auto-coder 0.1.296__py3-none-any.whl → 0.1.298__py3-none-any.whl

{auto_coder-0.1.296.dist-info → auto_coder-0.1.298.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: auto-coder
-Version: 0.1.296
+Version: 0.1.298
 Summary: AutoCoder: AutoCoder
 Author: allwefantasy
 Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
@@ -26,7 +26,8 @@ Requires-Dist: tabulate
 Requires-Dist: jupyter-client
 Requires-Dist: prompt-toolkit
 Requires-Dist: tokenizers
-Requires-Dist: byzerllm[saas] >=0.1.174
+Requires-Dist: aiofiles
+Requires-Dist: byzerllm[saas] >=0.1.176
 Requires-Dist: patch
 Requires-Dist: diff-match-patch
 Requires-Dist: GitPython

{auto_coder-0.1.296.dist-info → auto_coder-0.1.298.dist-info}/RECORD

@@ -1,7 +1,7 @@
 autocoder/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/auto_coder.py,sha256=zKqfMdm2F419hrNGaosW4kYJ3IbaxBKlpTlVl6JFWmE,65563
 autocoder/auto_coder_lang.py,sha256=Rtupq6N3_HT7JRhDKdgCBcwRaiAnyCOR_Gsp4jUomrI,3229
-autocoder/auto_coder_rag.py,sha256=yhwRh_TJZyBxBCmUusZ8h5guU42i0Z6UJ10mT0FH3Rc,33857
+autocoder/auto_coder_rag.py,sha256=5TtAfbEBwyt-cB4WcI8eQ1G3AuKij0056wFYRViDhLs,34036
 autocoder/auto_coder_rag_client_mcp.py,sha256=QRxUbjc6A8UmDMQ8lXgZkjgqtq3lgKYeatJbDY6rSo0,6270
 autocoder/auto_coder_rag_mcp.py,sha256=-RrjNwFaS2e5v8XDIrKR-zlUNUE8UBaeOtojffBrvJo,8521
 autocoder/auto_coder_runner.py,sha256=Q8dr9EsQob6xHZbVxopE30C4vuk5tDQ5lO8nGfx-1kM,101365
@@ -12,7 +12,7 @@ autocoder/chat_auto_coder_lang.py,sha256=-vyKq02RGn6N275YA06JZqXpfJkNq6PNjQ8wy4M
 autocoder/command_args.py,sha256=9aYJ-AmPxP1sQh6ciw04FWHjSn31f2W9afXFwo8wgx4,30441
 autocoder/lang.py,sha256=U6AjVV8Rs1uLyjFCZ8sT6WWuNUxMBqkXXIOs4S120uk,14511
 autocoder/models.py,sha256=AyoZ-Pzy0oyYUmWCxOIRiOImsqboSfRET7LO9-UOuxI,11172
-autocoder/version.py,sha256=ihc06tTR_94f1_ktAvLzOrOQwItQGilOeekInep7SCU,23
+autocoder/version.py,sha256=6WgD8AUxiPQngIKXonXPlISWgH75TnoidQDJCS5V0hc,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
@@ -31,7 +31,7 @@ autocoder/commands/auto_web.py,sha256=_449f4rCoRG7Sv0SB0hIBRFLPLPJ5DgWW4DlI22a3X
 autocoder/commands/tools.py,sha256=iBcGdTAOCoWk1wX05EbEfkzIEZiTMAKRthap0q1rM5E,21384
 autocoder/common/JupyterClient.py,sha256=O-wi6pXeAEYhAY24kDa0BINrLYvKS6rKyWe98pDClS0,2816
 autocoder/common/ShellClient.py,sha256=fM1q8t_XMSbLBl2zkCNC2J9xuyKN3eXzGm6hHhqL2WY,2286
-autocoder/common/__init__.py,sha256=LE-HHb_HwpWwRKwbwDlTqRzMV6nTzqF3RB55zVMDS3c,13656
+autocoder/common/__init__.py,sha256=pO198YBdywGYarPhT7ild3GcdvcFhLJTe63ilZR1d5E,13705
 autocoder/common/anything2images.py,sha256=0ILBbWzY02M-CiWB-vzuomb_J1hVdxRcenAfIrAXq9M,25283
 autocoder/common/anything2img.py,sha256=iZQmg8srXlD7N5uGl5b_ONKJMBjYoW8kPmokkG6ISF0,10118
 autocoder/common/audio.py,sha256=Kn9nWKQddWnUrAz0a_ZUgjcu4VUU_IcZBigT7n3N3qc,7439
@@ -123,7 +123,7 @@ autocoder/rag/document_retriever.py,sha256=5BDqKVJqLPScEnua5S5suXhWuCaALIfPf5obX
 autocoder/rag/lang.py,sha256=_jmUtxZDG1fmF4b2mhMJbYS1YQDb2ZE8nyAn5_vrvjA,3350
 autocoder/rag/llm_wrapper.py,sha256=Ht5GF5yJtrztoliujsZzx_ooWZmHkd5xLZKcGEiicZw,4303
 autocoder/rag/long_context_rag.py,sha256=6rqq0pvYe9N4TvyLwd2OB21ZUrPC4FfxZuks0weAz4A,41935
-autocoder/rag/qa_conversation_strategy.py,sha256=y1ms_bRNDSHKw_MmqXNkhmvwqak3cj4rikjlqyablIg,6052
+autocoder/rag/qa_conversation_strategy.py,sha256=1AcHV0MU00yTls20LlCPO-Un_OhSrr_p-H5lxLleAq4,6060
 autocoder/rag/rag_config.py,sha256=8LwFcTd8OJWWwi1_WY4IzjqgtT6RyE2j4PjxS5cCTDE,802
 autocoder/rag/rag_entry.py,sha256=6TKtErZ0Us9XSV6HgRKXA6yR3SiZGPHpynOKSaR1wgE,2463
 autocoder/rag/raw_rag.py,sha256=BOr0YGf3umjqXOIDVO1LXQ0bIHx8hzBdiubND2ezyxc,2946
@@ -140,11 +140,11 @@ autocoder/rag/utils.py,sha256=f21ybCAlYVgr3tJP9MkVoM9d82-uG5NHu2gsv2oaVBQ,4961
 autocoder/rag/variable_holder.py,sha256=PFvBjFcR7-fNDD4Vcsc8CpH2Te057vcpwJMxtrfUgKI,75
 autocoder/rag/cache/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 autocoder/rag/cache/base_cache.py,sha256=EaYYYbclMBvnlOUoM7qonnluwZX5oSvUjdvGvFun8_8,742
-autocoder/rag/cache/byzer_storage_cache.py,sha256=okmNUDRCDv81JOfBuTspmTxf8ltYmYctd4vrPdjUgOg,29239
+autocoder/rag/cache/byzer_storage_cache.py,sha256=S_afj4_fFt88A0KJ23N-Gc3mMwPTDjzv9E3J9O_IiT4,29319
 autocoder/rag/cache/cache_result_merge.py,sha256=VnTdbT2OMBmWl_83bqds97d9_M33IhPNX8tF7KH2GMM,10556
 autocoder/rag/cache/file_monitor_cache.py,sha256=OdSXTH3vo6inAzkN5d55I0RN03GUlSlnUEKmXpjFl78,9443
-autocoder/rag/cache/local_byzer_storage_cache.py,sha256=7_6zCRY3BiCM0ec3U96i1G4l2SzmAedaTfkArNkMfQU,31925
-autocoder/rag/cache/local_duckdb_storage_cache.py,sha256=xEU_GG0MwYZvg-_FhGQBp0O3UNefNqZen6DxD6bkx1E,27686
+autocoder/rag/cache/local_byzer_storage_cache.py,sha256=1xskK7X_hFEAsHHoT_F9lFYhQOTrpQtsFyFCIeI2Mvk,31964
+autocoder/rag/cache/local_duckdb_storage_cache.py,sha256=gTfAzFTfVLoi8cqxct_I3-ozdoZvKUVSiI1w8NxbjX8,28540
 autocoder/rag/cache/rag_file_meta.py,sha256=RQ3n4wfkHlB-1ljS3sFSi8ijbsUPeIqBSgjmmbRuwRI,20521
 autocoder/rag/cache/simple_cache.py,sha256=j9dxhei-Nwq9FJrrGOWhaDIDSb_Iz6JSojT1pelS9k4,13084
 autocoder/rag/loaders/__init__.py,sha256=EQHEZ5Cmz-mGP2SllUTvcIbYCnF7W149dNpNItfs0yE,304
@@ -183,9 +183,9 @@ 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-0.1.296.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
-auto_coder-0.1.296.dist-info/METADATA,sha256=XLlbDjwZ0CnGmPaWfozGzZMSKCJWvZn4wNBxuqz4vlo,2665
-auto_coder-0.1.296.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
-auto_coder-0.1.296.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
-auto_coder-0.1.296.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
-auto_coder-0.1.296.dist-info/RECORD,,
+auto_coder-0.1.298.dist-info/LICENSE,sha256=HrhfyXIkWY2tGFK11kg7vPCqhgh5DcxleloqdhrpyMY,11558
+auto_coder-0.1.298.dist-info/METADATA,sha256=COiMKzvXqK4P1_xs-qwuOdDDQZ3fLMinxhc0EpzJP6I,2689
+auto_coder-0.1.298.dist-info/WHEEL,sha256=GV9aMThwP_4oNCtvEC2ec3qUYutgWeAzklro_0m4WJQ,91
+auto_coder-0.1.298.dist-info/entry_points.txt,sha256=0nzHtHH4pNcM7xq4EBA2toS28Qelrvcbrr59GqD_0Ak,350
+auto_coder-0.1.298.dist-info/top_level.txt,sha256=Jqc0_uJSw2GwoFQAa9iJxYns-2mWla-9ok_Y3Gcznjk,10
+auto_coder-0.1.298.dist-info/RECORD,,

autocoder/auto_coder_rag.py

@@ -187,6 +187,13 @@ def main(input_args: Optional[List[str]] = None):
         help="The storage type of the RAG, duckdb or byzer-storage",
     )
 
+    build_index_parser.add_argument(
+        "--rag_index_build_workers",
+        type=int,
+        default=5,
+        help="The number of workers to build the RAG index",
+    )
+
     build_index_parser.add_argument(
         "--quick", action="store_true", help="Skip system initialization"
     )

autocoder/common/__init__.py

@@ -300,6 +300,7 @@ class AutoCoderArgs(pydantic.BaseModel):
     rag_duckdb_vector_dim: Optional[int] = 1024  # DuckDB 向量化存储的维度
     rag_duckdb_query_similarity: Optional[float] = 0.1  # DuckDB 向量化检索 相似度 阈值
     rag_duckdb_query_top_k: Optional[int] = 10000  # DuckDB 向量化检索 返回 TopK个结果(且大于相似度)
+    rag_index_build_workers: Optional[int] = 10
     # rag 本地图床地址
     local_image_host: Optional[str] = ""
     rag_recall_max_queries: Optional[int] = 5
@@ -377,7 +378,7 @@ class AutoCoderArgs(pydantic.BaseModel):
     rank_strategy: Optional[str] = "file"
 
     action: List[str] = []
-    enable_global_memory: Optional[bool] = True  
+    enable_global_memory: Optional[bool] = False  
     product_mode: Optional[str] = "lite"
 
     keep_reasoning_content: Optional[bool] = False

autocoder/rag/cache/byzer_storage_cache.py

@@ -132,6 +132,7 @@ class ByzerStorageCache(BaseCacheManager):
         self.path = path
         self.ignore_spec = ignore_spec
         self.required_exts = required_exts
+        self.extra_params = extra_params
         self.rag_build_name = extra_params.rag_build_name
         self.storage = ByzerStorage("byzerai_store", "rag", self.rag_build_name)
         self.queue = []
@@ -329,7 +330,7 @@ class ByzerStorageCache(BaseCacheManager):
             start_time = time.time()
             
             # Use more workers to process the smaller batches efficiently
-            max_workers = min(10, total_batches)  # Cap at 10 workers or total batch count
+            max_workers = min(self.extra_params.rag_index_build_workers, total_batches)  # Cap at 10 workers or total batch count
             logger.info(f"[BUILD CACHE] Using {max_workers} parallel workers for processing")
             
             with ThreadPoolExecutor(max_workers=max_workers) as executor:

autocoder/rag/cache/local_byzer_storage_cache.py

@@ -338,7 +338,7 @@ class LocalByzerStorageCache(BaseCacheManager):
             start_time = time.time()
 
             # Use more workers to process the smaller batches efficiently
-            max_workers = min(10, total_batches)  # Cap at 10 workers or total batch count
+            max_workers = min(self.extra_params.rag_index_build_workers, total_batches)  # Cap at 10 workers or total batch count
             logger.info(f"[BUILD CACHE] Using {max_workers} parallel workers for processing")
 
             with ThreadPoolExecutor(max_workers=max_workers) as executor:

autocoder/rag/cache/local_duckdb_storage_cache.py

@@ -133,15 +133,31 @@ class LocalDuckdbStorage:
         return reduced
 
     def _embedding(self, context: str, norm: bool = True, dim: int | None = None) -> List[float]:
-        embedding = self.llm.emb_query(context)[0].output
-
-        if dim:
-            embedding = self._apply_pca(embedding, target_dim=dim)  # 降维后形状 (1024,)
-
-        if norm:
-            embedding = embedding / np.linalg.norm(embedding)
-
-        return embedding.tolist()
+        max_retries = 3
+        retry_count = 0
+        
+        while retry_count < max_retries:
+            try:
+                embedding = self.llm.emb_query(context)[0].output
+                
+                if dim:
+                    embedding = self._apply_pca(embedding, target_dim=dim)  # 降维后形状 (1024,)
+                
+                if norm:
+                    embedding = embedding / np.linalg.norm(embedding)
+                
+                return embedding.tolist()
+            except Exception as e:
+                retry_count += 1
+                if retry_count >= max_retries:
+                    logger.error(f"Failed to get embedding after {max_retries} attempts: {str(e)}")
+                    raise
+                
+                # Sleep between 1-5 seconds before retrying
+                sleep_time = 1 + (retry_count * 1.5)
+                logger.warning(f"Embedding API call failed (attempt {retry_count}/{max_retries}). "
+                              f"Error: {str(e)}. Retrying in {sleep_time:.1f} seconds...")
+                time.sleep(sleep_time)
 
     def _initialize(self) -> None:
         if self.embed_dim is None:
@@ -452,9 +468,9 @@ class LocalDuckDBStorageCache(BaseCacheManager):
         self.write_cache()
 
         if items:
-            logger.info("[BUILD CACHE] Clearing existing cache from Byzer DuckDB Storage")
+            logger.info("[BUILD CACHE] Clearing existing cache from DuckDB  Storage")
             self.storage.truncate_table()
-            logger.info(f"[BUILD CACHE] Preparing to write to Byzer DuckDB Storage, "
+            logger.info(f"[BUILD CACHE] Preparing to write to DuckDB  Storage, "
                         f"total chunks: {len(items)}, total files: {len(files_to_process)}")
 
             # Use a fixed optimal batch size instead of dividing by worker count
@@ -464,12 +480,12 @@ class LocalDuckDBStorageCache(BaseCacheManager):
             total_batches = len(item_batches)
             completed_batches = 0
 
-            logger.info(f"[BUILD CACHE] Starting to write to Byzer Storage using {batch_size} items per batch, "
+            logger.info(f"[BUILD CACHE] Starting to write to DuckDB Storage using {batch_size} items per batch, "
                         f"total batches: {total_batches}")
             start_time = time.time()
 
-            # Use more workers to process the smaller batches efficiently
-            max_workers = min(10, total_batches)  # Cap at 10 workers or total batch count
+            # Use more workers to process the smaller batches efficiently            
+            max_workers = min(self.extra_params.rag_index_build_workers, total_batches)  # Cap at 10 workers or total batch count
             logger.info(f"[BUILD CACHE] Using {max_workers} parallel workers for processing")
 
             def batch_add_doc(_batch):

autocoder/rag/qa_conversation_strategy.py

@@ -63,7 +63,7 @@ class MultiRoundStrategy(QAConversationStrategy):
         {% if local_image_host %}
         4. 图片路径处理
         - 图片地址需返回绝对路径, 
-        - 对于Windows风格的路径，需要转换为Linux风格， 例如：C:\Users\user\Desktop\image.png 转换为 C:/Users/user/Desktop/image.png
+        - 对于Windows风格的路径，需要转换为Linux风格， 例如：C:\\Users\\user\\Desktop\\image.png 转换为 C:/Users/user/Desktop/image.png
         - 为请求图片资源 需增加 http://{{ local_image_host }}/static/ 作为前缀
         例如：/path/to/images/image.png， 返回 http://{{ local_image_host }}/static/path/to/images/image.png
         {% endif %}
@@ -115,7 +115,7 @@ class SingleRoundStrategy(QAConversationStrategy):
         {% if local_image_host %}
         4. 图片路径处理
         - 图片地址需返回绝对路径, 
-        - 对于Windows风格的路径，需要转换为Linux风格， 例如：C:\Users\user\Desktop\image.png 转换为 C:/Users/user/Desktop/image.png
+        - 对于Windows风格的路径，需要转换为Linux风格， 例如：C:\\Users\\user\\Desktop\\image.png 转换为 C:/Users/user/Desktop/image.png
         - 为请求图片资源 需增加 http://{{ local_image_host }}/static/ 作为前缀
         例如：/path/to/images/image.png， 返回 http://{{ local_image_host }}/static/path/to/images/image.png
         {% endif %}

autocoder/version.py

@@ -1 +1 @@
-__version__ = "0.1.296"
+__version__ = "0.1.298"