Upload 3 files

app.py

@@ -508,68 +508,59 @@ def process_text(text, model_type="bert"):
     return ent_text, rel_text, kg_text, f"{total_duration:.2f} 秒"
 
 # ======================== 知识图谱可视化 ========================
+import matplotlib.pyplot as plt
+import networkx as nx
+import tempfile
+import os
+import logging
+from matplotlib import font_manager
+
+# 这个函数用于查找并验证中文字体路径
+def find_chinese_font():
+    # 尝试查找 Noto Sans CJK 字体
+    font_paths = [
+        "/usr/share/fonts/truetype/noto/NotoSansCJK-Regular.ttc",  # Noto CJK 字体
+        "/usr/share/fonts/truetype/wqy/wqy-microhei.ttc"         # 微软雅黑
+    ]
+    
+    for font_path in font_paths:
+        if os.path.exists(font_path):
+            logging.info(f"Found font at {font_path}")
+            return font_path
+    
+    logging.error("No Chinese font found!")
+    return None
+
 def generate_kg_image(entities, relations):
+    """
+    中文知识图谱生成函数，支持自动匹配系统中的中文字体，避免中文显示为方框。
+    """
     try:
-        import matplotlib.pyplot as plt
-        import networkx as nx
-        import tempfile
-        import os
-        from matplotlib import font_manager
-        import matplotlib
-        
-        # === 1. 强制字体配置 ===
-        # 方法1：尝试使用系统字体
-        try:
-            # 重建字体缓存
-            matplotlib.font_manager._rebuild()
-            
-            # 查找所有可用字体
-            font_paths = []
-            for font in font_manager.fontManager.ttflist:
-                if 'Noto Sans CJK' in font.name or 'SimHei' in font.name:
-                    font_paths.append(font.fname)
-            
-            # 优先使用Noto Sans CJK SC
-            selected_font = None
-            for font_path in font_paths:
-                if 'NotoSansCJKsc' in font_path:
-                    selected_font = font_path
-                    break
-            
-            if selected_font:
-                font_prop = font_manager.FontProperties(fname=selected_font)
-                plt.rcParams['font.family'] = font_prop.get_name()
-                print(f"使用字体: {selected_font}")
-            else:
-                # 方法2：使用绝对路径字体（上传到项目）
-                font_path = os.path.join(os.path.dirname(__file__), 'fonts', 'SimHei.ttf')
-                if os.path.exists(font_path):
-                    font_prop = font_manager.FontProperties(fname=font_path)
-                    plt.rcParams['font.family'] = font_prop.get_name()
-                    print(f"使用本地字体: {font_path}")
-                else:
-                    # 方法3：最后手段 - 使用系统默认字体
-                    plt.rcParams['font.family'] = ['DejaVu Sans']
-                    print("警告：使用默认字体，中文可能显示为方框")
-            
-            plt.rcParams['axes.unicode_minus'] = False
-        except Exception as e:
-            print(f"字体配置错误: {e}")
+        # === 1. 确保使用合适的中文字体 ===
+        chinese_font = find_chinese_font()  # 调用查找字体函数
+        if chinese_font:
+            font_prop = font_manager.FontProperties(fname=chinese_font)
+            plt.rcParams['font.family'] = font_prop.get_name()
+        else:
+            # 如果字体路径未找到，使用默认字体（DejaVu Sans）
+            logging.warning("Using default font")
+            plt.rcParams['font.family'] = ['DejaVu Sans']
+
+        plt.rcParams['axes.unicode_minus'] = False
 
         # === 2. 创建图谱 ===
         G = nx.DiGraph()
         entity_colors = {
-            'PER': '#FF6B6B',  # 红色
-            'ORG': '#4ECDC4',  # 青色
-            'LOC': '#45B7D1',  # 蓝色
-            'TIME': '#96CEB4'   # 绿色
+            'PER': '#FF6B6B',  # 人物-红色
+            'ORG': '#4ECDC4',  # 组织-青色
+            'LOC': '#45B7D1',  # 地点-蓝色
+            'TIME': '#96CEB4'  # 时间-绿色
         }
 
-        # 简化标签（只显示中文名称）
         for entity in entities:
             G.add_node(
                 entity["text"],
-                label=entity["text"],  # 只显示中文文本
+                label=f"{entity['text']} ({entity['type']})",
                 color=entity_colors.get(entity['type'], '#D3D3D3')
             )
 
@@ -582,14 +573,14 @@ def generate_kg_image(entities, relations):
                 )
 
         # === 3. 绘图配置 ===
-        plt.figure(figsize=(12, 8), dpi=120)
-        pos = nx.spring_layout(G, k=0.8, seed=42)  # 增大k值避免重叠
+        plt.figure(figsize=(12, 8), dpi=150)
+        pos = nx.spring_layout(G, k=0.7, seed=42)
 
-        # 绘制元素
         nx.draw_networkx_nodes(
             G, pos,
             node_color=[G.nodes[n]['color'] for n in G.nodes],
-            node_size=800
+            node_size=800,
+            alpha=0.9
         )
         nx.draw_networkx_edges(
             G, pos,
@@ -599,35 +590,40 @@ def generate_kg_image(entities, relations):
             arrowsize=20
         )
 
-        # === 4. 手动添加文本标签 ===
-        for node, (x, y) in pos.items():
-            plt.text(x, y, 
-                    node,  # 直接使用中文文本
-                    fontsize=10,
-                    ha='center',
-                    va='center',
-                    bbox=dict(facecolor='white', alpha=0.7, edgecolor='none'))
+        node_labels = {n: G.nodes[n]['label'] for n in G.nodes}
+        nx.draw_networkx_labels(
+            G, pos,
+            labels=node_labels,
+            font_size=10,
+            font_family=font_prop.get_name() if chinese_font else 'SimHei',
+            font_weight='bold'
+        )
 
-        # 边标签（英文关系）
         edge_labels = nx.get_edge_attributes(G, 'label')
         nx.draw_networkx_edge_labels(
             G, pos,
             edge_labels=edge_labels,
-            font_size=8
+            font_size=8,
+            font_family=font_prop.get_name() if chinese_font else 'SimHei'
         )
 
         plt.axis('off')
-        
-        # === 5. 保存图片 ===
-        temp_dir = tempfile.mkdtemp()
+        plt.tight_layout()
+
+        # === 4. 保存图片 ===
+        temp_dir = tempfile.mkdtemp()  # 确保在 Docker 容器中有权限写入
         output_path = os.path.join(temp_dir, "kg.png")
+
+        # 打印路径以方便调试
+        logging.info(f"Saving graph image to {output_path}")
+
         plt.savefig(output_path, bbox_inches='tight', pad_inches=0.1)
         plt.close()
-        
+
         return output_path
 
     except Exception as e:
-        print(f"[ERROR] 图谱生成失败: {e}")
+        logging.error(f"[ERROR] 图谱生成失败: {str(e)}")
         return None
 
 

apt.txt

@@ -0,0 +1 @@
+fonts-noto-cjk