Python：解决将编码流作为Blob上传至Azure容器后仍编码的问题

我想用Python把StringIO或bytes类型的流上传到Azure容器作为Blob，但上传后内容似乎处于编码状态，尝试了以下代码仍未解决：

def upload_blob_stream(blob_service_client: BlobServiceClient, container_name,stream,encoding=None):
    blob_client = blob_service_client.get_blob_client(container=container_name, blob=log_blob_name)
    if encoding:
        data = stream.decode(encoding)
    else:
        data = stream
    input_stream = io.BytesIO(data)
    blob_client.upload_blob(input_stream, blob_type="BlockBlob")

upload_blob_stream(blob_service_client,container_name,stream)

问题核心是对StringIO和bytes两种流类型的处理逻辑错误，调整后即可解决：

关键修正逻辑 #

1. bytes类型流：直接用io.BytesIO包装后上传，无需解码操作
2. StringIO类型流：先将内部字符串编码为bytes，再用io.BytesIO包装

修改后的代码 #

import io
from azure.storage.blob import BlobServiceClient

def upload_blob_stream(blob_service_client: BlobServiceClient, container_name, blob_name, stream, encoding='utf-8'):
    blob_client = blob_service_client.get_blob_client(container=container_name, blob=blob_name)
    
    # 根据流类型做针对性处理
    if isinstance(stream, io.StringIO):
        # StringIO内容转bytes后包装为BytesIO
        content_bytes = stream.getvalue().encode(encoding)
        input_stream = io.BytesIO(content_bytes)
    elif isinstance(stream, bytes):
        # bytes直接包装为BytesIO
        input_stream = io.BytesIO(stream)
    else:
        # 兼容已有的BytesIO或其他可读取流
        input_stream = stream
    
    # 可选添加overwrite=True覆盖已有Blob
    blob_client.upload_blob(input_stream, blob_type="BlockBlob", overwrite=True)

# 调用示例
# 1. 上传bytes流
# upload_blob_stream(blob_service_client, "my-container", "test-bin.blob", b"raw byte content")
# 2. 上传StringIO流
# s_stream = io.StringIO("中文测试内容")
# upload_blob_stream(blob_service_client, "my-container", "test-txt.blob", s_stream)

额外说明 #

• 原代码中log_blob_name未定义，需改为传入的blob_name参数
• 若需要覆盖容器中已存在的同名Blob，添加overwrite=True参数即可

内容的提问来源于stack exchange，提问作者Asu