auto-coder 0.1.243__py3-none-any.whl → 0.1.244__py3-none-any.whl

{auto_coder-0.1.243.dist-info → auto_coder-0.1.244.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.243
+Version: 0.1.244
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
@@ -26,7 +26,7 @@ Requires-Dist: tabulate
 Requires-Dist: jupyter-client
 Requires-Dist: prompt-toolkit
 Requires-Dist: tokenizers
-Requires-Dist: byzerllm[saas] >=0.1.159
+Requires-Dist: byzerllm[saas] >=0.1.160
 Requires-Dist: patch
 Requires-Dist: diff-match-patch
 Requires-Dist: GitPython

{auto_coder-0.1.243.dist-info → auto_coder-0.1.244.dist-info}/RECORD

@@ -1,17 +1,17 @@
 autocoder/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-autocoder/auto_coder.py,sha256=iQeyrg5L87IgBLFJLZLIJetHFeMuT2uslRlzo_Vjq9s,61829
+autocoder/auto_coder.py,sha256=ghIEJ8NzcFr77fosIZhKxI5F55EZoU6DMHFFpw5OFso,62671
 autocoder/auto_coder_lang.py,sha256=Rtupq6N3_HT7JRhDKdgCBcwRaiAnyCOR_Gsp4jUomrI,3229
 autocoder/auto_coder_rag.py,sha256=illKgzP2bv-Tq50ujsofJnOHdI4pzr0ALtfR8NHHWdQ,22351
 autocoder/auto_coder_rag_client_mcp.py,sha256=WV7j5JUiQge0x4-B7Hp5-pSAFXLbvLpzQMcCovbauIM,6276
 autocoder/auto_coder_rag_mcp.py,sha256=-RrjNwFaS2e5v8XDIrKR-zlUNUE8UBaeOtojffBrvJo,8521
 autocoder/auto_coder_server.py,sha256=XU9b4SBH7zjPPXaTWWHV4_zJm-XYa6njuLQaplYJH_c,20290
 autocoder/benchmark.py,sha256=Ypomkdzd1T3GE6dRICY3Hj547dZ6_inqJbBJIp5QMco,4423
-autocoder/chat_auto_coder.py,sha256=s9uMjDQQawXEsq171GO7SVMO4fDjWd4xWT0KYO7nRp4,105295
-autocoder/chat_auto_coder_lang.py,sha256=V-VIieyKF5cwlK448B1V2LUbTdrU03tfgDrOk2aBvFk,14891
+autocoder/chat_auto_coder.py,sha256=RJC3Fh_9ywkllETWFgxapHK3aY0M9ORhRON9PNCiKZs,105338
+autocoder/chat_auto_coder_lang.py,sha256=pdGkEMR96TvjYIzgWWB51LYmBYSrS_pO3sjGH-muuDI,15052
 autocoder/command_args.py,sha256=9aYJ-AmPxP1sQh6ciw04FWHjSn31f2W9afXFwo8wgx4,30441
 autocoder/lang.py,sha256=U6AjVV8Rs1uLyjFCZ8sT6WWuNUxMBqkXXIOs4S120uk,14511
-autocoder/models.py,sha256=FlBrF6HhGao_RiCSgYhCmP7vs0KlG4hI_BI6dyZiL9s,5292
-autocoder/version.py,sha256=u0hWeuFclX3Z9nFe5oFsCdX854VeUHOs69Ggv1pvBvk,23
+autocoder/models.py,sha256=8Tabnz3wlVswaeTpfPxnBhN55enP_5sdAJIx4Z79nOo,5342
+autocoder/version.py,sha256=nyjL74IBCp8tNAtI88ddAEba5PkcN70aC5Kdv_PKe7w,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
@@ -28,7 +28,7 @@ autocoder/common/__init__.py,sha256=2isE_u4VgfogwmcUCnFcussVFlzeNOLHDMFm5z_axbU,
 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=9FBNhcl6Do4ICh-klevYsCTsDuy5kD99r8EE5Gs1QoM,12592
+autocoder/common/auto_coder_lang.py,sha256=rIr4WJCUHgwkUkGwAaAiEd5kNx1nYRvVArbISbCfXp0,13248
 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
@@ -46,7 +46,7 @@ autocoder/common/command_completer.py,sha256=SSeb8MDH0JPvfdyW-S2uaHnui4VBDfSQvQP
 autocoder/common/command_generator.py,sha256=v4LmU7sO-P7jEZIXCWHUC6P-vT7AvBi_x_PTwCqBAE8,1323
 autocoder/common/command_templates.py,sha256=mnB3n8i0yjH1mqzyClEg8Wpr9VbZV44kxky66Zu6OJY,8557
 autocoder/common/const.py,sha256=eTjhjh4Aj4CUzviJ81jaf3Y5cwqsLATySn2wJxaS6RQ,2911
-autocoder/common/files.py,sha256=uGpfKASYwIncK_Vt_e_FOjFlO5VyAQOnRJe2SFdSWrg,877
+autocoder/common/files.py,sha256=CguxG9digkWBJpRaILErZmL_G5ryPRahPmPFWGB7X18,1973
 autocoder/common/git_utils.py,sha256=btK45sxvfm4tX3fBRNUPRZoGQuZuOEQrWSAwLy1yoLw,23095
 autocoder/common/image_to_page.py,sha256=O0cNO_vHHUP-fP4GXiVojShmNqkPnZXeIyiY1MRLpKg,13936
 autocoder/common/interpreter.py,sha256=62-dIakOunYB4yjmX8SHC0Gdy2h8NtxdgbpdqRZJ5vk,2833
@@ -79,7 +79,7 @@ autocoder/dispacher/actions/plugins/action_translate.py,sha256=nVAtRSQpdGNmZxg1R
 autocoder/index/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/index/entry.py,sha256=KJaxqtaKgL27w8-j7OiAqI0anPpmrJSl7PkfeVF2ipE,11713
 autocoder/index/for_command.py,sha256=BFvljE4t6VaMBGboZAuhUCzVK0EitCy_n5D_7FEnihw,3204
-autocoder/index/index.py,sha256=8AcaELR1FS___7VlNyxPnJsDVQ4wjORbqXvcA6TifCE,20337
+autocoder/index/index.py,sha256=VjfcBYHywU4tjQTA7mpHfzRM8nBPhPHrUnkuBbsj6do,20409
 autocoder/index/symbols_utils.py,sha256=CjcjUVajmJZB75Ty3a7kMv1BZphrm-tIBAdOJv6uo-0,2037
 autocoder/index/types.py,sha256=a2s_KV5FJlq7jqA2ELSo9E1sjuLwDB-JJYMhSpzBAhU,596
 autocoder/index/filter/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -119,7 +119,7 @@ autocoder/rag/stream_event/event_writer.py,sha256=l7kq_LnDDE8E5dZ-73C7J2MgzSL7Wr
 autocoder/rag/stream_event/types.py,sha256=rtLwOE8rShmi1dJdxyBpAV5ZjLBGG9vptMiSzMxGuIA,318
 autocoder/regex_project/__init__.py,sha256=EBZeCL5ORyD_9_5u_UuG4s7XtpXOu0y1sWDmxWFtufE,6781
 autocoder/regexproject/__init__.py,sha256=cEr-ZOaQjLD5sx7T7F2DhD5ips03HcJ02rded9EpSXc,9693
-autocoder/suffixproject/__init__.py,sha256=cmP54Y01ditZ83tiJqw5wle0I-uJBC0aZbZ7lYNSVO8,11080
+autocoder/suffixproject/__init__.py,sha256=VcXjUbGf3uQrpoqVCItDvGG9DoeHJ_qEmghKwrVNw9w,11058
 autocoder/tsproject/__init__.py,sha256=yloVzkGLnbTd4Hcj9fMO-rcjNTTx4wI3Ga41LWOSYrY,11747
 autocoder/utils/__init__.py,sha256=KtcGElFNBgZPF7dEL8zF9JpXkCAjoyDrzaREJBhJrcs,994
 autocoder/utils/_markitdown.py,sha256=RU88qn4eZfYIy0GDrPxlI8oYXIypbi63VRJjdlnE0VU,47431
@@ -134,15 +134,15 @@ autocoder/utils/print_table.py,sha256=ZMRhCA9DD0FUfKyJBWd5bDdj1RrtPtgOMWSJwtvZcL
 autocoder/utils/queue_communicate.py,sha256=buyEzdvab1QA4i2QKbq35rG5v_9x9PWVLWWMTznWcYM,6832
 autocoder/utils/request_event_queue.py,sha256=r3lo5qGsB1dIjzVQ05dnr0z_9Z3zOkBdP1vmRciKdi4,2095
 autocoder/utils/request_queue.py,sha256=nwp6PMtgTCiuwJI24p8OLNZjUiprC-TsefQrhMI-yPE,3889
-autocoder/utils/rest.py,sha256=opE_kBEdNQdxh350M5lUTMk5TViRfpuKP_qWc0B1lks,8861
+autocoder/utils/rest.py,sha256=hLBhr78y-WVnV0oQf9Rxc22EwqF78KINkScvYa1MuYA,6435
 autocoder/utils/tests.py,sha256=BqphrwyycGAvs-5mhH8pKtMZdObwhFtJ5MC_ZAOiLq8,1340
 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=fcXusKEUKMu8WY9Y1_JL5aPkC-soKFxQcFAKThrNZoQ,13338
+autocoder/utils/auto_coder_utils/chat_stream_out.py,sha256=yzh3QYJdYC1CPcdhSTn0fmFHtqV8iSXeH0drV3HjN5A,9053
 autocoder/utils/chat_auto_coder_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-auto_coder-0.1.243.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
-auto_coder-0.1.243.dist-info/METADATA,sha256=zq_UlYzkagreMYmIrWVkCmf43Zr3_jIX1w577MlSXQE,2616
-auto_coder-0.1.243.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-auto_coder-0.1.243.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
-auto_coder-0.1.243.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
-auto_coder-0.1.243.dist-info/RECORD,,
+auto_coder-0.1.244.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
+auto_coder-0.1.244.dist-info/METADATA,sha256=EVj3424BKEVPU6GAnoHC-NQ4J2VQFT7kVt3ANVby28A,2616
+auto_coder-0.1.244.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+auto_coder-0.1.244.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
+auto_coder-0.1.244.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
+auto_coder-0.1.244.dist-info/RECORD,,

autocoder/auto_coder.py

@@ -692,6 +692,23 @@ def main(input_args: Optional[List[str]] = None):
                 )
                 llm.setup_sub_client("vl_model", vl_model)
 
+            if args.index_model:   
+                model_name = args.index_model.strip()
+                model_info = models_module.get_model_by_name(model_name)                
+                index_model = byzerllm.SimpleByzerLLM(default_model_name=model_name)
+                index_model.deploy(
+                    model_path="",
+                    pretrained_model_type="saas/openai",
+                    udf_name=model_name,
+                    infer_params={
+                        "saas.base_url": model_info["base_url"],
+                        "saas.api_key": model_info["api_key"],
+                        "saas.model": model_info["model_name"],
+                        "saas.is_reasoning": model_info["is_reasoning"]
+                    }
+                )
+                llm.setup_sub_client("index_model", index_model)    
+
             if args.sd_model:
                 model_name = args.sd_model.strip()
                 model_info = models_module.get_model_by_name(model_name)

autocoder/chat_auto_coder.py

@@ -1605,8 +1605,8 @@ def code_next(query: str):
         if os.path.exists(temp_yaml):
             os.remove(temp_yaml)
 
-    llm = byzerllm.ByzerLLM.from_default_model(
-        args.inference_model or args.model)
+    product_mode = conf.get("product_mode", "lite")
+    llm = get_single_llm(args.chat_model or args.model, product_mode=product_mode)
 
     auto_guesser = AutoGuessQuery(
         llm=llm, project_dir=os.getcwd(), skip_diff=True)

autocoder/chat_auto_coder_lang.py

@@ -1,4 +1,5 @@
 import locale
+from byzerllm.utils import format_str_jinja2
 
 MESSAGES = {
     "en": {
@@ -236,3 +237,7 @@ def get_system_language():
 def get_message(key):
     lang = get_system_language()
     return MESSAGES.get(lang, MESSAGES['en']).get(key, MESSAGES['en'][key])
+
+
+def get_message_with_format(msg_key: str, **kwargs):
+    return format_str_jinja2(get_message(msg_key), **kwargs)

autocoder/common/auto_coder_lang.py

@@ -1,9 +1,13 @@
 import locale
+from byzerllm.utils import format_str_jinja2
 
 MESSAGES = {
-"en": {
+    "en": {
+        "model_not_found": "Model {{model_name}} not found",
         "new_session_started": "New session started. Previous chat history has been archived.",    
         "memory_save_success": "✅ Saved to your memory",
+        "file_decode_error": "Failed to decode file: {{file_path}}. Tried encodings: {{encodings}}",
+        "file_write_error": "Failed to write file: {{file_path}}. Error: {{error}}",
         "index_file_too_large": "⚠️ File {{ file_path }} is too large ({{ file_size }} > {{ max_length }}), splitting into chunks...",
         "index_update_success": "✅ Successfully updated index for {{ file_path }} (md5: {{ md5 }}) in {{ duration }}s",
         "index_build_error": "❌ Error building index for {{ file_path }}: {{ error }}",
@@ -73,8 +77,11 @@ MESSAGES = {
         "git_init_required": "⚠️ auto_merge only applies to git repositories.\n\nPlease try using git init in the source directory:\n\n```shell\ncd {{ source_dir }}\ngit init.\n```\n\nThen run auto - coder again.\nError: {{ error }}"
     },
     "zh": {
+        "model_not_found": "未找到模型: {{model_name}}",
         "new_session_started": "新会话已开始。之前的聊天历史已存档。",
         "memory_save_success": "✅ 已保存到您的记忆中",
+        "file_decode_error": "无法解码文件: {{file_path}}。尝试的编码: {{encodings}}",
+        "file_write_error": "无法写入文件: {{file_path}}. 错误: {{error}}",
         "index_file_too_large": "⚠️ 文件 {{ file_path }} 过大 ({{ file_size }} > {{ max_length }}), 正在分块处理...",
         "index_update_success": "✅ 成功更新 {{ file_path }} 的索引 (md5: {{ md5 }}), 耗时 {{ duration }} 秒",
         "index_build_error": "❌ 构建 {{ file_path }} 索引时出错: {{ error }}",
@@ -156,3 +163,6 @@ def get_system_language():
 def get_message(key):
     lang = get_system_language()
     return MESSAGES.get(lang, MESSAGES['en']).get(key, MESSAGES['en'][key])
+
+def get_message_with_format(msg_key: str, **kwargs):
+    return format_str_jinja2(get_message(msg_key), **kwargs)

autocoder/common/files.py

@@ -1,3 +1,6 @@
+from autocoder.common.auto_coder_lang import get_message_with_format
+from typing import List, Dict, Union
+
 def read_file(file_path):
     """Read a file with automatic encoding detection.
     
@@ -23,4 +26,33 @@ def read_file(file_path):
         except UnicodeDecodeError:
             continue
             
-    raise ValueError(f"无法解码文件: {file_path}。尝试的编码: {', '.join(encodings)}")
+    raise ValueError(get_message_with_format("file_decode_error", 
+        file_path=file_path, 
+        encodings=", ".join(encodings)))
+
+
+
+def save_file(file_path: str, content: Union[str, List[str]]) -> None:
+    """Save content to a file using UTF-8 encoding.
+    
+    Args:
+        file_path (str): Path to the file to write
+        content (Union[str, List[str]]): Content to write to the file. 
+            Can be a string or list of strings (will be joined with newlines)
+        
+    Raises:
+        IOError: If the file cannot be written
+        TypeError: If content is neither str nor List[str]
+    """
+    try:
+        with open(file_path, 'w', encoding='utf-8') as f:
+            if isinstance(content, str):
+                f.write(content)
+            elif isinstance(content, list):
+                f.write('\n'.join(content))
+            else:
+                raise TypeError("Content must be either str or List[str]")
+    except IOError as e:
+        raise IOError(get_message_with_format("file_write_error",
+            file_path=file_path,
+            error=str(e)))

autocoder/index/index.py

@@ -130,6 +130,7 @@ class IndexManager:
         如果有符号，按如下格式返回:
 
         ```
+        用途：主要用于提供自动实现函数模板的功能。
         {符号类型}: {符号名称}, {符号名称}, ...
         ```
 

autocoder/models.py

@@ -36,6 +36,22 @@ default_models_list = [
     }
 ]
 
+def process_api_key_path(base_url: str) -> str:
+    """
+    从 base_url 中提取 host 部分并处理特殊字符
+    例如: https://api.example.com:8080/v1 -> api.example.com_8080
+    """
+    if not base_url:
+        return ""
+    
+    parsed = urlparse(base_url)
+    host = parsed.netloc
+    
+    # 将冒号替换为下划线
+    host = host.replace(":", "_")
+    
+    return host
+
 def load_models() -> List[Dict]:
     """
     Load models from ~/.auto-coder/keys/models.json and merge with default_models_list.
@@ -73,7 +89,7 @@ def load_models() -> List[Dict]:
             api_key_file = os.path.join(api_key_dir, model["api_key_path"])
             if os.path.exists(api_key_file):
                 with open(api_key_file, "r") as f:
-                    model["api_key"] = f.read()
+                    model["api_key"] = f.read()                   
     return target_models
 
 def save_models(models: List[Dict]) -> None:
@@ -85,22 +101,6 @@ def save_models(models: List[Dict]) -> None:
         json.dump(models, f, indent=2, ensure_ascii=False)
 
 
-def process_api_key_path(base_url: str) -> str:
-    """
-    从 base_url 中提取 host 部分并处理特殊字符
-    例如: https://api.example.com:8080/v1 -> api.example.com_8080
-    """
-    if not base_url:
-        return ""
-    
-    parsed = urlparse(base_url)
-    host = parsed.netloc
-    
-    # 将冒号替换为下划线
-    host = host.replace(":", "_")
-    
-    return host
-
 def get_model_by_name(name: str) -> Dict:
     """
     根据模型名称查找模型
@@ -109,7 +109,8 @@ def get_model_by_name(name: str) -> Dict:
     v = [m for m in models if m["name"] == name.strip()]
     
     if len(v) == 0:
-        raise Exception(f"Model {name} not found")
+        from autocoder.common.auto_coder_lang import get_message_with_format
+        raise Exception(get_message_with_format("model_not_found", model_name=name))
     return v[0]
 
 def update_model_with_api_key(name: str, api_key: str) -> Dict:
@@ -135,9 +136,8 @@ def update_model_with_api_key(name: str, api_key: str) -> Dict:
                     
     if not found_model:
         return None
-        
-    # 从 base_url 中提取并处理 host
-    api_key_path = process_api_key_path(found_model["base_url"])
+            
+    api_key_path = name
     if api_key_path:
         found_model["api_key_path"] = api_key_path
         

autocoder/suffixproject/__init__.py

@@ -36,9 +36,8 @@ class SuffixProject:
         self.target_file = args.target_file
         self.project_type = args.project_type
         self.suffixs = [
-            f".{suffix.strip()}" if not suffix.startswith(".") else suffix.strip()
-            for suffix in self.project_type.split(",")
-            if suffix.strip() != ""
+            suffix.strip() if suffix.startswith(".") else f".{suffix.strip()}"
+            for suffix in self.project_type.split(",") if suffix.strip()
         ]
         self.file_filter = file_filter
         self.sources = []

autocoder/utils/auto_coder_utils/chat_stream_out.py

@@ -11,256 +11,132 @@ from autocoder.utils.request_queue import request_queue
 import time
 
 MAX_HISTORY_LINES = 40  # 最大保留历史行数
-LAYOUT_TYPES = Literal["vertical", "horizontal"]
 
-class StreamController:
-    def __init__(self, layout_type: LAYOUT_TYPES = "vertical", console: Optional[Console] = None):
-        self.console = console or Console(force_terminal=True, color_system="auto", height=24)  # 设置默认高度
-        self.layout = Layout()
-        self.queue = Queue()
+class StreamRenderer:
+    def __init__(self, title: str):
+        self.title = title
+        self.content = ""
         self.lock = Lock()
-        self.running = True
-        self.workers = []
-        self.layout_type = layout_type
-        self.stream_count = 0
-
-    def _create_stream_panel(self, idx: int) -> Layout:
-        """创建流面板布局"""
-        # 计算安全高度
-        current_height = self.console.height or 24  # 默认24行防止获取失败
-        safe_height = max(min(50, current_height // 2 - 4), 5)  # 限制最小高度为5行
-        
-        # 使用整数设置 Layout 的 size
-        panel = Layout(name=f"stream-{idx}", size=safe_height)
-        
-        panel.update(
-            Panel(
-                Markdown(""),
-                title=f"Stream {idx + 1}",
-                border_style="green",
-                height=safe_height  # 确保数值有效
-            )
-        )
-        return panel
-
-    def prepare_layout(self, count: int):
-        """准备动态布局结构"""
-        self.stream_count = count
-        
-        # 创建一个主布局容器
-        streams_layout = Layout(name="streams")
-        
-        # 创建所有流的布局
-        stream_layouts = []
-        for i in range(count):
-            stream_layout = Layout(name=f"stream-{i}")
-            panel = self._create_stream_panel(i)
-            stream_layout.update(panel)
-            stream_layouts.append(stream_layout)
-        
-        # 将所有流添加到主布局
-        if stream_layouts:
-            streams_layout.update(stream_layouts[0])
-            for i in range(1, len(stream_layouts)):
-                if self.layout_type == "vertical":
-                    streams_layout.split_column(stream_layouts[i])
-                elif self.layout_type == "horizontal":
-                    streams_layout.split_row(stream_layouts[i])
-                else:
-                    streams_layout.split_column(stream_layouts[i])
+        self.is_complete = False
         
-        # header 与 streams 布局分开
-        self.layout.split(
-            Layout(name="header", size=1),
-            streams_layout
-        )
-
-    def update_panel(self, idx: int, content: str, final: bool = False):
-        """线程安全的面板更新方法"""
+    def update(self, content: str):
         with self.lock:
-            # 计算安全高度
-            safe_height = min(50, self.console.height // 2 - 4)
+            self.content += content
             
-            if final:
-                new_panel = Panel(
-                    Markdown(content),
-                    title=f"Final Stream {idx+1}",
-                    border_style="blue",
-                    height=safe_height
-                )
-            else:
-                new_panel = Panel(
-                    Markdown(content),
-                    title=f"Stream {idx+1}",
-                    border_style="green",
-                    height=safe_height
-                )
+    def get_content(self) -> str:
+        with self.lock:
+            return self.content
+            
+    def complete(self):
+        with self.lock:
+            self.is_complete = True
 
-            panel_name = f"stream-{idx}"
-            streams_layout = self.layout["streams"]
+class MultiStreamRenderer:
+    def __init__(self, stream_titles: List[str], layout: str = "horizontal", console: Optional[Console] = None):
+        """
+        Initialize multi-stream renderer
+        
+        Args:
+            stream_titles: List of titles for each stream
+            layout: "horizontal" or "vertical"
+            console: Rich console instance
+        """
+        if console is None:
+            console = Console(force_terminal=True, color_system="auto")
             
-            # 递归查找目标布局
-            def find_layout(layout, name):
-                if layout.name == name:
-                    return layout
-                for child in layout.children:
-                    result = find_layout(child, name)
-                    if result:
-                        return result
-                return None
+        self.console = console
+        self.layout_type = layout
+        self.streams = [StreamRenderer(title) for title in stream_titles]
+        self.layout = Layout()
+        
+        # Create named layouts for each stream
+        self.stream_layouts = [Layout(name=f"stream{i}") for i in range(len(stream_titles))]
+        
+        # Configure layout
+        if layout == "horizontal":
+            self.layout.split_row(*self.stream_layouts)
+        else:
+            self.layout.split_column(*self.stream_layouts)
             
-            # 查找并更新目标布局
-            target_layout = find_layout(streams_layout, panel_name)
-            if target_layout:
-                target_layout.update(new_panel)
-            else:
-                import logging
-                logging.warning(f"未找到布局 {panel_name}，无法更新面板。")
+    def _process_stream(self, 
+                       stream_idx: int, 
+                       stream_generator: Generator[Tuple[str, Dict[str, Any]], None, None]):
+        """Process a single stream in a separate thread"""
+        stream = self.streams[stream_idx]
+        try:
+            for content, meta in stream_generator:
+                if content:
+                    stream.update(content)
+        finally:
+            stream.complete()
 
-def stream_worker(
-    idx: int,
-    generator: Generator[Tuple[str, Dict[str, Any]], None, None],
-    controller: StreamController,
-    request_id: Optional[str] = None
-) -> Tuple[str, Optional[Dict[str, Any]]]:
-    """单个流处理工作线程"""
-    lines_buffer = []
-    current_line = ""
-    assistant_response = ""
-    last_meta = None
-    
-    try:
-        for res in generator:
-            content, meta = res
-            last_meta = meta
+    def render_streams(self, 
+                      stream_generators: List[Generator[Tuple[str, Dict[str, Any]], None, None]]) -> List[str]:
+        """
+        Render multiple streams simultaneously
+        
+        Args:
+            stream_generators: List of stream generators to render
             
-            assistant_response += content
-            display_delta = meta.reasoning_content or content
-
-            parts = (current_line + display_delta).split("\n")
-            if len(parts) > 1:
-                lines_buffer.extend(parts[:-1])
-                if len(lines_buffer) > MAX_HISTORY_LINES:
-                    del lines_buffer[0:len(lines_buffer) - MAX_HISTORY_LINES]
+        Returns:
+            List of final content from each stream
+        """
+        assert len(stream_generators) == len(self.streams), "Number of generators must match number of streams"
+        
+        # Start processing threads
+        threads = []
+        for i, generator in enumerate(stream_generators):
+            thread = Thread(target=self._process_stream, args=(i, generator))
+            thread.daemon = True
+            thread.start()
+            threads.append(thread)
             
-            current_line = parts[-1]
-            display_content = "\n".join(lines_buffer[-MAX_HISTORY_LINES:] + [current_line])
+        try:
+            with Live(self.layout, console=self.console, refresh_per_second=10) as live:
+                while any(not stream.is_complete for stream in self.streams):
+                    # Update all panels
+                    for i, stream in enumerate(self.streams):
+                        panel = Panel(
+                            Markdown(stream.get_content() or "Waiting..."),
+                            title=stream.title,
+                            border_style="green" if not stream.is_complete else "blue"
+                        )
+                        
+                        # Update appropriate layout section
+                        self.stream_layouts[i].update(panel)
+                        
+                    time.sleep(0.1)  # Prevent excessive CPU usage
+                    
+        except KeyboardInterrupt:
+            print("\nStopping streams...")
             
-            controller.queue.put((idx, display_content, False))
+        # Wait for all threads to complete
+        for thread in threads:
+            thread.join()
             
-            if request_id and request_queue:
-                request_queue.add_request(
-                    request_id,
-                    RequestValue(
-                        value=StreamValue(value=[content]),
-                        status=RequestOption.RUNNING,
-                    ),
-                )
-
-        if current_line:
-            lines_buffer.append(current_line)
-        controller.queue.put((idx, assistant_response, True))
-        return assistant_response, last_meta
-    
-    except Exception as e:
-        error_content = f"Error: {str(e)}"
-        controller.queue.put((idx, error_content, True))
-        if request_id and request_queue:
-            request_queue.add_request(
-                request_id,
-                RequestValue(
-                    value=StreamValue(value=[str(e)]), 
-                    status=RequestOption.FAILED
-                ),
-            )
-        return assistant_response, last_meta
-    finally:
-        if request_id and request_queue:
-            request_queue.add_request(
-                request_id,
-                RequestValue(
-                    value=StreamValue(value=[""]), 
-                    status=RequestOption.COMPLETED
-                ),
-            )
+        return [stream.get_content() for stream in self.streams]
 
 def multi_stream_out(
     stream_generators: List[Generator[Tuple[str, Dict[str, Any]], None, None]],
-    request_ids: Optional[List[str]] = None,
-    console: Optional[Console] = None,
-    layout_type: LAYOUT_TYPES = "vertical"
-) -> List[Tuple[str, Optional[Dict[str, Any]]]]:
+    titles: List[str],
+    layout: str = "horizontal",
+    console: Optional[Console] = None
+) -> List[str]:
     """
-    多流并行输出处理器
+    Render multiple streams with Rich
     
     Args:
-        stream_generators: 流处理器列表
-        request_ids: 对应请求ID列表
-        console: Rich Console对象
-        layout_type: 布局类型 vertical/horizontal
+        stream_generators: List of stream generators
+        titles: List of titles for each stream
+        layout: "horizontal" or "vertical"
+        console: Optional Rich console instance
         
     Returns:
-        List[Tuple[str, Dict]]: 各流的处理结果
+        List of final content from each stream
     """
-    # 确保使用统一的console实例
-    if console is None:
-        console = Console(force_terminal=True, color_system="auto", height=24)
-    
-    # 初始化控制器
-    controller = StreamController(layout_type, console=console)
-    stream_count = len(stream_generators)
-    controller.prepare_layout(stream_count)
-    
-    # 启动工作线程
-    results = [None] * stream_count
-    threads = []
-    
-    # 创建工作线程
-    def worker_target(idx: int, gen: Generator[Tuple[str, Dict[str, Any]], None, None]):
-        req_id = request_ids[idx] if request_ids and idx < len(request_ids) else None
-        results[idx] = stream_worker(idx, gen, controller, req_id)
-    
-    # 启动所有工作线程
-    for idx, gen in enumerate(stream_generators):
-        t = Thread(target=worker_target, args=(idx, gen))
-        t.start()
-        threads.append(t)
-    
-    # 主渲染线程
-    try:
-        with Live(
-            controller.layout, 
-            console=console or controller.console,
-            refresh_per_second=10,
-            screen=True
-        ) as live:
-            while controller.running:
-                updated = False
-                try:
-                    while True:  # 处理队列中的所有更新
-                        idx, content, final = controller.queue.get_nowait()
-                        controller.update_panel(idx, content, final)
-                        updated = True
-                except Empty:
-                    pass
-                
-                if updated:
-                    live.refresh()
-                
-                # 检查线程是否全部完成
-                if all(not t.is_alive() for t in threads):
-                    break
-                
-                time.sleep(0.1)
-                
-    finally:
-        controller.running = False
-        for t in threads:
-            t.join()
+    renderer = MultiStreamRenderer(titles, layout, console)
+    return renderer.render_streams(stream_generators)
 
-    # 确保最后一次刷新
-    (console or controller.console).print(controller.layout)        
-    return results
 
 def stream_out(
     stream_generator: Generator[Tuple[str, Dict[str, Any]], None, None],

autocoder/utils/rest.py

@@ -1,16 +1,22 @@
 import requests
 from bs4 import BeautifulSoup
-from typing import List,Dict,Type,Optional
+from typing import List,Dict,Union,Optional
 from autocoder.common import SourceCode
 import byzerllm
-from bs4 import BeautifulSoup
 from loguru import logger
 import os
 from pathlib import Path
 from autocoder.common import files as FileUtils
+import traceback
+from autocoder.rag.loaders import (
+    extract_text_from_pdf,
+    extract_text_from_docx,
+    extract_text_from_ppt,
+    extract_text_from_excel
+)
 
 class HttpDoc:
-    def __init__(self, args, llm: byzerllm.ByzerLLM,urls:Optional[List[str]]=None):
+    def __init__(self, args, llm: Union[byzerllm.ByzerLLM, byzerllm.SimpleByzerLLM],urls:Optional[List[str]]=None):
         self.args = args
         urls_from_args = self.args.urls
         if urls_from_args:
@@ -41,104 +47,50 @@ class HttpDoc:
         {{ html }}
                 
         输出的内容请以 "<MARKER></MARKER> 标签对包裹。
-        """
-
-    def is_binary_file(self,filepath):            
-        try:
-            with open(filepath, 'rb') as file:
-                chunk = file.read(1024*8)  # Read first 1024 bytes
-                if b'\x00' in chunk:  # Binary files often contain null bytes
-                    return True
-                # Attempt to decode as UTF-8 (or any encoding you expect your text files to be in)
-                chunk.decode('utf-8')
-                return False
-        except UnicodeDecodeError:
-            return True   
+        """    
 
-    def get_file_extractor(self):        
+    def _process_local_file(self, file_path: str) -> List[SourceCode]:
+        """统一处理本地文件，返回标准化的 SourceCode 列表"""
+        results = []
         try:
-            from llama_index.core.readers.base import BaseReader
-            from fsspec import AbstractFileSystem
-            from llama_index.core.schema import Document
-            from llama_index.core.readers.file.base import get_default_fs
-            from llama_index.readers.file import (
-                DocxReader,
-                EpubReader,
-                HWPReader,
-                ImageReader,
-                IPYNBReader,
-                MarkdownReader,
-                MboxReader,
-                PandasCSVReader,
-                PDFReader,
-                PptxReader,
-                VideoAudioReader,
-            )  # pants: no-infer-dep                                
-        except ImportError as e:
-            raise ImportError(f"`llama-index-readers-file` package not found. {e}")
+            ext = os.path.splitext(file_path)[1].lower()
+                        
+            # 分发到不同 loader
+            if ext == '.pdf':
+                content = extract_text_from_pdf(file_path)
+                results.append(SourceCode(module_name=file_path, source_code=content))
+            elif ext == '.docx':
+                content = extract_text_from_docx(file_path)
+                results.append(SourceCode(module_name=file_path, source_code=content))
+            elif ext in ('.pptx', '.ppt'):
+                for slide_id, slide_content in extract_text_from_ppt(file_path):
+                    results.append(SourceCode(module_name=f"{file_path}#{slide_id}", source_code=slide_content))
+            elif ext in ('.xlsx', '.xls'):
+                for sheet_name, sheet_content in extract_text_from_excel(file_path):
+                    results.append(SourceCode(module_name=f"{file_path}#{sheet_name}", source_code=sheet_content))
+            else:
+                content = FileUtils.read_file(file_path)
+                results.append(SourceCode(module_name=file_path, source_code=content))
 
-        default_file_reader_cls: Dict[str, BaseReader] = {
-            ".hwp": HWPReader(),
-            ".pdf": PDFReader(return_full_document=True),
-            ".docx": DocxReader(),
-            # ".pptx": PptxReader(),
-            # ".ppt": PptxReader(),
-            # ".pptm": PptxReader(),
-            # ".jpg": ImageReader(),
-            # ".png": ImageReader(),
-            # ".jpeg": ImageReader(),
-            # ".mp3": VideoAudioReader(),
-            # ".mp4": VideoAudioReader(),
-            # ".csv": PandasCSVReader(),
-            ".epub": EpubReader(),            
-            ".mbox": MboxReader(),
-            ".ipynb": IPYNBReader(),            
-        }
-        return default_file_reader_cls
+        except Exception as e:
+            logger.error(f"Failed to process {file_path}: {str(e)}")
+            traceback.print_exc()
+        
+        return results
 
     def crawl_urls(self) -> List[SourceCode]:
-        source_codes = []        
+        source_codes = []
         for url in self.urls:
-            if not url.startswith("http://") and not url.startswith("https://"):
+            if not url.startswith(("http://", "https://")):
                 try:
-                 from llama_index.core import SimpleDirectoryReader
-                 exts = self.get_file_extractor()
-                 documents = []   
-
-                 def process_single_file(file_path: str,skip_binary_file_test:bool=False): 
-                    temp_documents = []
-                    ext = os.path.splitext(file_path)[1].lower()                    
-                    if  not skip_binary_file_test and self.is_binary_file(file_path):
-                        logger.warning(f"Skipping binary file: {file_path}")
-                        return temp_documents
-                    
-                    if ext not in exts.keys():                                                
-                        main_content = FileUtils.read_file(file_path)
-                        source_code = SourceCode(module_name=file_path, source_code=main_content)
-                        source_codes.append(source_code)                                   
+                    if os.path.isdir(url):
+                        for root, _, files in os.walk(url, followlinks=True):
+                            for file in files:
+                                source_codes.extend(self._process_local_file(os.path.join(root, file)))
                     else:
-                        temp_documents = SimpleDirectoryReader(input_files=[url],file_extractor=exts).load_data()  
-                    return temp_documents                                    
-
-                 if os.path.isdir(url):
-                    for root, dirs, files in os.walk(url,followlinks=True):
-                        dirs[:] = [d for d in dirs if d not in ['.git',"node_modules"]]  # Exclude .git directory
-                        for file in files:
-                            file_path = os.path.join(root, file)                            
-                            documents.extend(process_single_file(file_path))
-                    
-                 else:                    
-                    documents.extend(process_single_file(url,skip_binary_file_test=True))
-                    
-                 for document in documents:
-                    source_code = SourceCode(module_name=document.metadata["file_path"], source_code=document.get_content())
-                    source_codes.append(source_code)
-
-                except ImportError as e:
-                    logger.warning(f"Failed to import llama_index. Please install it using 'pip install llama_index' {e}")
-                    main_content = FileUtils.read_file(url)
-                    source_code = SourceCode(module_name=url, source_code=main_content)
-                    source_codes.append(source_code)                                                
+                        source_codes.extend(self._process_local_file(url))
+                except Exception as e:
+                    logger.error(f"Error accessing path {url}: {str(e)}")
             else:
                 if self.args.urls_use_model:
                     from autocoder.common.screenshots import gen_screenshots

autocoder/version.py

@@ -1 +1 @@
-__version__ = "0.1.243"
+__version__ = "0.1.244"