chore: bump task worker version comment
Force new git SHA to avoid CI scientific notation bug. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
178
backend/scripts/test-treez-selectors.ts
Normal file
178
backend/scripts/test-treez-selectors.ts
Normal file
@@ -0,0 +1,178 @@
|
||||
/**
|
||||
* Analyze all product element structures to find all selector patterns
|
||||
*/
|
||||
|
||||
import puppeteer, { Page } from 'puppeteer';
|
||||
|
||||
const STORE_ID = 'best';
|
||||
|
||||
async function sleep(ms: number): Promise<void> {
|
||||
return new Promise(resolve => setTimeout(resolve, ms));
|
||||
}
|
||||
|
||||
async function bypassAgeGate(page: Page): Promise<void> {
|
||||
const ageGate = await page.$('[data-testid="age-gate-modal"]');
|
||||
if (ageGate) {
|
||||
const btn = await page.$('[data-testid="age-gate-submit-button"]');
|
||||
if (btn) await btn.click();
|
||||
await sleep(2000);
|
||||
}
|
||||
}
|
||||
|
||||
async function main() {
|
||||
console.log('Analyzing Treez product selectors...\n');
|
||||
|
||||
const browser = await puppeteer.launch({
|
||||
headless: true,
|
||||
args: ['--no-sandbox', '--disable-setuid-sandbox'],
|
||||
});
|
||||
|
||||
const page = await browser.newPage();
|
||||
await page.setViewport({ width: 1920, height: 1080 });
|
||||
|
||||
await page.setRequestInterception(true);
|
||||
page.on('request', (req) => {
|
||||
if (['image', 'font', 'media'].includes(req.resourceType())) {
|
||||
req.abort();
|
||||
} else {
|
||||
req.continue();
|
||||
}
|
||||
});
|
||||
|
||||
const url = `https://${STORE_ID}.treez.io/onlinemenu/brands?customerType=ADULT`;
|
||||
await page.goto(url, { waitUntil: 'networkidle2', timeout: 60000 });
|
||||
await sleep(3000);
|
||||
await bypassAgeGate(page);
|
||||
await sleep(2000);
|
||||
|
||||
// Analyze product elements
|
||||
const analysis = await page.evaluate(() => {
|
||||
const products = document.querySelectorAll('[class*="product_product__"]');
|
||||
const results: {
|
||||
hasName: number;
|
||||
hasPrice: number;
|
||||
noName: number;
|
||||
noPrice: number;
|
||||
nameClasses: Record<string, number>;
|
||||
priceClasses: Record<string, number>;
|
||||
sampleNoNameHTML: string[];
|
||||
sampleWithNameHTML: string[];
|
||||
h5Count: number;
|
||||
h4Count: number;
|
||||
h3Count: number;
|
||||
allHeadingsWithName: number;
|
||||
} = {
|
||||
hasName: 0,
|
||||
hasPrice: 0,
|
||||
noName: 0,
|
||||
noPrice: 0,
|
||||
nameClasses: {},
|
||||
priceClasses: {},
|
||||
sampleNoNameHTML: [],
|
||||
sampleWithNameHTML: [],
|
||||
h5Count: 0,
|
||||
h4Count: 0,
|
||||
h3Count: 0,
|
||||
allHeadingsWithName: 0,
|
||||
};
|
||||
|
||||
products.forEach((el, i) => {
|
||||
// Check current selectors
|
||||
const nameEl = el.querySelector('[class*="product__name"], [class*="name__"]');
|
||||
const priceEl = el.querySelector('[class*="price"]');
|
||||
|
||||
if (nameEl) {
|
||||
results.hasName++;
|
||||
const cls = nameEl.className?.toString?.() || '';
|
||||
results.nameClasses[cls] = (results.nameClasses[cls] || 0) + 1;
|
||||
} else {
|
||||
results.noName++;
|
||||
if (results.sampleNoNameHTML.length < 3) {
|
||||
results.sampleNoNameHTML.push(el.innerHTML.slice(0, 500));
|
||||
}
|
||||
}
|
||||
|
||||
if (priceEl) {
|
||||
results.hasPrice++;
|
||||
} else {
|
||||
results.noPrice++;
|
||||
}
|
||||
|
||||
// Check for headings that might contain names
|
||||
const h5 = el.querySelector('h5');
|
||||
const h4 = el.querySelector('h4');
|
||||
const h3 = el.querySelector('h3');
|
||||
if (h5) results.h5Count++;
|
||||
if (h4) results.h4Count++;
|
||||
if (h3) results.h3Count++;
|
||||
|
||||
// Any heading with text
|
||||
const anyHeading = el.querySelector('h1, h2, h3, h4, h5, h6');
|
||||
if (anyHeading?.textContent?.trim()) {
|
||||
results.allHeadingsWithName++;
|
||||
}
|
||||
});
|
||||
|
||||
return results;
|
||||
});
|
||||
|
||||
console.log('Product Analysis:');
|
||||
console.log(`Total products: ${analysis.hasName + analysis.noName}`);
|
||||
console.log(`With name (current selector): ${analysis.hasName}`);
|
||||
console.log(`Without name (current selector): ${analysis.noName}`);
|
||||
console.log(`With price: ${analysis.hasPrice}`);
|
||||
console.log(`\nHeading counts:`);
|
||||
console.log(` H5: ${analysis.h5Count}`);
|
||||
console.log(` H4: ${analysis.h4Count}`);
|
||||
console.log(` H3: ${analysis.h3Count}`);
|
||||
console.log(` Any heading with text: ${analysis.allHeadingsWithName}`);
|
||||
|
||||
console.log('\nName classes found:');
|
||||
Object.entries(analysis.nameClasses).forEach(([cls, count]) => {
|
||||
console.log(` (${count}x) ${cls.slice(0, 80)}`);
|
||||
});
|
||||
|
||||
console.log('\n--- Sample products WITHOUT name selector ---');
|
||||
analysis.sampleNoNameHTML.forEach((html, i) => {
|
||||
console.log(`\n[Sample ${i + 1}]:`);
|
||||
console.log(html);
|
||||
});
|
||||
|
||||
// Try different selectors
|
||||
console.log('\n\n--- Testing alternative selectors ---');
|
||||
|
||||
const altResults = await page.evaluate(() => {
|
||||
const products = document.querySelectorAll('[class*="product_product__"]');
|
||||
const tests: Record<string, number> = {};
|
||||
|
||||
const selectors = [
|
||||
'h5',
|
||||
'h4',
|
||||
'h3',
|
||||
'[class*="heading"]',
|
||||
'[class*="title"]',
|
||||
'[class*="name"]',
|
||||
'a[href*="/product/"]',
|
||||
'.product_product__name__JcEk0',
|
||||
'[class*="ProductCard"]',
|
||||
];
|
||||
|
||||
selectors.forEach(sel => {
|
||||
let count = 0;
|
||||
products.forEach(el => {
|
||||
if (el.querySelector(sel)) count++;
|
||||
});
|
||||
tests[sel] = count;
|
||||
});
|
||||
|
||||
return tests;
|
||||
});
|
||||
|
||||
Object.entries(altResults).forEach(([sel, count]) => {
|
||||
console.log(` ${sel}: ${count} products`);
|
||||
});
|
||||
|
||||
await browser.close();
|
||||
}
|
||||
|
||||
main().catch(console.error);
|
||||
Reference in New Issue
Block a user