gecko-dev/intl/l10n/Localization.jsm
Zibi Braniecki a715714c3c Bug 1438613 - Remove explicit observer unregistering from Fluent. r=mossop
MozReview-Commit-ID: 3SavwzU9ZQD

--HG--
extra : rebase_source : 97a5a8d661d93ac3a35635bab527990b47d3b35c
2018-02-15 11:06:52 -08:00

401 lines
12 KiB
JavaScript

/* vim: set ts=2 et sw=2 tw=80 filetype=javascript: */
/* Copyright 2017 Mozilla Foundation and others
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
/* fluent@0.6.3 */
/* eslint no-console: ["error", { allow: ["warn", "error"] }] */
/* global console */
const { XPCOMUtils } = ChromeUtils.import("resource://gre/modules/XPCOMUtils.jsm", {});
const { L10nRegistry } = ChromeUtils.import("resource://gre/modules/L10nRegistry.jsm", {});
const { Services } = ChromeUtils.import("resource://gre/modules/Services.jsm", {});
const { AppConstants } = ChromeUtils.import("resource://gre/modules/AppConstants.jsm", {});
/*
* CachedIterable caches the elements yielded by an iterable.
*
* It can be used to iterate over an iterable many times without depleting the
* iterable.
*/
class CachedIterable {
/**
* Create an `CachedIterable` instance.
*
* @param {Iterable} iterable
* @returns {CachedIterable}
*/
constructor(iterable) {
if (Symbol.asyncIterator in Object(iterable)) {
this.iterator = iterable[Symbol.asyncIterator]();
} else if (Symbol.iterator in Object(iterable)) {
this.iterator = iterable[Symbol.iterator]();
} else {
throw new TypeError("Argument must implement the iteration protocol.");
}
this.seen = [];
}
[Symbol.iterator]() {
const { seen, iterator } = this;
let cur = 0;
return {
next() {
if (seen.length <= cur) {
seen.push(iterator.next());
}
return seen[cur++];
}
};
}
[Symbol.asyncIterator]() {
const { seen, iterator } = this;
let cur = 0;
return {
async next() {
if (seen.length <= cur) {
seen.push(await iterator.next());
}
return seen[cur++];
}
};
}
/**
* This method allows user to consume the next element from the iterator
* into the cache.
*/
touchNext() {
const { seen, iterator } = this;
if (seen.length === 0 || seen[seen.length - 1].done === false) {
seen.push(iterator.next());
}
}
}
/**
* The default localization strategy for Gecko. It comabines locales
* available in L10nRegistry, with locales requested by the user to
* generate the iterator over MessageContexts.
*
* In the future, we may want to allow certain modules to override this
* with a different negotitation strategy to allow for the module to
* be localized into a different language - for example DevTools.
*/
function defaultGenerateMessages(resourceIds) {
const appLocales = Services.locale.getAppLocalesAsLangTags();
return L10nRegistry.generateContexts(appLocales, resourceIds);
}
/**
* The `Localization` class is a central high-level API for vanilla
* JavaScript use of Fluent.
* It combines language negotiation, MessageContext and I/O to
* provide a scriptable API to format translations.
*/
class Localization {
/**
* @param {Array<String>} resourceIds - List of resource IDs
* @param {Function} generateMessages - Function that returns a
* generator over MessageContexts
*
* @returns {Localization}
*/
constructor(resourceIds, generateMessages = defaultGenerateMessages) {
this.resourceIds = resourceIds;
this.generateMessages = generateMessages;
this.ctxs = new CachedIterable(this.generateMessages(this.resourceIds));
}
/**
* Format translations and handle fallback if needed.
*
* Format translations for `keys` from `MessageContext` instances on this
* DOMLocalization. In case of errors, fetch the next context in the
* fallback chain.
*
* @param {Array<Array>} keys - Translation keys to format.
* @param {Function} method - Formatting function.
* @returns {Promise<Array<string|Object>>}
* @private
*/
async formatWithFallback(keys, method) {
const translations = [];
for await (let ctx of this.ctxs) {
// This can operate on synchronous and asynchronous
// contexts coming from the iterator.
if (typeof ctx.then === "function") {
ctx = await ctx;
}
const missingIds = keysFromContext(method, ctx, keys, translations);
if (missingIds.size === 0) {
break;
}
if (AppConstants.NIGHTLY_BUILD) {
const locale = ctx.locales[0];
const ids = Array.from(missingIds).join(", ");
console.warn(`Missing translations in ${locale}: ${ids}`);
}
}
return translations;
}
/**
* Format translations into {value, attrs} objects.
*
* The fallback logic is the same as in `formatValues` but the argument type
* is stricter (an array of arrays) and it returns {value, attrs} objects
* which are suitable for the translation of DOM elements.
*
* docL10n.formatMessages([
* ['hello', { who: 'Mary' }],
* ['welcome', undefined]
* ]).then(console.log);
*
* // [
* // { value: 'Hello, Mary!', attrs: null },
* // { value: 'Welcome!', attrs: { title: 'Hello' } }
* // ]
*
* Returns a Promise resolving to an array of the translation strings.
*
* @param {Array<Array>} keys
* @returns {Promise<Array<{value: string, attrs: Object}>>}
* @private
*/
formatMessages(keys) {
return this.formatWithFallback(keys, messageFromContext);
}
/**
* Retrieve translations corresponding to the passed keys.
*
* A generalized version of `DOMLocalization.formatValue`. Keys can
* either be simple string identifiers or `[id, args]` arrays.
*
* docL10n.formatValues([
* ['hello', { who: 'Mary' }],
* ['hello', { who: 'John' }],
* ['welcome']
* ]).then(console.log);
*
* // ['Hello, Mary!', 'Hello, John!', 'Welcome!']
*
* Returns a Promise resolving to an array of the translation strings.
*
* @param {Array<Array>} keys
* @returns {Promise<Array<string>>}
*/
formatValues(keys) {
return this.formatWithFallback(keys, valueFromContext);
}
/**
* Retrieve the translation corresponding to the `id` identifier.
*
* If passed, `args` is a simple hash object with a list of variables that
* will be interpolated in the value of the translation.
*
* docL10n.formatValue(
* 'hello', { who: 'world' }
* ).then(console.log);
*
* // 'Hello, world!'
*
* Returns a Promise resolving to the translation string.
*
* Use this sparingly for one-off messages which don't need to be
* retranslated when the user changes their language preferences, e.g. in
* notifications.
*
* @param {string} id - Identifier of the translation to format
* @param {Object} [args] - Optional external arguments
* @returns {Promise<string>}
*/
async formatValue(id, args) {
const [val] = await this.formatValues([[id, args]]);
return val;
}
/**
* Register weak observers on events that will trigger cache invalidation
*/
registerObservers() {
Services.obs.addObserver(this, "intl:app-locales-changed", true);
}
/**
* Default observer handler method.
*
* @param {String} subject
* @param {String} topic
* @param {Object} data
*/
observe(subject, topic, data) {
switch (topic) {
case "intl:app-locales-changed":
this.onLanguageChange();
break;
default:
break;
}
}
/**
* This method should be called when there's a reason to believe
* that language negotiation or available resources changed.
*/
onLanguageChange() {
this.ctxs = new CachedIterable(this.generateMessages(this.resourceIds));
}
}
Localization.prototype.QueryInterface = XPCOMUtils.generateQI([
Ci.nsISupportsWeakReference
]);
/**
* Format the value of a message into a string.
*
* This function is passed as a method to `keysFromContext` and resolve
* a value of a single L10n Entity using provided `MessageContext`.
*
* If the function fails to retrieve the entity, it will return an ID of it.
* If formatting fails, it will return a partially resolved entity.
*
* In both cases, an error is being added to the errors array.
*
* @param {MessageContext} ctx
* @param {Array<Error>} errors
* @param {string} id
* @param {Object} args
* @returns {string}
* @private
*/
function valueFromContext(ctx, errors, id, args) {
const msg = ctx.getMessage(id);
return ctx.format(msg, args, errors);
}
/**
* Format all public values of a message into a { value, attrs } object.
*
* This function is passed as a method to `keysFromContext` and resolve
* a single L10n Entity using provided `MessageContext`.
*
* The function will return an object with a value and attributes of the
* entity.
*
* If the function fails to retrieve the entity, the value is set to the ID of
* an entity, and attrs to `null`. If formatting fails, it will return
* a partially resolved value and attributes.
*
* In both cases, an error is being added to the errors array.
*
* @param {MessageContext} ctx
* @param {Array<Error>} errors
* @param {String} id
* @param {Object} args
* @returns {Object}
* @private
*/
function messageFromContext(ctx, errors, id, args) {
const msg = ctx.getMessage(id);
const formatted = {
value: ctx.format(msg, args, errors),
attrs: null,
};
if (msg.attrs) {
formatted.attrs = [];
for (const name in msg.attrs) {
const value = ctx.format(msg.attrs[name], args, errors);
if (value !== null) {
formatted.attrs.push({ name, value });
}
}
}
return formatted;
}
/**
* This function is an inner function for `Localization.formatWithFallback`.
*
* It takes a `MessageContext`, list of l10n-ids and a method to be used for
* key resolution (either `valueFromContext` or `messageFromContext`) and
* optionally a value returned from `keysFromContext` executed against
* another `MessageContext`.
*
* The idea here is that if the previous `MessageContext` did not resolve
* all keys, we're calling this function with the next context to resolve
* the remaining ones.
*
* In the function, we loop over `keys` and check if we have the `prev`
* passed and if it has an error entry for the position we're in.
*
* If it doesn't, it means that we have a good translation for this key and
* we return it. If it does, we'll try to resolve the key using the passed
* `MessageContext`.
*
* In the end, we fill the translations array, and return the Set with
* missing ids.
*
* See `Localization.formatWithFallback` for more info on how this is used.
*
* @param {Function} method
* @param {MessageContext} ctx
* @param {Array<string>} keys
* @param {{Array<{value: string, attrs: Object}>}} translations
*
* @returns {Set<string>}
* @private
*/
function keysFromContext(method, ctx, keys, translations) {
const messageErrors = [];
const missingIds = new Set();
keys.forEach((key, i) => {
if (translations[i] !== undefined) {
return;
}
if (ctx.hasMessage(key[0])) {
messageErrors.length = 0;
translations[i] = method(ctx, messageErrors, key[0], key[1]);
// XXX: Report resolver errors
} else {
missingIds.add(key[0]);
}
});
return missingIds;
}
this.Localization = Localization;
var EXPORTED_SYMBOLS = ["Localization"];