vault backup: 2025-12-21 20:42:49

quiz/IMPLEMENTATION_SUMMARY.md

@@ -1,143 +0,0 @@
-# Auto-Import Implementation Summary
-
-## What was implemented:
-
-### 1. Dependencies (pyproject.toml)
-- Created `pyproject.toml` in `quiz/` directory
-- Added Django and watchdog dependencies
-- Configured for `uv` package manager
-
-### 2. Question Importer (quiz/utils/importer.py)
-- `ImportStats` class: Tracks import statistics including:
-  - Total files found
-  - MCQ vs non-MCQ questions
-  - Questions with answers vs TODO placeholders
-  - Created vs updated counts
-  - Per-folder completion statistics
-- `parse_markdown_question()`: Parses markdown files to extract:
-  - Question text
-  - Options (supports both `- A:` and `- A` formats)
-  - Correct answer from spoiler blocks
-  - Skips questions with "TODO" answers
-- `import_question_file()`: Imports a single question file
-- `import_questions()`: Bulk import with statistics
-- `delete_question_by_path()`: Removes deleted questions from database
-
-### 3. File System Watcher (quiz/utils/watcher.py)
-- `QuestionFileHandler`: Handles file system events with:
-  - 2-second debounce for file changes
-  - Auto-import on file create/modify
-  - Auto-delete on file removal
-  - Real-time console feedback
-- `QuestionWatcher`: Main watcher class using watchdog Observer
-- `start_watcher_thread()`: Starts watcher in background daemon thread
-  - Performs initial full import on startup
-  - Displays comprehensive statistics
-  - Continues watching for changes
-
-### 4. Django Integration
-- Updated `QuizAppConfig.ready()` in `apps.py`:
-  - Automatically starts watcher thread on Django startup
-  - Runs in daemon thread (doesn't block shutdown)
-  - Only runs in main process (not reloader)
-- Updated `import_questions` management command:
-  - Uses refactored importer
-  - Shows detailed statistics output
-- Added `QUESTION_WATCH_PATH` setting to `settings.py`
-
-## Current Statistics:
-```
-Total .md files found:     312
-MCQ questions found:       162
-Non-MCQ skipped:           152
-Questions with answers:    6
-Questions with TODO:       154
-Overall completion:        3.7%
-
-Completion by Exam Folder:
-2022-01-15             2/ 25 MCQ (  8.0%)
-2022-06-01             4/ 19 MCQ ( 21.1%)
-2023-01-11             0/ 17 MCQ (  0.0%)
-2023-05-31             0/ 10 MCQ (  0.0%)
-2024-01-10             0/ 14 MCQ (  0.0%)
-2024-05-29             0/ 14 MCQ (  0.0%)
-2025-01-15             0/ 16 MCQ (  0.0%)
-2025-02-08             0/ 16 MCQ (  0.0%)
-2025-06-03             0/ 16 MCQ (  0.0%)
-2025-08-08             0/ 15 MCQ (  0.0%)
-```
-
-## How it works:
-
-1. **On Django startup**:
-   - Background thread starts automatically
-   - Performs initial import of all questions
-   - Displays comprehensive statistics
-   - Begins watching for file changes
-
-2. **When you edit a question in Obsidian**:
-   - Watcher detects file change
-   - Waits 2 seconds (debounce multiple saves)
-   - Automatically imports/updates the question
-   - Shows console feedback
-
-3. **When you delete a question file**:
-   - Watcher detects deletion
-   - Removes question from database
-
-4. **Manual import**:
-   - Run: `python3 manage.py import_questions`
-   - Shows same detailed statistics
-
-## Future considerations:
-
-### Multi-select questions support:
-Currently the `Question.correct_answer` field is `max_length=1`, which only supports single answers. Many questions have "Välj två alternativ" and answers like "B och D" or "B, D".
-
-To support multi-select:
-1. Update `Question.correct_answer` to `max_length=50`
-2. Create Django migration
-3. Update `parse_markdown_question()` to extract multiple letters (e.g., "B och D" → "B,D")
-4. Update `views.py` answer validation to compare sorted comma-separated values
-5. Update quiz UI to allow selecting multiple options
-
-### Answer format normalization:
-Need to standardize multi-select answer format in Obsidian:
-- Current: "B och D", "B, D", "BD"
-- Recommended: "B,D" (sorted, comma-separated, no spaces)
-
-### Question types not yet supported:
-- `frågetyp/hotspot`: Image-based clickable questions
-- `frågetyp/dnd-text`: Drag-and-drop text matching
-- `frågetyp/textfält`: Free text input questions
-- `frågetyp/sammansatt`: Multi-part questions
-
-These are currently skipped during import.
-
-## Testing the implementation:
-
-1. **Start Django server**:
-   ```bash
-   cd /Users/johandahlin/dev/medical-notes/quiz
-   python3 manage.py runserver
-   ```
-   You'll see the initial import statistics on startup.
-
-2. **Test auto-import**:
-   - Open a question in Obsidian with "TODO" in spoiler block
-   - Replace "TODO" with a letter (e.g., "B")
-   - Save the file
-   - Check Django console for auto-import message
-
-3. **Test manual import**:
-   ```bash
-   python3 manage.py import_questions
-   ```
-
-4. **Check database**:
-   ```bash
-   python3 manage.py shell
-   >>> from quiz.models import Question
-   >>> Question.objects.count()
-   ```
-

quiz/quiz/admin.py

@@ -0,0 +1,165 @@
+from django.contrib import admin
+from django.utils.html import format_html
+from .models import User, Question, Option, QuizResult
+
+
+class OptionInline(admin.TabularInline):
+    """Inline admin for question options"""
+    model = Option
+    extra = 0
+    fields = ['letter', 'text']
+    ordering = ['letter']
+
+
+@admin.register(Question)
+class QuestionAdmin(admin.ModelAdmin):
+    """Admin interface for Questions"""
+    list_display = ['id', 'question_preview', 'correct_answer', 'option_count', 'file_source', 'updated_at']
+    list_filter = ['created_at', 'updated_at']
+    search_fields = ['text', 'file_path', 'correct_answer']
+    readonly_fields = ['file_path', 'file_mtime', 'created_at', 'updated_at', 'formatted_mtime']
+    fieldsets = [
+        ('Question Content', {
+            'fields': ['text', 'correct_answer']
+        }),
+        ('File Tracking', {
+            'fields': ['file_path', 'file_mtime', 'formatted_mtime'],
+            'classes': ['collapse']
+        }),
+        ('Timestamps', {
+            'fields': ['created_at', 'updated_at'],
+            'classes': ['collapse']
+        }),
+    ]
+    inlines = [OptionInline]
+
+    def question_preview(self, obj):
+        """Show question text preview"""
+        return obj.text[:60] + '...' if len(obj.text) > 60 else obj.text
+    question_preview.short_description = 'Question'
+
+    def option_count(self, obj):
+        """Show number of options"""
+        return obj.options.count()
+    option_count.short_description = '# Options'
+
+    def file_source(self, obj):
+        """Show file path with folder highlight"""
+        if obj.file_path:
+            parts = obj.file_path.split('/')
+            if len(parts) > 1:
+                folder = parts[-2]
+                filename = parts[-1]
+                return format_html('<span style="color: #666;">{}/</span><strong>{}</strong>', folder, filename)
+        return obj.file_path or '-'
+    file_source.short_description = 'Source File'
+
+    def formatted_mtime(self, obj):
+        """Show formatted modification time"""
+        if obj.file_mtime:
+            from datetime import datetime
+            dt = datetime.fromtimestamp(obj.file_mtime)
+            return dt.strftime('%Y-%m-%d %H:%M:%S')
+        return '-'
+    formatted_mtime.short_description = 'File Modified'
+
+
+@admin.register(Option)
+class OptionAdmin(admin.ModelAdmin):
+    """Admin interface for Options"""
+    list_display = ['id', 'question_preview', 'letter', 'text_preview', 'is_correct']
+    list_filter = ['letter']
+    search_fields = ['text', 'question__text']
+    readonly_fields = ['question']
+
+    def question_preview(self, obj):
+        """Show question preview"""
+        return obj.question.text[:40] + '...'
+    question_preview.short_description = 'Question'
+
+    def text_preview(self, obj):
+        """Show option text preview"""
+        return obj.text[:50] + '...' if len(obj.text) > 50 else obj.text
+    text_preview.short_description = 'Option Text'
+
+    def is_correct(self, obj):
+        """Highlight if this is the correct answer"""
+        if obj.question.correct_answer and obj.letter in obj.question.correct_answer:
+            return format_html('<span style="color: green; font-weight: bold;">✓ Correct</span>')
+        return format_html('<span style="color: #999;">-</span>')
+    is_correct.short_description = 'Status'
+
+
+@admin.register(User)
+class UserAdmin(admin.ModelAdmin):
+    """Admin interface for Users"""
+    list_display = ['id', 'session_preview', 'result_count', 'score_percentage', 'created_at']
+    list_filter = ['created_at']
+    search_fields = ['session_key']
+    readonly_fields = ['session_key', 'created_at', 'full_session_key']
+    fieldsets = [
+        ('User Info', {
+            'fields': ['full_session_key', 'created_at']
+        }),
+    ]
+
+    def session_preview(self, obj):
+        """Show session key preview"""
+        return f"{obj.session_key[:12]}..."
+    session_preview.short_description = 'Session'
+
+    def result_count(self, obj):
+        """Show number of quiz results"""
+        return obj.results.count()
+    result_count.short_description = '# Answers'
+
+    def score_percentage(self, obj):
+        """Show score percentage"""
+        total = obj.results.count()
+        if total == 0:
+            return '-'
+        correct = obj.results.filter(is_correct=True).count()
+        percentage = (correct / total * 100)
+        color = 'green' if percentage >= 70 else 'orange' if percentage >= 50 else 'red'
+        return format_html(
+            '<span style="color: {}; font-weight: bold;">{:.1f}%</span> ({}/{})',
+            color, percentage, correct, total
+        )
+    score_percentage.short_description = 'Score'
+
+    def full_session_key(self, obj):
+        """Show full session key"""
+        return obj.session_key
+    full_session_key.short_description = 'Full Session Key'
+
+
+@admin.register(QuizResult)
+class QuizResultAdmin(admin.ModelAdmin):
+    """Admin interface for Quiz Results"""
+    list_display = ['id', 'user_preview', 'question_preview', 'selected_answer', 'correct_answer', 'result_status', 'answered_at']
+    list_filter = ['is_correct', 'answered_at']
+    search_fields = ['user__session_key', 'question__text']
+    readonly_fields = ['user', 'question', 'selected_answer', 'is_correct', 'answered_at']
+
+    def user_preview(self, obj):
+        """Show user session preview"""
+        return f"{obj.user.session_key[:8]}..."
+    user_preview.short_description = 'User'
+
+    def question_preview(self, obj):
+        """Show question preview"""
+        return obj.question.text[:40] + '...'
+    question_preview.short_description = 'Question'
+
+    def correct_answer(self, obj):
+        """Show correct answer"""
+        return obj.question.correct_answer
+    correct_answer.short_description = 'Correct'
+
+    def result_status(self, obj):
+        """Show visual result status"""
+        if obj.is_correct:
+            return format_html('<span style="color: green; font-weight: bold;">✓ Correct</span>')
+        return format_html('<span style="color: red; font-weight: bold;">✗ Wrong</span>')
+    result_status.short_description = 'Result'
+

quiz/quiz/apps.py

@@ -1,4 +1,5 @@
 import os
+import sys
 from django.apps import AppConfig
 
 
@@ -12,7 +13,10 @@ class QuizAppConfig(AppConfig):
         Starts the auto-import watcher in a background thread.
         """
         # Only run in the main process (not in reloader process)
-        if os.environ.get('RUN_MAIN') == 'true' or os.environ.get('RUN_MAIN') is None:
+        # And not during management commands (to avoid database locking)
+        is_management_command = any(cmd in sys.argv for cmd in ['import_questions', 'migrate', 'makemigrations', 'shell'])
+
+        if not is_management_command and (os.environ.get('RUN_MAIN') == 'true' or os.environ.get('RUN_MAIN') is None):
             from quiz.utils.watcher import start_watcher_thread
             start_watcher_thread()
 

quiz/quiz/management/commands/import_questions.py

@@ -26,12 +26,16 @@ class Command(BaseCommand):
 
         stats = import_questions(folder, folder)
 
-        # Output statistics
-        self.stdout.write(stats.format_output())
-
-        if stats.errors > 0:
-            self.stdout.write(self.style.WARNING(f'Completed with {stats.errors} errors'))
+        # Only show full statistics if there were changes
+        output = stats.format_output(show_if_no_changes=False)
+        if output:
+            self.stdout.write(output)
+            if stats.errors > 0:
+                self.stdout.write(self.style.WARNING(f'Completed with {stats.errors} errors'))
+            else:
+                self.stdout.write(self.style.SUCCESS('Import completed successfully!'))
         else:
-            self.stdout.write(self.style.SUCCESS('Import completed successfully!'))
+            # No changes, show brief message
+            self.stdout.write(self.style.SUCCESS(f'✓ All {stats.total_files} files up to date, no changes needed'))
 
 

quiz/quiz/migrations/0002_alter_question_correct_answer.py

@@ -0,0 +1,18 @@
+# Generated by Django 6.0 on 2025-12-21 19:25
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('quiz', '0001_initial'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='question',
+            name='correct_answer',
+            field=models.CharField(max_length=50),
+        ),
+    ]

quiz/quiz/migrations/0003_question_file_mtime.py

@@ -0,0 +1,18 @@
+# Generated by Django 6.0 on 2025-12-21 19:32
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('quiz', '0002_alter_question_correct_answer'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='question',
+            name='file_mtime',
+            field=models.FloatField(blank=True, null=True),
+        ),
+    ]

quiz/quiz/models.py

@@ -12,7 +12,8 @@ class User(models.Model):
 class Question(models.Model):
     file_path = models.CharField(max_length=500, unique=True)
     text = models.TextField()
-    correct_answer = models.CharField(max_length=1)
+    correct_answer = models.CharField(max_length=50)  # Support multi-select answers like "A,B,C"
+    file_mtime = models.FloatField(null=True, blank=True)  # Track file modification time
     created_at = models.DateTimeField(auto_now_add=True)
     updated_at = models.DateTimeField(auto_now=True)
 

quiz/quiz/urls.py

@@ -1,8 +1,10 @@
+from django.contrib import admin
 from django.urls import path
 
 from .views import index, get_next_question, submit_answer, stats
 
 urlpatterns = [
+    path('admin/', admin.site.urls),
     path('', index, name='index'),
     path('next/', get_next_question, name='next_question'),
     path('submit/', submit_answer, name='submit_answer'),

quiz/quiz/utils/importer.py

@@ -24,8 +24,20 @@ class ImportStats:
             'todo': 0
         })
 
-    def format_output(self) -> str:
-        """Format statistics for console output"""
+    def has_changes(self) -> bool:
+        """Check if there were any actual changes"""
+        return self.created > 0 or self.updated > 0 or self.errors > 0
+
+    def format_output(self, show_if_no_changes: bool = True) -> str:
+        """
+        Format statistics for console output
+
+        Args:
+            show_if_no_changes: If False, returns empty string when no changes
+        """
+        if not show_if_no_changes and not self.has_changes():
+            return ""
+
         lines = []
         lines.append("\n" + "="*70)
         lines.append("QUESTION IMPORT STATISTICS")
@@ -71,18 +83,30 @@ def parse_markdown_question(file_path: Path, content: str) -> Tuple[bool, dict]:
     """
     lines = content.split('\n')
 
-    # Check for MCQ tags in frontmatter
-    is_mcq = False
+    # Check for question tags in frontmatter
+    # Accept: frågetyp/mcq, frågetyp/scq, frågetyp/textalternativ, frågetyp/textfält
+    is_question = False
+    question_type = None
     in_frontmatter = False
+
     for line in lines:
         if line.strip() == '---':
             in_frontmatter = not in_frontmatter
             continue
-        if in_frontmatter and ('frågetyp/mcq' in line or 'frågetyp/scq' in line):
-            is_mcq = True
+        if in_frontmatter and 'frågetyp/' in line:
+            is_question = True
+            # Extract question type
+            if 'frågetyp/mcq' in line:
+                question_type = 'mcq'
+            elif 'frågetyp/scq' in line:
+                question_type = 'scq'
+            elif 'frågetyp/textalternativ' in line:
+                question_type = 'textalternativ'
+            elif 'frågetyp/textfält' in line:
+                question_type = 'textfält'
             break
 
-    if not is_mcq:
+    if not is_question:
         return False, {}
 
     # Extract question text (first non-empty line after frontmatter)
@@ -110,9 +134,34 @@ def parse_markdown_question(file_path: Path, content: str) -> Tuple[bool, dict]:
     if not question_text:
         return True, {}
 
-    # Extract options (pattern: "- A:" or "- A" or just "- A:")
+    # Extract options (pattern: "- A:" or "- A" for MCQ, or text for textalternativ)
     options_data = []
+    in_frontmatter = False
+    frontmatter_done = False
+    in_spoiler = False
+
     for line in lines:
+        # Track frontmatter to skip it
+        if line.strip() == '---':
+            if not in_frontmatter:
+                in_frontmatter = True
+            else:
+                in_frontmatter = False
+                frontmatter_done = True
+            continue
+
+        # Skip frontmatter and spoiler blocks
+        if in_frontmatter or not frontmatter_done:
+            continue
+
+        if line.strip().startswith('```spoiler-block:'):
+            in_spoiler = True
+            continue
+        if in_spoiler:
+            if line.strip() == '```':
+                in_spoiler = False
+            continue
+
         # Match "- A: text" or "- A: " or just "- A"
         match = re.match(r'^-\s*([A-Z]):\s*(.*)$', line.strip())
         if not match:
@@ -123,14 +172,29 @@ def parse_markdown_question(file_path: Path, content: str) -> Tuple[bool, dict]:
             letter = match.group(1)
             text = match.group(2) if len(match.groups()) > 1 else ""
             options_data.append((letter, text.strip()))
+        else:
+            # For textalternativ, options might be plain text items
+            if question_type in ['textalternativ', 'textfält'] and line.strip().startswith('-') and not line.strip().startswith('--'):
+                # Extract text after dash
+                option_text = line.strip()[1:].strip()
+                # Skip if it's a sub-question marker like "a)" or "b)"
+                if option_text and not re.match(r'^[a-z]\)', option_text):
+                    # Use incrementing letters for text options
+                    letter = chr(ord('A') + len(options_data))
+                    options_data.append((letter, option_text))
 
-    if len(options_data) < 2:
+    # For text-based questions, options are optional
+    if question_type in ['textfält'] and len(options_data) == 0:
+        # Create a dummy option for text field questions
+        options_data = [('A', '')]
+    elif len(options_data) < 2 and question_type in ['mcq', 'scq']:
         return True, {}
 
     # Extract answer from spoiler block
     correct_answer = None
     has_answer = False
     in_spoiler = False
+    answer_lines = []
 
     for line in lines:
         if line.strip().startswith('```spoiler-block:'):
@@ -140,27 +204,69 @@ def parse_markdown_question(file_path: Path, content: str) -> Tuple[bool, dict]:
             if line.strip() == '```':
                 break
             stripped = line.strip()
-            if stripped and stripped != 'TODO':
-                # Extract single letter answer (e.g., "B" or "F")
-                answer_match = re.match(r'^([A-Z])$', stripped)
-                if answer_match:
-                    correct_answer = answer_match.group(1)
-                    has_answer = True
-                break
-            elif stripped == 'TODO':
-                break
+            if stripped:
+                answer_lines.append(stripped)
+
+    # Process collected answer lines
+    if answer_lines:
+        full_answer = ' '.join(answer_lines)
+
+        # Check for TODO
+        if 'TODO' in full_answer.upper():
+            has_answer = False
+        else:
+            has_answer = True
+
+            # For MCQ/SCQ: Extract capital letters
+            if question_type in ['mcq', 'scq']:
+                letters = re.findall(r'\b([A-Z])\b', full_answer)
+                if letters:
+                    correct_answer = ','.join(sorted(set(letters)))
+            else:
+                # For text-based questions: Store the full answer text
+                correct_answer = full_answer[:200]  # Limit to 200 chars for database field
 
     return True, {
         'text': question_text,
         'options': options_data,
         'correct_answer': correct_answer,
-        'has_answer': has_answer
+        'has_answer': has_answer,
+        'question_type': question_type
     }
 
 
-def import_question_file(file_path: Path, base_path: Path, stats: ImportStats):
-    """Import a single question file"""
+def import_question_file(file_path: Path, base_path: Path, stats: ImportStats, force: bool = False):
+    """
+    Import a single question file, checking modification time to avoid unnecessary updates.
+
+    Args:
+        file_path: Path to the question file
+        base_path: Base path for relative calculations
+        stats: ImportStats object to track statistics
+        force: If True, import regardless of mtime (for initial import)
+    """
     try:
+        # Get file modification time
+        file_mtime = file_path.stat().st_mtime
+
+        # Calculate path relative to project root
+        from django.conf import settings
+        project_root = settings.BASE_DIR.parent
+        try:
+            file_path_str = str(file_path.relative_to(project_root))
+        except ValueError:
+            file_path_str = str(file_path.relative_to(base_path))
+
+        # Check if file has changed by comparing mtime
+        if not force:
+            try:
+                existing_question = Question.objects.get(file_path=file_path_str)
+                if existing_question.file_mtime and existing_question.file_mtime >= file_mtime:
+                    # File hasn't changed, skip
+                    return 'skipped_unchanged'
+            except Question.DoesNotExist:
+                pass  # New file, will import
+
         content = file_path.read_text(encoding='utf-8')
         is_mcq, question_data = parse_markdown_question(file_path, content)
 
@@ -171,31 +277,30 @@ def import_question_file(file_path: Path, base_path: Path, stats: ImportStats):
 
         if not is_mcq:
             stats.non_mcq_skipped += 1
-            return
+            return 'skipped_not_mcq'
 
         stats.mcq_questions += 1
         stats.by_folder[folder_name]['mcq'] += 1
 
         if not question_data or not question_data.get('text'):
             stats.non_mcq_skipped += 1
-            return
+            return 'skipped_invalid'
 
         if not question_data['has_answer']:
             stats.questions_with_todo += 1
             stats.by_folder[folder_name]['todo'] += 1
-            return  # Skip questions without answers
+            return 'skipped_todo'
 
         stats.questions_with_answers += 1
         stats.by_folder[folder_name]['answered'] += 1
 
-        # Import to database
-        file_path_str = str(file_path.relative_to(base_path.parent))
-
+        # Import to database with mtime tracking
         question, created = Question.objects.update_or_create(
             file_path=file_path_str,
             defaults={
                 'text': question_data['text'],
                 'correct_answer': question_data['correct_answer'],
+                'file_mtime': file_mtime,  # Track modification time
             }
         )
 
@@ -206,21 +311,29 @@ def import_question_file(file_path: Path, base_path: Path, stats: ImportStats):
 
         # Update options
         question.options.all().delete()
+        # Deduplicate options by letter (keep first occurrence)
+        seen_letters = set()
         for letter, text in question_data['options']:
-            Option.objects.create(question=question, letter=letter, text=text)
+            if letter not in seen_letters:
+                Option.objects.create(question=question, letter=letter, text=text)
+                seen_letters.add(letter)
+
+        return 'imported' if created else 'updated'
 
     except Exception as e:
         stats.errors += 1
         print(f"Error importing {file_path}: {e}")
+        return 'error'
 
 
-def import_questions(folder_path: Path, base_path: Path = None) -> ImportStats:
+def import_questions(folder_path: Path, base_path: Path = None, force: bool = False) -> ImportStats:
     """
     Import all questions from a folder.
 
     Args:
         folder_path: Path to the folder containing question markdown files
         base_path: Base path for relative path calculations (defaults to folder_path)
+        force: If True, import all files regardless of mtime (for initial import)
 
     Returns:
         ImportStats object with import statistics
@@ -232,17 +345,22 @@ def import_questions(folder_path: Path, base_path: Path = None) -> ImportStats:
 
     for md_file in folder_path.rglob('*.md'):
         stats.total_files += 1
-        import_question_file(md_file, base_path, stats)
+        import_question_file(md_file, base_path, stats, force=force)
 
     return stats
 
 
-def delete_question_by_path(file_path: Path, base_path: Path):
+def delete_question_by_path(file_path: Path):
     """Delete a question from the database by file path"""
     try:
-        file_path_str = str(file_path.relative_to(base_path.parent))
-        Question.objects.filter(file_path=file_path_str).delete()
-        print(f"Deleted question: {file_path_str}")
+        from django.conf import settings
+        project_root = settings.BASE_DIR.parent
+        file_path_str = str(file_path.relative_to(project_root))
+        deleted_count, _ = Question.objects.filter(file_path=file_path_str).delete()
+        if deleted_count > 0:
+            print(f"[Auto-delete] ✓ Deleted question: {file_path.name}")
+        return deleted_count > 0
     except Exception as e:
-        print(f"Error deleting question {file_path}: {e}")
+        print(f"[Auto-delete] ✗ Error deleting question {file_path}: {e}")
+        return False
 

quiz/quiz/utils/watcher.py

@@ -8,7 +8,7 @@ from quiz.utils.importer import import_question_file, delete_question_by_path, I
 
 
 class QuestionFileHandler(FileSystemEventHandler):
-    """Handle file system events for question markdown files"""
+    """Handle file system events for question markdown files with mtime checking"""
 
     def __init__(self, base_path: Path, watch_path: Path):
         super().__init__()
@@ -18,29 +18,38 @@ class QuestionFileHandler(FileSystemEventHandler):
         self.debounce_seconds = 2
         self.lock = threading.Lock()
 
-    def _debounced_import(self, file_path: Path):
-        """Import file after debounce delay"""
+    def _debounced_import(self, file_path: Path, event_type: str):
+        """Import file after debounce delay, checking mtime for actual changes"""
         time.sleep(self.debounce_seconds)
 
         with self.lock:
             if file_path in self.pending_events:
                 del self.pending_events[file_path]
 
-        if file_path.exists():
-            print(f"\n[Auto-import] Processing: {file_path.name}")
-            stats = ImportStats()
-            import_question_file(file_path, self.watch_path, stats)
+        if not file_path.exists():
+            return
 
-            if stats.created > 0:
-                print(f"[Auto-import] ✓ Created question from {file_path.name}")
-            elif stats.updated > 0:
-                print(f"[Auto-import] ✓ Updated question from {file_path.name}")
-            elif stats.questions_with_todo > 0:
-                print(f"[Auto-import] ⊘ Skipped {file_path.name} (TODO answer)")
-            elif stats.non_mcq_skipped > 0:
-                print(f"[Auto-import] ⊘ Skipped {file_path.name} (not MCQ)")
+        # Import with mtime checking (force=False means only import if changed)
+        stats = ImportStats()
+        result = import_question_file(file_path, self.watch_path, stats, force=False)
 
-    def _handle_file_change(self, file_path: Path):
+        # Provide feedback based on result
+        if result == 'imported':
+            print(f"\n[Auto-import] ✓ Created: {file_path.name}")
+        elif result == 'updated':
+            print(f"\n[Auto-import] ✓ Updated: {file_path.name}")
+        elif result == 'skipped_unchanged':
+            # File hasn't actually changed (same mtime), no output
+            pass
+        elif result == 'skipped_todo':
+            print(f"\n[Auto-import] ⊘ Skipped: {file_path.name} (TODO answer)")
+        elif result == 'skipped_not_mcq':
+            # Silently skip non-MCQ files
+            pass
+        elif result == 'error':
+            print(f"\n[Auto-import] ✗ Error: {file_path.name}")
+
+    def _handle_file_change(self, file_path: Path, event_type: str = 'modified'):
         """Handle file creation or modification with debouncing"""
         if not file_path.suffix == '.md':
             return
@@ -51,26 +60,29 @@ class QuestionFileHandler(FileSystemEventHandler):
                 self.pending_events[file_path].cancel()
 
             # Schedule new import
-            timer = threading.Timer(self.debounce_seconds, self._debounced_import, args=[file_path])
+            timer = threading.Timer(
+                self.debounce_seconds,
+                self._debounced_import,
+                args=[file_path, event_type]
+            )
             self.pending_events[file_path] = timer
             timer.start()
 
     def on_created(self, event: FileSystemEvent):
         """Handle file creation"""
         if not event.is_directory:
-            self._handle_file_change(Path(event.src_path))
+            self._handle_file_change(Path(event.src_path), 'created')
 
     def on_modified(self, event: FileSystemEvent):
         """Handle file modification"""
         if not event.is_directory:
-            self._handle_file_change(Path(event.src_path))
+            self._handle_file_change(Path(event.src_path), 'modified')
 
     def on_deleted(self, event: FileSystemEvent):
         """Handle file deletion"""
         if not event.is_directory and event.src_path.endswith('.md'):
             file_path = Path(event.src_path)
-            print(f"\n[Auto-import] Deleting: {file_path.name}")
-            delete_question_by_path(file_path, self.watch_path)
+            delete_question_by_path(file_path)
 
 
 class QuestionWatcher:
@@ -116,10 +128,16 @@ def start_watcher_thread():
             print(f"[QuestionWatcher] Warning: Watch path does not exist: {watch_path}")
             return
 
-        # Initial import
-        print("\n[QuestionWatcher] Starting initial import...")
-        stats = import_questions(watch_path, watch_path)
-        print(stats.format_output())
+        # Initial import with mtime checking (force=False to only import changed files)
+        print("\n[QuestionWatcher] Checking for changes...")
+        stats = import_questions(watch_path, watch_path, force=False)
+
+        # Only show stats if there were changes
+        output = stats.format_output(show_if_no_changes=False)
+        if output:
+            print(output)
+        else:
+            print(f"[QuestionWatcher] ✓ All files up to date")
 
         # Start watching for changes
         watcher = QuestionWatcher(watch_path, watch_path)

quiz/settings.py

@@ -17,8 +17,11 @@ ALLOWED_HOSTS = os.environ.get('ALLOWED_HOSTS', '*').split(',')
 
 # Application definition
 INSTALLED_APPS = [
+    'django.contrib.admin',
+    'django.contrib.auth',
     'django.contrib.contenttypes',
     'django.contrib.sessions',
+    'django.contrib.messages',
     'django.contrib.staticfiles',
     'quiz.apps.QuizAppConfig',
 ]
@@ -28,6 +31,8 @@ MIDDLEWARE = [
     'django.contrib.sessions.middleware.SessionMiddleware',
     'django.middleware.common.CommonMiddleware',
     'django.middleware.csrf.CsrfViewMiddleware',
+    'django.contrib.auth.middleware.AuthenticationMiddleware',
+    'django.contrib.messages.middleware.MessageMiddleware',
     'quiz.middleware.LazyAuthMiddleware',
 ]
 
@@ -41,6 +46,7 @@ TEMPLATES = [{
         'context_processors': [
             'django.template.context_processors.debug',
             'django.template.context_processors.request',
+            'django.contrib.auth.context_processors.auth',
             'django.contrib.messages.context_processors.messages',
         ],
     },