auto-coder 0.1.266__py3-none-any.whl → 0.1.268__py3-none-any.whl

{auto_coder-0.1.266.dist-info → auto_coder-0.1.268.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.266
+Version: 0.1.268
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence

{auto_coder-0.1.266.dist-info → auto_coder-0.1.268.dist-info}/RECORD

@@ -6,12 +6,12 @@ autocoder/auto_coder_rag_client_mcp.py,sha256=QRxUbjc6A8UmDMQ8lXgZkjgqtq3lgKYeat
 autocoder/auto_coder_rag_mcp.py,sha256=-RrjNwFaS2e5v8XDIrKR-zlUNUE8UBaeOtojffBrvJo,8521
 autocoder/auto_coder_server.py,sha256=6YQweNEKUrGAZ3yPvw8_qlNZJYLVSVUXGrn1K6udLts,20413
 autocoder/benchmark.py,sha256=Ypomkdzd1T3GE6dRICY3Hj547dZ6_inqJbBJIp5QMco,4423
-autocoder/chat_auto_coder.py,sha256=fX_VusqljMF6Mq4TIrnFEehksHlJzc_g9S-5ctOXVbk,114803
+autocoder/chat_auto_coder.py,sha256=X6jlyvABAPNLAx2cnuV8O6XjKPqX6Kl7nv6iEuALm-4,115960
 autocoder/chat_auto_coder_lang.py,sha256=ShOQVOnMA-WlT-fB9OrOer-xQkbcWxJGl-WMPuZcUkM,19572
 autocoder/command_args.py,sha256=9aYJ-AmPxP1sQh6ciw04FWHjSn31f2W9afXFwo8wgx4,30441
 autocoder/lang.py,sha256=U6AjVV8Rs1uLyjFCZ8sT6WWuNUxMBqkXXIOs4S120uk,14511
 autocoder/models.py,sha256=PlG1tKHSHwB57cKLOl5gTl5yTzFUDzCgeHPJU3N9F6Q,9106
-autocoder/version.py,sha256=HfowxCQGSKrxyoyS64Th4zdCNDXYWxSKkcfQj7AFt24,23
+autocoder/version.py,sha256=5v4jQYuNlNWVifBwTlRwUJGC6N6-LuHjzLm1nfAkQQM,23
 autocoder/agent/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/agent/auto_demand_organizer.py,sha256=NWSAEsEk94vT3lGjfo25kKLMwYdPcpy9e-i21txPasQ,6942
 autocoder/agent/auto_filegroup.py,sha256=CW7bqp0FW1GIEMnl-blyAc2UGT7O9Mom0q66ITz1ckM,6635
@@ -24,7 +24,7 @@ autocoder/agent/planner.py,sha256=SZTSZHxHzDmuWZo3K5fs79RwvJLWurg-nbJRRNbX65o,91
 autocoder/agent/project_reader.py,sha256=tWLaPoLw1gI6kO_NzivQj28KbobU2ceOLuppHMbfGl8,18234
 autocoder/chat/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/commands/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-autocoder/commands/auto_command.py,sha256=dNtoWuPgZvZxJn7kt_7-TUnAJT4dUte9IEVcCmeG4mw,51906
+autocoder/commands/auto_command.py,sha256=G4SAmwKhoOxNS2JztUlXCyS3PWhXFhyLK_s22PtltwM,52894
 autocoder/commands/tools.py,sha256=xDhGD1jRN67fGfeck33pM74TxXlAMeo49S3Q9K-VKco,20107
 autocoder/common/JupyterClient.py,sha256=O-wi6pXeAEYhAY24kDa0BINrLYvKS6rKyWe98pDClS0,2816
 autocoder/common/ShellClient.py,sha256=fM1q8t_XMSbLBl2zkCNC2J9xuyKN3eXzGm6hHhqL2WY,2286
@@ -32,8 +32,8 @@ autocoder/common/__init__.py,sha256=eH5NyEsiCzJaNzsxI-Y8TJbB5kpFzuwBgbTrWz9Uthk,
 autocoder/common/anything2images.py,sha256=0ILBbWzY02M-CiWB-vzuomb_J1hVdxRcenAfIrAXq9M,25283
 autocoder/common/anything2img.py,sha256=4TREa-sOA-iargieUy7MpyCYVUE-9Mmq0wJtwomPqnE,7662
 autocoder/common/audio.py,sha256=Kn9nWKQddWnUrAz0a_ZUgjcu4VUU_IcZBigT7n3N3qc,7439
-autocoder/common/auto_coder_lang.py,sha256=z_tI1eNJQiSRy_RL-x4KeQDByGvUE4OJxthoTkBXKyQ,29125
-autocoder/common/auto_configure.py,sha256=tdEwfycZUjomZAgps1GOCtocYEtfuUgRksYPFHBP_bs,12211
+autocoder/common/auto_coder_lang.py,sha256=74VM_RXNrUXEEyuIr6pXf5xbWQId3ZWhx9NDzQbGquw,29450
+autocoder/common/auto_configure.py,sha256=y0fTplwUdtzVF91iOdZh8zl10MdYFjAquO_Kmj1BWqs,12300
 autocoder/common/buildin_tokenizer.py,sha256=L7d5t39ZFvUd6EoMPXUhYK1toD0FHlRH1jtjKRGokWU,1236
 autocoder/common/chunk_validation.py,sha256=BrR_ZWavW8IANuueEE7hS8NFAwEvm8TX34WnPx_1hs8,3030
 autocoder/common/cleaner.py,sha256=NU72i8C6o9m0vXExab7nao5bstBUsfJFcj11cXa9l4U,1089
@@ -53,7 +53,7 @@ autocoder/common/command_templates.py,sha256=BrmRwOJnyuMETY8v5AQw9D00UQb7ql00BiG
 autocoder/common/conf_import_export.py,sha256=w__WsIobe6nmsGns2pV-laU7R5ZvtQNuIbXebxhbY7A,3967
 autocoder/common/conf_validator.py,sha256=EzSmadpZ22D9e8iWmfeWodUeYJt0IgMoaAOmCleXliI,8795
 autocoder/common/const.py,sha256=eTjhjh4Aj4CUzviJ81jaf3Y5cwqsLATySn2wJxaS6RQ,2911
-autocoder/common/context_pruner.py,sha256=yaDAMX8Zgfgl666gp1Rwxd7I7jmZXtEKaf3_TebJ0I4,12021
+autocoder/common/context_pruner.py,sha256=zLY7VllI3XG8icv8WtvDyV7GLOV1qnO3eBCY32ueH4U,12513
 autocoder/common/conversation_pruner.py,sha256=mdMpTpTdPJl8f0UjC1TGKRiYtDc1o6QQD0nYPR9yp1c,5628
 autocoder/common/files.py,sha256=2-9CJwOZtyWkk2TQM4gPSkpJ3_cwb-l_3sdsCd1H5GQ,3380
 autocoder/common/git_utils.py,sha256=qeuF_IB3G3M72asHxWokROU3hINCuFA1nar-UtF9wIU,26022
@@ -93,9 +93,9 @@ autocoder/dispacher/actions/plugins/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQ
 autocoder/dispacher/actions/plugins/action_regex_project.py,sha256=AqGIkjbqV1eOS3vBoZUTSOpyOlkv1p5h35mI2Kcvekw,6906
 autocoder/dispacher/actions/plugins/action_translate.py,sha256=GEn7dZA22jy5WyzINomjmzzB795p2Olg-CJla97lRF8,7744
 autocoder/index/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-autocoder/index/entry.py,sha256=2XH5X46nG8g-Q7RVIo9BIXZNP44pPjGk48XzqFqJ5GM,13405
+autocoder/index/entry.py,sha256=WsJrOf586GAiHacFCdcXQx5MmGKQXIIeEqo5h46AHpE,13474
 autocoder/index/for_command.py,sha256=BFvljE4t6VaMBGboZAuhUCzVK0EitCy_n5D_7FEnihw,3204
-autocoder/index/index.py,sha256=lott5kcNN0Q5iyfl2dgkOSeDnsMpODKVWV3H54ISaiI,25507
+autocoder/index/index.py,sha256=1ZNw4s_Fq_ARqjnfU0Vmo1mvdzM6sjaLspjcfinTEh4,27350
 autocoder/index/symbols_utils.py,sha256=_EP7E_qWXxluAxq3FGZLlLfdrfwx3FmxCdulI8VGuac,2244
 autocoder/index/types.py,sha256=a2s_KV5FJlq7jqA2ELSo9E1sjuLwDB-JJYMhSpzBAhU,596
 autocoder/index/filter/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -161,9 +161,9 @@ autocoder/utils/types.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/utils/auto_coder_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/utils/auto_coder_utils/chat_stream_out.py,sha256=lkJ_A-sYU36JMzjFWkk3pR6uos8oZHYt9GPsPe_CPAo,11766
 autocoder/utils/chat_auto_coder_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-auto_coder-0.1.266.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
-auto_coder-0.1.266.dist-info/METADATA,sha256=d5ObaF7aDGHJmHEh9RKQH1UK286V-EPZ8G0tFrthE0Q,2616
-auto_coder-0.1.266.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-auto_coder-0.1.266.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
-auto_coder-0.1.266.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
-auto_coder-0.1.266.dist-info/RECORD,,
+auto_coder-0.1.268.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
+auto_coder-0.1.268.dist-info/METADATA,sha256=0SYv8MrsdLqjdHnrRix68X6NoyysothptSo8AfVHghE,2616
+auto_coder-0.1.268.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+auto_coder-0.1.268.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
+auto_coder-0.1.268.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
+auto_coder-0.1.268.dist-info/RECORD,,

autocoder/chat_auto_coder.py

@@ -133,8 +133,7 @@ commands = [
     "/index/build",
     "/index/export",
     "/index/import",
-    "/exclude_dirs",
-    "/exclude_files",
+    "/exclude_files",        
     "/help",
     "/shell",
     "/voice_input",
@@ -148,6 +147,7 @@ commands = [
     "/auto",
     "/conf/export",
     "/conf/import",
+    "/exclude_dirs",
 ]
 
 
@@ -1507,7 +1507,7 @@ def coding(query: str):
             "silence": conf.get("silence", "true") == "true",
             "include_project_structure": conf.get("include_project_structure", "true")
             == "true",
-            "exclude_files": conf.get("exclude_files", []),
+            "exclude_files": memory.get("exclude_files", []),
         }
 
         yaml_config["context"] = ""        
@@ -1636,7 +1636,7 @@ def chat(query: str):
         "skip_build_index": conf.get("skip_build_index", "true") == "true",
         "skip_confirm": conf.get("skip_confirm", "true") == "true",
         "silence": conf.get("silence", "true") == "true",
-        "exclude_files": conf.get("exclude_files", []),
+        "exclude_files": memory.get("exclude_files", []),
     }
 
     current_files = memory["current_files"]["files"] + get_llm_friendly_package_docs(
@@ -2284,10 +2284,44 @@ def exclude_dirs(dir_names: List[str]):
     save_memory()
     completer.refresh_files()
 
-def exclude_files(file_patterns: List[str]):
+def exclude_files(query: str):
     result_manager = ResultManager()
     printer = Printer()
-    new_file_patterns = file_patterns
+    if "/list" in query:
+        query = query.replace("/list", "", 1).strip()
+        existing_file_patterns = memory.get("exclude_files", []) 
+        console = Console()
+        # 打印表格
+        table = Table(title="Exclude Files")
+        table.add_column("File Pattern")
+        for file_pattern in existing_file_patterns:
+            table.add_row(file_pattern)
+        console.print(table)
+        result_manager.add_result(content=f"Exclude files: {existing_file_patterns}",meta={
+            "action": "exclude_files",
+            "input": {
+                "query": query
+            }
+        })
+        return
+
+    if "/drop" in query:
+        query = query.replace("/drop", "", 1).strip()
+        existing_file_patterns = memory.get("exclude_files", [])    
+        existing_file_patterns.remove(query.strip())
+        memory["exclude_files"] = existing_file_patterns
+        save_memory()
+        completer.refresh_files()
+        result_manager.add_result(content=f"Dropped exclude files: {query}",meta={
+            "action": "exclude_files",
+            "input": {
+                "query": query
+            }
+        })
+        return
+    
+    new_file_patterns = query.strip().split(",")
+    
     existing_file_patterns = memory.get("exclude_files", [])    
     file_patterns_to_add = [f for f in new_file_patterns if f not in existing_file_patterns]
 
@@ -2383,7 +2417,7 @@ def help(query: str):
     args = get_final_config()
     product_mode = memory.get("product_mode", "lite")
     llm = get_single_llm(args.chat_model or args.model, product_mode=product_mode)
-    auto_config_tuner = ConfigAutoTuner(llm=llm, memory_config=MemoryConfig(memory=memory, save_memory_func=save_memory))
+    auto_config_tuner = ConfigAutoTuner(args=args, llm=llm, memory_config=MemoryConfig(memory=memory, save_memory_func=save_memory))
     auto_config_tuner.tune(AutoConfigRequest(query=query))    
 
 @run_in_raw_thread()
@@ -2924,9 +2958,8 @@ def main():
                 exclude_dirs(dir_names)
 
             elif user_input.startswith("/exclude_files"):
-                file_patterns = user_input[len(
-                    "/exclude_files"):].strip().split(",")
-                exclude_files(file_patterns)
+                query = user_input[len("/exclude_files"):].strip()                
+                exclude_files(query)
 
             elif user_input.startswith("/ask"):
                 query = user_input[len("/ask"):].strip()

autocoder/commands/auto_command.py

@@ -213,7 +213,11 @@ class CommandAutoTuner:
         通过 get_project_structure 来获取项目结构，然后通过 get_project_map 来获取你想看的某个文件的用途，符号列表，最后再通过 read_files/read_file_with_keyword_ranges 函数来读取文件内容,确认对应的功能是否在相关的文件里。
         5. 调用 coding 函数的时候，尽可能多的 @文件和@@符号，让需求更加清晰明了，建议多描述具体怎么完成对应的需求。
         6. 对于代码需求设计，尽可能使用 chat 函数。
-        7. 如果成功执行了 coding 函数，最好再调用一次 chat("/review /commit")
+        7. 如果成功执行了 coding 函数，最好再调用一次 chat("/review /commit")    
+        8. 我们所有的对话不能超过 {{ conversation_safe_zone_tokens }} 个tokens,当你读取索引文件 (get_project_map) 的时候，你可以看到
+        每个文件的tokens数，你可以根据这个信息来决定如何读取这个文件。比如对于很小的文件，那么可以直接全部读取，
+        而对于分析一个超大文件推荐组合 read_files 带上 line_ranges 参数来读取，或者组合 read_file_withread_file_with_keyword_ranges 等来读取，
+        每个函数你还可以使用多次来获取更多信息。
         </function_combination_readme>
 
 
@@ -262,7 +266,8 @@ class CommandAutoTuner:
             "current_conf": json.dumps(self.memory_config.memory["conf"], indent=2),        
             "env_info": env_info,
             "shell_type": shells.get_terminal_name(),
-            "shell_encoding": shells.get_terminal_encoding()
+            "shell_encoding": shells.get_terminal_encoding(),
+            "conversation_safe_zone_tokens": self.args.conversation_prune_safe_zone_tokens
         }
     
     @byzerllm.prompt()
@@ -274,7 +279,7 @@ class CommandAutoTuner:
         
         <function_result>
         {{ result }}
-        </function_result>
+        </function_result>                
 
         请根据命令执行结果以及前面的对话，返回下一个函数。
         
@@ -283,7 +288,8 @@ class CommandAutoTuner:
         2. 你最多尝试 {{ auto_command_max_iterations }} 次，如果 {{ auto_command_max_iterations }} 次都没有满足要求，则不要返回任何函数，确保 suggestions 为空。
         '''   
         return {
-            "auto_command_max_iterations": self.args.auto_command_max_iterations
+            "auto_command_max_iterations": self.args.auto_command_max_iterations,
+            "conversation_safe_zone_tokens": self.args.conversation_prune_safe_zone_tokens
         } 
     
     def analyze(self, request: AutoCommandRequest) -> AutoCommandResponse:
@@ -302,8 +308,7 @@ class CommandAutoTuner:
             # 提取 JSON 并转换为 AutoCommandResponse
             try:
                 response = to_model(content, AutoCommandResponse)  
-                if response.suggestions:
-                    print(response)
+                if response.suggestions:                    
                     command = response.suggestions[0].command
                     parameters = response.suggestions[0].parameters                    
                     if parameters:
@@ -507,11 +512,21 @@ class CommandAutoTuner:
           如果你检测到用户的coding执行结果，缺少必要的文件修改，可以尝试使用该函数先添加文件再执行coding。
         </description>
         <usage>
-         该方法只有一个参数 args，args 是一个列表，列表的元素是字符串。会包含子指令，例如 
-         
-         add_files(args=["/refresh"]) 
-        
-         会刷新文件列表。下面是常见的子指令：
+         该方法只有一个参数 args，args 是一个列表，列表的元素是字符串。
+
+         如果没有包含子指令，单纯的添加文件，那么 args 列表的元素是文件路径，注意我们需要使用绝对路径。
+
+         使用例子：
+
+         add_files(args=["/absolute/path/to/file1.py"])
+
+         也支持glob 语法，例如：
+
+         add_files(args=["**/*.py"])
+
+         这样会把项目根目录下的所有.py文件添加到活跃区，尽量确保少的添加文件。
+
+         如果是有子指令，参考下面是常见的子指令说明。
 
          ## /refresh 刷新文件列表
          刷新文件列表
@@ -757,6 +772,8 @@ class CommandAutoTuner:
          coding(query="@auth.py 添加JWT认证")
          coding(query="@@login 优化错误处理")
          coding(query="<img>design/flow.png</img> 实现这个流程图的功能")
+
+         在使用 coding 函数时，建议通过 ask_user 来确认是否执行 coding 函数，除非用户明确说不要询问，直接执行。
         </usage>
         </command>
 
@@ -974,23 +991,7 @@ class CommandAutoTuner:
          感兴趣，可以配合 read_files 函数来读取文件内容，从而帮你做更好的决策
              
         </usage>
-        </command>
-
-        <command>
-        <name>get_related_files</name>
-        <description>根据类名、函数名或文件用途描述，返回项目中相关的文件。</description>
-        <usage>
-         该命令接受一个参数 query，为要查询的符号或描述字符串。
-         
-         使用例子：
-         
-         get_related_files(query="用户登录功能")
-         
-         注意：
-         - 返回值为逗号分隔的文件路径列表
-         - 只能返回已被索引的文件
-        </usage>
-        </command>
+        </command>        
 
         <command>
         <name>get_project_map</name>
@@ -1170,15 +1171,23 @@ class CommandAutoTuner:
         <name>exclude_files</name>
         <description>排除指定文件。</description>
         <usage>
-         该命令接受一个参数 file_patterns, 为要排除的文件模式字符串列表。
+         该命令接受一个参数 query, 为要排除的文件模式字符串,多个文件模式用逗号分隔。
          
          使用例子,比如你想要排除 package-lock.json 文件，你可以这样调用：
         
-         exclude_files(file_patterns=["regex://.*/package-lock\.json"])
+         exclude_files(query="regex://.*/package-lock\.json")
          
          注意：
          - 文件模式字符串必须以 regex:// 开头
          - regex:// 后面部分是标准的正则表达式         
+
+         也支持子命令：
+         /list 列出当前排除的文件模式
+         /drop 删除指定的文件模式
+
+         使用例子：
+         exclude_files(query="/list")
+         exclude_files(query="/drop regex://.*/package-lock\.json")
         </usage>
         </command>
         </commands>

autocoder/common/auto_coder_lang.py

@@ -3,6 +3,7 @@ from byzerllm.utils import format_str_jinja2
 
 MESSAGES = {
     "en": {
+        "file_scored_message": "File scored: {{file_path}} - Score: {{score}}",
         "invalid_file_pattern": "Invalid file pattern: {{file_pattern}}. e.g. regex://.*/package-lock\\.json",
         "config_validation_error": "Config validation error: {{error}}",
         "invalid_boolean_value": "Value '{{value}}' is not a valid boolean(true/false)",
@@ -159,9 +160,11 @@ MESSAGES = {
         "index_export_success": "Index exported successfully: {{path}}",
         "index_import_success": "Index imported successfully: {{path}}",
         "edits_title": "edits",
-        "diff_blocks_title":"diff blocks"
+        "diff_blocks_title":"diff blocks",
+        "index_exclude_files_error": "index filter exclude files fail: {{ error }}"
     },
     "zh": {
+        "file_scored_message": "文件评分: {{file_path}} - 分数: {{score}}",
         "invalid_file_pattern": "无效的文件模式: {{file_pattern}}. 例如: regex://.*/package-lock\\.json",
         "conf_not_found": "未找到配置文件: {{path}}",
         "conf_import_success": "成功导入配置: {{path}}",
@@ -317,6 +320,7 @@ MESSAGES = {
         "index_import_success": "索引导入成功: {{path}}",
         "edits_title": "编辑块",
         "diff_blocks_title": "差异块",
+        "index_exclude_files_error": "索引排除文件时出错: {{error}}"
     }}
 
 

autocoder/common/auto_configure.py

@@ -14,6 +14,8 @@ from byzerllm.utils.str2model import to_model
 from autocoder.utils.auto_coder_utils.chat_stream_out import stream_out
 from autocoder.common.result_manager import ResultManager
 from autocoder.utils import llms as llms_utils
+from autocoder.common import AutoCoderArgs
+
 logger = logging.getLogger(__name__)
 
 class ConfigMessage(BaseModel):
@@ -120,9 +122,10 @@ class AutoConfigResponse(BaseModel):
     reasoning: str = ""        
 
 class ConfigAutoTuner:
-    def __init__(self, llm: Union[byzerllm.ByzerLLM, byzerllm.SimpleByzerLLM], memory_config: MemoryConfig):
+    def __init__(self,args: AutoCoderArgs, llm: Union[byzerllm.ByzerLLM, byzerllm.SimpleByzerLLM], memory_config: MemoryConfig):
         self.llm = llm
         self.memory_config = memory_config
+        self.args = args
             
 
     def configure(self, conf: str, skip_print: bool = False) -> None:

autocoder/common/context_pruner.py

@@ -9,11 +9,15 @@ from autocoder.index.types import VerifyFileRelevance
 import byzerllm
 from concurrent.futures import ThreadPoolExecutor, as_completed
 
+from autocoder.common.printer import Printer
+from autocoder.common.auto_coder_lang import get_message_with_format
+
 class PruneContext:
     def __init__(self, max_tokens: int, args: AutoCoderArgs, llm: Union[byzerllm.ByzerLLM, byzerllm.SimpleByzerLLM]):
         self.max_tokens = max_tokens
         self.args = args
         self.llm = llm
+        self.printer = Printer()
 
     def _delete_overflow_files(self, file_paths: List[str]) -> List[SourceCode]:
         """直接删除超出 token 限制的文件"""
@@ -201,7 +205,7 @@ class PruneContext:
         total_tokens,sources = self._count_tokens(file_paths)
         if total_tokens <= self.max_tokens:
             return sources
-
+        # print(f"total_tokens: {total_tokens} {self.max_tokens}, 进行策略: {strategy}")
         if strategy == "score":            
             return self._score_and_filter_files(file_paths, conversations)
         if strategy == "delete":
@@ -281,9 +285,15 @@ class PruneContext:
         with ThreadPoolExecutor() as executor:
             futures = [executor.submit(_score_file, file_path) for file_path in file_paths]
             for future in as_completed(futures):
-                result = future.result()
-                print(f"score file {result['file_path']} {result['score']}")
+                result = future.result()                
                 if result:
+                    self.printer.print_str_in_terminal(
+                        get_message_with_format(
+                            "file_scored_message",
+                            file_path=result["file_path"],
+                            score=result["score"]
+                        )
+                    )
                     scored_files.append(result)
 
         # 第二步：按分数从高到低排序

autocoder/index/entry.py

@@ -128,8 +128,9 @@ def build_index_and_filter_files(
                 model_name = "unknown(without default model name)" 
             printer.print_in_terminal("normal_filter_start", style="blue",model_name=model_name)
             normal_filter = NormalFilter(index_manager,stats,sources)
+            normal_filter_result = normal_filter.filter(index_manager.read_index(),args.query)
             # Merge normal filter results into final_files
-            final_files.update(normal_filter.filter(index_manager.read_index(),args.query))
+            final_files.update(normal_filter_result.files)
         
 
     def display_table_and_get_selections(data):
@@ -212,7 +213,7 @@ def build_index_and_filter_files(
 
     # Phase 6: File selection and limitation
     printer.print_in_terminal("phase6_file_selection")
-    phase_start = time.monotonic()
+    phase_start = time.monotonic()    
 
     if args.index_filter_file_num > 0:
         logger.info(
@@ -278,7 +279,7 @@ def build_index_and_filter_files(
             event=CommunicateEvent(
                 event_type=CommunicateEventType.CODE_INDEX_FILTER_FILE_SELECTED.value,
                 data=json.dumps([
-                    (file.file_path, file.reason)
+                    (file["file_path"], file.reason)
                     for file in final_files.values()
                     if file.file_path in depulicated_sources
                 ])

autocoder/index/index.py

@@ -12,6 +12,7 @@ from autocoder.index.symbols_utils import (
 from autocoder.privacy.model_filter import ModelPathFilter
 from concurrent.futures import ThreadPoolExecutor, as_completed
 import threading
+import re
 
 import byzerllm
 import hashlib
@@ -27,6 +28,8 @@ from autocoder.index.types import (
 from autocoder.common.global_cancel import global_cancel
 from autocoder.utils.llms import get_llm_names
 from autocoder.rag.token_counter import count_tokens
+
+
 class IndexManager:
     def __init__(
         self, llm: byzerllm.ByzerLLM, sources: List[SourceCode], args: AutoCoderArgs
@@ -52,12 +55,14 @@ class IndexManager:
             self.index_filter_llm = llm
 
         self.llm = llm
-        
+
         # Initialize model filters
         if self.index_llm:
-            self.index_model_filter = ModelPathFilter.from_model_object(self.index_llm, args)
+            self.index_model_filter = ModelPathFilter.from_model_object(
+                self.index_llm, args)
         if self.index_filter_llm:
-            self.index_filter_model_filter = ModelPathFilter.from_model_object(self.index_filter_llm, args)
+            self.index_filter_model_filter = ModelPathFilter.from_model_object(
+                self.index_filter_llm, args)
         self.args = args
         self.max_input_length = (
             args.index_model_max_input_length or args.model_max_input_length
@@ -68,7 +73,6 @@ class IndexManager:
         if not os.path.exists(self.index_dir):
             os.makedirs(self.index_dir)
 
-    
     @byzerllm.prompt()
     def verify_file_relevance(self, file_content: str, query: str) -> str:
         """
@@ -201,12 +205,12 @@ class IndexManager:
         if current_chunk:
             chunks.append("\n".join(current_chunk))
         return chunks
-    
+
     def should_skip(self, file_path: str):
         ext = os.path.splitext(file_path)[1].lower()
         if ext in [".md", ".html", ".txt", ".doc", ".pdf"]:
             return True
-            
+
         # Check model filter restrictions
         if self.index_model_filter and not self.index_model_filter.is_accessible(file_path):
             self.printer.print_in_terminal(
@@ -216,10 +220,10 @@ class IndexManager:
                 model_name=",".join(get_llm_names(self.index_llm))
             )
             return True
-            
+
         return False
 
-    def build_index_for_single_source(self, source: SourceCode):   
+    def build_index_for_single_source(self, source: SourceCode):
         if global_cancel.requested:
             return None
 
@@ -251,13 +255,13 @@ class IndexManager:
 
             start_time = time.monotonic()
             source_code = source.source_code
-            
+
             # 统计token和成本
             total_input_tokens = 0
             total_output_tokens = 0
             total_input_cost = 0.0
             total_output_cost = 0.0
-            
+
             if count_tokens(source.source_code) > self.args.conversation_prune_safe_zone_tokens:
                 self.printer.print_in_terminal(
                     "index_file_too_large",
@@ -276,34 +280,40 @@ class IndexManager:
                         self.index_llm).with_meta(meta_holder).run(source.module_name, chunk)
                     time.sleep(self.anti_quota_limit)
                     symbols.append(chunk_symbols)
-                    
+
                     if meta_holder.get_meta():
                         meta_dict = meta_holder.get_meta()
-                        total_input_tokens += meta_dict.get("input_tokens_count", 0)
-                        total_output_tokens += meta_dict.get("generated_tokens_count", 0)
-                        
+                        total_input_tokens += meta_dict.get(
+                            "input_tokens_count", 0)
+                        total_output_tokens += meta_dict.get(
+                            "generated_tokens_count", 0)
+
                 symbols = "\n".join(symbols)
             else:
                 meta_holder = byzerllm.MetaHolder()
                 symbols = self.get_all_file_symbols.with_llm(
                     self.index_llm).with_meta(meta_holder).run(source.module_name, source_code)
                 time.sleep(self.anti_quota_limit)
-                
+
                 if meta_holder.get_meta():
                     meta_dict = meta_holder.get_meta()
-                    total_input_tokens += meta_dict.get("input_tokens_count", 0)
-                    total_output_tokens += meta_dict.get("generated_tokens_count", 0)
-            
+                    total_input_tokens += meta_dict.get(
+                        "input_tokens_count", 0)
+                    total_output_tokens += meta_dict.get(
+                        "generated_tokens_count", 0)
+
             # 计算总成本
             for name in model_names:
                 info = model_info_map.get(name, {})
-                total_input_cost += (total_input_tokens * info.get("input_price", 0.0)) / 1000000
-                total_output_cost += (total_output_tokens * info.get("output_price", 0.0)) / 1000000
-            
+                total_input_cost += (total_input_tokens *
+                                     info.get("input_price", 0.0)) / 1000000
+                total_output_cost += (total_output_tokens *
+                                      info.get("output_price", 0.0)) / 1000000
+
             # 四舍五入到4位小数
             total_input_cost = round(total_input_cost, 4)
             total_output_cost = round(total_output_cost, 4)
-            
+
             self.printer.print_in_terminal(
                 "index_update_success",
                 style="green",
@@ -340,9 +350,44 @@ class IndexManager:
             "generated_tokens_cost": total_output_cost
         }
 
+    def parse_exclude_files(self, exclude_files):
+        if not exclude_files:
+            return []
+
+        if isinstance(exclude_files, str):
+            exclude_files = [exclude_files]
+
+        exclude_patterns = []
+        for pattern in exclude_files:
+            if pattern.startswith("regex://"):
+                pattern = pattern[8:]
+                exclude_patterns.append(re.compile(pattern))
+            elif pattern.startswith("human://"):
+                pattern = pattern[8:]
+                v = (
+                    self.generate_regex_pattern.with_llm(self.llm)
+                    .with_extractor(self.extract_regex_pattern)
+                    .run(desc=pattern)
+                )
+                if not v:
+                    raise ValueError(
+                        "Fail to generate regex pattern, try again.")
+                exclude_patterns.append(re.compile(v))
+            else:
+                raise ValueError(
+                    "Invalid exclude_files format. Expected 'regex://<pattern>' or 'human://<description>' "
+                )
+        return exclude_patterns
+
+    def filter_exclude_files(self, file_path, exclude_patterns):
+        for pattern in exclude_patterns:
+            if pattern.search(file_path):
+                return True
+        return False
+
     def build_index(self):
         if os.path.exists(self.index_file):
-            with open(self.index_file, "r",encoding="utf-8") as file:
+            with open(self.index_file, "r", encoding="utf-8") as file:
                 index_data = json.load(file)
         else:
             index_data = {}
@@ -351,14 +396,27 @@ class IndexManager:
         keys_to_remove = []
         for file_path in index_data:
             if not os.path.exists(file_path):
-                keys_to_remove.append(file_path)                
-        
+                keys_to_remove.append(file_path)
+
+        # 删除被排除的文件
+        try:
+            exclude_patterns = self.parse_exclude_files(self.args.exclude_files)
+            for file_path in index_data:
+                if self.filter_exclude_files(file_path, exclude_patterns):
+                    keys_to_remove.append(file_path)
+        except Exception as e:
+            self.printer.print_in_terminal(
+                "index_exclude_files_error",
+                style="red",
+                error=str(e)
+            )
+
         # 删除无效条目并记录日志
         for key in set(keys_to_remove):
             if key in index_data:
                 del index_data[key]
                 self.printer.print_in_terminal(
-                    "index_file_removed", 
+                    "index_file_removed",
                     style="yellow",
                     file_path=key
                 )
@@ -388,7 +446,7 @@ class IndexManager:
                     for line in v:
                         new_v.append(line[line.find(":"):])
                     source_code = "\n".join(new_v)
-                
+
                 md5 = hashlib.md5(source_code.encode("utf-8")).hexdigest()
                 if (
                     source.module_name not in index_data
@@ -397,7 +455,8 @@ class IndexManager:
                     wait_to_build_files.append(source)
 
             # Remove duplicates based on module_name
-            wait_to_build_files = list({source.module_name: source for source in wait_to_build_files}.values())
+            wait_to_build_files = list(
+                {source.module_name: source for source in wait_to_build_files}.values())
 
             counter = 0
             num_files = len(wait_to_build_files)
@@ -433,16 +492,17 @@ class IndexManager:
                     index_data[module_name] = result
                     updated_sources.append(module_name)
                     if len(updated_sources) > 5:
-                        with open(self.index_file, "w",encoding="utf-8") as file:
-                            json.dump(index_data, file, ensure_ascii=False, indent=2)
+                        with open(self.index_file, "w", encoding="utf-8") as file:
+                            json.dump(index_data, file,
+                                      ensure_ascii=False, indent=2)
                         updated_sources = []
-        
+
         # 如果 updated_sources 或 keys_to_remove 有值，则保存索引文件
         if updated_sources or keys_to_remove:
-            with open(self.index_file, "w",encoding="utf-8") as file:
+            with open(self.index_file, "w", encoding="utf-8") as file:
                 json.dump(index_data, file, ensure_ascii=False, indent=2)
 
-            print("")    
+            print("")
             self.printer.print_in_terminal(
                 "index_file_saved",
                 style="green",
@@ -461,14 +521,14 @@ class IndexManager:
         if not os.path.exists(self.index_file):
             return []
 
-        with open(self.index_file, "r",encoding="utf-8") as file:
+        with open(self.index_file, "r", encoding="utf-8") as file:
             return file.read()
 
     def read_index(self) -> List[IndexItem]:
         if not os.path.exists(self.index_file):
             return []
 
-        with open(self.index_file, "r",encoding="utf-8") as file:
+        with open(self.index_file, "r", encoding="utf-8") as file:
             index_data = json.load(file)
 
         index_items = []
@@ -572,7 +632,7 @@ class IndexManager:
             {file.file_path: file for file in all_results}.values())
         return FileList(file_list=all_results)
 
-    def _query_index_with_thread(self, query, func):        
+    def _query_index_with_thread(self, query, func):
         all_results = []
         lock = threading.Lock()
         completed_threads = 0
@@ -582,7 +642,7 @@ class IndexManager:
             nonlocal completed_threads
             result = self._get_target_files_by_query.with_llm(
                 self.llm).with_return_type(FileList).run(chunk, query)
-            
+
             if result is not None:
                 with lock:
                     all_results.extend(result.file_list)
@@ -708,4 +768,3 @@ class IndexManager:
 
         请确保结果的准确性和完整性，包括所有可能相关的文件。
         """
-

autocoder/version.py

@@ -1 +1 @@
-__version__ = "0.1.266"
+__version__ = "0.1.268"