feat: add course csv export endpoint for publisher reports

[rjv31]

No description provided.

[Copilot]

Pull request overview

Adds a new CSV export endpoint on the Courses API to support publisher reporting use cases, along with test coverage validating filtering/permissions and CSV formatting.

Changes:

• Added a csv collection action to CourseViewSet that returns a CSV download of filtered courses.
• Implemented CSV column mapping and publisher URL/status display transformations for exported rows.
• Added a dedicated test suite validating headers, filtering parity, partner scoping, and auth/permission behavior.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
course_discovery/apps/api/v1/views/courses.py	Adds the csv export action and CSV row/header construction logic for course exports.
course_discovery/apps/api/v1/tests/test_views/test_courses.py	Adds tests to verify CSV export behavior, filtering, formatting, and permissions.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Suppressed comments (2)

course_discovery/apps/api/v1/views/courses.py:218

• StreamingHttpResponse is constructed from a fully materialized csv_lines list, so the response isn’t actually streamed and memory usage scales with the full export size. Yielding rows from a generator keeps memory bounded while preserving the same output.

        csv_lines = [self._csv_line(self.EXPORT_CSV_HEADERS)]
        for serialized_course in csv_serialized:
            course_metadata = course_metadata_by_uuid.get(serialized_course.get('uuid'), {})
            csv_lines.append(
                self._csv_line(
                    self._csv_row(
                        serialized_course,
                        course_metadata.get('publisher_url', ''),
                        course_metadata.get('organization_key', ''),
                        course_metadata.get('project_coordinator', ''),
                    )
                )
            )

        response = StreamingHttpResponse(csv_lines, content_type='text/csv; charset=utf-8')
        response['Content-Disposition'] = 'attachment; filename="publisher_courses.csv"'
        return response

course_discovery/apps/api/v1/views/courses.py:119

• CSV formula-injection escaping only checks the first character, so values with leading whitespace (e.g., " \t=SUM(…)" or " -1+2") won’t be escaped and can still be interpreted as formulas in spreadsheet tools. Consider checking the first non-whitespace character instead while preserving the original value.

        def escape_csv_cell(value):
            if isinstance(value, str) and value and value[0] in ('=', '+', '-', '@'):
                return "'" + value
            return value

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Suppressed comments (1)

course_discovery/apps/api/v1/views/courses.py:195

• filtered_courses = list(filtered_queryset) and self.get_serializer(filtered_courses, many=True).data fully materialize both the queryset (including all prefetched relations from get_queryset) and the entire serialized payload in memory before streaming. For large exports this defeats the main scalability benefit of StreamingHttpResponse and can cause high memory usage/time-to-first-byte.

Consider introducing a dedicated, lightweight export queryset/serializer (only fields needed for CSV) and streaming rows without building the full serialized list in memory (e.g., by iterating courses in chunks and writing rows per chunk).

        filtered_courses = list(filtered_queryset)

        csv_serialized = self.get_serializer(filtered_courses, many=True).data

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Suppressed comments (1)

course_discovery/apps/api/v1/views/courses.py:192

• _iterate_queryset_in_chunks uses queryset slicing with an ever-increasing OFFSET. For large exports this degrades significantly (OFFSET scans get slower as the offset grows), which undermines the intent of streaming the response. A more scalable approach is to stream primary keys via .iterator() and refetch each chunk using pk__in, preserving the queryset's existing filters/orderings while avoiding large OFFSETs.

    @staticmethod
    def _iterate_queryset_in_chunks(queryset, chunk_size=500):
        """Efficiently iterate through a queryset in bounded chunks."""
        offset = 0
        while True:
            chunk = list(queryset[offset:offset + chunk_size])
            if not chunk:
                break
            yield chunk
            offset += chunk_size