How could I understand the format of your data set and get the multi view graph matrix of java code

[tmsxk]

Hi, sorry if I bother you. I want to run your model using my own java data set, but I don't know how to get the same format of your data set from my own source code data set. I have download both your data file data.zip and the original tl-codesum dataset, but the data format between your data.zip and the original tl-codesum dataset is very different. Therefore, I want to know how you processed your java data set and got the format in data.zip. I would be very grateful if you can share your relevant scripts or codes. In addition, I also read your astruc.py code, but this is for the ast of the python code, I want to know how you get the mutil view graph and the corresponding matrix of your java code fragment. If you can share these codes or scripts, or provide some guidance on how to perform these operations, it can really help me a lot

[gingasan]

Sorry for our late.
There are little unified toolkits or platforms to process codes of various languages efficiently. Recently, we work on a more general tool in our leisure. You can refer to https://github.com/2001-SamLiu/Jsonl-Generator-for-AST. We haven't test the training performances of models when using this tool.