From b45a6e4f115b186135f8649430c59af63655e479 Mon Sep 17 00:00:00 2001
From: Mrna <jan.mrna@siemens.com>
Date: Mon, 3 Nov 2025 15:40:34 +0100
Subject: [PATCH] Web: link and serve the file

---
 main.py              | 19 ++++++++++++++++++-
 templates/index.html |  4 +++-
 2 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/main.py b/main.py
index 6dfb164..b764c8a 100644
--- a/main.py
+++ b/main.py
@@ -182,7 +182,7 @@ def query(db_path: str, query_text: str):
 def start_web_server(db_path: str, host: str = "127.0.0.1", port: int = 5000):
     """Start a web server for the semantic search tool."""
     try:
-        from flask import Flask, request, jsonify, render_template
+        from flask import Flask, request, jsonify, render_template, send_file
     except ImportError:
         print("❌ Flask not found. Please install it first:")
         print("   pip install flask")
@@ -201,6 +201,22 @@ def start_web_server(db_path: str, host: str = "127.0.0.1", port: int = 5000):
     def index():
         return render_template("index.html", results=None)
     
+    @app.route('/file/<path:document_path>')
+    def serve_file(document_path):
+        """Serve PDF files directly."""
+        try:
+            file_path = Path(document_path)
+            if not file_path.exists():
+                return jsonify({'error': 'File not found'}), 404
+            
+            # Check if it's a PDF file for security
+            if file_path.suffix.lower() != '.pdf':
+                return jsonify({'error': 'Only PDF files are allowed'}), 403
+            
+            return send_file(file_path, as_attachment=False)
+        except Exception as e:
+            return jsonify({'error': str(e)}), 500
+    
     @app.route('/api/search', methods=['POST'])
     def search():
         try:
@@ -221,6 +237,7 @@ def start_web_server(db_path: str, host: str = "127.0.0.1", port: int = 5000):
                 formatted_results.append({
                     'distance': float(distance),
                     'document': record.document.name,
+                    'document_path': str(record.document),  # Full path for the link
                     'page': record.page,
                     'chunk': record.chunk,
                     'text': ' '.join(record.text[:300].split())  # Clean and truncate text
diff --git a/templates/index.html b/templates/index.html
index 9dc6302..bb74687 100644
--- a/templates/index.html
+++ b/templates/index.html
@@ -12,6 +12,8 @@
             .result-header { font-weight: bold; color: #333; margin-bottom: 10px; }
             .result-text { background: #f9f9f9; padding: 10px; border-radius: 3px; }
             .distance { color: #666; font-size: 0.9em; }
+            .document-link { color: #007cba; text-decoration: none; }
+            .document-link:hover { text-decoration: underline; }
             .no-results { text-align: center; color: #666; margin: 40px 0; }
             .loading { text-align: center; color: #007cba; margin: 20px 0; }
         </style>
@@ -56,7 +58,7 @@
                     resultsDiv.innerHTML = data.results.map((result, i) => `
                         <div class="result">
                             <div class="result-header">
-                                Result ${i + 1} - ${result.document} 
+                                Result ${i + 1} - <a href="/file/${encodeURIComponent(result.document_path)}" class="document-link" target="_blank">${result.document}</a>
                                 <span class="distance">(Distance: ${result.distance.toFixed(4)})</span>
                             </div>
                             <div>Page: ${result.page}, Chunk: ${result.chunk}</div>