@hanivanrizky/nestjs-html-parser
Version:
⚠️ ARCHIVED - This package is no longer maintained. Please migrate to @hanivanrizky/nestjs-xpath-parser (https://github.com/Hanivan/nestjs-xpath-parser)
591 lines • 23.4 kB
JavaScript
;
Object.defineProperty(exports, "__esModule", { value: true });
exports.ThreadIdExtractorPipe = exports.RegexReplacePipe = exports.RegexExtractionPipe = exports.QueryRemoverPipe = exports.ParseAsUrlPipe = exports.NumNormalizePipe = exports.mappingTransform = exports.DateFormatPipe = void 0;
exports.scrapeBmwSgForum = scrapeBmwSgForum;
const __1 = require("../");
/**
* BMW-SG Forum Scraper Demo
*
* This example demonstrates a full forum crawling process based on the real BMW-SG forum payload.
* It loads HTML from https://www.bmw-sg.com/forums/forums/introduction-greetings.24/page-5
* and applies extraction schemas matching the inspiration payload structure.
*/
const payload = {
data: {
endPoint: 'https://www.bmw-sg.com/forums/forums/introduction-greetings.24/page-5',
patternReply: [
{
key: 'container',
pattern: "//article[contains(@class,'message--post')]",
returnType: 'text',
meta: { isContainer: 'true' },
},
{ key: 'author', pattern: './@data-author', returnType: 'text' },
{
key: 'replyDate',
pattern: './/time[@datetime]/@datetime',
returnType: 'text',
pipe: [
{
locale: 'en',
type: 'date-format',
format: 'YYYY-MM-DDTHH:mm:ssZZ',
timezone: 'asia/singapore',
},
],
},
{
key: 'replyId',
pattern: './@data-content',
returnType: 'text',
pipe: [
{
type: 'regex-replace',
regex: '^post-',
textReplacement: '',
flag: 'g',
},
],
},
{
key: 'replyHtml',
pattern: ".//article[contains(@class,'message-body')]",
returnType: 'html',
},
{
key: 'replyText',
pattern: ".//article[contains(@class,'message-body')]/descendant-or-self::node()/text()[normalize-space()]",
returnType: 'text',
meta: { multiple: 'true' },
},
{
key: 'nextPage',
pattern: "//ul[contains(@class,'pageNav-main')]/li/a[@href and not(@href='#')]",
returnType: 'text',
meta: { isPage: 'true' },
},
{
key: 'urlReply',
pattern: ".//div[contains(@class,'message-attribution-opposite')]//a/@href",
returnType: 'text',
pipe: [{ type: 'parse-as-url' }],
},
],
interval: 45,
priority: 'secondary',
page: 5,
patternPost: [
{
key: 'THREAD_ID_PATTERN',
pattern: './@class',
scope: 'thread',
pipe: [{ type: 'regex-extraction', regex: '\\w*$', flag: 'g' }],
},
{
key: 'THREAD_NODE_PATTERN',
pattern: "//div[contains(@class, 'structItem--thread')]",
scope: 'index-page',
meta: { isThreadNode: 'true' },
},
{
key: 'THREAD_TITLE_PATTERN',
pattern: ".//div[@class='structItem-title']/a/text()[normalize-space()]",
scope: 'thread',
},
{
key: 'THREAD_LINK_PATTERN',
pattern: ".//div[@class='structItem-title']/a/@href",
scope: 'thread',
pipe: [{ type: 'query-remover', removed: ['s'] }],
},
{
key: 'THREAD_REPLIES_PATTERN',
pattern: './div[3]/dl[1]/dd/text()',
scope: 'thread',
pipe: [{ type: 'num-normalize' }],
},
{
key: 'THREAD_VIEWS_PATTERN',
pattern: './div[3]/dl[2]/dd/text()',
scope: 'thread',
pipe: [{ type: 'num-normalize' }],
},
{
key: 'THREAD_LAST_POST_PATTERN',
pattern: './div[4]//time/@datetime',
scope: 'thread',
pipe: [
{
locale: 'en',
type: 'date-format',
format: 'YYYY-MM-DDTHH:mm:ssZZ',
timezone: 'Asia/Singapore',
},
],
},
{
key: 'PAGINATION_NODE_PATTERN',
pattern: "(//div[contains(@class, 'pageNav')])[1]/ul/li",
scope: 'index-page',
meta: {
isPaginateNode: 'true',
pagePatterns: { url: './a/@href', text: './a/text()' },
},
},
{
key: 'SECTION_TITLE',
pattern: '//head/title/text()',
scope: 'index-page',
},
],
maxItterateTRPage: 10,
subForumId: 433,
maxItteratePage: 10,
langCode: 'en',
countryCode: 'sg',
numItterate: 4,
numRetry: 0,
origin: 'bmw-sg.com',
mediaId: 902,
timeout: 45,
engine: 'html',
timecheck: 1757488584,
},
};
// Pipe classes following inspiration structure
class ParseAsUrlPipe {
type = 'parse-as-url';
baseUrl;
transform(url) {
if (!this.baseUrl) {
throw new Error('BaseURL is required for ParseAsUrlPipe');
}
try {
return new URL(url, this.baseUrl).toString();
}
catch (error) {
throw new Error(`Invalid URL: ${url} with baseUrl: ${this.baseUrl}`);
}
}
}
exports.ParseAsUrlPipe = ParseAsUrlPipe;
class RegexReplacePipe {
type = 'regex-replace';
baseUrl;
regex = '';
textReplacement = '';
flag = 'g';
transform(val) {
if (typeof val === 'string') {
const flag = Array.isArray(this.flag) ? this.flag.join(',') : this.flag;
const result = val.replace(new RegExp(this.regex, flag), this.textReplacement);
return result;
}
else {
return val;
}
}
}
exports.RegexReplacePipe = RegexReplacePipe;
class RegexExtractionPipe {
type = 'regex-extraction';
baseUrl;
regex = '';
flag = 'g';
transform(val) {
if (typeof val === 'string') {
const flag = Array.isArray(this.flag) ? this.flag.join(',') : this.flag;
const match = val.match(new RegExp(this.regex, flag));
return match ? match[0] : '';
}
else {
return val;
}
}
}
exports.RegexExtractionPipe = RegexExtractionPipe;
class QueryRemoverPipe {
type = 'query-remover';
baseUrl;
removed = [];
transform(url) {
try {
const urlObj = new URL(url);
this.removed.forEach((param) => {
urlObj.searchParams.delete(param);
});
return urlObj.toString();
}
catch (error) {
return url; // Return original if not a valid URL
}
}
}
exports.QueryRemoverPipe = QueryRemoverPipe;
class NumNormalizePipe {
type = 'num-normalize';
baseUrl;
transform(numString) {
if (typeof numString !== 'string') {
return numString;
}
const val = numString?.toLowerCase().replace(new RegExp(',', 'g'), '.');
let resVal = parseFloat(val);
if (val.endsWith('k') || val.endsWith('rb')) {
resVal *= 1000;
}
else if (val.endsWith('m')) {
resVal *= 1000000;
}
if (isNaN(resVal)) {
return 0;
}
return Math.round(resVal);
}
}
exports.NumNormalizePipe = NumNormalizePipe;
class DateFormatPipe {
type = 'date-format';
baseUrl;
locale = 'en';
format = 'YYYY-MM-DDTHH:mm:ssZZ';
timezone = 'Asia/Singapore';
transform(dateString) {
// Simple date parser that converts to Unix timestamp
if (typeof dateString !== 'string' || !dateString.trim()) {
return Date.now() / 1000;
}
const date = new Date(dateString);
return isNaN(date.getTime()) ? Date.now() / 1000 : date.getTime() / 1000;
}
}
exports.DateFormatPipe = DateFormatPipe;
class ThreadIdExtractorPipe {
type = 'thread-id-extractor';
baseUrl;
transform(url) {
if (typeof url !== 'string')
return url;
const match = url.match(/\.(\d+)\//);
return match ? match[1] : url;
}
}
exports.ThreadIdExtractorPipe = ThreadIdExtractorPipe;
// Mapping function to convert raw pipes to object-based transform format (DRY principle)
const mappingTransform = (rawPipes, withItsPayload) => {
if (!rawPipes || !Array.isArray(rawPipes))
return [];
return rawPipes.map((pipe) => {
// Handle different pipe types based on inspiration CleanerType enum
switch (pipe.type) {
case 'parse-as-url':
return {
class: ParseAsUrlPipe,
};
case 'query-remover':
return {
class: QueryRemoverPipe,
payload: {
removed: pipe.removed || withItsPayload?.removed || [],
},
};
case 'num-normalize':
return {
class: NumNormalizePipe,
};
case 'date-format':
return {
class: DateFormatPipe,
payload: {
locale: pipe.locale || 'en',
format: pipe.format || 'YYYY-MM-DDTHH:mm:ssZZ',
timezone: pipe.timezone || 'Asia/Singapore',
},
};
case 'regex-replace':
return {
class: RegexReplacePipe,
payload: {
regex: pipe.regex || '',
textReplacement: pipe.textReplacement || '',
flag: pipe.flag || 'g',
},
};
case 'regex-extraction':
return {
class: RegexExtractionPipe,
payload: {
regex: pipe.regex || '',
flag: pipe.flag || 'g',
},
};
default:
console.warn(`Unknown pipe type: ${pipe.type}`);
// Return a no-op transform instead of null
return {
class: class NoOpPipe {
transform(val) {
return val;
}
},
};
}
}); // Don't filter, always return valid transforms
};
exports.mappingTransform = mappingTransform;
async function scrapeBmwSgForum(verbose = false) {
const parser = new __1.HtmlParserService();
console.log('🏎️ BMW-SG Forum Scraper Demo');
console.log('='.repeat(50));
// inspiration payload configuration
const endPoint = payload.data.endPoint;
const baseUrl = parser.getOrigin(endPoint);
console.log(`🔗 Target URL: ${endPoint}`);
console.log(`🏠 Base URL: ${baseUrl}`);
console.log();
try {
console.log('📥 Fetching HTML from BMW-SG forum...');
// Fetch the HTML content from the real BMW-SG forum
const response = await parser.fetchHtml(endPoint, {
headers: {
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36',
},
});
console.log(`✅ HTML fetched successfully (${response.data.length} characters)`);
console.log(`📊 Response status: ${response.status}`);
console.log();
// 1. Thread extraction based on inspiration patternPost
console.log('📝 Thread Extraction (patternPost)');
console.log('-'.repeat(40));
// Build thread schema from payload patterns
const getPatternByKey = (key) => payload.data.patternPost.find((p) => p.key === key);
const threadSchema = {
// THREAD_ID_PATTERN - Extract numeric ID from URL
threadId: {
selector: './/div[@class="structItem-title"]/a/@href',
type: 'xpath',
transform: [{ class: ThreadIdExtractorPipe }],
},
// THREAD_TITLE_PATTERN
threadTitle: {
selector: getPatternByKey('THREAD_TITLE_PATTERN')?.pattern ||
'.//div[@class="structItem-title"]/a/text()[normalize-space()]',
type: 'xpath',
},
// THREAD_LINK_PATTERN
threadLink: {
selector: getPatternByKey('THREAD_LINK_PATTERN')?.pattern ||
'.//div[@class="structItem-title"]/a/@href',
type: 'xpath',
transform: [
{ class: ParseAsUrlPipe }, // Convert to absolute URL
...mappingTransform(getPatternByKey('THREAD_LINK_PATTERN')?.pipe || [
{ type: 'query-remover', removed: ['s'] },
]),
],
},
// THREAD_REPLIES_PATTERN - From structItem-cell--meta
threadReplies: {
selector: './/div[contains(@class,"structItem-cell--meta")]//dl[dt="Replies"]/dd/text()',
type: 'xpath',
transform: mappingTransform([{ type: 'num-normalize' }]),
},
// THREAD_VIEWS_PATTERN - From structItem-cell--meta
threadViews: {
selector: './/div[contains(@class,"structItem-cell--meta")]//dl[dt="Views"]/dd/text()',
type: 'xpath',
transform: mappingTransform([{ type: 'num-normalize' }]),
},
// THREAD_LAST_POST_PATTERN - From structItem-cell--latest (updated for real HTML structure)
threadLastPost: {
selector: './/div[contains(@class,"structItem-cell--latest")]//time/@datetime',
type: 'xpath',
transform: mappingTransform(getPatternByKey('THREAD_LAST_POST_PATTERN')?.pipe || [
{
type: 'date-format',
locale: 'en',
format: 'YYYY-MM-DDTHH:mm:ssZZ',
timezone: 'Asia/Singapore',
},
]),
},
};
// Extract threads using THREAD_NODE_PATTERN
const threadNodePattern = payload.data.patternPost.find((p) => p.key === 'THREAD_NODE_PATTERN')
?.pattern || '//div[contains(@class, "structItem--thread")]';
const threads = parser.extractStructuredList(response.data, threadNodePattern, // THREAD_NODE_PATTERN
threadSchema, 'xpath', {
verbose,
baseUrl,
});
console.log(`✅ Extracted ${threads.length} threads:`);
console.log();
threads.forEach((thread, index) => {
console.log(`📄 Thread ${index + 1}:`);
console.log(` ID: ${thread.threadId}`);
console.log(` Title: ${thread.threadTitle}`);
console.log(` Link: ${thread.threadLink}`);
console.log(` Replies: ${thread.threadReplies}`);
console.log(` Views: ${thread.threadViews}`);
const timestamp = thread.threadLastPost && !isNaN(thread.threadLastPost)
? thread.threadLastPost
: Date.now() / 1000;
const lastPostDate = new Date(timestamp * 1000);
console.log(` Last Post: ${lastPostDate.toISOString()}`);
console.log();
});
// 2. Reply extraction based on inspiration patternReply
console.log('💬 Reply Extraction (patternReply)');
console.log('-'.repeat(40));
// Build reply schema from payload patterns
const getReplyPatternByKey = (key) => payload.data.patternReply.find((p) => p.key === key);
const replySchema = {
// Author
author: {
selector: getReplyPatternByKey('author')?.pattern || './@data-author',
type: 'xpath',
},
// Reply date
replyDate: {
selector: getReplyPatternByKey('replyDate')?.pattern ||
'.//time[@datetime]/@datetime',
type: 'xpath',
transform: mappingTransform(getReplyPatternByKey('replyDate')?.pipe || [
{
type: 'date-format',
locale: 'en',
format: 'YYYY-MM-DDTHH:mm:ssZZ',
timezone: 'Asia/Singapore',
},
]),
},
// Reply ID
replyId: {
selector: getReplyPatternByKey('replyId')?.pattern || './@data-content',
type: 'xpath',
transform: mappingTransform(getReplyPatternByKey('replyId')?.pipe || [
{
type: 'regex-replace',
regex: '^post-',
textReplacement: '',
flag: 'g',
},
]),
},
// Reply HTML
replyHtml: {
selector: getReplyPatternByKey('replyHtml')?.pattern ||
'.//article[contains(@class,"message-body")]',
type: 'xpath',
raw: true, // Return HTML instead of text
},
// Reply text (multiple text nodes)
replyText: {
selector: getReplyPatternByKey('replyText')?.pattern ||
'.//article[contains(@class,"message-body")]/descendant-or-self::node()/text()[normalize-space()]',
type: 'xpath',
multiple: true, // Extract multiple text nodes
},
// URL reply
urlReply: {
selector: getReplyPatternByKey('urlReply')?.pattern ||
'.//div[contains(@class,"message-attribution-opposite")]//a/@href',
type: 'xpath',
transform: mappingTransform(getReplyPatternByKey('urlReply')?.pipe || [{ type: 'parse-as-url' }]),
},
};
// Extract replies using container pattern from payload
const replyContainerPattern = getReplyPatternByKey('container')?.pattern ||
'//article[contains(@class,"message--post")]';
const replies = parser.extractStructuredList(response.data, replyContainerPattern, // container pattern
replySchema, 'xpath', {
verbose,
baseUrl,
});
console.log(`✅ Extracted ${replies.length} replies:`);
console.log();
replies.slice(0, 3).forEach((reply, index) => {
// Show first 3 replies
console.log(`💭 Reply ${index + 1}:`);
console.log(` Author: ${reply.author}`);
const replyTimestamp = reply.replyDate && !isNaN(reply.replyDate)
? reply.replyDate
: Date.now() / 1000;
const replyDate = new Date(replyTimestamp * 1000);
console.log(` Date: ${replyDate.toISOString()}`);
console.log(` ID: ${reply.replyId}`);
console.log(` Text (first 100 chars): ${reply.replyText ? reply.replyText.join(' ').substring(0, 100) : 'N/A'}...`);
console.log(` URL: ${reply.urlReply}`);
console.log();
});
// 3. Pagination and section info
console.log('📄 Page Information');
console.log('-'.repeat(40));
// Extract section title using payload pattern
const sectionTitlePattern = getPatternByKey('SECTION_TITLE')?.pattern || '//head/title/text()';
const sectionTitle = parser.extractSingle(response.data, sectionTitlePattern, 'xpath');
console.log(`📋 Section Title: ${sectionTitle}`);
// Extract pagination info using payload pattern
const nextPagePattern = getReplyPatternByKey('nextPage')?.pattern ||
'//ul[contains(@class,"pageNav-main")]/li/a[@href and not(@href="#")]/@href';
const nextPageUrls = parser.extractMultiple(response.data, nextPagePattern, 'xpath', undefined, {
baseUrl,
transform: mappingTransform([{ type: 'parse-as-url' }]),
});
console.log(`🔗 Available page URLs: ${nextPageUrls.length}`);
nextPageUrls.slice(0, 5).forEach((url, index) => {
console.log(` ${index + 1}. ${url}`);
});
// 4. Summary statistics
console.log();
console.log('📊 Scraping Summary');
console.log('-'.repeat(40));
console.log(`🎯 Endpoint: ${endPoint}`);
console.log(`📝 Threads extracted: ${threads.length}`);
console.log(`💬 Replies extracted: ${replies.length}`);
console.log(`📄 Pagination URLs: ${nextPageUrls.length}`);
console.log(`📋 Section: ${sectionTitle}`);
console.log(`🏷️ Media ID: ${payload.data.mediaId}`);
console.log(`🌐 Origin: ${payload.data.origin}`);
console.log(`⏱️ Interval: ${payload.data.interval}s`);
console.log(`📄 Page: ${payload.data.page}`);
console.log(`🔧 Engine: ${payload.data.engine}`);
// Show some pipe transformation examples
console.log();
console.log('🔧 Live Pipe Transformations');
console.log('-'.repeat(40));
if (threads.length > 0) {
const sampleThread = threads[0];
console.log('📝 Thread ID extraction:');
console.log(` Raw class: "${sampleThread.threadId}"`);
console.log(` Regex pattern: "\\w*$" → Extract last word`);
console.log();
}
if (replies.length > 0) {
const sampleReply = replies[0];
console.log('💭 Reply ID transformation:');
console.log(` Raw data-content: "post-${sampleReply.replyId}"`);
console.log(` Regex replace: "^post-" → "" (remove prefix)`);
console.log(` Result: "${sampleReply.replyId}"`);
console.log();
}
}
catch (error) {
console.error('❌ Error during BMW-SG forum scraping:', error);
throw error;
}
}
// Run the scraper if this file is executed directly
if (require.main === module) {
(async () => {
try {
await scrapeBmwSgForum(false);
console.log('🎉 BMW-SG forum scraping completed successfully!');
}
catch (error) {
console.error('💥 Scraping failed:', error);
process.exit(1);
}
})();
}
//# sourceMappingURL=bmw-sg.com.js.map