menxli / lires Goto Github PK

Parsing many bibtex on initializing database (DataPoint.loadInfo) can be slow due to too many entries,
May speed it up with multiprocessing.

Currently, the project planning are scheduled in readme file,
Use github issue would be a better way to keep track of the thoughts and progress.

Establish a machenism to share datapoints between libraries

Reality:

Expected：

And also the website cannot be opened.
What I exec:
docker run -d -p 9096:8080 -v /lires_data/:/root/.Lires --restart=always --gpus all -e OPENAI_API_BASE=xxxxxx -e OPENAI_API_KEY=xxxxx --name lrs lires:latest

To visualize the document feature in frontend.
Enable interactions with the plot, possibly use plotly.js?

Maybe change doc feature on modifications:

• Addition
• Deletion
• Adding / Changing abstract
• Adding / Changing document
• AI summarize

Resbibman is not a good name ...

Need a short, concise, insightful, marvellous, ground breaking name!

The bibtex special characters and curly braces are not handled correctly,
for example, a bibtex author entry of:

@article{sochorova2021practical,
    author = "Sochorov{\'a}, {\v{S}}{\'a}rka and Jamri{\v{s}}ka, Ond{\v{r}}ej",
    title = "Practical pigment mixing for digital painting",
    journal = "ACM Transactions on Graphics (TOG)",
    volume = "40",
    number = "6",
    pages = "1--11",
    year = "2021",
    publisher = "ACM New York, NY, USA"
}

will not be rendered correctly.
Need to find a solution.

Vector database offers better performace and control over vector operations, and are more suitable for large scale data.
Planning to use it to substitute current pytorch-based management system.

Currently the bibtex has to be entered manually,
which is inconvenient when BibTeX or any source of citations is not available.
Particularly problematic when trying to save PDF slides or webpages.

Need to find a better solution to address this issue.

Will deprecate PyQt UI in the future and use webui as the only UI.
i.e. Move all PyQt import into resbibman.gui and make it an optional submodule

Add an account associated level-of-reading record

纯新手，开始以为这几个命令是都要一次性运行的：

registry            Start lires registry server
    server              Start lires main server
    ai                  Start lires ai server
    log                 Start lires log server
    feed                Start lires feed server

后来发现一次似乎只要运行一个就行，不知道这样理解对不对。
运行第一行lires registry，基本上是一路缺啥pip啥，最后提示Starting service at: http://127.0.0.1:8700。我是在局域网中的一个Ubuntu服务器上远程安装的，所以使用了http://局域网服务器ip:8700 尝试打开网页，没成功，是要用nginx做反向代理吗？

一路pip后，运行lires server时，提示 File "/home/user/.local/lib/python3.11/site-packages/starlette/staticfiles.py", line 59, in __init__ raise RuntimeError(f"Directory '{directory}' does not exist") RuntimeError: Directory 'static/' does not exist，不知道该如何修改