discord-bot-rss-feed/app/models/feed-data.js

117 lines
3.7 KiB
JavaScript

const DiscordUtil = require("../../discord-bot-core").util;
const Camo = require("camo");
const Config = require("../config.json");
const Dns = require("dns"); //for host resolution checking
const Url = require("url"); //for url parsing
const { promisify } = require("util");
const FeedReadPromise = promisify(require("feed-read")); //for extracing new links from RSS feeds
const DnsResolvePromise = promisify(Dns.resolve);
const GetUrls = require("get-urls"); //for extracting urls from messages
module.exports = class FeedData extends Camo.EmbeddedDocument {
constructor() {
super();
this.feedID = String;
this.url = String;
this.channelID = String;
this.roleID = String;
this.cachedLinks = [String];
this.maxCacheSize = Number;
}
cache(...elements) {
Array.prototype.push.apply(
this.cachedLinks,
elements
.map(el => normaliseUrl(el))
.filter(el => !this.cachedLinks.includes(el))
);
//seeing as new links come in at the end of the array, we need to remove the old links from the beginning
this.cachedLinks.splice(0, this.cachedLinks.length - this.maxCacheSize);
}
updatePastPostedLinks(guild) {
const channel = guild.channels.get(this.channelID);
if (!channel)
return Promise.reject("Channel not found!");
return new Promise((resolve, reject) => {
channel.fetchMessages({ limit: 100 })
.then(messages => {
/* we want to push the links in oldest first, but discord.js returns messages newest first, so we need to reverse them
* discord.js returns a map, and maps don't have .reverse methods, hence needing to spread the elements into an array first */
[...messages.values()].reverse().forEach(m => this.cache(...GetUrls(m.content)));
resolve();
})
.catch(reject);
});
}
fetchLatest(guild) {
const dnsPromise = DnsResolvePromise(Url.parse(this.url).host).then(() => this._doFetchRSS(guild));
dnsPromise.catch(err => DiscordUtil.dateError("Connection error: Can't resolve host", err.message || err));
return dnsPromise;
}
toString() {
const blacklist = ["cachedLinks", "maxCacheSize"];
return `\`\`\`JavaScript\n ${JSON.stringify(this, (k, v) => !blacklist.find(x => x === k) ? v : undefined, "\t")} \`\`\``;
}
_doFetchRSS(guild) {
const feedPromise = FeedReadPromise(this.url).then(articles => this._processLatestArticle(guild, articles));
feedPromise.catch(err => DiscordUtil.dateDebugError([`Error reading feed ${this.url}`, err]));
return feedPromise;
}
_processLatestArticle(guild, articles) {
if (articles.length === 0 || !articles[0].link)
return;
const latest = normaliseUrl(articles[0].link);
if (this.cachedLinks.includes(latest))
return;
this.cache(latest);
const channel = guild.channels.get(this.channelID),
role = guild.roles.get(this.roleID);
channel.send((role || "") + formatPost(articles[0]))
.catch(err => DiscordUtil.dateDebugError(`Error posting in ${channel.id}: ${err.message || err}`));
}
};
function formatPost(article) {
let message = "";
if (article.title)
message += `\n**${article.title}**`;
if (article.content)
message += article.content.length > Config.charLimit ? "\nArticle content too long for a single Discord message!" : `\n${article.content}`;
if (article.link)
message += `\n\n${normaliseUrl(article.link)}`;
return message;
}
function normaliseUrl(url) {
url = url.replace("https://", "http://"); //hacky way to treat http and https the same
const parsedUrl = Url.parse(url);
if (parsedUrl.host && parsedUrl.host.includes("youtube.com")) {
const videoIDParam = (parsedUrl.query || "").split("&").find(x => x.startsWith("v="));
if (videoIDParam) {
const videoID = videoIDParam.substring(videoIDParam.indexOf("=") + 1, videoIDParam.length);
url = "http://youtu.be/" + videoID;
}
}
return url;
}