bandcamp-fetch
Version:
Scrape Bandcamp content
190 lines • 8.06 kB
JavaScript
;
var __classPrivateFieldGet = (this && this.__classPrivateFieldGet) || function (receiver, state, kind, f) {
if (kind === "a" && !f) throw new TypeError("Private accessor was defined without a getter");
if (typeof state === "function" ? receiver !== state || !f : !state.has(receiver)) throw new TypeError("Cannot read private member from an object whose class did not declare it");
return kind === "m" ? f : kind === "a" ? f.call(receiver) : f ? f.value : state.get(receiver);
};
var _a, _SearchResultsParser_matchSearchItemTypeWithResult;
Object.defineProperty(exports, "__esModule", { value: true });
const cheerio_1 = require("cheerio");
const Parse_js_1 = require("../utils/Parse.js");
const SearchAPI_js_1 = require("./SearchAPI.js");
const VALID_RESULT_TYPES = ['artist', 'label', 'album', 'track', 'fan'];
class SearchResultsParser {
static parseResults(html, opts) {
const $ = (0, cheerio_1.load)(html);
const resultsList = $('li.searchresult');
const items = [];
resultsList.each((i, resultListItem) => {
resultListItem = $(resultListItem);
const resultInfo = resultListItem.find('.result-info');
// Common info
const resultType = resultInfo
.children('.itemtype')
.text()
.trim()
.toLowerCase();
const imgSrc = $('.art img', resultListItem).attr('src');
const heading = $('.heading a', resultInfo);
const name = heading.text().trim();
const url = resultInfo.find('.itemurl').text().trim();
const imageUrl = (0, Parse_js_1.reformatImageUrl)(imgSrc, resultType === 'album' || resultType === 'track' ?
opts.albumImageFormat
: opts.artistImageFormat);
if (!__classPrivateFieldGet(this, _a, "m", _SearchResultsParser_matchSearchItemTypeWithResult).call(this, opts.itemType, resultType) ||
!name ||
!url) {
return true;
}
// Other info
let location, genre, tags, artist, numTracks, duration, releaseDate, album;
resultInfo
.find('.subhead, .genre, .tags, .released, .length')
.each((i, info) => {
info = $(info);
if (info.hasClass('subhead')) {
if (resultType === 'artist' || resultType === 'label') {
location = info.text().trim();
}
else if (resultType === 'album' || resultType === 'track') {
const infoText = info.text();
artist = (0, Parse_js_1.substrAfter)(infoText, 'by ')?.trim();
if (resultType === 'track') {
album = (0, Parse_js_1.substrBefore)(infoText, ' by');
if (album) {
album = (0, Parse_js_1.substrAfter)(album, 'from ')?.trim();
}
}
}
return true;
}
if (info.hasClass('genre')) {
genre = (0, Parse_js_1.substrAfter)(info.text(), 'genre: ')?.trim();
return true;
}
if (info.hasClass('tags')) {
tags = (0, Parse_js_1.substrAfter)(info.text(), 'tags:');
if (tags) {
tags = (0, Parse_js_1.stripLineBreaks)((0, Parse_js_1.stripMultipleWhitespaces)(tags)).trim();
}
return true;
}
if (info.hasClass('released')) {
releaseDate = (0, Parse_js_1.substrAfter)(info.text(), 'released ')?.trim();
return true;
}
if (info.hasClass('length')) {
const lengthParts = info.text().split(',');
const tracksText = lengthParts[0];
const minutesText = lengthParts[1];
const numTracksStr = tracksText ? (0, Parse_js_1.substrBefore)(tracksText, 'tracks') : null;
if (numTracksStr) {
numTracks = parseInt(numTracksStr, 10);
}
const durationStr = minutesText ? (0, Parse_js_1.substrBefore)(minutesText, 'minutes') : null;
if (durationStr) {
duration = parseInt(durationStr, 10) * 60;
}
}
});
const base = {
name: heading.text().trim(),
url: resultInfo.find('.itemurl').text().trim() || ''
};
if (imageUrl) {
base.imageUrl = imageUrl;
}
if (resultType === 'artist') {
const artist = {
type: 'artist',
...base
};
if (location)
artist.location = location;
if (genre)
artist.genre = genre;
if (tags)
artist.tags = tags;
items.push(artist);
}
else if (resultType === 'label') {
const label = {
type: 'label',
...base
};
if (location)
label.location = location;
items.push(label);
}
else if (resultType === 'album') {
const album = {
type: 'album',
...base
};
if (artist)
album.artist = artist;
if (numTracks)
album.numTracks = numTracks;
if (duration)
album.duration = duration;
if (releaseDate)
album.releaseDate = releaseDate;
if (tags)
album.tags = tags;
items.push(album);
}
else if (resultType === 'track') {
const track = {
type: 'track',
...base
};
if (artist)
track.artist = artist;
if (album)
track.album = album;
if (releaseDate)
track.releaseDate = releaseDate;
if (tags)
track.tags = tags;
items.push(track);
}
else if (resultType === 'fan') {
const fan = {
type: 'fan',
...base
};
if (genre)
fan.genre = genre;
items.push(fan);
}
});
let totalPages = parseInt($('.pagelist').find('.pagenum').last().text(), 10);
if (isNaN(totalPages)) {
totalPages = 1;
}
return {
items,
totalPages
};
}
}
_a = SearchResultsParser, _SearchResultsParser_matchSearchItemTypeWithResult = function _SearchResultsParser_matchSearchItemTypeWithResult(searchType, resultType) {
if (searchType === SearchAPI_js_1.SearchItemType.All &&
VALID_RESULT_TYPES.includes(resultType)) {
return true;
}
switch (searchType) {
case SearchAPI_js_1.SearchItemType.ArtistsAndLabels:
return resultType === 'artist' || resultType === 'label';
case SearchAPI_js_1.SearchItemType.Albums:
return resultType === 'album';
case SearchAPI_js_1.SearchItemType.Tracks:
return resultType === 'track';
case SearchAPI_js_1.SearchItemType.Fans:
return resultType === 'fan';
default:
return false;
}
};
exports.default = SearchResultsParser;
//# sourceMappingURL=SearchResultsParser.js.map