Local Anime Tracker

追番手账 — 你的本地番剧档案馆

21,087 titles · 1975–2026 · All stored locally

中文文档

---

What if you could own every anime season from the past 50 years — searchable, sortable, and yours to keep?

Local Anime Tracker pulls data from Bangumi, organizes it by Year → Season → Title, and gives you a gorgeous Web UI to browse, track, and export your anime universe. No cloud, no account, no tracking — just you and 21,000+ anime on your own machine.

Highlights

• One-command collection — Pull 50 years of anime data from Bangumi API, with resume support
• Dual-theme Web UI — Cyberpunk dark mode with cyan-purple glow, or a fresh light mode with sky-blue & pink
• Smart browsing — Filter by year/season/type/status, sort by 9 fields, instant search in Japanese & Chinese
• Watch tracking — Mark titles as Want to Watch / Watching / Watched / On Hold / Dropped — persisted locally
• Excel export — Color-coded multi-sheet workbook, one sheet per year, high-score highlights
• All local — JSON cache + SQLite-free, your data never leaves your machine

Quick Start

Note: Bangumi API requires network access. Use a VPN if you're outside China.

# Install dependencies
pip install flask requests openpyxl pandas

# Launch the Web UI
python app.py

Open http://localhost:5000 and you're in.

Web UI Tour

Dashboard

Stats at a glance: total count, year range, coverage rates, plus a bar chart of anime per year and donut charts for season & category distribution.

Browse

The heart of the app. Filter by year, season (January / April / July / October), type (TV/OVA/Movie/WEB), and watch status. Dropdowns auto-trigger search — no click needed. Sort by any of 9 fields with ascending/descending toggle.

Every row has an inline watch-status selector. Changes are saved to the local cache instantly.

Collect

Set a year range, hit start, and watch the progress bar fill in real-time via SSE. Supports resume — if interrupted, re-run picks up where it left off.

Export

One click to generate a full Excel chronology. Or split by year into separate files. Download previous exports anytime.

Dual Themes

Mode	Vibe	Colors
Dark (default)	Cyberpunk / Hacker	Cyan-blue #7aa2f7 + Purple #bb9af7 on deep black
Light	Clean & Modern	Sky-blue #5bcffa + Pink #f5abb9 on white

Toggle with the moon/sun button in the sidebar. The sidebar itself collapses and expands — your layout preference persists across sessions.

CLI

For when you'd rather type than click:

# Collect everything (1975–2026)
python run_chronology.py collect

# Collect a specific range
python run_chronology.py collect --start-year 2020 --end-year 2026

# One year only
python run_chronology.py collect --year 2024

# Force re-fetch (ignore cache)
python run_chronology.py collect --no-cache

# Enrich with details (broadcast day, studio)
python run_chronology.py collect --enrich

# Export to Excel
python run_chronology.py export

# Export by year (one file per year)
python run_chronology.py export --by-year

# Quick sanity check
python run_chronology.py test

Excel Output

A single workbook with 54 sheets:

• Overview — Per-year anime count across all four seasons
• 1975–2026 — One sheet per year, grouped by season with color-coded headers
• All Data — Flat table for sorting and filtering

Season headers use color coding: blue for January, green for April, yellow for July, orange for October. Titles rated ≥8.0 get a green highlight.

Season	Months
January	Jan–Mar
April	Apr–Jun
July	Jul–Sep
October	Oct–Dec

Data Schema

Each record contains:

Field	Description
name_jp	Original Japanese title
name_cn	Chinese translated title
category	TV / OVA / Movie / WEB
episodes	Episode count
air_date	Broadcast start date
air_weekday	Broadcast weekday
studio	Production studio
score	Bangumi rating
watch_status	Your personal status

Project Structure

├── app.py                    # Flask Web UI
├── config.py                 # Configuration & constants
├── bangumi_collector.py      # Bangumi API client
├── infobox_parser.py         # Infobox field parser
├── season_builder.py         # Season-based collection orchestrator
├── chronology_store.py       # Data storage & caching
├── chronology_exporter.py    # Excel export engine
├── run_chronology.py         # CLI entry point
├── templates/                # Jinja2 templates
├── static/style.css          # Dual-theme stylesheet
├── data/bangumi/             # Cached data (JSON)
└── output/                   # Exported Excel files

Data Source

All data comes from the Bangumi API:

• GET /v0/subjects?type=2&year=Y&month=M — Browse anime by month
• GET /v0/subjects/{id} — Fetch detail (broadcast day, studio, etc.)

Requests are rate-limited to 1.5s intervals with automatic 429 retry. A full collection takes ~30 minutes.

License

MIT