WIP Modernize

- switch to type: module
- replace all CJS require/module.exports with ES6 import/expor

Author: bhousel

index.mjs

@@ -1,3 +1,3 @@
-export { default as matcher } from './lib/matcher.js';
-export { default as simplify } from './lib/simplify.js';
-export { default as stemmer } from './lib/stemmer.js';
+export { Matcher } from './lib/matcher.js';
+export { simplify } from './lib/simplify.js';
+export { stemmer } from './lib/stemmer.js';

lib/file_tree.js

@@ -1,18 +1,21 @@
-const colors = require('colors/safe');
-const fs = require('fs-extra');
-const glob = require('glob');
-const idgen = require('./idgen.js');
-const JSON5 = require('json5');
-const sortObject = require('./sort_object.js');
-const stringify = require('@aitodotai/json-stringify-pretty-compact');
-const withLocale = require('locale-compare')('en-US');
+import colors from 'colors/safe.js';
+import fs from 'fs-extra';
+import glob from 'glob';
+import JSON5 from 'json5';
+import localeCompare from 'locale-compare';
+import stringify from '@aitodotai/json-stringify-pretty-compact';
+
+import { idgen } from './idgen.js';
+import { sortObject } from './sort_object.js';
+import { validate } from './validate.js';
+
+const withLocale = localeCompare('en-US');
 
 // metadata about the trees
-const trees = require('../config/trees.json').trees;
+const trees = JSON5.parse(fs.readFileSync('./config/trees.json', 'utf8')).trees;
 
 // validate the files as we read them
-const validate = require('./validate.js');
-const categoriesSchema = require('../schema/categories.json');
+const categoriesSchema = JSON5.parse(fs.readFileSync('./schema/categories.json', 'utf8'));
 
 // The code in here
 //  - validates data on read, generating any missing data
@@ -32,7 +35,9 @@ const categoriesSchema = require('../schema/categories.json');
 //   },
 
 
-exports.read = (cache, loco) => {
+export let fileTree = {
+
+read: (cache, loco) => {
   cache = cache || {};
   cache.id = cache.id || new Map();
   cache.path = cache.path || {};
@@ -156,10 +161,10 @@ exports.read = (cache, loco) => {
   });
 
   return cache;
-};
+},
 
 
-exports.write = (cache) => {
+write: (cache) => {
   cache = cache || {};
   cache.path = cache.path || {};
 
@@ -305,10 +310,10 @@ exports.write = (cache) => {
     if (typeof val !== 'string') return val;
     return val.trim().toLowerCase();
   }
-};
+},
 
 
-exports.expandTemplates = (cache, loco) => {
+expandTemplates: (cache, loco) => {
   cache = cache || {};
   cache.id = cache.id || new Map();
   cache.path = cache.path || {};
@@ -406,4 +411,6 @@ exports.expandTemplates = (cache, loco) => {
   });
 
   return cache;
+}
+
 };

lib/idgen.js

@@ -1,10 +1,11 @@
-const crypto = require('crypto');
-const simplify = require('./simplify.js');
+import crypto from 'node:crypto';
+import { simplify } from './simplify.js';
+
 
 // We want the identifiers to be useable in url strings and other places,
 // and avoid any unicode or right-to-left surprises,
 // so limit them to /^\w+$/  (only [A-Za-z0-9_] characters)
-module.exports = (item, tkv, locationID) => {
+export function idgen(item, tkv, locationID) {
   let name;
 
   const parts = tkv.split('/', 3);     // tkv = "tree/key/value"
@@ -53,4 +54,4 @@ module.exports = (item, tkv, locationID) => {
   } else {
     return null;
   }
-};
+}

lib/matcher.js

@@ -1,11 +1,17 @@
-const simplify = require('./simplify.js');
-const matchGroups = require('../config/matchGroups.json').matchGroups;
-const genericWords = require('../config/genericWords.json').genericWords;
-const trees = require('../config/trees.json').trees;
-const whichPolygon = require('which-polygon');
+import whichPolygon from 'which-polygon';
+import { simplify } from './simplify.js';
 
+// This will not work in the browser :(
+// We may be able to switch to `import`, but:
+// - for node, with node with --experimental-json-modules  https://stackoverflow.com/a/59758333/7620
+// - for browser, both esbuild and rollup should be able to bundle a .json import
+import fs from 'fs';
+const matchGroups = JSON.parse(fs.readFileSync('./config/matchGroups.json', 'utf8')).matchGroups;
+const genericWords = JSON.parse(fs.readFileSync('./config/genericWords.json', 'utf8')).genericWords;
+const trees = JSON.parse(fs.readFileSync('./config/trees.json', 'utf8')).trees;
 
-module.exports = () => {
+
+export function Matcher() {
 
   // The `_matchIndex` is a specialized structure that allows us to quickly answer
   //   _"Given a [key/value tagpair, name, location], what canonical items (brands etc) can match it?"_
@@ -521,4 +527,4 @@ module.exports = () => {
 
 
   return matcher;
-};
+}

lib/simplify.js

@@ -1,8 +1,8 @@
-const diacritics = require('diacritics');
+import diacritics from 'diacritics';
 
 // remove spaces, punctuation, diacritics
 // for punction see https://stackoverflow.com/a/21224179
-module.exports = (str) => {
+export function simplify(str) {
   if (typeof str !== 'string') return '';
 
   return diacritics.remove(
@@ -12,4 +12,4 @@ module.exports = (str) => {
       .replace(/[\s\-=_!"#%'*{},.\/:;?\(\)\[\]@\\$\^*+<>«»~`’\u00a1\u00a7\u00b6\u00b7\u00bf\u037e\u0387\u055a-\u055f\u0589\u05c0\u05c3\u05c6\u05f3\u05f4\u0609\u060a\u060c\u060d\u061b\u061e\u061f\u066a-\u066d\u06d4\u0700-\u070d\u07f7-\u07f9\u0830-\u083e\u085e\u0964\u0965\u0970\u0af0\u0df4\u0e4f\u0e5a\u0e5b\u0f04-\u0f12\u0f14\u0f85\u0fd0-\u0fd4\u0fd9\u0fda\u104a-\u104f\u10fb\u1360-\u1368\u166d\u166e\u16eb-\u16ed\u1735\u1736\u17d4-\u17d6\u17d8-\u17da\u1800-\u1805\u1807-\u180a\u1944\u1945\u1a1e\u1a1f\u1aa0-\u1aa6\u1aa8-\u1aad\u1b5a-\u1b60\u1bfc-\u1bff\u1c3b-\u1c3f\u1c7e\u1c7f\u1cc0-\u1cc7\u1cd3\u2000-\u206f\u2cf9-\u2cfc\u2cfe\u2cff\u2d70\u2e00-\u2e7f\u3001-\u3003\u303d\u30fb\ua4fe\ua4ff\ua60d-\ua60f\ua673\ua67e\ua6f2-\ua6f7\ua874-\ua877\ua8ce\ua8cf\ua8f8-\ua8fa\ua92e\ua92f\ua95f\ua9c1-\ua9cd\ua9de\ua9df\uaa5c-\uaa5f\uaade\uaadf\uaaf0\uaaf1\uabeb\ufe10-\ufe16\ufe19\ufe30\ufe45\ufe46\ufe49-\ufe4c\ufe50-\ufe52\ufe54-\ufe57\ufe5f-\ufe61\ufe68\ufe6a\ufe6b\ufeff\uff01-\uff03\uff05-\uff07\uff0a\uff0c\uff0e\uff0f\uff1a\uff1b\uff1f\uff20\uff3c\uff61\uff64\uff65]+/g,'')
       .toLowerCase()
   );
-};
+}

lib/sort_object.js

@@ -1,8 +1,9 @@
-const withLocale = require('locale-compare')('en-US');
+import localeCompare from 'locale-compare';
+const withLocale = localeCompare('en-US');
 
 // Returns an object with sorted keys and sorted values.
 // (This is useful for file diffing)
-module.exports = (obj) => {
+export function sortObject(obj) {
   if (!obj) return null;
 
   let sorted = {};
@@ -21,4 +22,4 @@ module.exports = (obj) => {
       return withLocale(a, b);
     }
   }
-};
+}

lib/stemmer.js

@@ -1,8 +1,8 @@
-const simplify = require('./simplify.js');
+import { simplify } from './simplify.js';
 
 // Removes noise from the name so that we can compare
 // similar names for catching duplicates.
-module.exports = (str) => {
+export function stemmer(str) {
   if (typeof str !== 'string') return '';
 
   const noise = [
@@ -17,4 +17,4 @@ module.exports = (str) => {
 
   str = noise.reduce((acc, regex) => acc.replace(regex, ''), str);
   return simplify(str);
-};
+}

lib/validate.js

@@ -1,8 +1,9 @@
-const colors = require('colors/safe');
-const Validator = require('jsonschema').Validator;
+import colors from 'colors/safe.js';
+import jsonschema from 'jsonschema';
 
 // Perform JSON Schema validation
-module.exports = (fileName, object, schema) => {
+export function validate(fileName, object, schema) {
+  const Validator = jsonschema.Validator;
   const v = new Validator();
   const validationErrors = v.validate(object, schema, { nestedErrors: true }).errors;
   if (validationErrors.length) {
@@ -25,4 +26,4 @@ module.exports = (fileName, object, schema) => {
     console.error();
     process.exit(1);
   }
-};
+}

lib/write_file_with_meta.js

@@ -1,8 +1,8 @@
-const crypto = require('crypto');
-const fs = require('fs');
-const JSON5 = require('json5');
-const packageJSON = require('../package.json');
+import crypto from 'node:crypto';
+import fs from 'node:fs';
+import JSON5 from 'json5';
 
+const packageJSON = JSON.parse(fs.readFileSync('./package.json', 'utf8'));
 const URLRoot = 'https://raw.githubusercontent.com/osmlab/name-suggestion-index/main';
 
 //
@@ -13,7 +13,7 @@ const URLRoot = 'https://raw.githubusercontent.com/osmlab/name-suggestion-index/
 // `file` = the path to the file
 // `contents` = should be stringified json containing an object {}
 //
-module.exports = (file, contents) => {
+export function writeFileWithMeta(file, contents) {
   // Load the previous file
   let previous = { _meta: { } };
   try {
@@ -39,4 +39,4 @@ module.exports = (file, contents) => {
     // Stick metadata at the beginning of the file in the most hacky way possible
     fs.writeFileSync(file, contents.replace(/^\{/, '{' + meta));
   }
-};
+}

package.json

@@ -52,8 +52,13 @@
     "transit",
     "wikidata"
   ],
+  "type": "module",
   "main": "dist/index.js",
-  "module": "index.mjs",
+  "module": "./index.mjs",
+  "exports": {
+    "import": "./index.mjs",
+    "require": "./dist/index.js"
+  },
   "nsiguide": "docs/index.html",
   "targets": {
     "nsiguide": {
@@ -78,7 +83,7 @@
     "build": "run-s build:features build:index",
     "build:features": "node scripts/build_features.js",
     "build:index": "node scripts/build_index.js",
-    "jest": "jest --coverage",
+    "jest": "node --experimental-vm-modules node_modules/.bin/jest --no-cache --coverage",
     "lint": "eslint scripts/*.js lib/*.js",
     "test": "run-s lint build jest",
     "validate": "node scripts/validate.js",
@@ -125,20 +130,23 @@
     "rollup": "^2.36.2",
     "safe-regex": "^2.1.1",
     "shelljs": "^0.8.0",
-    "whatwg-fetch": "^3.5.0",
-    "xmlbuilder2": "^2.1.2"
-  },
-  "optionalDependencies": {
     "twitter": "^1.7.1",
+    "whatwg-fetch": "^3.5.0",
     "wikibase-edit": "^4.7.3",
-    "wikibase-sdk": "^7.7.1"
+    "wikibase-sdk": "^7.7.1",
+    "xmlbuilder2": "^2.1.2"
   },
   "engines": {
-    "node": ">=10"
+    "node": "^12.20.0 || ^14.13.1 || >=16.0.0"
+  },
+  "jest": {
+    "moduleFileExtensions": ["ts", "js", "json"],
+    "transform": {},
+    "verbose": true
   },
   "babel": {
     "presets": [
       "@babel/preset-react"
     ]
   }
-}
+}