Local ranking and search toolkit for MoonBit packages, combining dependency, release, and download signals.

moonbit

mooncakes

ranking

Download zip

Author

Luna-Flow

Version

0.2.0

License

Apache-2.0

Repository

github.com/Luna-Flow/mooncake_impact_factor

Last updated

3 days ago

Downloads

#MOONCAKE IMPACT FACTOR

mooncake_impact_factor ranks the packages of the MoonBit registry by how much the ecosystem relies on them, the way a journal impact factor measures how often a journal is cited. It counts the packages whose latest release depends on each package (those of the same owner at a quarter), the recent ones among them, downloads and release age; it places every package by position, grade (S to D, shares of the registry) and 30-day momentum (New, Rising, Stable, Cooling), and serves the rankings at https://impact-factor.lunaflow.cn, rebuilt every day. The rules are a MoonBit library, so the same numbers can be computed in any MoonBit program, with the same bits on every backend.

This README describes version 0.2.0.

#Install

moon add Luna-Flow/mooncake-impact-factor@0.2.0

Import the package you need in moon.pkg:

import {
  "Luna-Flow/mooncake-impact-factor/score",
}

#Example

test "score a registry" {
  // 20 dependents of other owners (4 of them recent), 3 of the same owner,
  // 300 downloads, released 40 days ago; and a package without any signal.
  let busy : @score.Signals = {
    external_dependents: 20,
    self_dependents: 3,
    recent_external_dependents: 4,
    recent_self_dependents: 0,
    downloads: 300,
    days_since_release: 40,
  }
  let idle : @score.Signals = {
    ..busy,
    external_dependents: 0,
    self_dependents: 0,
    recent_external_dependents: 0,
    downloads: 0,
  }
  let snapshots = @score.score_population([busy, idle][:], [Some(busy), None][:])
  inspect(snapshots[0].rank_label, content="S")
  inspect(snapshots[0].momentum_label, content="Stable")
  inspect(snapshots[1].momentum_label, content="New")
}

The score is for dependents, recent dependents (both counting those of the same owner at ), downloads and a release-recency multiplier falling from at 30 days to at one year.

#Packages

Package	Purpose
score	Impact score and its breakdown, rank positions, grades, momentum labels, scored populations.
metrics	Registry signals: latest releases, current dependents, recent dependents, the signals of 30 days ago, and the scores of the whole registry.
cli	JavaScript command build-index that runs metrics on a registry file; used by the Python index builder.
query	Search query language: query tree, expression parser and serializer, flat parameters, labels, sort keys, paging.
query_sql	Compiles search requests to SQLite statements for the dynamic site.
static_search	Search engine of the static site: evaluation, relevance, sorting, paging.

Outside MoonBit, scripts/build_index.py moves the registry index and the download counts into the MoonBit computation and its report into SQLite, scripts/export_static_json.py exports the database for the static site, and app/, frontend/src/ and lib/ hold the Next.js application. Its look and its localisation follow the Luna-Flow documentation site: the same tokens and page chrome, pages under /en/, /zh-cn/ and /ja/, and interface strings translated through gettext catalogs in web/i18n.

#Run the application

moon update                        # refresh ~/.moon/registry/index/user
python3 scripts/build_index.py --db data/mooncake.db --refresh-downloads
npm install
MOONCAKE_DB_PATH=data/mooncake.db npm run dev -- --hostname 127.0.0.1 --port 3000

Use --skip-mooncakes-downloads to build offline from the cached counts, or --downloads-json <file> to override download counts. just dev runs the whole sequence. For the static site, run npm run build:static-data, npm run build:static and npm run serve:static.

The rankings come from the registry index and the download counts of mooncakes.io; they measure reliance, not quality, and are not an authoritative ranking of the ecosystem.

#Toolchain

MoonBit with moonc 0.10 or later (moon.mod / moon.pkg manifests).
Node.js 20.16, 22.3 or later, and npm.
Python 3.

#Documentation

The manual is published at https://lunaflow.cn/en/mooncake_impact_factor/ with Chinese and Japanese translations. Its English source is doc/manual/index.md: API references, tutorials and design notes for every package, a getting-started guide and an architecture guide. The score design derives the formula, the grades and the momentum rules, the metrics design defines the signals, and the static_search design the browser ranking and how it differs from the server. Changes between versions are listed in CHANGELOG.md.

#Contributing

See CONTRIBUTING.md for the workflow, the checks to run before a pull request and the release procedure.

#License

Apache-2.0. See LICENSE.

Site source Report issue Packages Build queue Skills Statistics