Help: How to obtain the content of the file uploaded at the Start node

[Baijianlu]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• Please do not modify this template :) and fill in all the required fields.

1. Is this request related to a challenge you're experiencing? Tell me about your story.

Version

dify 1.2.0

Cloud or Self Hosted

Self Hosted, Ubuntu system

Description of the problem

I need to extract the content of a file uploaded at the Start node. Since the file is not a typical doc type file, I have to write a customer plugin to process it. However, I met problems when try to obtain the content of the uploaded file in Plugin daemon. The "file" being transferred into the plugin is actually a "dify_plugin.file.File" object, it doesn't contain the real content of the file, instead, it contains the description and some "properties" of the file, such as filename, extension, size, url, etc. I noticed I could get the content of the file using the "blob" function of class File, but when I used it, it reported the following error:

"Run failed: url should start with http:// or https://"

Since in blob function, it uses "httpx" to extract the content from file url, that's why it requires url start with http:// or https://. However, I checked the "url" of my file object, it started with "/file/****". I think this is because I host the dify in my own computer, while dify doesn't produce the correct url start with http.
I don't know if this is a bug or something? It seems using blob function with correct url as input is the only way to get the content of uploaded file. Do anyone know other method to get the content of the file?

2. Additional context or comments

No response

[vvhr]

try to using file donwload api

[Baijianlu]

Could you please explain more? which kind of download api I should use?

[Baijianlu]

Does anyone meet the same problem?

[ma642]

Actually the 'dify_plugin.file.File.blob' download file from remote after you have set FILES_URL/INTERNAL_FILES_URL in .env file. Remember expose the 5001 port in docker-compse.yaml if in plugin debugging mode.