PowderCoatingLogix/scripts/Web Scraping/backup/Get-Product-Info.ps1

# Crawl and Index Prismatic Colors - Known-Good Style JSON.ps1
#
# Rollback to the earlier working browser pattern:
# - Playwright Chromium
# - Full Chrome-style User-Agent
# - JSON output
# - Structured price tiers
# - Color matches from #collection-list
#
# First-time setup:
#   Set-ExecutionPolicy -Scope Process -ExecutionPolicy Bypass
#   .\Crawl-and-Index-Prismatic-colors-known-good-json.ps1 -InstallPlaywright
#
# Normal run:
#   .\Crawl-and-Index-Prismatic-colors-known-good-json.ps1
#
# Watch browser:
#   .\Crawl-and-Index-Prismatic-colors-known-good-json.ps1 -Headed

param(
    [switch]$InstallPlaywright,
    [switch]$Headed
)

$ErrorActionPreference = "Stop"

function Ensure-NodeAvailable {
    if (-not (Get-Command node -ErrorAction SilentlyContinue)) {
        throw "Node.js is required. Install Node.js LTS from https://nodejs.org/"
    }

    if (-not (Get-Command npm -ErrorAction SilentlyContinue)) {
        throw "npm is required. It usually comes with Node.js."
    }
}

function Install-PlaywrightIfNeeded {
    param([bool]$Requested)

    Ensure-NodeAvailable

    if ($Requested -or -not (Test-Path ".\node_modules\playwright")) {
        Write-Host "Installing Playwright package locally..."
        npm init -y | Out-Null
        npm install playwright | Out-Null

        Write-Host "Installing Playwright Chromium browser..."
        npx playwright install chromium
    }
}

function Write-NodeScraper {
    # Single-quoted here-string prevents PowerShell from interpreting JavaScript regex/template strings.
    $js = @'
const fs = require("fs");
const { chromium } = require("playwright");

const headed = process.argv.includes("--headed");

const productUrls = [
  "https://www.prismaticpowders.com/shop/powder-coating-colors/PSS-11248/high-gloss-black"
];

const outputJson = "prismatic_powders.json";

function clean(text) {
  return (text || "").replace(/\s+/g, " ").trim();
}

function absoluteUrl(baseUrl, maybeUrl) {
  if (!maybeUrl) return "";
  try {
    return new URL(maybeUrl, baseUrl).href;
  } catch {
    return maybeUrl;
  }
}

function unique(items) {
  return [...new Set(items.filter(Boolean).map(clean).filter(Boolean))];
}

async function getLinkByText(page, patterns) {
  const links = await page.locator("a").evaluateAll((anchors) =>
    anchors.map(a => ({
      text: (a.innerText || a.textContent || "").replace(/\s+/g, " ").trim(),
      href: a.getAttribute("href") || ""
    }))
  );

  for (const link of links) {
    if (patterns.some(p => new RegExp(p, "i").test(link.text))) {
      return absoluteUrl(page.url(), link.href);
    }
  }

  return "";
}

function parsePriceTiers(plainText) {
  const priceMatches = [...plainText.matchAll(/(\d+\s*-\s*\d+\s*lbs|\d+\s*\+\s*lbs)\s*\$([\d.]+)/gi)];

  return priceMatches.map(m => {
    const rangeText = clean(m[1]);
    const price = parseFloat(m[2]);

    let min = null;
    let max = null;

    const rangeMatch = rangeText.match(/(\d+)\s*-\s*(\d+)/);
    if (rangeMatch) {
      min = parseInt(rangeMatch[1], 10);
      max = parseInt(rangeMatch[2], 10);
    }

    const plusMatch = rangeText.match(/(\d+)\s*\+/);
    if (plusMatch) {
      min = parseInt(plusMatch[1], 10);
      max = null;
    }

    return {
      min,
      max,
      price
    };
  });
}

async function getSampleImageUrl(page) {
  const imageUrls = await page.locator("img").evaluateAll((imgs) =>
    imgs.map(img =>
      img.currentSrc ||
      img.src ||
      img.getAttribute("src") ||
      img.getAttribute("data-src") ||
      ""
    ).filter(Boolean)
  );

  return (
    imageUrls.find(src => /images\.nicindustries\.com/i.test(src) && !/thumbnail/i.test(src)) ||
    imageUrls.find(src => /images\.nicindustries\.com/i.test(src)) ||
    imageUrls.find(src => /prismatic|powder|color/i.test(src)) ||
    ""
  );
}

async function parseProduct(page, url) {
  console.log(`Scraping ${url}`);

  const response = await page.goto(url, {
    waitUntil: "domcontentloaded",
    timeout: 60000
  });

  await page.waitForTimeout(3000);

  const status = response ? response.status() : 0;
  const pageTitle = clean(await page.title().catch(() => ""));
  const plainText = clean(await page.locator("body").innerText().catch(() => ""));

  console.log(`HTTP status: ${status}`);
  console.log(`Page title: ${pageTitle}`);

  // Do not silently output a fake product if blocked.
  if (status === 403 || /^403 Forbidden$/i.test(pageTitle) || /^403 Forbidden$/i.test(plainText)) {
    throw new Error("403 Forbidden returned by site.");
  }

  const title = clean(await page.locator("h1").first().innerText().catch(() => ""));

  const skuMatch = plainText.match(/Item:\s*([A-Z0-9-]+)/i);
  const sku = skuMatch ? skuMatch[1] : "";

  const descMatch = plainText.match(/Description:\s*(.*?)(WARNING:|What does this match\?|$)/is);
  const description = descMatch ? clean(descMatch[1]) : "";

  const priceTiers = parsePriceTiers(plainText);

  const safetyDataSheetUrl = await getLinkByText(page, ["Safety Data Sheet", "\\bSDS\\b"]);
  const applicationGuideUrl = await getLinkByText(page, ["Application Guide"]);
  const technicalDataSheetUrl = await getLinkByText(page, ["Tech Data Sheet", "Technical Data Sheet", "\\bTDS\\b"]);
  const sampleImageUrl = await getSampleImageUrl(page);

  return {
    sku,
    color_name: title,
    description,
    price_tiers: priceTiers,
    safety_data_sheet_url: safetyDataSheetUrl,
    technical_data_sheet_url: technicalDataSheetUrl,
    application_guide_url: applicationGuideUrl,
    sample_image_url: sampleImageUrl,
    product_url: url,
    scraped_at: new Date().toISOString()
  };
}

(async () => {
  const browser = await chromium.launch({
    headless: !headed
  });

  const context = await browser.newContext({
    userAgent: "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
    viewport: { width: 1365, height: 900 },
    locale: "en-US",
    timezoneId: "America/New_York"
  });

  const page = await context.newPage();

  const results = [];
  const errors = [];

  for (const url of productUrls) {
    try {
      const row = await parseProduct(page, url);
      results.push(row);
      await page.waitForTimeout(3000);
    } catch (err) {
      console.warn(`Failed ${url}: ${err.message}`);
      errors.push({
        product_url: url,
        error: err.message,
        scraped_at: new Date().toISOString()
      });
    }
  }

  await browser.close();

  // If you prefer only the array, change this to JSON.stringify(results, null, 2)
  const output = {
    results,
    errors
  };

  fs.writeFileSync(outputJson, JSON.stringify(output, null, 2), "utf8");

  console.log(`Done. Output: ${outputJson}`);
})();
'@

    Set-Content -Path ".\prismatic-browser-scraper.js" -Value $js -Encoding UTF8
}

try {
    Install-PlaywrightIfNeeded -Requested:$InstallPlaywright
    Write-NodeScraper

    Write-Host "Running browser scraper..."

    if ($Headed) {
        node .\prismatic-browser-scraper.js --headed
    }
    else {
        node .\prismatic-browser-scraper.js
    }
}
catch {
    Write-Error $_.Exception.Message
    exit 1
}