[Bug] `--language <value>` option in `joern-parse` is useless.

[jiezhuzzz]

Describe the bug

--language <value> option in joern-parse is useless. joern-parse will stick to file extension when parsing files.

To Reproduce Steps to reproduce the behavior:

1. create a file code.c

   #include<stdio.h>
   int main() {
   int a = 1;
   printf("a = %d", a);
   }

2. run joern-parse --language newc code.c and joern-export --repr all -o withc
3. rename the code file as mv code.c code.java. Export graph as before: joern-parse --language newc code.java, and joern-export --repr all -o withjava
4. you will find two graph dot file are totally different

Expected behavior joern-parse will ignore the file extension and use --language to determine which frontend to be invoked.

Desktop (please complete the following information):

• OS: Fedora Linux 40
• Joern Version: HEAD+20240318-1648
• Java version: 21.0.3

[max-leuthaeuser]

It works as expected. If you try to joern-parse a .java file with --language newc you will end up with a mostly empty CPG because c2cpg wont touch any .java file. Thats why your dot files are different.

The --language flag itself selects the frontend to run but that frontend will only handle files related to the language.

[jiezhuzzz]

Understood, thanks for your reply! I thought that could be a bug rather than a default behavior.