AnswerDotAI / AnswerDotAI/fastprogress
python-fasthtml dependency pulls in SQLite C bindings, breaking installs on some platforms
- Lenguaje dominante
- Jupyter Notebook
- Estrellas
- 1.1k
- Forks
- 103
- Merge medio
- 1 min
- PR fusionados (30 d)
- 1
Descripción
fastprogress 1.1.0 switched from raw HTML strings to fasthtml FT components for Jupyter rendering. The problem is the transitive dep chain this introduces:
```
fastprogress -> python-fasthtml -> fastlite -> apswutils -> apsw
```
apsw is a SQLite C extension that needs pre-built wheels or a compiler. Poetry can't find wheels for some platforms, so `poetry install` just fails -- which blocks anything that depends on fastprogress (in my case, blackjax -- It seems the blackjax team is also tracking this on their end: https://github.com/blackjax-devs/blackjax/issues/812).
fastprogress only uses 9 tag constructors (Div, Progress, Table, Tr, Td, Th, Thead, Tbody, P) but python-fasthtml also drags in starlette, uvicorn, httpx, oauthlib, beautifulsoup4, etc. That's a lot of weight for a progress bar.
The old f-string approach in core.py still exists and worked fine. Could the fasthtml dep be made optional with a fallback to the f-string versions? Or even just vendor the few FT helpers that are actually needed?
My workaround for this is to pin `fastprogress<1.1` in my project.
Guía de contribución
Línea de trabajo
Comienza con la configuración de dependencias utilizada por Poetry y la implementación existente de f-string en core.py; sigue cómo python-fasthtml introduce la cadena de apsw. Se considera terminado cuando las plataformas afectadas pueden instalar fastprogress sin la dependencia de SQLite C que falla y la representación del progreso en Jupyter sigue funcionando.
Escrito por el modelo de indexación a partir del texto del issue.
Evaluación
- Stack tecnológico
- jupyter-notebook, python
- Área
- build-system
- Tipo de issue
- Error
- Dificultad
- 4/5
- Tiempo estimado
- 3-5 días
- Estado de actividad
- Estancado
- Claridad
- Bastante claro
- Aptitud para principiantes
- 38/100