gphotos-scraper
Version:
A tool to extract public url and metadata from shared album
172 lines • 7.49 kB
JavaScript
;
var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
return new (P || (P = Promise))(function (resolve, reject) {
function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
step((generator = generator.apply(thisArg, _arguments || [])).next());
});
};
Object.defineProperty(exports, "__esModule", { value: true });
exports.getAlbum = exports.getData = void 0;
const exceptions_1 = require("../exceptions");
const axiosClient_1 = require("./axiosClient");
const RPCID = "snAcKc";
const BASE_URL = "https://photos.google.com/_/PhotosUi/data/batchexecute";
// Pre-compiled Regex for better performance
const RE_CANONICAL = /<link\s+rel="canonical"\s+href="([^"]+)"/;
const RE_KEY = /key=([^"&]+)/;
const getData = (id, key, token) => __awaiter(void 0, void 0, void 0, function* () {
var _a;
const queryData = [[RPCID, JSON.stringify([id, token, null, key])]];
const params = new URLSearchParams();
params.append("f.req", JSON.stringify([queryData]));
let data;
try {
const res = yield axiosClient_1.apiClient.post(BASE_URL, params, {
headers: {
"Content-Type": "application/x-www-form-urlencoded;charset=UTF-8",
},
});
data = res.data.replace(/^\)\]\}'/, ""); // Faster regex replacement
// DEBUG: Log raw RPC response
const fs = require('fs');
fs.writeFileSync('tmp/debug_rpc_album.txt', data);
}
catch (error) {
console.error((error === null || error === void 0 ? void 0 : error.message) || error);
throw error;
}
const list = JSON.parse(data);
const reponseData = (_a = list.find((item) => (item === null || item === void 0 ? void 0 : item[1]) === RPCID)) === null || _a === void 0 ? void 0 : _a[2];
if (!reponseData)
return {};
const parsedResponse = JSON.parse(reponseData);
const albumItems = (parsedResponse === null || parsedResponse === void 0 ? void 0 : parsedResponse[1]) || [];
const next = parsedResponse === null || parsedResponse === void 0 ? void 0 : parsedResponse[2];
const albumDetails = (parsedResponse === null || parsedResponse === void 0 ? void 0 : parsedResponse[3]) || [];
const [albumId, albumTitle, albumDates, downloadUrl, albumCover, albumOwner] = albumDetails;
const sharedUrl = albumDetails === null || albumDetails === void 0 ? void 0 : albumDetails[32];
const createdAt = albumDates === null || albumDates === void 0 ? void 0 : albumDates[4];
const updatedAt = albumDates === null || albumDates === void 0 ? void 0 : albumDates[8];
const coverUrl = albumCover === null || albumCover === void 0 ? void 0 : albumCover[0];
const coverWidth = albumCover === null || albumCover === void 0 ? void 0 : albumCover[1];
const coverHeight = albumCover === null || albumCover === void 0 ? void 0 : albumCover[2];
const ownerId = albumOwner === null || albumOwner === void 0 ? void 0 : albumOwner[1];
const ownerDetails = (albumOwner === null || albumOwner === void 0 ? void 0 : albumOwner[11]) || [];
const [ownerFullname, , ownerGenre, ownerName] = ownerDetails;
const ownerPhoto = albumOwner === null || albumOwner === void 0 ? void 0 : albumOwner[3];
const items = albumItems.map((item) => {
const [id, itemData] = item;
const [url] = itemData;
return { id, url };
});
return {
id: albumId,
key,
token: next,
title: albumTitle,
url: sharedUrl,
downloadUrl,
createdAt,
updatedAt,
cover: {
url: coverUrl,
width: coverWidth,
height: coverHeight,
},
owner: {
id: ownerId,
fullname: ownerFullname,
name: ownerName,
genre: ownerGenre,
photo: ownerPhoto,
},
items,
};
});
exports.getData = getData;
const getAlbum = (url) => __awaiter(void 0, void 0, void 0, function* () {
var _a;
let data = null;
try {
const res = yield axiosClient_1.apiClient.get(url);
data = res.data;
}
catch (error) {
if (((_a = error === null || error === void 0 ? void 0 : error.response) === null || _a === void 0 ? void 0 : _a.status) === 404 || (error === null || error === void 0 ? void 0 : error.code) === "ERR_BAD_REQUEST") {
throw new exceptions_1.AlbumNotFoundException();
}
throw error;
}
// Extract Metadata using Regex (No Cheerio)
const canonicalMatch = data.match(RE_CANONICAL);
const canonicalUrl = canonicalMatch ? canonicalMatch[1] : "";
if (!canonicalUrl)
throw new exceptions_1.AlbumNotFoundException();
const albumUrlObj = new URL(canonicalUrl);
const pathParts = albumUrlObj.pathname.split("/");
const id = pathParts[pathParts.length - 1] || pathParts[pathParts.length - 2]; // Handle trailing slash
// Key might be in the original URL or need to be extracted from response if redirected
// But usually it's in the query param of the canonical or initial URL.
// We'll trust the canonical for ID, but for key we check the URL params of the Input URL or the Canonical.
let key = albumUrlObj.searchParams.get("key");
if (!key) {
// Fallback: try to find key in the HTML if not in URL
// (The original code assumed it was in the canonical URL params)
// If the original logic worked with `albumUrl.searchParams.get('key')`, we stick to it.
// But let's check the input `url` too just in case.
try {
const inputUrlObj = new URL(url);
key = inputUrlObj.searchParams.get("key");
}
catch (_b) { }
}
// Last resort regex if key is still missing (sometimes encoded in JS vars)
if (!key) {
const keyMatch = data.match(RE_KEY);
if (keyMatch)
key = keyMatch[1];
}
if (!id || !key)
throw new exceptions_1.AlbumNotFoundException();
let album;
let token;
let pageCount = 0;
const MAX_PAGES = 100; // Safety limit
do {
// Safety break
if (pageCount++ > MAX_PAGES)
break;
const parts = yield (0, exports.getData)(id, key, token);
if (!album) {
// set once on first page
if (!parts.id) {
// If getData failed to parse on first try
throw new exceptions_1.AlbumNotFoundException();
}
const { title, owner, cover, downloadUrl, createdAt, updatedAt } = parts;
album = {
key,
id,
url,
title,
token,
owner,
cover,
downloadUrl,
createdAt,
updatedAt,
items: [],
};
}
token = parts.token;
if (parts.items) {
album.items.push(...parts.items);
}
} while (token);
return album;
});
exports.getAlbum = getAlbum;
//# sourceMappingURL=getAlbum.js.map