Currently diffs are done parsing the text provided by the AI and getting the String of the parsed text; this may differ from the original text.
For example:
Original class:
package io.github.jeddict.ai.agent;
public class HelloWorld {
public void main(String[] args) {
System.out.println("Hello world");
}
}
Ask AI to reformat it in a single line. Answer from AI:
package io.github.jeddict.ai.agent; public class HelloWorld { public static void main(String[] args) { System.out.println("Hello world"); } }
Then right click on the text and select "Diff HelloWorld":
Somehow more importantly, the code produced for the diff does not contain other relevant text like package declaration, comments, copyright header etc.
Currently diffs are done parsing the text provided by the AI and getting the String of the parsed text; this may differ from the original text.
For example:
Original class:
Ask AI to reformat it in a single line. Answer from AI:
Then right click on the text and select "Diff HelloWorld":
Somehow more importantly, the code produced for the diff does not contain other relevant text like package declaration, comments, copyright header etc.