fix: Fix submission file filtering when '*' whitelist pattern is specified in the assignment

Author: natkam

grader_service/autograding/git_manager.py

@@ -109,6 +109,9 @@ def _commit_files(self, filenames: List[str], output_path: str) -> None:
             self.log.info("No files to commit.")
             return
 
+        # Make sure we do not commit the gradebook.json
+        filenames = [f for f in filenames if f != "gradebook.json"]
+
         self._run_git([self.git_executable, "add", "--", *filenames], output_path)
         self._run_git(
             [self.git_executable, "commit", "-m", self.submission.commit_hash], output_path

grader_service/convert/converters/base.py

@@ -236,18 +236,9 @@ def init_destination(self) -> bool:
                 return False
         return True
 
-    def copy_unmatched_files(self, gb: Gradebook):
-        """
-        Copy the files from source to the output directory that match the allowed file patterns,
-        excluding files and directories that match any of the ignore patterns.
-        :return: None
-        """
-        dst = self._output_directory
-        src = self._input_directory
-
+    def get_include_patterns(self, gb: Gradebook) -> List[str]:
+        """Get glob patterns specifying which submission files to copy."""
         allowed_files = self._assignment_settings.allowed_files
-        ignore_patterns = self.ignore  # List of patterns to ignore
-
         if allowed_files is None:
             self.log.info(
                 "No additional file patterns specified; only copying files included "
@@ -257,6 +248,19 @@ def copy_unmatched_files(self, gb: Gradebook):
         else:
             self.log.info(f"Found additional file patterns: {allowed_files}")
             files_patterns = allowed_files + gb.get_extra_files()
+        return files_patterns
+
+    def copy_unmatched_files(self, gb: Gradebook):
+        """
+        Copy the files from source to the output directory that match the allowed file patterns,
+        excluding files and directories that match any of the ignore patterns.
+        :return: None
+        """
+        dst = self._output_directory
+        src = self._input_directory
+
+        ignore_patterns = self.ignore  # List of patterns to ignore
+        files_patterns = self.get_include_patterns(gb)
 
         copied_files = []
 
@@ -285,9 +289,6 @@ def is_ignored(file_path):
             ]  # Modify dirs in-place to ignore unwanted dirs
 
             for file in files:
-                if file == "gradebook.json":
-                    continue
-
                 abs_file_path = os.path.join(root, file)
                 rel_file_path = os.path.relpath(abs_file_path, src)
 

grader_service/convert/converters/generate_feedback.py

@@ -10,7 +10,7 @@
 from grader_service.convert import utils
 from grader_service.convert.converters.base import BaseConverter
 from grader_service.convert.converters.baseapp import ConverterApp
-from grader_service.convert.gradebook.gradebook import MissingEntry
+from grader_service.convert.gradebook.gradebook import Gradebook
 from grader_service.convert.preprocessors import GetGrades
 
 
@@ -55,17 +55,21 @@ def __init__(
         self.update_config(c)
         self.force = True  # always overwrite generated assignments
 
-    def convert_single_notebook(self, notebook_filename: str) -> None:
-        """Generate feedback for a single notebook.
+    def init_notebooks(self) -> None:
+        # We only generate feedback for the notebooks for which there is a gradebook entry,
+        # and ignore e.g. additional notebooks created by the student, because feedback
+        # generation would fail for them anyway.
+        json_path = os.path.join(self._output_directory, "gradebook.json")
+        with Gradebook(json_path) as gb:
+            self.notebooks = [f"{nb_id}.ipynb" for nb_id in gb.model.notebook_id_set]
 
-        We ignore any notebooks for which there are no gradebook entries
-        (e.g. additional notebooks created by the student), because feedback
-        generation would fail for them anyway.
+    def get_include_patterns(self, gb: Gradebook) -> list[str]:
+        """Get glob patterns specifying for which submission files to generate feedback.
+
+        In case of feedback generation, it can only be done for the notebooks
+        which were included in the original assignment.
         """
-        try:
-            super().convert_single_notebook(notebook_filename)
-        except MissingEntry:
-            self.log.info("Skipping notebook %s", notebook_filename)
+        return self.notebooks
 
 
 class GenerateFeedbackApp(ConverterApp):