tiferet-streamlit 0.1.0

A Streamlit extension for the Tiferet Framework

tiferet-streamlit

A Streamlit extension for the Tiferet Framework — build multi-page Streamlit applications with Domain-Driven Design.

Installation

pip install tiferet-streamlit

Quick Start

import streamlit as st
from tiferet_streamlit import StreamlitApp, ViewContext

class HomeView(ViewContext):
    def init_state(self):
        self.session.set('count', 0)

    def render(self):
        count = self.session.get('count')
        st.title('Home')
        st.write(f'Count: {count}')
        if st.button('Increment'):
            self.session.set('count', count + 1)
            st.rerun()

app = StreamlitApp()
app.load_app_service()
app.run('my_interface', pages={'/': HomeView})

Core Concepts

ViewContext

The code-behind for a Streamlit page. Manages state via SessionCacheContext, dispatches Tiferet features via AppInterfaceContext, and defines UI through render().

• init_state() — Called once on first construction. Override to set initial state.
• dispatch(feature_id, headers=None, **data) — Execute a Tiferet feature.
• render() — Override to define Streamlit widgets.
• __call__() — Makes the view callable for st.Page composition.

ViewComponent

A lightweight, prop-driven sub-component with parent ViewContext access.

from tiferet_streamlit import ViewComponent

class Counter(ViewComponent):
    def render(self, label='Count', start=0):
        count = self.ctx.session.get('count') or start
        st.write(f'{label}: {count}')

SessionCacheContext

Cache backed by st.session_state with namespace isolation for multi-page apps.

from tiferet_streamlit import SessionCacheContext

cache = SessionCacheContext(namespace='my_view')
cache.set('key', 'value')
cache.get('key')  # 'value'

Multi-Page Applications

Use PageContext to register multiple views with routes:

app.run('my_interface', pages={
    '/': HomeView,
    '/about': AboutView,
    '/settings': SettingsView,
})

Config-Driven Pages

Define pages as Page domain objects for YAML-driven configuration:

from tiferet_streamlit import Page

pages = [
    Page(route='/', title='Home', icon='🏠',
         view_module_path='app.views.home', view_class_name='HomeView'),
    Page(route='/about', title='About', icon='ℹ️',
         view_module_path='app.views.about', view_class_name='AboutView'),
]

app.run('my_interface', page_configs=pages)

Feature Dispatch

Views dispatch Tiferet features for backend logic:

class CalcView(ViewContext):
    def render(self):
        a = st.number_input('a')
        b = st.number_input('b')
        if st.button('Add'):
            result = self.dispatch('calc.add', a=a, b=b)
            st.write(f'Result: {result}')

API Reference

Class	Module	Description
Page	domain.view	Page configuration domain object
ViewService	interfaces.view	Abstract service for page management
SessionCacheContext	contexts.session	Session-state-backed cache with namespacing
ViewContext	contexts.view	Page code-behind with lifecycle management
ViewComponent	contexts.view	Prop-driven sub-component
PageContext	contexts.page	Multi-page navigation manager
StreamlitBuilder	builders.main	Application entry point (also aliased as StreamlitApp)

Development

# Clone and set up
git clone https://github.com/greatstrength/tiferet-streamlit.git
cd tiferet-streamlit
python -m venv .venv
source .venv/bin/activate
pip install -e .[test]

# Run tests
pytest --verbose

License

MIT