Plotly Dash：如何重现 dcc.Upload 的 'content' 输出？ （即 base64 编码的字符串）

Question

我无法重现 dcc.Upload 组件的 content-属性 的准确输出。

如果我将文件 my_excel.xlsx 上传到 dcc.Upload 组件，我的回调函数会收到一个“base64 编码字符串”（根据 dcc.Upload documentation）。我不知道如何在没有 dcc.Upload 组件的情况下重现完全相同的字符串 （我想使用单元测试的输出）。

我目前的做法：

import base64
import io
import pandas as pd

# This is what I try to reproduce the output of the dcc.Upload Component
with open('tests/data/my_excel.xlsx', 'rb') as file:
    raw_data = file.read()
    
# raw_data is the output I receive from the dcc.Upload Component

# these steps are raise no Error with the output of dcc.Upload
_, content_string = raw_data.split(',') # this Fails
decoded = base64.b64decode(content_string)
df = pd.read_excel(io.BytesIO(decoded))

我收到错误 TypeError: a bytes-like object is required, not 'str'。

如果我添加

raw_data = base64.b64encode(raw_data)

在 raw_data.split(',') 之前，我得到了同样的错误。

如何在没有 dcc.Upload 组件的情况下获得完全相同的“base64 编码字符串”？

Answer 1

我找不到一个函数来重现 dcc.Upload 的 contents 属性，但能够手动创建 dcc.Upload.[=20 的输出=]

从 documentation 我们有：

contents is a base64 encoded string that contains the files contents [...] Property accept (string; optional): Allow specific types of files. See https://github.com/okonet/attr-accept for more information. Keep in mind that mime type determination is not reliable across platforms. CSV files, for example, are reported as text/plain under macOS but as application/vnd.ms-excel under Windows. In some cases there might not be a mime type set at all.

检查 contents 字符串，发现它由两个字符串组成：

content_type, content_string = contents.split(',')

进一步检查显示：
content_type: 包含文件
的mime类型信息 content_string: 文件的base64编码内容

import base64
import io
import pandas as pd
import magic

filepath = 'tests/data/my_excel.xlsx'

# Reproduce output of dcc.Upload Component
with open(filepath, "rb") as file:
    decoded = file.read()
content_bytes = base64.b64encode(decoded)
content_string = content_bytes.decode("utf-8")

mime = magic.Magic(mime=True)
mime_type = mime.from_file(filepath)
content_type = "".join(["data:", mime_type, ";base64"])

contents = "".join([content_type, ",", content_string])

# and now revert: convert contents to binary file stream
content_type, content_string = contents.split(",")
decoded = base64.b64decode(content_string)
df = pd.read_excel(io.BytesIO(decoded))