无法在 azure 函数中使用 pandas 读取 csv 和 python

Question

我在 Azure function python 中创建了一个 Azure Blob 存储触发器。一个 CSV 文件添加到 Blob 存储中，我尝试用 pandas 读取它。

import logging
import pandas as pd

import azure.functions as func


def main(myblob: func.InputStream):
    logging.info(f"Python blob trigger function processed blob \n"
                 f"Name: {myblob.name}\n"
                 f"Blob Size: {myblob.length} bytes")

    df_new = pd.read_csv(myblob)
    print(df_new.head())

如果我将myblob传递给pd.read_csv ，那么我会得到UnsupportedOperation: read1

Python blob trigger function processed blob 
Name: samples-workitems/Data_26112022_080027.csv
Blob Size: None bytes
[2022-11-27T16:19:25.650Z] Executed 'Functions.BlobTrigger1' (Failed, Id=2df388f5-a8dc-4554-80fa-f809cfaeedfe, Duration=1472ms)
[2022-11-27T16:19:25.655Z] System.Private.CoreLib: Exception while executing function: Functions.BlobTrigger1. System.Private.CoreLib: Result: Failure
Exception: UnsupportedOperation: read1

如果我通过myblob.read() ，

df_new = pd.read_csv(myblob.read())

它给出TypeError: Expected file path name or file-like object, got <class 'bytes'> type

Python blob trigger function processed blob 
Name: samples-workitems/Data_26112022_080027.csv
Blob Size: None bytes
[2022-11-27T16:09:56.513Z] Executed 'Functions.BlobTrigger1' (Failed, Id=e3825c28-7538-4e30-bad2-2526f9811697, Duration=1468ms)
[2022-11-27T16:09:56.518Z] System.Private.CoreLib: Exception while executing function: Functions.BlobTrigger1. System.Private.CoreLib: Result: Failure
Exception: TypeError: Expected file path name or file-like object, got <class 'bytes'> type

来自Azure 函数文档：

InputStream 是 File-like object 表示输入 blob。

来自Pandas read_csv 文档：

read_csv 采用 filepath_or_bufferstr，路径 object 或类似文件 object

所以从技术上讲，我应该阅读这个 object。我在这里缺少什么拼图？

Answer 1

如果你参考这篇文章，它说这段代码可以工作。 但建议用于较小的文件，因为整个文件进入 memory。不建议用于较大的文件。

import logging
import pandas as pd

import azure.functions as func
from io import BytesIO

def main(myblob: func.InputStream):
    logging.info(f"Python blob trigger function processed blob \n"
                 f"Name: {myblob.name}\n"
                 f"Blob Size: {myblob.length} bytes")
    df_new = pd.read_csv(BytesIO(myblob.read()))
    print(df_new.head())