Merge pull request #12 from transitive-bullshit/feature/sklearn-1.5.2

Author: transitive-bullshit

.eslintrc.json

@@ -0,0 +1,10 @@
+{
+  "root": true,
+  "extends": ["@fisch0920/eslint-config/node"],
+  "ignorePatterns": ["packages/sklearn/src/generated"],
+  "rules": {
+    "unicorn/no-array-reduce": "off",
+    "unicorn/filename-case": "off",
+    "array-callback-return": "off"
+  }
+}

.github/workflows/test.yml

@@ -9,42 +9,20 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        node-version:
-          - 19
-          - 18
-          - 16
+        node-version: [lts/*]
         python-version:
-          - '3.11'
+          - '3.13'
 
     steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Install Node.js
-        uses: actions/setup-node@v3
-        with:
-          node-version: ${{ matrix.node-version }}
-
-      - name: Install pnpm
-        uses: pnpm/action-setup@v2
-        id: pnpm-install
+      - uses: actions/checkout@v4
+      - uses: pnpm/action-setup@v4
         with:
-          version: 8
+          version: 9
           run_install: false
-
-      - name: Get pnpm store directory
-        id: pnpm-cache
-        shell: bash
-        run: |
-          echo "STORE_PATH=$(pnpm store path)" >> $GITHUB_OUTPUT
-
-      - uses: actions/cache@v3
-        name: Setup pnpm cache
+      - uses: actions/setup-node@v4
         with:
-          path: ${{ steps.pnpm-cache.outputs.STORE_PATH }}
-          key: ${{ runner.os }}-pnpm-store-${{ hashFiles('**/pnpm-lock.yaml') }}
-          restore-keys: |
-            ${{ runner.os }}-pnpm-store-
+          node-version: ${{ matrix.node-version }}
+          cache: 'pnpm'
 
       - name: Install Python
         uses: actions/setup-python@v4
@@ -56,7 +34,7 @@ jobs:
         run: pip install numpy scikit-learn
 
       - name: Install Node.js dependencies
-        run: pnpm install --frozen-lockfile
+        run: pnpm install --frozen-lockfile --strict-peer-dependencies
 
       - name: Run test
-        run: pnpm run test
+        run: pnpm test

.npmrc

@@ -1,2 +1,3 @@
 enable-pre-post-scripts=true
 ignore-workspace-root-check=true
+package-manager-strict=false

.prettierignore

@@ -6,4 +6,4 @@ node_modules/
 .vercel/
 third-party/
 out/
-packages/sklearn/src/generated/
+packages/sklearn/src/generated/

.prettierrc

@@ -0,0 +1,11 @@
+{
+  "singleQuote": true,
+  "jsxSingleQuote": true,
+  "semi": false,
+  "useTabs": false,
+  "tabWidth": 2,
+  "bracketSpacing": true,
+  "bracketSameLine": false,
+  "arrowParens": "always",
+  "trailingComma": "none"
+}

.prettierrc.cjs

@@ -1,7 +1,6 @@
 import fs from 'node:fs/promises'
 import path from 'node:path'
 
-import mkdir from 'mkdirp'
 import pMap from 'p-map'
 
 import { formatSource } from '../src/formatter'
@@ -13,7 +12,7 @@ import {
 
 async function main() {
   const outDir = path.join('packages', 'sklearn', 'src', 'generated')
-  await mkdir(outDir)
+  await fs.mkdir(outDir, { recursive: true })
 
   // 487 total definitions
   // 228 functions
@@ -24,8 +23,8 @@ async function main() {
   // console.log(JSON.stringify(scikitDocUrls, null, 2))
 
   const sources = scikitDocUrls.filter((url) => {
-    const nameParts = url.split('/').slice(-1)[0].split('.')
-    const name = nameParts.slice(-2)[0]
+    const nameParts = url.split('/').at(-1).split('.')
+    const name = nameParts.at(-2)
     if (name.toLowerCase() === name) {
       // TODO: likely a function
       // console.log('skipping', nameParts.slice(0, -1).join('.'))
@@ -38,12 +37,14 @@ async function main() {
   // return
 
   // const sources = [
-  //   'https://scikit-learn.org/stable/modules/generated/sklearn.metrics.average_precision_score.html'
+  //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html'
+  //   'https://scikit-learn.org/stable/modules/generated/sklearn.utils.metadata_routing.MetadataRouter.html'
+  //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.metrics.average_precision_score.html',
   //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.manifold.TSNE.html',
   //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.cluster.KMeans.html',
   //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.MinMaxScaler.html',
   //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html',
-  //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html'
+  //   // 'https://scikit-learn.org/stable/modules/generated/sklearn.calibration.CalibratedClassifierCV.html'
   // ]
 
   console.log(`\nfetching ${sources.length} docs...\n`)
@@ -74,12 +75,12 @@ async function main() {
   console.log(`\nprocessing ${docs.length} docs...\n`)
   // console.log(
   //   JSON.stringify(
-  //     docs.map((doc) => doc.name.split('.').slice(-1)[0]),
+  //     docs.map((doc) => doc.name.split('.').at(-1)),
   //     null,
   //     2
   //   )
   // )
-  // // console.log(JSON.stringify(docs[0], null, 2))
+  // console.log(JSON.stringify(docs[0], null, 2))
 
   const errors: string[] = []
   const generatedDirs: any = {}
@@ -99,10 +100,10 @@ async function main() {
           const fileName = `${doc.name}.ts`
           const namespaceDirs = doc.namespace.split('.').slice(1)
           const destDir = path.join(outDir, ...namespaceDirs)
-          await mkdir(destDir)
+          await fs.mkdir(destDir, { recursive: true })
 
           const filePath = path.join(destDir, fileName)
-          await fs.writeFile(filePath, source, 'utf-8')
+          await fs.writeFile(filePath, source, 'utf8')
 
           let generatedDir = generatedDirs
           for (const dir of namespaceDirs) {
@@ -158,7 +159,7 @@ async function main() {
       const source = await formatSource(indexFileSource)
       const fileName = 'index.ts'
       const filePath = path.join(destDir, fileName)
-      await fs.writeFile(filePath, source, 'utf-8')
+      await fs.writeFile(filePath, source, 'utf8')
       console.log(filePath)
     }
 
@@ -179,7 +180,7 @@ If you found this project helpful, please consider [donating to the official sci
 
       const fileName = 'readme.md'
       const filePath = path.join(destDir, fileName)
-      await fs.writeFile(filePath, readmeFileSource, 'utf-8')
+      await fs.writeFile(filePath, readmeFileSource, 'utf8')
       console.log(filePath)
     }
   }
@@ -195,4 +196,4 @@ If you found this project helpful, please consider [donating to the official sci
   )
 }
 
-main()
+await main()

bin/generate-sklearn.ts

@@ -3,7 +3,7 @@ import * as sklearn from 'sklearn'
 async function main() {
   const py = await sklearn.createPythonBridge()
 
-  const v = [
+  const _v = [
     [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
     [0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
     [3, 4, 5, 6, 7, 8, 9, 10, 11, 12],
@@ -52,4 +52,4 @@ async function main() {
   await py.disconnect()
 }
 
-main()
+await main()

bin/playground.ts

@@ -2,13 +2,12 @@ import fs from 'node:fs/promises'
 import path from 'node:path'
 
 import { globby } from 'globby'
-import mkdir from 'mkdirp'
 import pMap from 'p-map'
 
 async function main() {
   const srcDocsDir = path.join('packages', 'sklearn', 'docs')
   const outDir = path.join('docs', 'pages', 'docs')
-  await mkdir(outDir)
+  await fs.mkdir(outDir, { recursive: true })
 
   const docs = (
     await globby('**/*.md', {
@@ -45,7 +44,7 @@ async function main() {
     async (doc) => {
       console.log(`processing ${doc.relativePath}`)
 
-      let out = await fs.readFile(doc.srcPath, 'utf-8')
+      let out = await fs.readFile(doc.srcPath, 'utf8')
 
       out =
         out
@@ -54,11 +53,16 @@ async function main() {
             'Defined in: $1'
           )
           // TODO: handle relative links
-          .replaceAll(/\[([^\]]+)\]\((\.\.[^\)]+)\)/g, '$1')
+          .replaceAll(/\[([^\]]+)]\((\.\.[^)]+)\)/g, '$1')
           .replaceAll('[Readme](readme.md)', '')
+          .replaceAll(/#+ Parameters/g, '**Parameters**')
+          .replaceAll(/#+ Returns\n\n/gm, '**Returns** ')
+          .replaceAll(/#+ Defined in\n\n/gm, '**Defined in** ')
+          .replaceAll(/#+ Get Signature/g, '**Get Signature**')
+          .replaceAll(/#+ Set Signature/g, '**Set Signature**')
           .trim() + '\n'
 
-      if (doc.relativePath === 'modules.md') {
+      if (doc.relativePath === 'globals.md') {
         out = out.replaceAll(/### Variables\b/g, '### Constants')
       }
 
@@ -74,62 +78,55 @@ async function main() {
         metaMap.variables.push(doc.relativePath)
       }
 
-      await mkdir(path.dirname(doc.destPath))
-      await fs.writeFile(doc.destPath, out, 'utf-8')
+      await fs.mkdir(path.dirname(doc.destPath), { recursive: true })
+      await fs.writeFile(doc.destPath, out, 'utf8')
     },
     {
       concurrency: 8
     }
   )
 
   {
-    // top-level nextra _meta.json file
+    // top-level nextra _meta.ts file
     const docsMeta = {
-      modules: 'Table of Contents',
+      globals: 'Table of Contents',
       classes: 'Classes',
       functions: 'Functions',
-      interfaces: {
-        title: 'Interfaces',
+      'type-aliases': 'Types',
+      variables: 'Constants',
+      README: {
         display: 'hidden'
-      },
-      types: 'Types',
-      variables: 'Constants'
+      }
     }
 
     await fs.writeFile(
-      path.join(outDir, '_meta.json'),
-      JSON.stringify(docsMeta, null, 2),
-      'utf-8'
+      path.join(outDir, '_meta.ts'),
+      `export default ${JSON.stringify(docsMeta, null, 2)}`,
+      'utf8'
     )
   }
 
-  // sub nextra _meta.json files
+  // sub nextra _meta.ts files
   await pMap(
     Object.keys(metaMap),
     async (key) => {
       const values = metaMap[key]
-        .map((value) => value.split('/').slice(-1)[0].split('.')[0].trim())
+        .map((value) => value.split('/').at(-1).split('.')[0].trim())
         .filter(Boolean)
         .sort((a, b) => a.localeCompare(b))
 
       if (!values.length) {
         return
       }
 
-      const meta = values.reduce(
-        (acc, value) => ({
-          ...acc,
-          [value]: value
-        }),
-        {}
-      )
+      const meta = Object.fromEntries(values.map((value) => [value, value]))
 
       const destDir = path.join(outDir, key)
 
       await fs.writeFile(
-        path.join(path.join(destDir, '_meta.json')),
-        JSON.stringify(meta, null, 2),
-        'utf-8'
+        path.join(path.join(destDir, '_meta.ts')),
+        `export default ${JSON.stringify(meta, null, 2)}`,
+        'utf8'
       )
     },
     {
@@ -138,4 +135,4 @@ async function main() {
   )
 }
 
-main()
+await main()

bin/transform-docs.ts

@@ -0,0 +1,21 @@
+# Contributing
+
+## TODO
+
+## Node.js
+
+- install pnpm
+
+## Python
+
+- install poetry
+
+https://scipy.github.io/devdocs/building/index.html#system-level-dependencies
+
+```sh
+brew install scipy
+```
+
+```sh
+poetry install
+```