package crawler

import (
	"context"
	"errors"
	"fmt"
	"log"
	"math/rand"
	"regexp"
	"strings"
	"time"

	"github.com/gocolly/colly/v2"
)

// ... userAgents, referers, regex vars ...
var userAgents = []string{
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36",
	"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36",
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Firefox/117.0",
}
var referers = []string{
	"https://www.google.com/",
	"https://www.hermes.com/us/en/",
}
var blockRegex = regexp.MustCompile(`(?i)Access\s+blocked`)

func init() {
	rand.New(rand.NewSource(time.Now().UnixNano()))
}

func Scrape(ctx context.Context, url string) (map[string]Bag, error) {
	var blockedError error
	var responseBody []byte // Variable to store the decompressed body
	c := colly.NewCollector()

	c.Limit(&colly.LimitRule{
		DomainGlob:  "*",
		RandomDelay: 2 * time.Second,
	})

	c.OnRequest(func(r *colly.Request) {
		r.Headers.Set("User-Agent", userAgents[rand.Intn(len(userAgents))])
		r.Headers.Set("Referer", referers[rand.Intn(len(referers))])
		// By default, Colly adds the correct Accept-Encoding header.
		// We don't need to set it manually.
	})

	// The OnResponse callback now just saves the body for later.
	c.OnResponse(func(r *colly.Response) {
		responseBody = r.Body
	})

	foundBags := make(map[string]Bag)
	skuRegex := regexp.MustCompile(`-H\w+$`)

	c.OnHTML("div.product-grid-list-item", func(e *colly.HTMLElement) {
		// ... HTML parsing logic remains the same ...
		idStr := e.Attr("id")
		skuMatch := skuRegex.FindString(idStr)
		if skuMatch == "" {
			return
		}
		sku := strings.TrimPrefix(skuMatch, "-")

		unavailableIndicator := e.DOM.Find("span:contains('Unavailable')")
		isAvailable := unavailableIndicator.Length() == 0

		bag := Bag{
			SKU:          sku,
			Name:         e.ChildText("span.product-title"),
			URL:          e.Request.AbsoluteURL(e.ChildAttr("a.product-item-name", "href")),
			ImageURL:     "https:" + e.ChildAttr("img[id^='img-']", "src"),
			Availability: isAvailable,
		}

		if bag.Name != "" && bag.URL != "" {
			foundBags[sku] = bag
		}
	})

	c.OnScraped(func(r *colly.Response) {
		// The response body is now guaranteed to be decompressed and readable.
		if isDebug, ok := ctx.Value(DebugContextKey).(bool); ok && isDebug {
			log.Printf("[DEBUG] Raw HTML response received:\n%s\n", string(responseBody))
		}

		// Check for the block message on the readable content.
		if blockRegex.Match(responseBody) {
			blockedError = errors.New("request failed: hit anti-crawling wall (message found in page)")
		}
	})

	c.OnError(func(r *colly.Response, err error) {
		log.Printf("❌ Colly request error for %s (Status: %d): %v", r.Request.URL, r.StatusCode, err)
	})

	err := c.Visit(url)

	if blockedError != nil {
		return nil, blockedError
	}

	if err != nil {
		return nil, fmt.Errorf("could not complete visit to %s: %w", url, err)
	}

	return foundBags, nil
}