Morning
- Uploaded a series of Python source code for web scraping and API calling, which were previously used by the MEM project, to the GitLab on DaoCloud.
- Additionally, successfully utilized n8n to request the interface of FastGPT for searching purposes only
- This interface will only return the search results, without engaging in conversation.
Afternoon
- Originally, the plan was to call the n8n pipeline on the apitable widget.
- However, it turned out to be unnecessary as operations can be directly performed within the apitable widget.
- Therefore, an attempt was made to convert the HTTP request from n8n into a format that could be executed on the apitable widget.
- However, there was difficulty in running it on the apitable widget.
- It was discovered that the header was missing the ‘Content-Type: application/json’ parameter.
- Strangely, in order to run successfully, n8n required the deletion of this parameter, while the apitable widget still needed it.
- This only involves establishing interaction with the fastgpt knowledge base; consideration needs to be given to the user interface aspect.
- Initially, the idea was to automatically query the knowledge base based on the content in a selected cell.
- However, it was realized that this was unnecessary and instead, users could input their query directly.
- Later, after reading a Linux book for a while, learned the basic usage of two commands - mount and umount.
- However, practical implementation has not been extensively explored yet.