fix(operation): report actual file total_lines instead of content slice line count

[tusharmath]

Summary

Fix total_lines attribute in file read output to report the actual total line count of the file rather than the number of lines in the returned content slice.

Context

When reading a file with an explicit line range or when truncation is applied, the total_lines XML attribute was computed as content.lines().count() — i.e., the number of lines in the returned content rather than the total number of lines in the full file. This caused the attribute to be misleading: an agent reading lines 2–3 of a 5-line file would see total_lines="3" instead of total_lines="5", and a truncated read of a 200-line file would report total_lines="1".

The FileInfo struct already carries the correct total_lines value populated during the read operation, so the fix is straightforward.

Changes

• Replaced content.lines().count() with output.info.total_lines when emitting the total_lines XML attribute in operation.rs
• Updated two insta snapshots to reflect the corrected values:
  • fs_read_with_explicit_range: total_lines now correctly shows 5 (full file) instead of 3 (lines in the range)
  • fs_read_with_truncation_path: total_lines now correctly shows 200 (full file) instead of 1 (lines in the truncated output)

Testing

cargo insta test --accept -p forge_app

Both snapshot tests now assert the correct total_lines value that reflects the actual file size.