Wrap file contents in XML tags

[Rolland-He]

• helpers/template_utils.py:
  === {filename} === -> <submission file="correct_submission.py">

Example:

Compare the student's code and solution code...

The student's submission file is correct_submission.py.
The instructor's solution file is solution.py.

Files to Reference:
<submission file="correct_submission.py">
(Line 1) def fizzbuzz(n: int) -> list:
...
</submission>

<solution file="solution.py">
(Line 1) def fizzbuzz(n: int) -> list:
...
</solution>

test_output is also included here.

• integration_test.py: Update tests.

I have also cleaned the commit history.

[wkukka1]

Great job, @Rolland-He! I’ve added a couple of minor suggestions around code style.

[wkukka1]

Duplicated logic for wrapping lines with XML tags, both for PDFs and regular text files. Consider refactoring the repeated line formatting into a helper function—e.g., _wrap_lines_with_xml(lines, tag_name, filename)

[Rolland-He]

It was actually part of the original code, I’ve moved it into a helper function already. I’ll work on simplifying the logic further. Thanks for pointing that out!

[wkukka1]

should be Optional[Path] for all these args

[Rolland-He]

Fixed.

[wkukka1]

Looks Good! @Rolland-He

[david-yz-liu]

Use the attribute name filename instead of file, here and throughout all XML tags that have to do with files

[Rolland-He]

Fixed.

[david-yz-liu]

Okay looking at this more carefully, I'm realizing that actually for text extracted from PDFs I don't think we need to do the line numbering (which was really to support precise annotations for code). I'm not sure we need to call this function at all when the given text is from a PDF, and instead we can just pass the raw text directly as the file contents to the LLM.

[Rolland-He]

Fixed!