raccoon / README.md
grapplerulrich's picture
Update streamlit version for spaces
9642bab unverified
---
title: Raccoon
emoji: 🦝
colorFrom: blue
colorTo: indigo
sdk: streamlit
sdk_version: 1.10.0
python_version: 3.9
app_file: app.py
pinned: false
license: mit
---
# Raccoon
## Installation
It is recommend to use virtual environment using [`venv`](https://docs.python.org/3/library/venv.html).
The fol
- If using Apple Silicon install rust `curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh` and `brew install cmake`
- Create the virtual envoirnment: `python3 -m venv .venv`
- Activate the virtual envoirnment: `source .venv/bin/activate`
- To deactive the virtual envoirnment run `deactivate` within the virtual envoirnment.
- Install the required packages: `.venv/bin/pip install -r requirements.txt`
- `.venv/bin/pip install -e .`
- [Create a custom search engine in Google](https://programmablesearchengine.google.com/controlpanel/all).
- Create a API for the custom search engine.
- Add the custom search engine key and PI key to `.streamlit/secrets.toml`.
```toml
google_search_api_key = "api-key"
google_search_engine_id = "search-engine-id"
```
- To start the interface: `streamlit run app.py`
### Todo
- [ ] Improve fetched content.
- [x] Fix issue of duplicate content extracted by beautifulsoup.
- [x] Exclude code from content
- [x] Find sentences that contain the search keywords.
- [ ] Find sentences that contain the search keywords taking into account different spellings health care vs healthcare.
- [ ] Get some content from every search result.
- [ ] Div's with text & tags. Extract text from tags and then decompose the tags. Keep order of content and no duplicates.
- [ ] Summarization requires truncation. Find solution where not needed.
- [ ] Support German content with language switcher.
- [ ] Improve queries to include more keywords (Expand abrivations & define context)
- [ ] Control the number of results from the UI.
- [ ] Control summary length via settings: https://docs.streamlit.io/library/advanced-features/session-state