fiuai-s3 0.2.1__py3-none-any.whl → 0.3.1__py3-none-any.whl

fiuai_s3/minio/minio_storage.py

@@ -7,10 +7,12 @@
 
 from io import BytesIO
 import logging
-from typing import List, Optional
+from typing import List, Optional, Dict
 from minio import Minio
 from minio.error import S3Error
+from minio.commonconfig import Tags
 from ..object_storage import ObjectStorage, StorageConfig
+from ..type import DocFileObject, DocFileType       
 
 logger = logging.getLogger(__name__)
 
@@ -42,7 +44,7 @@ class MinioStorage(ObjectStorage):
         # 确保bucket存在
         if not self.client.bucket_exists(self.bucket_name):
             self.client.make_bucket(self.bucket_name)
-            logger.info(f"创建bucket: {self.bucket_name}")
+            logger.info(f"create bucket: {self.bucket_name}")
         
     def _format_endpoint(self, endpoint: str) -> str:
         """格式化endpoint，确保符合MinIO要求
@@ -65,7 +67,7 @@ class MinioStorage(ObjectStorage):
         
         # 验证格式
         if '/' in endpoint:
-            raise ValueError("MinIO endpoint不能包含路径，格式应为: host:port")
+            raise ValueError("MinIO endpoint can not contain path, format should be: host:port")
             
         return endpoint
         
@@ -86,10 +88,10 @@ class MinioStorage(ObjectStorage):
                 data=BytesIO(data),
                 length=len(data)
             )
-            logger.info(f"文件上传成功: {object_key}")
+            logger.info(f"file upload success: {object_key}")
             return True
         except S3Error as e:
-            logger.error(f"文件上传失败: {str(e)}")
+            logger.error(f"file upload failed: {str(e)}")
             return False
             
     def download_file(self, object_key: str) -> bytes:
@@ -108,7 +110,7 @@ class MinioStorage(ObjectStorage):
             )
             return response.read()
         except S3Error as e:
-            logger.error(f"文件下载失败: {str(e)}")
+            logger.error(f"file download failed: {str(e)}")
             return None
             
     def delete_file(self, object_key: str) -> bool:
@@ -125,10 +127,10 @@ class MinioStorage(ObjectStorage):
                 bucket_name=self.bucket_name,
                 object_name=object_key
             )
-            logger.info(f"文件删除成功: {object_key}")
+            logger.info(f"file delete success: {object_key}")
             return True
         except S3Error as e:
-            logger.error(f"文件删除失败: {str(e)}")
+            logger.error(f"file delete failed: {str(e)}")
             return False
             
     def list_files(self, prefix: Optional[str] = None) -> List[str]:
@@ -150,7 +152,7 @@ class MinioStorage(ObjectStorage):
                 files.append(obj.object_name)
             return files
         except S3Error as e:
-            logger.error(f"列出文件失败: {str(e)}")
+            logger.error(f"list files failed: {str(e)}")
             return [] 
 
     def _build_doc_path(self, filename: str, auth_tenant_id: Optional[str], auth_company_id: Optional[str], doc_id: Optional[str]) -> str:
@@ -161,24 +163,26 @@ class MinioStorage(ObjectStorage):
             raise ValueError("auth_tenant_id、auth_company_id、doc_id 不能为空")
         return f"{tenant_id}/{company_id}/{docid}/{filename}"
 
-    def upload_doc_file(self, filename: str, data: bytes, tags: Optional[dict] = None, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> bool:
+    def upload_doc_file(self, filename: str, data: bytes, tags: Optional[Dict[str, str]] = None, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> bool:
         try:
             object_key = self._build_doc_path(filename, auth_tenant_id, auth_company_id, doc_id)
-            extra_headers = None
-            extra_tags = None
-            if tags:
-                extra_tags = "&".join([f"{k}={v}" for k, v in tags.items()])
+            print(f"tags is {tags}")
+            # 将tags转换为Tags对象
+            tags_obj = Tags()
+            for k, v in tags.items():
+                tags_obj[k] = v
+            
             self.client.put_object(
                 bucket_name=self.bucket_name,
                 object_name=object_key,
                 data=BytesIO(data),
                 length=len(data),
-                metadata=tags if tags else None
+                tags=tags_obj
             )
-            logger.info(f"单据文件上传成功: {object_key}")
+            logger.info(f"doc file upload success: {object_key}")
             return True
         except S3Error as e:
-            logger.error(f"单据文件上传失败: {str(e)}")
+            logger.error(f"doc file upload failed: {str(e)}")
             return False
 
     def download_doc_file(self, filename: str, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> bytes:
@@ -190,10 +194,10 @@ class MinioStorage(ObjectStorage):
             )
             return response.read()
         except S3Error as e:
-            logger.error(f"单据文件下载失败: {str(e)}")
+            logger.error(f"doc file download failed: {str(e)}")
             return None
 
-    def list_doc_files(self, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> list:
+    def list_doc_files(self, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> List[DocFileObject]:
         try:
             tenant_id = auth_tenant_id or self.auth_tenant_id
             company_id = auth_company_id or self.auth_company_id
@@ -201,14 +205,27 @@ class MinioStorage(ObjectStorage):
             if not (tenant_id and company_id and docid):
                 raise ValueError("auth_tenant_id、auth_company_id、doc_id 不能为空")
             prefix = f"{tenant_id}/{company_id}/{docid}/"
-            files = []
+            files: List[DocFileObject] = []
             objects = self.client.list_objects(
                 bucket_name=self.bucket_name,
                 prefix=prefix
             )
             for obj in objects:
-                files.append(obj.object_name.split(prefix, 1)[-1])
+                file_path = obj.object_name
+                file_name = file_path.split("/")[-1]
+                # 获取tag
+                tags = {}
+                try:
+                    tag_res = self.client.get_object_tags(self.bucket_name, file_path)
+                    tags = tag_res if tag_res else {}
+                except Exception as tag_e:
+                    logger.warning(f"get object tag failed: {file_path}, {str(tag_e)}")
+                # 推断文件类型
+                files.append(DocFileObject(
+                    file_name=file_name,
+                    tags=tags
+                ))
             return files
         except S3Error as e:
-            logger.error(f"列出单据文件失败: {str(e)}")
-            return [] 
+            logger.error(f"list doc files failed: {str(e)}")
+            return []

fiuai_s3/object_storage.py

@@ -10,6 +10,7 @@ from typing import Optional, List, Dict, Any
 from pydantic import BaseModel, Field, ConfigDict
 import logging
 from uuid import uuid4
+from .type import DocFileObject
 
 logger = logging.getLogger(__name__)
 
@@ -96,7 +97,7 @@ class ObjectStorage(ABC):
         pass
 
     @abstractmethod
-    def upload_doc_file(self, filename: str, data: bytes, tags: Optional[Dict[str, str]] = None, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> bool:
+    def upload_doc_file(self, filename: str, data: bytes, tags: Optional[Dict[str, Any]] = None, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> bool:
         """
         上传单据文件，自动拼接存储路径并打tag
         Args:
@@ -126,7 +127,7 @@ class ObjectStorage(ABC):
         pass
 
     @abstractmethod
-    def list_doc_files(self, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> List[str]:
+    def list_doc_files(self, auth_tenant_id: Optional[str] = None, auth_company_id: Optional[str] = None, doc_id: Optional[str] = None) -> List[DocFileObject]:
         """
         列出单据下所有文件
         Args:

fiuai_s3/type.py

@@ -0,0 +1,31 @@
+        # -- coding: utf-8 --
+# Project: fiuai-s3
+# Created Date: 1700-01-01
+# Author: liming
+# Email: lmlala@aliyun.com
+# Copyright (c) 2025 FiuAI
+
+from enum import Enum
+from pydantic import BaseModel
+from typing import Optional, Dict
+
+class DocFileType(Enum):
+    """
+    文档文件类型
+    """
+    PDF = "pdf"
+    OFD = "ofd"
+    XML = "xml"
+    DOCTYPE = "doctype"
+    TABLE = "table"
+    TEXT = "text"
+    IMAGE = "image"
+    AUDIO = "audio"
+    ARCHIVE = "archive"
+
+class DocFileObject(BaseModel):
+    """
+    文档文件对象
+    """
+    file_name: str
+    tags: Optional[Dict[str, str]] = None

{fiuai_s3-0.2.1.dist-info → fiuai_s3-0.3.1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fiuai-s3
-Version: 0.2.1
+Version: 0.3.1
 Summary: 一个支持阿里云OSS和MinIO的对象存储抽象包
 Project-URL: Homepage, https://github.com/fiuai-sz/fiuai-s3
 Project-URL: Repository, https://github.com/fiuai-sz/fiuai-s3.git

fiuai_s3-0.3.1.dist-info/RECORD

@@ -0,0 +1,11 @@
+fiuai_s3/__init__.py,sha256=SZNl4Ohwf19PLyjRjIASpzhg3rvutd1WbLpblz_8G3s,312
+fiuai_s3/object_storage.py,sha256=wuDaMo5Nw4Aus6v7Vye3K-K-j2vmFPdgNi3zmyGBtb8,8555
+fiuai_s3/type.py,sha256=3LdjA9wO9IKKfKONLuI0SQ9FVSablhDnRphJqpDosRQ,608
+fiuai_s3/alicloud/__init__.py,sha256=mmrCrFp5DzRF5fViJDq7_LpsqCViwTPXOPz4qoaSscI,218
+fiuai_s3/alicloud/alicloud_storage.py,sha256=d24AFYDs7RsTsfL9IiVOxYlGyLzsaUa_WUYD5VPLP4s,6035
+fiuai_s3/minio/__init__.py,sha256=hOmpUkTq8TPgYOxfeOMcWq1_YsJ2r8Zf9VTX3w_v1Lk,209
+fiuai_s3/minio/minio_storage.py,sha256=SU743Nr7w8HMran-xkccEoeecSNKYySVNCGjo462-Lk,8476
+fiuai_s3-0.3.1.dist-info/METADATA,sha256=9LO5NJMY2Xp8nBNWyaxJ2HWJKHH_vlZ6Kr9xVyQTj9E,17609
+fiuai_s3-0.3.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+fiuai_s3-0.3.1.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+fiuai_s3-0.3.1.dist-info/RECORD,,

fiuai_s3-0.2.1.dist-info/RECORD

@@ -1,10 +0,0 @@
-fiuai_s3/__init__.py,sha256=SZNl4Ohwf19PLyjRjIASpzhg3rvutd1WbLpblz_8G3s,312
-fiuai_s3/object_storage.py,sha256=mBOQeSB5y98RBaugg2tJ0guZWBNycJaPYh1kEaPTq24,8513
-fiuai_s3/alicloud/__init__.py,sha256=mmrCrFp5DzRF5fViJDq7_LpsqCViwTPXOPz4qoaSscI,218
-fiuai_s3/alicloud/alicloud_storage.py,sha256=d24AFYDs7RsTsfL9IiVOxYlGyLzsaUa_WUYD5VPLP4s,6035
-fiuai_s3/minio/__init__.py,sha256=hOmpUkTq8TPgYOxfeOMcWq1_YsJ2r8Zf9VTX3w_v1Lk,209
-fiuai_s3/minio/minio_storage.py,sha256=79WkyPGIBYH_fMbYdukj2YIGMxkGEDYWpsGRhqqLgHQ,7772
-fiuai_s3-0.2.1.dist-info/METADATA,sha256=Z88cGuggf7YU-JAyHkB-DmlJvHpARhn5LNznZehXV9Y,17609
-fiuai_s3-0.2.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-fiuai_s3-0.2.1.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-fiuai_s3-0.2.1.dist-info/RECORD,,