Fix Gradio File component and NLTK data path

Dockerfile

@@ -38,25 +38,27 @@ RUN useradd -m -u 1000 user && \
 # Copy requirements first to leverage Docker cache
 COPY --chown=user:user requirements.txt .
 
-# Install Python dependencies and NLTK data as root
+# Install Python dependencies
 USER root
 
-# Create NLTK data directory with proper permissions
-RUN mkdir -p /usr/share/nltk_data/tokenizers \
-    && chmod -R 777 /usr/share/nltk_data
+# Create a directory for NLTK data with proper permissions
+RUN mkdir -p /usr/local/share/nltk_data \
+    && chmod -R 777 /usr/local/share/nltk_data
 
-# Install NLTK and download data
-RUN pip install --no-cache-dir -r requirements.txt \
-    && python -c "import nltk; nltk.download('punkt', download_dir='/usr/share/nltk_data')" \
-    && python -c "import nltk; nltk.download('stopwords', download_dir='/usr/share/nltk_data')" \
-    && python -c "import nltk; nltk.download('wordnet', download_dir='/usr/share/nltk_data')" \
-    && python -c "import nltk; nltk.download('averaged_perceptron_tagger', download_dir='/usr/share/nltk_data')" \
-    && chmod -R 755 /usr/share/nltk_data
+# Install Python dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Download NLTK data as root
+RUN python -c "import nltk; nltk.download('punkt', download_dir='/usr/local/share/nltk_data')" \
+    && python -c "import nltk; nltk.download('stopwords', download_dir='/usr/local/share/nltk_data')" \
+    && python -c "import nltk; nltk.download('wordnet', download_dir='/usr/local/share/nltk_data')" \
+    && python -c "import nltk; nltk.download('averaged_perceptron_tagger', download_dir='/usr/local/share/nltk_data')" \
+    && chmod -R 755 /usr/local/share/nltk_data
 
 # Set NLTK_DATA environment variable
-ENV NLTK_DATA=/usr/share/nltk_data
+ENV NLTK_DATA=/usr/local/share/nltk_data
 
-# Switch back to non-root user
+# Switch to non-root user
 USER user
 
 # Copy application files

app.py

@@ -43,7 +43,7 @@ BASE_DIR = Path("/home/user/app/data")
 UPLOAD_DIR = BASE_DIR / "uploads"
 PROCESSED_DIR = BASE_DIR / "processed"
 # Use system NLTK data directory that we'll populate in the Dockerfile
-NLTK_DATA_DIR = Path("/usr/share/nltk_data")
+NLTK_DATA_DIR = Path("/usr/local/share/nltk_data")
 
 # Ensure directories exist with proper permissions
 for directory in [BASE_DIR, UPLOAD_DIR, PROCESSED_DIR]:
@@ -212,7 +212,7 @@ with gr.Blocks() as demo:
     gr.Markdown("# ParseAI PDF 분석 서비스")
     
     with gr.Tab("PDF 업로드"):
-        file_input = gr.File(type="file", file_types=[".pdf"])
+        file_input = gr.File(type="filepath", file_types=["pdf"])
         upload_button = gr.Button("업로드")
         summary_output = gr.Textbox(label="요약")