Parse plural-forms header into JS function, drop make-plural dependency

Author: eemeli

README.md

@@ -22,7 +22,7 @@ object destructuring and arrow functions, you'll want to use a transpiler for th
 
 ```js
 const { parsePo, parseMo } = require('gettext-to-messageformat')
-const { headers, translations } = parsePo(`
+const { headers, pluralFunction, translations } = parsePo(`
 # Examples from http://pology.nedohodnik.net/doc/user/en_US/ch-poformat.html
 # Note that the given plural-form is incomplete
 msgid ""
@@ -48,7 +48,7 @@ msgstr "Nema zvezde po imenu %(starname)s."
 `)
 
 const MessageFormat = require('messageformat')
-const mf = new MessageFormat(headers.language)
+const mf = new MessageFormat({ [headers.language]: pluralFunction })
 const messages = mf.compile(translations)
 
 messages['Time: %1 second']([1])
@@ -80,20 +80,32 @@ the parser, including the following fields:
   If no context is set, by default this top-level key is not included unless
   `forceContext` is set to `true`.
 
-- `pluralCategories` (array of strings) – If the Language header is not set in
-  the input, or if its Plural-Forms `nplurals` value is not 1, 2, or 6, this
-  needs to be set to the pluralization category names to be used for the input
-  enumerated categories if any message includes a plural form.
+- `pluralFunction` (function) – If your input file does not include a Plural-Forms
+  header, or if for whatever reason you'd prefer to use your own, set this to be
+  a stringifiable function that takes in a single variable, and returns the
+  appropriate pluralisation category. Following the model used internally in
+  [messageformat], the function variable should also include `cardinal` as a
+  member array of its possible categories, in the order corresponding to the
+  gettext pluralisation categories. This is relevant if you'd like to avoid the
+  `new Function` constructor otherwise used to generate `pluralFunction`, or to
+  allow for more fine-tuned categories than gettext allows, e.g. differentiating
+  between the categories of `'1.0'` and `'1'`.
 
 - `verbose` (boolean, default `false`) – If set to `true`, missing translations
   will cause warnings.
 
 For more options, take a look at the [source](./index.js).
 
-Both functions return an object `{ headers, translations }` where `headers`
-contains the raw contents of the input file's headers, with keys lower-cased, and
-`translations` is an object containing the MessageFormat strings keyed by their
-`msgid` and if used, `msgctxt`.
+Both functions return an object containing the following fields:
+
+- `headers` (object) – The raw contents of the input file's headers, with keys
+  lower-cased
+- `pluralFunction` (function) – An appropriate pluralisation function to use for
+  the output translations, suitable to be used directly by [messageformat]. May
+  be `null` if none was set in `options` and if the input did not include a
+  Plural-Forms header.
+- `translations` (object) – An object containing the MessageFormat strings keyed
+  by their `msgid` and if used, `msgctxt`.
 
 [messageformat]: https://messageformat.github.io/
 [gettext-parser]: https://github.com/smhg/gettext-parser

index.js

@@ -1,10 +1,10 @@
 const { mo, po } = require('gettext-parser')
-const pluralCategories = require('make-plural/umd/pluralCategories')
+const getPluralFunction = require('./plural-forms')
 
 const defaultOptions = {
   defaultCharset: null,
   forceContext: false,
-  pluralCategories: null,
+  pluralFunction: null,
   pluralVariablePattern: /%(?:\((\w+)\))?\w/,
   replacements: [
     {
@@ -32,22 +32,8 @@ const defaultOptions = {
   verbose: false
 }
 
-const getPluralCategories = ({ language, 'plural-forms': pluralForms }) => {
-  if (language) {
-    const pc = pluralCategories[language.replace(/[-_].*/, '')]
-    if (pc) return pc.cardinal
-  }
-  const m = pluralForms && pluralForms.match(/^nplurals=(\d);/)
-  switch (m && m[1]) {
-    case '1': return ['other']
-    case '2': return ['one', 'other']
-    case '6': return ['zero', 'one', 'two', 'few', 'many', 'other']
-    default: return null
-  }
-}
-
 const getMessageFormat = (
-  { pluralCategories, pluralVariablePattern, replacements, verbose },
+  { pluralFunction, pluralVariablePattern, replacements, verbose },
   { msgid, msgid_plural, msgstr }
 ) => {
   if (!msgid || !msgstr) return null
@@ -56,8 +42,8 @@ const getMessageFormat = (
     msgstr[0] = msgid
   }
   if (msgid_plural) {
-    if (!pluralCategories) throw new Error('Plural categories not identified')
-    for (let i = 1; i < pluralCategories.length; ++i) {
+    if (!pluralFunction) throw new Error('Plural-Forms not defined')
+    for (let i = 1; i < pluralFunction.cardinal.length; ++i) {
       if (!msgstr[i]) {
         if (verbose) console.warn('Plural translation not found:', msgid, i)
         msgstr[i] = msgid_plural
@@ -73,7 +59,7 @@ const getMessageFormat = (
   if (msgid_plural) {
     const m = msgid_plural.match(pluralVariablePattern)
     const pv = m && m[1] || '0'
-    const pc = pluralCategories.map((c, i) => `${c}{${msgstr[i]}}`)
+    const pc = pluralFunction.cardinal.map((c, i) => `${c}{${msgstr[i]}}`)
     return `{${pv}, plural, ${pc.join(' ')}}`
   }
   return msgstr[0]
@@ -82,7 +68,9 @@ const getMessageFormat = (
 const convert = (parse, input, options) => {
   options = Object.assign(defaultOptions, options)
   const { headers, translations } = parse(input, options.defaultCharset)
-  if (!options.pluralCategories) options.pluralCategories = getPluralCategories(headers)
+  if (!options.pluralFunction) {
+    options.pluralFunction = getPluralFunction(headers['plural-forms'])
+  }
   let hasContext = false
   for (const context in translations) {
     if (context) hasContext = true
@@ -95,6 +83,7 @@ const convert = (parse, input, options) => {
   }
   return {
     headers,
+    pluralFunction: options.pluralFunction,
     translations: hasContext || options.forceContext ? translations : translations['']
   }
 }

package.json

@@ -6,14 +6,14 @@
     "messageformat",
     "gettext",
     "po",
+    "plural-forms",
     "i18n"
   ],
   "main": "index.js",
   "repository": "https://github.com/eemeli/gettext-to-messageformat",
   "author": "Eemeli Aro <[email protected]>",
   "license": "MIT",
   "dependencies": {
-    "gettext-parser": "^1.3.0",
-    "make-plural": "^4.0.1"
+    "gettext-parser": "^1.3.0"
   }
 }

plural-forms.js

@@ -0,0 +1,49 @@
+/*
+Examples from https://www.gnu.org/software/gettext/manual/html_node/Plural-forms.html
+
+Note: true == 1, false == 0
+
+Plural-Forms: nplurals=1; plural=0;
+Plural-Forms: nplurals=2; plural=n != 1;
+Plural-Forms: nplurals=2; plural=n>1;
+Plural-Forms: nplurals=2; plural=n == 1 ? 0 : 1;
+Plural-Forms: nplurals=3; plural=n%10==1 && n%100!=11 ? 0 : n != 0 ? 1 : 2;
+Plural-Forms: nplurals=3; plural=n==1 ? 0 : n==2 ? 1 : 2;
+Plural-Forms: nplurals=3; plural=n==1 ? 0 : (n==0 || (n%100 > 0 && n%100 < 20)) ? 1 : 2;
+Plural-Forms: nplurals=3; plural=n%10==1 && n%100!=11 ? 0 : n%10>=2 && (n%100<10 || n%100>=20) ? 1 : 2;
+Plural-Forms: nplurals=4; plural=n%100==1 ? 0 : n%100==2 ? 1 : n%100==3 || n%100==4 ? 2 : 3;
+Plural-Forms: nplurals=6; plural=n==0 ? 0 : n==1 ? 1 : n==2 ? 2 : n%100>=3 && n%100<=10 ? 3 : n%100>=11 ? 4 : 5;
+*/
+
+const validPlural = (plural) => {
+  const words = plural.match(/\b\w+\b/g)
+  if (!words) return null
+  for (let i = 0; i < words.length; ++i) {
+    const word = words[i]
+    if (word !== 'n' && isNaN(Number(word))) return null
+  }
+  return plural.trim()
+}
+
+const getPluralFunction = (pluralForms) => {
+  if (!pluralForms) return null
+  let nplurals
+  let plural
+  pluralForms.split(';').forEach(part => {
+    const m = part.match(/^\s*(\w+)\s*=(.*)/)
+    switch (m && m[1]) {
+      case 'nplurals':
+        nplurals = Number(m[2])
+        break
+      case 'plural':
+        plural = validPlural(m[2])
+        break
+    }
+  })
+  if (!nplurals || !plural) throw new Error('Invalid plural-forms: ' + pluralForms)
+  const pluralFunc = new Function('n', `return 'p' + Number(${plural})`)
+  pluralFunc.cardinal = new Array(nplurals).fill().map((_, i) => 'p' + i)
+  return pluralFunc
+}
+
+module.exports = getPluralFunction