Fishing Analyzer

Fishing Analyzer now runs as a lightweight server-rendered dashboard with:

• FastAPI
• Jinja2 templates
• HTMX for partial updates
• Tailwind (CDN)

A legacy Dash implementation is still present under src/fishing_analyzer/apps/.

Why this stack

• Python-first and minimal operational overhead
• No SPA build pipeline required
• Easy server-side rendering and progressive enhancement with HTMX
• Good fit for dashboard-only use cases without auth/admin complexity

Repository Layout

• src/fishing_analyzer/web/: FastAPI app, templates, dashboard service
• src/fishing_analyzer/data/: fish dataset and legacy environment loaders
• src/fishing_analyzer/apps/: legacy Dash pages
• docker/: container Dockerfile and entrypoint
• docs/: development and release notes
• tests/: unit tests

Prerequisites

• Python 3.11+
• uv

Setup

just setup

Run (FastAPI dashboard)

just dev

Open: http://localhost:8085

Legacy Dash mode

just dev-legacy

Development Commands

just format
just lint
just typecheck
just test
just check
just ci

Docker

just docker-up
just docker-down

By default Docker runs the FastAPI dashboard and does not require MongoDB.

Optional Mongo Integration

If you want to run legacy Mongo-backed imports in container mode:

• set RUN_WITH_MONGO=true
• set RUN_DATA_IMPORT=true
• provide MONGODB_URI

Release Workflow

• CI entry: .github/workflows/ci.yml
• Reusable jobs: lint, typecheck, tests, build, binary
• Release jobs: PyPI/TestPyPI + binaries

Tag-based release:

git tag vX.Y.Z
git push origin vX.Y.Z

Troubleshooting

• ModuleNotFoundError: run just setup again.
• If template changes do not show: restart just dev.
• If lockfile drift happens: run uv lock and commit uv.lock.