Merge pull request #16 from malkoG/feature/curation

아티클 큐레이션 기능

Author: darjeeling

.env.sample

@@ -0,0 +1,2 @@
+OPENAI_API_KEY=sk-proj-xxxxx
+GEMINI_API_KEY=asdasd

docker-compose.yml

@@ -15,6 +15,8 @@ services:
       - ./pythonkr_backend:/app/pythonkr_backend
     ports:
       - "8080:8080"
+    env_file:
+      - .env
     depends_on:
       - db
     environment:
@@ -23,4 +25,4 @@ services:
       POSTGRES_PASSWORD: pktesting
 
 volumes:
-  postgres_data:
+  postgres_data:

pyproject.toml

@@ -5,11 +5,18 @@ description = "Add your description here"
 readme = "README.md"
 requires-python = ">=3.12"
 dependencies = [
+    "beautifulsoup4>=4.13.3",
     "django>=5.1.7",
     "gunicorn>=23.0.0",
     "httpx>=0.28.1",
+    "llm>=0.24.2",
+    "llm-gemini>=0.18.1",
+    "lxml>=5.3.2",
     "markdown>=3.7",
     "psycopg[binary]>=3.2.5",
+    "readtime>=3.0.0",
+    "requests>=2.32.3",
+    "tiktoken>=0.9.0",
     "wagtail>=6.4.1",
     "wagtail-bakery>=0.8.0",
 ]

pythonkr_backend/curation/__init__.py

@@ -0,0 +1,88 @@
+from django.contrib import admin, messages
+from .models import Article, Category # Or combine imports
+
+
+@admin.register(Category)
+class CategoryAdmin(admin.ModelAdmin):
+    list_display = ('name', 'slug')
+    search_fields = ('name',)
+    prepopulated_fields = {'slug': ('name',)} # Auto-populate slug from name
+
+
+@admin.action(description="Fetch content, summarize, and translate selected articles")
+def summarize_selected_articles(modeladmin, request, queryset):
+    success_count = 0
+    errors = []
+    
+    for article in queryset:
+        result = article.fetch_and_summarize()
+        if result.startswith("Error"):
+            errors.append(f"{article.url}: {result}")
+        else:
+            success_count += 1
+    
+    if success_count > 0:
+        modeladmin.message_user(
+            request, 
+            f"Successfully processed {success_count} article(s) (fetch, summarize, translate).", 
+            messages.SUCCESS
+        )
+    
+    if errors:
+        error_message = "Errors encountered:\n" + "\n".join(errors)
+        modeladmin.message_user(request, error_message, messages.WARNING)
+
+
+
+
+@admin.register(Article)
+class ArticleAdmin(admin.ModelAdmin):
+    list_display = ('url', 'title', 'display_categories', 'summary_preview', 'summary_ko_preview', 'reading_time_minutes', 'updated_at', 'created_at')
+    list_filter = ('categories', 'created_at', 'updated_at')
+    search_fields = ('url', 'title', 'summary', 'summary_ko', 'categories__name')
+    readonly_fields = ('created_at', 'updated_at', 'summary', 'summary_ko', 'reading_time_minutes')
+    actions = [summarize_selected_articles]
+    filter_horizontal = ('categories',)
+    
+    fieldsets = (
+        ('Article Information', {
+            'fields': ('url', 'title', 'categories')
+        }),
+        ('Generated Content', {
+            'fields': ('summary', 'summary_ko', 'reading_time_minutes'),
+            'classes': ('collapse',)
+        }),
+        ('Metadata', {
+            'fields': ('created_at', 'updated_at'),
+            'classes': ('collapse',)
+        }),
+    )
+    
+    @admin.display(description='Categories')
+    def display_categories(self, obj):
+        """Displays categories as a comma-separated string in the list view."""
+        if obj.categories.exists():
+            return ", ".join([category.name for category in obj.categories.all()])
+        return '-' # Or None, or empty string
+        
+    def get_readonly_fields(self, request, obj=None):
+        # Make 'categories' always read-only as it's set by the LLM
+        readonly = list(super().get_readonly_fields(request, obj))
+        if 'categories' not in readonly:
+             readonly.append('categories')
+        return readonly
+    
+    @admin.display(description='Summary Preview')
+    def summary_preview(self, obj):
+        if obj.summary:
+            preview = obj.summary[:100]
+            return f"{preview}..." if len(obj.summary) > 100 else preview
+        return "No summary available"
+        
+    @admin.display(description='Korean Summary Preview')
+    def summary_ko_preview(self, obj):
+         if obj.summary_ko:
+            if obj.summary_ko.startswith("Translation Error"): return obj.summary_ko
+            preview = obj.summary_ko[:50]
+            return f"{preview}..." if len(obj.summary_ko) > 50 else preview
+         return "No Korean summary"

pythonkr_backend/curation/admin.py

@@ -0,0 +1,6 @@
+from django.apps import AppConfig
+
+
+class CurationConfig(AppConfig):
+    default_auto_field = 'django.db.models.BigAutoField'
+    name = 'curation'

pythonkr_backend/curation/apps.py

@@ -0,0 +1,25 @@
+# Generated by Django 5.2 on 2025-04-20 05:42
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    initial = True
+
+    dependencies = [
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='Article',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('url', models.URLField(help_text='The unique URL of the article.', max_length=2048, unique=True)),
+                ('title', models.CharField(blank=True, help_text='Article title (can be fetched automatically or entered manually).', max_length=512)),
+                ('summary', models.TextField(blank=True, help_text='AI-generated summary of the article.')),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+                ('updated_at', models.DateTimeField(auto_now=True)),
+            ],
+        ),
+    ]

pythonkr_backend/curation/migrations/0001_initial.py

@@ -0,0 +1,18 @@
+# Generated by Django 5.2 on 2025-04-20 06:20
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('curation', '0001_initial'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='article',
+            name='reading_time_minutes',
+            field=models.PositiveIntegerField(blank=True, help_text='Estimated reading time in minutes.', null=True),
+        ),
+    ]

pythonkr_backend/curation/migrations/0002_article_reading_time_minutes.py

@@ -0,0 +1,18 @@
+# Generated by Django 5.2 on 2025-04-20 06:23
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('curation', '0002_article_reading_time_minutes'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='article',
+            name='summary_ko',
+            field=models.TextField(blank=True, help_text='Korean translation of the summary (via OpenAI).'),
+        ),
+    ]

pythonkr_backend/curation/migrations/0003_article_summary_ko.py

@@ -0,0 +1,18 @@
+# Generated by Django 5.2 on 2025-04-20 06:34
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('curation', '0003_article_summary_ko'),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name='article',
+            name='reading_time_minutes',
+            field=models.PositiveIntegerField(blank=True, help_text='Estimated reading time in minutes (based on full article content).', null=True),
+        ),
+    ]

pythonkr_backend/curation/migrations/0004_alter_article_reading_time_minutes.py

@@ -0,0 +1,25 @@
+# Generated by Django 5.2 on 2025-04-20 07:11
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('curation', '0004_alter_article_reading_time_minutes'),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='Category',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text="The name of the category (e.g., 'Web Development', 'LLM').", max_length=100, unique=True)),
+                ('slug', models.SlugField(blank=True, help_text='A URL-friendly slug for the category.', max_length=100, unique=True)),
+            ],
+            options={
+                'verbose_name_plural': 'Categories',
+                'ordering': ['name'],
+            },
+        ),
+    ]

pythonkr_backend/curation/migrations/0005_category.py

@@ -0,0 +1,18 @@
+# Generated by Django 5.2 on 2025-04-20 07:25
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('curation', '0005_category'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='article',
+            name='categories',
+            field=models.ManyToManyField(blank=True, help_text='Select one or more categories for this article.', related_name='articles', to='curation.category'),
+        ),
+    ]