F95API/app/scripts/scraper.js

415 lines
13 KiB
JavaScript
Raw Normal View History

2020-10-31 15:00:26 +00:00
"use strict";
// Public modules from npm
const cheerio = require("cheerio");
2020-12-07 16:07:36 +00:00
const {DateTime} = require("luxon");
2020-10-31 15:00:26 +00:00
// Modules from file
const { fetchHTML, getUrlRedirect } = require("./network-helper.js");
const shared = require("./shared.js");
const GameInfo = require("./classes/game-info.js");
const f95Selector = require("./constants/css-selector.js");
/**
* @protected
* Get information from the game's main page.
* @param {String} url URL of the game/mod to extract data from
* @return {Promise<GameInfo>} Complete information about the game you are
2020-11-09 16:53:29 +00:00
* looking for or `null` if is impossible to parse information
2020-10-31 15:00:26 +00:00
*/
module.exports.getGameInfo = async function (url) {
shared.logger.info("Obtaining game info");
// Fetch HTML and prepare Cheerio
const html = await fetchHTML(url);
2020-12-12 11:39:16 +00:00
if(!html) return null;
2020-10-31 15:00:26 +00:00
const $ = cheerio.load(html);
const body = $("body");
const mainPost = $(f95Selector.GS_POSTS).first();
// Extract data
const titleData = extractInfoFromTitle(body);
const tags = extractTags(body);
const prefixesData = parseGamePrefixes(body);
const src = extractPreviewSource(body);
const changelog = extractChangelog(mainPost);
2020-10-31 15:00:26 +00:00
const structuredData = extractStructuredData(body);
2020-11-09 16:53:29 +00:00
// Sometimes the JSON-LD are not set, especially in low-profile game
if(!structuredData) return null;
const parsedInfos = parseMainPostText(structuredData.description);
const overview = getOverview(structuredData.description, prefixesData.mod);
2020-10-31 15:00:26 +00:00
// Obtain the updated URL
const redirectUrl = await getUrlRedirect(url);
// Fill in the GameInfo element with the information obtained
const info = new GameInfo();
2020-11-07 17:16:20 +00:00
info.id = extractIDFromURL(url);
2020-10-31 15:00:26 +00:00
info.name = titleData.name;
info.author = titleData.author;
info.isMod = prefixesData.mod;
info.engine = prefixesData.engine;
info.status = prefixesData.status;
2020-10-31 15:00:26 +00:00
info.tags = tags;
info.url = redirectUrl;
info.language = parsedInfos.Language;
2020-10-31 15:00:26 +00:00
info.overview = overview;
info.supportedOS = parsedInfos.SupportedOS;
info.censored = parsedInfos.Censored;
info.lastUpdate = parsedInfos.LastUpdate;
info.previewSrc = src;
info.changelog = changelog;
2020-10-31 15:00:26 +00:00
info.version = titleData.version;
shared.logger.info(`Founded data for ${info.name}`);
return info;
};
//#region Private methods
/**
* @private
* Parse the game prefixes obtaining the engine used,
* the advancement status and if the game is actually a game or a mod.
* @param {cheerio.Cheerio} body Page `body` selector
2020-11-02 09:01:39 +00:00
* @returns {Object.<string, object>} Dictionary of values with keys `engine`, `status`, `mod`
*/
function parseGamePrefixes(body) {
shared.logger.trace("Parsing prefixes...");
// Local variables
let mod = false,
engine = null,
status = null;
// Obtain the title prefixes
const prefixeElements = body.find(f95Selector.GT_TITLE_PREFIXES);
prefixeElements.each(function parseGamePrefix(idx, el) {
// Obtain the prefix text
let prefix = cheerio.load(el).text().trim();
// Remove the square brackets
prefix = prefix.replace("[", "").replace("]", "");
// Check what the prefix indicates
if (isEngine(prefix)) engine = prefix;
else if (isStatus(prefix)) status = prefix;
else if (isMod(prefix)) mod = true;
});
// If the status is not set, then the game in in development (Ongoing)
status = !status ? "Ongoing" : status; // status ?? "Ongoing";
return {
engine,
status,
mod
};
}
/**
* @private
* Extracts all the possible informations from the title.
2020-10-31 15:00:26 +00:00
* @param {cheerio.Cheerio} body Page `body` selector
2020-11-02 09:01:39 +00:00
* @returns {Object.<string, string>} Dictionary of values with keys `name`, `author`, `version`
2020-10-31 15:00:26 +00:00
*/
function extractInfoFromTitle(body) {
shared.logger.trace("Extracting information from title...");
2020-10-31 15:00:26 +00:00
const title = body
.find(f95Selector.GT_TITLE)
.text()
.trim();
// From the title we can extract: Name, author and version
// [PREFIXES] TITLE [VERSION] [AUTHOR]
const matches = title.match(/\[(.*?)\]/g);
2020-10-31 15:00:26 +00:00
// Get the title name
let name = title;
matches.forEach(function replaceElementsInTitle(e) {
name = name.replace(e, "");
2020-10-31 15:00:26 +00:00
});
name = name.trim();
// The version is the penultimate element.
// If the matches are less than 2, than the title
// is malformes and only the author is fetched
// (usually the author is always present)
let version = null;
2020-11-02 16:26:06 +00:00
if (matches.length >= 2) {
2020-11-02 17:47:24 +00:00
// The regex [[\]]+ remove the square brackets
2020-11-02 16:26:06 +00:00
version = matches[matches.length - 2].replace(/[[\]]+/g, "").trim();
// Remove the trailing "v"
if (version[0] === "v") version = version.replace("v", "");
2020-11-02 16:26:06 +00:00
}
2020-11-02 17:47:24 +00:00
// Last element (the regex [[\]]+ remove the square brackets)
const author = matches[matches.length - 1].replace(/[[\]]+/g, "").trim();
2020-10-31 15:00:26 +00:00
return {
name,
version,
author,
};
}
/**
* @private
* Gets the tags used to classify the game.
* @param {cheerio.Cheerio} body Page `body` selector
* @returns {String[]} List of tags
*/
function extractTags(body) {
shared.logger.trace("Extracting tags...");
2020-10-31 15:00:26 +00:00
// Get the game tags
const tagResults = body.find(f95Selector.GT_TAGS);
return tagResults.map(function parseGameTags(idx, el) {
return cheerio.load(el).text().trim();
2020-10-31 15:00:26 +00:00
}).get();
}
/**
* @private
* Gets the URL of the image used as a preview.
* @param {cheerio.Cheerio} body Page `body` selector
* @returns {String} URL of the image
2020-10-31 15:00:26 +00:00
*/
function extractPreviewSource(body) {
shared.logger.trace("Extracting image preview source...");
const image = body.find(f95Selector.GT_IMAGES);
2020-10-31 15:00:26 +00:00
// The "src" attribute is rendered only in a second moment,
// we need the "static" src value saved in the attribute "data-src"
const source = image ? image.attr("data-src") : null;
return source;
}
/**
* @private
* Gets the changelog of the latest version.
* @param {cheerio.Cheerio} mainPost main post selector
* @returns {String} Changelog of the last version or `null` if no changelog is fetched
*/
function extractChangelog(mainPost) {
shared.logger.trace("Extracting last changelog...");
2020-11-20 09:23:06 +00:00
// Obtain the changelog for ALL the versions
let changelog = mainPost.find(f95Selector.GT_LAST_CHANGELOG).text().trim();
2020-11-20 09:23:06 +00:00
// Parse the latest changelog
const endChangelog = changelog.indexOf("\nv"); // \n followed by version (v)
if (endChangelog !== -1) changelog = changelog.substring(0, endChangelog + 1);
// Clean changelog
changelog = changelog.replace("Spoiler", "");
2020-11-20 09:23:06 +00:00
changelog = changelog.replace(/\n+/g, "\n"); // Multiple /n
changelog = changelog.trim();
2020-11-19 19:45:05 +00:00
// Delete the version at the start of the changelog
const firstNewLine = changelog.indexOf("\n");
const supposedVersion = changelog.substring(0, firstNewLine);
if (supposedVersion[0] === "v") changelog = changelog.substring(firstNewLine).trim();
// Return changelog
return changelog ? changelog : null;
2020-10-31 15:00:26 +00:00
}
/**
* @private
* Process the main post text to get all the useful
* information in the format *DESCRIPTOR : VALUE*.
2020-11-02 09:01:39 +00:00
* Gets "standard" values such as: `Language`, `SupportedOS`, `Censored`, and `LastUpdate`.
* All non-canonical values are instead grouped together as a dictionary with the key `Various`.
2020-10-31 15:00:26 +00:00
* @param {String} text Structured text of the post
2020-11-02 09:01:39 +00:00
* @returns {Object.<string, object>} Dictionary of information
2020-10-31 15:00:26 +00:00
*/
function parseMainPostText(text) {
shared.logger.trace("Parsing main post raw text...");
const data = {};
2020-10-31 15:00:26 +00:00
// The information searched in the game post are one per line
const splittedText = text.split("\n");
for (const line of splittedText) {
if (!line.includes(":")) continue;
// Create pair key/value
const splitted = line.split(":");
const key = splitted[0].trim().toUpperCase().replace(/ /g, "_"); // Uppercase to avoid mismatch
const value = splitted[1].trim();
// Add pair to the dict if valid
if (value !== "") data[key] = value;
}
// Parse the standard pairs
const parsedDict = {};
// Check if the game is censored
if (data.CENSORED) {
const censored = data.CENSORED.toUpperCase() === "NO" ? false : true;
parsedDict["Censored"] = censored;
delete data.CENSORED;
2020-10-31 15:00:26 +00:00
}
// Last update of the main post
2020-12-07 16:07:36 +00:00
if (data.UPDATED && DateTime.fromISO(data.UPDATED).isValid) {
parsedDict["LastUpdate"] = new Date(data.UPDATED);
delete data.UPDATED;
}
2020-12-07 16:07:36 +00:00
else if (data.THREAD_UPDATED && DateTime.fromISO(data.THREAD_UPDATED).isValid) {
parsedDict["LastUpdate"] = new Date(data.THREAD_UPDATED);
delete data.THREAD_UPDATED;
}
else parsedDict["LastUpdate"] = null;
// Parse the supported OS
if (data.OS) {
const listOS = [];
// Usually the string is something like "Windows, Linux, Mac"
const splitted = data.OS.split(",");
splitted.forEach(function (os) {
listOS.push(os.trim());
});
parsedDict["SupportedOS"] = listOS;
delete data.OS;
}
// Rename the key for the language
if (data.LANGUAGE) {
parsedDict["Language"] = data.LANGUAGE;
delete data.LANGUAGE;
}
// What remains is added to a sub dictionary
parsedDict["Various"] = data;
return parsedDict;
2020-10-31 15:00:26 +00:00
}
2020-12-12 11:39:16 +00:00
/**
* @private
* Parse a JSON-LD element.
* @param {cheerio.Element} element
*/
function parseScriptTag(element) {
// Get the element HTML
const html = cheerio.load(element).html().trim();
// Obtain the JSON-LD
const data = html
.replace("<script type=\"application/ld+json\">", "")
.replace("</script>", "");
// Convert the string to an object
const json = JSON.parse(data);
// Return only the data of the game
if (json["@type"] === "Book") return json;
}
2020-10-31 15:00:26 +00:00
/**
* @private
* Extracts and processes the JSON-LD values found at the bottom of the page.
* @param {cheerio.Cheerio} body Page `body` selector
2020-11-02 09:01:39 +00:00
* @returns {Object.<string, string>} JSON-LD or `null` if no valid JSON is found
2020-10-31 15:00:26 +00:00
*/
function extractStructuredData(body) {
shared.logger.trace("Extracting JSON-LD data...");
2020-12-12 11:39:16 +00:00
// Fetch the JSON-LD data
const structuredDataElements = body.find(f95Selector.GT_JSONLD);
2020-12-12 11:39:16 +00:00
// Parse the data
const json = structuredDataElements.map((idx, el) => parseScriptTag(el)).get();
2020-11-09 16:53:29 +00:00
return json.lenght !== 0 ? json[0] : null;
2020-10-31 15:00:26 +00:00
}
/**
* @private
* Get the game description from its web page.
* Different processing depending on whether the game is a mod or not.
* @param {String} text Structured text extracted from the game's web page
* @param {Boolean} mod Specify if it is a game or a mod
2020-11-02 09:21:36 +00:00
* @returns {String} Game description
2020-10-31 15:00:26 +00:00
*/
function getOverview(text, mod) {
shared.logger.trace("Extracting game overview...");
2020-11-02 09:21:36 +00:00
2020-10-31 15:00:26 +00:00
// Get overview (different parsing for game and mod)
const overviewEndIndex = mod ? text.indexOf("Updated") : text.indexOf("Thread Updated");
return text.substring(0, overviewEndIndex).replace("Overview:\n", "").trim();
}
/**
* @private
* Check if the prefix is a game's engine.
* @param {String} prefix Prefix to check
* @return {Boolean}
*/
function isEngine(prefix) {
const engines = toUpperCaseArray(Object.values(shared.engines));
2020-10-31 15:00:26 +00:00
return engines.includes(prefix.toUpperCase());
}
/**
* @private
* Check if the prefix is a game's status.
* @param {String} prefix Prefix to check
* @return {Boolean}
*/
function isStatus(prefix) {
const statuses = toUpperCaseArray(Object.values(shared.statuses));
2020-10-31 15:00:26 +00:00
return statuses.includes(prefix.toUpperCase());
}
/**
* @private
* Check if the prefix indicates a mod.
* @param {String} prefix Prefix to check
* @return {Boolean}
*/
function isMod(prefix) {
const modPrefixes = ["MOD", "CHEAT MOD"];
return modPrefixes.includes(prefix.toUpperCase());
}
2020-11-07 17:16:20 +00:00
/**
* @private
* Extracts the game's unique ID from the game's URL.
* @param {String} url Game's URL
* @return {Number} Game's ID
*/
function extractIDFromURL(url) {
// URL are in the format https://f95zone.to/threads/GAMENAME-VERSION-DEVELOPER.ID/
2020-11-30 12:22:35 +00:00
// or https://f95zone.to/threads/ID/
2020-12-03 09:11:48 +00:00
const match = url.match(/([0-9]+)(?=\/|\b)(?!-|\.)/);
2020-11-30 12:22:35 +00:00
if(!match) return -1;
2020-11-07 17:16:20 +00:00
// Parse and return number
2020-11-30 12:22:35 +00:00
return parseInt(match[0], 10);
2020-11-07 17:16:20 +00:00
}
2020-10-31 15:00:26 +00:00
/**
* @private
* Makes an array of strings uppercase.
* @param {String[]} a
*/
function toUpperCaseArray(a) {
/**
* Makes a string uppercase.
* @param {String} s
* @returns {String}
*/
function toUpper(s) {
return s.toUpperCase();
}
return a.map(toUpper);
}
//#endregion Private methods