feat: Enhance Docker setup and application status reporting with build date and container info

Dockerfile

@@ -1,41 +1,38 @@
-# Stage 1: Build dependencies
+# Use multi-stage build for a smaller final image
 FROM python:3.13 AS builder
 
 WORKDIR /app
 
+# Copy requirements first for better layer caching
 COPY requirements.txt /app/
 RUN pip install --no-cache-dir -r requirements.txt
 
-# Stage 2: Final image
+# Second stage for the actual runtime
 FROM python:3.13.2-slim
 
 WORKDIR /app
 
-# Copy installed dependencies
+# Copy installed packages from builder stage
 COPY --from=builder /usr/local/lib/python3.13/site-packages /usr/local/lib/python3.13/site-packages
 COPY --from=builder /usr/local/bin /usr/local/bin
 
-# Copy application files correctly
+# Copy application code
 COPY ./app /app/app
-COPY ./frontend /app/frontend
 COPY ./VERSION /app/VERSION
-COPY ./LICENSE /app/LICENSE
+COPY ./frontend /app/frontend
 
-# Copy build script and generate build date
+# Create runtime_info directory
-COPY ./docker/build-scripts/save-build-date.sh /tmp/
+RUN mkdir -p /app/runtime_info
-RUN mkdir -p /app/docker/build-scripts/ && \
-    cp /tmp/save-build-date.sh /app/docker/build-scripts/ && \
-    chmod +x /tmp/save-build-date.sh && \
-    /tmp/save-build-date.sh
 
-# Set build date as environment variable
+# Create necessary directories
-#ENV BUILD_DATE=$(cat /app/BUILD_DATE)
+RUN mkdir -p /workdir
 
-# Set Python path explicitly
+# Set environment variables
 ENV PYTHONPATH=/app
+ENV PYTHONUNBUFFERED=1
 
-# Expose API port
+# Expose the port the app runs on
 EXPOSE 8000
-WORKDIR /app
 
+# Default command
 CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000"]

app/views/general.py

@@ -1,18 +1,19 @@
 """
 General routes for the application homepage and basic pages.
 """
-from fastapi import Request, HTTPException
+from fastapi import Request, HTTPException, Depends
 from fastapi.responses import FileResponse
 from pathlib import Path
 from datetime import date
+from sqlalchemy.orm import Session
 
-from app.views.base import APIRouter, templates, require_login
+from app.views.base import APIRouter, templates, require_login, get_db
 from app.utils.config_validator import get_provider_status, validate_storage_configs
 
 router = APIRouter()
 
 @router.get("/", include_in_schema=False)
-async def serve_index(request: Request):
+async def serve_index(request: Request, db: Session = Depends(get_db)):
     """Serve the index/home page."""
     # Get provider information from config validator
     providers = get_provider_status()
@@ -27,9 +28,20 @@ async def serve_index(request: Request):
                                                                     's3', 'ftp', 'webdav', 
                                                                     'google_drive', 'onedrive'])
     
+    # Query the actual file count from the database
+    processed_files = 0
+    try:
+        # Import the model here to avoid circular imports
+        from app.models import FileRecord
+        processed_files = db.query(FileRecord).count()
+    except Exception as e:
+        # Log error but continue (don't break the page if DB query fails)
+        from app.views.base import logger
+        logger.error(f"Error counting files: {str(e)}")
+    
     # Create stats object to pass to the template
     stats = {
-        "processed_files": 0,  # Placeholder - would need actual DB query
+        "processed_files": processed_files,
         "active_integrations": configured_providers,
         "storage_targets": configured_storage_targets
     }

app/views/status.py

@@ -3,6 +3,8 @@ Status and configuration views for the application.
 """
 from fastapi import Request
 from datetime import datetime
+import os
+import subprocess
 
 from app.views.base import APIRouter, templates, require_login, settings
 
@@ -19,14 +21,78 @@ async def status_dashboard(request: Request):
     # Get provider status
     providers = get_provider_status()
     
+    # Get build date from settings
+    build_date = getattr(settings, 'build_date', 'Unknown')
+    
+    # Try to get container information
+    container_info = {}
+    try:
+        # Check for Docker environment
+        if os.path.exists('/.dockerenv'):
+            # We're inside a Docker container
+            container_info['is_docker'] = True
+            
+            # Try to get container ID
+            try:
+                with open('/proc/self/cgroup', 'r') as f:
+                    for line in f:
+                        if 'docker' in line:
+                            container_id = line.split('/')[-1].strip()
+                            container_info['id'] = container_id[:12]  # Short ID format
+                            break
+            except Exception:
+                container_info['id'] = 'Unknown'
+            
+            # Try to get Git commit SHA from runtime info
+            try:
+                # First check runtime info directory
+                if os.path.exists('/app/runtime_info/GIT_SHA'):
+                    with open('/app/runtime_info/GIT_SHA', 'r') as f:
+                        git_sha = f.read().strip()
+                # Then try environment variable
+                else:
+                    git_sha = os.environ.get('GIT_COMMIT_SHA', '')
+                
+                # If still not found, try the original file location
+                if not git_sha and os.path.exists('/.git-commit-sha'):
+                    with open('/.git-commit-sha', 'r') as f:
+                        git_sha = f.read().strip()
+                
+                container_info['git_sha'] = git_sha[:7] if git_sha and git_sha != 'unknown' else 'Unknown'
+            except Exception:
+                container_info['git_sha'] = 'Unknown'
+            
+            # Try to get runtime information
+            try:
+                if os.path.exists('/app/runtime_info/RUNTIME_INFO'):
+                    with open('/app/runtime_info/RUNTIME_INFO', 'r') as f:
+                        container_info['runtime_info'] = f.read().strip()
+            except Exception:
+                pass
+        else:
+            container_info['is_docker'] = False
+            
+            # If not in Docker, try to get Git info directly
+            try:
+                git_sha = subprocess.check_output(['git', 'rev-parse', 'HEAD'], 
+                                                stderr=subprocess.DEVNULL, 
+                                                text=True).strip()[:7]
+                container_info['git_sha'] = git_sha
+            except (subprocess.SubprocessError, FileNotFoundError):
+                container_info['git_sha'] = 'Unknown'
+    except Exception:
+        container_info = {'is_docker': False, 'id': 'Unknown', 'git_sha': 'Unknown'}
+    
     return templates.TemplateResponse(
         "status_dashboard.html",
         {
             "request": request, 
             "providers": providers,
             "app_version": settings.version,
+            "build_date": build_date,
             "debug_enabled": getattr(settings, 'debug', False),
-            "last_check": datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+            "last_check": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
+            "container_info": container_info
         }
     )
 

docker-compose.yaml

@@ -10,7 +10,7 @@ services:
     working_dir: /workdir
 
     # We'll run uvicorn from the container's /app code
-    command: ["sh", "-c", "cd /app && uvicorn app.main:app --host 0.0.0.0 --port 8000 --proxy-headers"]
+    command: ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000", "--proxy-headers"]
 
     # Environment variables
     environment:
@@ -50,7 +50,7 @@ services:
       - redis
       - gotenberg
 
-    # Mount the shared directory (and optionally your code if you want dev mode)
+    # Mount the shared directory
     volumes:
       - /var/docparse/workdir:/workdir
 
@@ -59,7 +59,6 @@ services:
     container_name: gotenberg
     restart: always
 
-
   redis:
     image: redis:alpine
     container_name: document_redis

docker/build-scripts/save-build-date.sh

@@ -1,11 +0,0 @@
-#!/bin/bash
-
-# Get current date in Month DD, YYYY format (e.g., May 15, 2024)
-BUILD_DATE=$(date +"%B %d, %Y")
-
-# Save it to the BUILD_DATE file
-echo $BUILD_DATE > /app/BUILD_DATE
-
-# Also set it as an environment variable
-echo "Setting BUILD_DATE=$BUILD_DATE"
-export BUILD_DATE

frontend/templates/status_dashboard.html

@@ -14,10 +14,20 @@
     </p>
     <div class="bg-blue-100 border-l-4 border-blue-500 text-blue-700 p-4 my-4" role="alert">
       <p><strong>App Version:</strong> {{ app_version }}</p>
+      <p><strong>Build Date:</strong> {{ build_date }}</p>
       <p><strong>Debug Mode:</strong> {{ "Enabled" if debug_enabled else "Disabled" }}</p>
       {% if last_check %}
       <p><strong>Last Check:</strong> {{ last_check }}</p>
       {% endif %}
+      {% if container_info.is_docker %}
+      <p><strong>Container ID:</strong> {{ container_info.id }}</p>
+      {% endif %}
+      {% if container_info.git_sha and container_info.git_sha != 'Unknown' %}
+      <p><strong>Git Commit:</strong> {{ container_info.git_sha }}</p>
+      {% endif %}
+      {% if container_info.runtime_info %}
+      <p><strong>Container Started:</strong> {{ container_info.runtime_info }}</p>
+      {% endif %}
     </div>
   </div>