverages radare2 plugin to, say decompile functions. In addition to having the ESIL representation norm/non-norm...
There are currently limitations on what function data can be saved. For instance, I was ran `bin2ml extract -f 080fb32c3766e0595cc0b2019c4687db -o scanOSData/ -m bytes`. This command will result in the...
I propose implementing the extraction of [EMBER](https://arxiv.org/pdf/1804.04637) features, a widely-used benchmark originally designed for PE (Portable Executable) files. While some EMBER features are PE-specific, others are format-agnostic and could benefit...
Refactoring of function `bytes` and `bytes-masked` extraction workflow to prevent unnecessary memory usage.
In PR #62 it was mentioned: > In the future is may be worth looking into setup_function_list and check if we can avoid running the same command twice when finfo...
Previously, if bin2ml extraction of function-separated data files (e.g. bytes) was be cancelled/aborted half way, restarting it would mean extracting all function data from the beginning, including the files that...
There is an error log file created when not a single function was processed successfully in `bytes`, `bytes-masked`, and `func-cfg` modes. This error log file would be created even if...
FuncInfo extraction jobs would parse the output of `aflj` with JSON to then serialise it again within the `write_to_json` function. This PR takes the String output of `aflj` in extract_function_info...
There is a bug upstream in radare2. I registered [issue 24712 on radare2's repository](https://github.com/radareorg/radare2/issues/24712) documenting it in more detail. In few words, the command `p8fm` can return a byte mask...
Various changes improve the performance of bin2ml when extracting function-level information. These performance issues were barely noticeable on small files, but binaries with tens of thousands of functions would consume...