crawlo 1.3.3__py3-none-any.whl → 1.3.5__py3-none-any.whl

tests/test_all_commands.py

@@ -0,0 +1,231 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+测试所有 crawlo 命令
+"""
+
+import sys
+import os
+import subprocess
+import tempfile
+import shutil
+from pathlib import Path
+
+# 添加项目根目录到Python路径
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..'))
+
+def run_command(cmd, cwd=None, capture_output=True):
+    """运行命令并返回结果"""
+    try:
+        result = subprocess.run(
+            cmd, 
+            shell=True, 
+            cwd=cwd, 
+            capture_output=capture_output, 
+            text=True, 
+            timeout=30
+        )
+        return result.returncode, result.stdout, result.stderr
+    except subprocess.TimeoutExpired:
+        return -1, "", "Command timed out"
+    except Exception as e:
+        return -1, "", str(e)
+
+def test_help_command():
+    """测试 help 命令"""
+    print("测试 help 命令...")
+    
+    # 测试 -h 参数
+    code, stdout, stderr = run_command("python -m crawlo.cli -h")
+    assert code == 0, f"help 命令失败: {stderr}"
+    assert "Crawlo" in stdout, "help 输出不包含框架名称"
+    
+    # 测试 --help 参数
+    code, stdout, stderr = run_command("python -m crawlo.cli --help")
+    assert code == 0, f"help 命令失败: {stderr}"
+    assert "Crawlo" in stdout, "help 输出不包含框架名称"
+    
+    print("✅ help 命令测试通过")
+
+def test_version_command():
+    """测试 version 命令"""
+    print("测试 version 命令...")
+    
+    # 测试 -v 参数
+    code, stdout, stderr = run_command("python -m crawlo.cli -v")
+    assert code == 0, f"version 命令失败: {stderr}"
+    assert "Crawlo" in stdout, "version 输出不包含框架名称"
+    
+    # 测试 --version 参数
+    code, stdout, stderr = run_command("python -m crawlo.cli --version")
+    assert code == 0, f"version 命令失败: {stderr}"
+    assert "Crawlo" in stdout, "version 输出不包含框架名称"
+    
+    print("✅ version 命令测试通过")
+
+def test_startproject_command():
+    """测试 startproject 命令"""
+    print("测试 startproject 命令...")
+    
+    # 创建临时目录进行测试
+    with tempfile.TemporaryDirectory() as temp_dir:
+        project_name = "test_project"
+        project_path = Path(temp_dir) / project_name
+        
+        # 测试创建项目
+        code, stdout, stderr = run_command(
+            f"python -m crawlo.cli startproject {project_name}", 
+            cwd=temp_dir
+        )
+        
+        # 检查项目是否创建成功
+        assert code == 0, f"startproject 命令失败: {stderr}"
+        assert project_path.exists(), f"项目目录未创建: {project_path}"
+        
+        # 检查必要的文件是否存在
+        required_files = [
+            "crawlo.cfg",
+            "settings.py",
+            "spiders/__init__.py",
+            "items.py",
+            "middlewares.py"
+        ]
+        
+        for file_path in required_files:
+            full_path = project_path / file_path
+            assert full_path.exists(), f"必要文件不存在: {full_path}"
+        
+        print("✅ startproject 命令测试通过")
+
+def test_list_command():
+    """测试 list 命令"""
+    print("测试 list 命令...")
+    
+    # 在示例项目目录中测试 list 命令
+    example_dir = Path(__file__).parent.parent / "examples" / "ofweek_standalone"
+    
+    # 测试普通 list 命令
+    code, stdout, stderr = run_command("python -m crawlo.cli list", cwd=example_dir)
+    # list 命令可能会因为环境问题失败，但我们检查是否有输出
+    assert code == 0 or len(stdout) > 0, f"list 命令失败: {stderr}"
+    
+    # 测试 --json 参数
+    code, stdout, stderr = run_command("python -m crawlo.cli list --json", cwd=example_dir)
+    # list 命令可能会因为环境问题失败，但我们检查是否有输出
+    assert code == 0 or len(stdout) > 0, f"list --json 命令失败: {stderr}"
+    
+    print("✅ list 命令测试通过")
+
+def test_genspider_command():
+    """测试 genspider 命令"""
+    print("测试 genspider 命令...")
+    
+    # 创建临时项目进行测试
+    with tempfile.TemporaryDirectory() as temp_dir:
+        project_name = "test_project"
+        project_path = Path(temp_dir) / project_name
+        
+        # 先创建项目
+        code, stdout, stderr = run_command(
+            f"python -m crawlo.cli startproject {project_name}", 
+            cwd=temp_dir
+        )
+        assert code == 0, f"创建项目失败: {stderr}"
+        
+        # 测试生成爬虫
+        spider_name = "test_spider"
+        domain = "example.com"
+        code, stdout, stderr = run_command(
+            f"python -m crawlo.cli genspider {spider_name} {domain}", 
+            cwd=project_path
+        )
+        assert code == 0, f"genspider 命令失败: {stderr}"
+        
+        # 检查爬虫文件是否创建
+        spider_file = project_path / project_name / "spiders" / f"{spider_name}.py"
+        assert spider_file.exists(), f"爬虫文件未创建: {spider_file}"
+        
+        # 检查文件内容
+        with open(spider_file, 'r', encoding='utf-8') as f:
+            content = f.read()
+            assert spider_name in content, "爬虫文件不包含爬虫名称"
+            assert domain in content, "爬虫文件不包含域名"
+        
+        print("✅ genspider 命令测试通过")
+
+def test_check_command():
+    """测试 check 命令"""
+    print("测试 check 命令...")
+    
+    # 在示例项目目录中测试 check 命令
+    example_dir = Path(__file__).parent.parent / "examples" / "ofweek_standalone"
+    
+    # 测试普通 check 命令
+    code, stdout, stderr = run_command("python -m crawlo.cli check", cwd=example_dir)
+    # check 命令可能会因为环境问题失败，但我们检查是否有输出
+    assert code == 0 or len(stdout) > 0, f"check 命令失败: {stderr}"
+    
+    print("✅ check 命令测试通过")
+
+def test_stats_command():
+    """测试 stats 命令"""
+    print("测试 stats 命令...")
+    
+    # 在示例项目目录中测试 stats 命令
+    example_dir = Path(__file__).parent.parent / "examples" / "ofweek_standalone"
+    
+    # 测试普通 stats 命令
+    code, stdout, stderr = run_command("python -m crawlo.cli stats", cwd=example_dir)
+    # stats 命令可能会因为没有统计数据而返回非0，但我们检查是否有输出
+    assert code == 0 or len(stdout) > 0, f"stats 命令失败: {stderr}"
+    
+    print("✅ stats 命令测试通过")
+
+def main():
+    """主函数"""
+    print("开始测试所有 crawlo 命令...")
+    print("=" * 50)
+    
+    try:
+        # 测试 help 命令
+        test_help_command()
+        print()
+        
+        # 测试 version 命令
+        test_version_command()
+        print()
+        
+        # 测试 startproject 命令
+        test_startproject_command()
+        print()
+        
+        # 测试 genspider 命令
+        test_genspider_command()
+        print()
+        
+        # 测试 list 命令
+        test_list_command()
+        print()
+        
+        # 测试 check 命令
+        test_check_command()
+        print()
+        
+        # 测试 stats 命令
+        test_stats_command()
+        print()
+        
+        print("=" * 50)
+        print("所有命令测试通过！")
+        
+    except Exception as e:
+        print("=" * 50)
+        print(f"测试失败: {e}")
+        import traceback
+        traceback.print_exc()
+        return 1
+    
+    return 0
+
+if __name__ == "__main__":
+    sys.exit(main())

tests/test_all_redis_key_configs.py

@@ -1,146 +1,146 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-"""
-所有Redis Key配置测试脚本
-用于验证所有配置文件是否符合新的Redis key命名规范
-"""
-import sys
-import os
-import re
-
-# 添加项目根目录到路径
-sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..'))
-
-
-def test_all_redis_key_configs():
-    """测试所有Redis key配置"""
-    print("测试所有Redis key配置...")
-    
-    try:
-        # 检查示例项目配置文件
-        example_projects = [
-            "examples/books_distributed/books_distributed/settings.py",
-            "examples/api_data_collection/api_data_collection/settings.py",
-            "examples/telecom_licenses_distributed/telecom_licenses_distributed/settings.py"
-        ]
-        
-        for project_config in example_projects:
-            print(f"   检查 {project_config}...")
-            if not os.path.exists(project_config):
-                print(f"配置文件不存在: {project_config}")
-                return False
-                
-            with open(project_config, 'r', encoding='utf-8') as f:
-                content = f.read()
-            
-            # 检查是否移除了旧的REDIS_KEY配置
-            if re.search(r'REDIS_KEY\s*=', content) and 'crawlo:{PROJECT_NAME}:filter:fingerprint' not in content:
-                print(f"{project_config}中仍然存在旧的REDIS_KEY配置")
-                return False
-                
-            # 检查是否添加了新的注释
-            if 'crawlo:{PROJECT_NAME}:filter:fingerprint' not in content:
-                print(f"{project_config}中缺少新的Redis key命名规范注释")
-                return False
-                
-            print(f"      {project_config}符合新的Redis key命名规范")
-        
-        # 检查模板文件
-        template_file = "crawlo/templates/project/settings.py.tmpl"
-        print(f"   检查 {template_file}...")
-        if not os.path.exists(template_file):
-            print(f"模板文件不存在: {template_file}")
-            return False
-            
-        with open(template_file, 'r', encoding='utf-8') as f:
-            template_content = f.read()
-        
-        # 检查是否移除了旧的REDIS_KEY配置
-        if "REDIS_KEY = f'{{project_name}}:fingerprint'" in template_content:
-            print("模板文件中仍然存在旧的REDIS_KEY配置")
-            return False
-            
-        # 检查是否添加了新的注释
-        if '# crawlo:{project_name}:filter:fingerprint (请求去重)' not in template_content:
-            print("模板文件中缺少请求去重的Redis key命名规范注释")
-            return False
-            
-        if '# crawlo:{project_name}:item:fingerprint (数据项去重)' not in template_content:
-            print("模板文件中缺少数据项去重的Redis key命名规范注释")
-            return False
-            
-        print(f"      {template_file}符合新的Redis key命名规范")
-        
-        # 检查mode_manager.py
-        mode_manager_file = "crawlo/mode_manager.py"
-        print(f"   检查 {mode_manager_file}...")
-        if not os.path.exists(mode_manager_file):
-            print(f"文件不存在: {mode_manager_file}")
-            return False
-            
-        with open(mode_manager_file, 'r', encoding='utf-8') as f:
-            mode_manager_content = f.read()
-        
-        # 检查是否移除了旧的REDIS_KEY配置
-        if "'REDIS_KEY': f'{project_name}:fingerprint'" in mode_manager_content:
-            print("mode_manager.py中仍然存在旧的REDIS_KEY配置")
-            return False
-            
-        # 检查是否添加了新的注释
-        if 'crawlo:{project_name}:filter:fingerprint (请求去重)' not in mode_manager_content:
-            print("mode_manager.py中缺少新的Redis key命名规范注释")
-            return False
-            
-        print(f"      {mode_manager_file}符合新的Redis key命名规范")
-        
-        # 检查默认设置文件
-        default_settings_file = "crawlo/settings/default_settings.py"
-        print(f"   检查 {default_settings_file}...")
-        if not os.path.exists(default_settings_file):
-            print(f"文件不存在: {default_settings_file}")
-            return False
-            
-        with open(default_settings_file, 'r', encoding='utf-8') as f:
-            default_settings_content = f.read()
-        
-        # 检查是否移除了旧的REDIS_KEY配置
-        if re.search(r'REDIS_KEY\s*=\s*.*fingerprint', default_settings_content):
-            print("默认设置文件中仍然存在旧的REDIS_KEY配置")
-            return False
-            
-        print(f"      {default_settings_file}符合新的Redis key命名规范")
-        
-        print("所有Redis key配置测试通过！")
-        return True
-        
-    except Exception as e:
-        print(f"测试过程中发生错误: {e}")
-        return False
-
-
-def main():
-    """主测试函数"""
-    print("开始所有Redis key配置测试...")
-    print("=" * 50)
-    
-    try:
-        success = test_all_redis_key_configs()
-        
-        print("=" * 50)
-        if success:
-            print("所有测试通过！所有配置文件符合新的Redis key命名规范")
-        else:
-            print("测试失败，请检查配置文件")
-            return 1
-            
-    except Exception as e:
-        print("=" * 50)
-        print(f"测试过程中发生异常: {e}")
-        return 1
-    
-    return 0
-
-
-if __name__ == "__main__":
-    exit_code = main()
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+所有Redis Key配置测试脚本
+用于验证所有配置文件是否符合新的Redis key命名规范
+"""
+import sys
+import os
+import re
+
+# 添加项目根目录到路径
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..'))
+
+
+def test_all_redis_key_configs():
+    """测试所有Redis key配置"""
+    print("测试所有Redis key配置...")
+    
+    try:
+        # 检查示例项目配置文件
+        example_projects = [
+            "examples/books_distributed/books_distributed/settings.py",
+            "examples/api_data_collection/api_data_collection/settings.py",
+            "examples/telecom_licenses_distributed/telecom_licenses_distributed/settings.py"
+        ]
+        
+        for project_config in example_projects:
+            print(f"   检查 {project_config}...")
+            if not os.path.exists(project_config):
+                print(f"配置文件不存在: {project_config}")
+                return False
+                
+            with open(project_config, 'r', encoding='utf-8') as f:
+                content = f.read()
+            
+            # 检查是否移除了旧的REDIS_KEY配置
+            if re.search(r'REDIS_KEY\s*=', content) and 'crawlo:{PROJECT_NAME}:filter:fingerprint' not in content:
+                print(f"{project_config}中仍然存在旧的REDIS_KEY配置")
+                return False
+                
+            # 检查是否添加了新的注释
+            if 'crawlo:{PROJECT_NAME}:filter:fingerprint' not in content:
+                print(f"{project_config}中缺少新的Redis key命名规范注释")
+                return False
+                
+            print(f"      {project_config}符合新的Redis key命名规范")
+        
+        # 检查模板文件
+        template_file = "crawlo/templates/project/settings.py.tmpl"
+        print(f"   检查 {template_file}...")
+        if not os.path.exists(template_file):
+            print(f"模板文件不存在: {template_file}")
+            return False
+            
+        with open(template_file, 'r', encoding='utf-8') as f:
+            template_content = f.read()
+        
+        # 检查是否移除了旧的REDIS_KEY配置
+        if "REDIS_KEY = f'{{project_name}}:fingerprint'" in template_content:
+            print("模板文件中仍然存在旧的REDIS_KEY配置")
+            return False
+            
+        # 检查是否添加了新的注释
+        if '# crawlo:{project_name}:filter:fingerprint (请求去重)' not in template_content:
+            print("模板文件中缺少请求去重的Redis key命名规范注释")
+            return False
+            
+        if '# crawlo:{project_name}:item:fingerprint (数据项去重)' not in template_content:
+            print("模板文件中缺少数据项去重的Redis key命名规范注释")
+            return False
+            
+        print(f"      {template_file}符合新的Redis key命名规范")
+        
+        # 检查mode_manager.py
+        mode_manager_file = "crawlo/mode_manager.py"
+        print(f"   检查 {mode_manager_file}...")
+        if not os.path.exists(mode_manager_file):
+            print(f"文件不存在: {mode_manager_file}")
+            return False
+            
+        with open(mode_manager_file, 'r', encoding='utf-8') as f:
+            mode_manager_content = f.read()
+        
+        # 检查是否移除了旧的REDIS_KEY配置
+        if "'REDIS_KEY': f'{project_name}:fingerprint'" in mode_manager_content:
+            print("mode_manager.py中仍然存在旧的REDIS_KEY配置")
+            return False
+            
+        # 检查是否添加了新的注释
+        if 'crawlo:{project_name}:filter:fingerprint (请求去重)' not in mode_manager_content:
+            print("mode_manager.py中缺少新的Redis key命名规范注释")
+            return False
+            
+        print(f"      {mode_manager_file}符合新的Redis key命名规范")
+        
+        # 检查默认设置文件
+        default_settings_file = "crawlo/settings/default_settings.py"
+        print(f"   检查 {default_settings_file}...")
+        if not os.path.exists(default_settings_file):
+            print(f"文件不存在: {default_settings_file}")
+            return False
+            
+        with open(default_settings_file, 'r', encoding='utf-8') as f:
+            default_settings_content = f.read()
+        
+        # 检查是否移除了旧的REDIS_KEY配置
+        if re.search(r'REDIS_KEY\s*=\s*.*fingerprint', default_settings_content):
+            print("默认设置文件中仍然存在旧的REDIS_KEY配置")
+            return False
+            
+        print(f"      {default_settings_file}符合新的Redis key命名规范")
+        
+        print("所有Redis key配置测试通过！")
+        return True
+        
+    except Exception as e:
+        print(f"测试过程中发生错误: {e}")
+        return False
+
+
+def main():
+    """主测试函数"""
+    print("开始所有Redis key配置测试...")
+    print("=" * 50)
+    
+    try:
+        success = test_all_redis_key_configs()
+        
+        print("=" * 50)
+        if success:
+            print("所有测试通过！所有配置文件符合新的Redis key命名规范")
+        else:
+            print("测试失败，请检查配置文件")
+            return 1
+            
+    except Exception as e:
+        print("=" * 50)
+        print(f"测试过程中发生异常: {e}")
+        return 1
+    
+    return 0
+
+
+if __name__ == "__main__":
+    exit_code = main()
     sys.exit(exit_code)