package main

import (
	"bufio"
	"crypto/md5"
	"encoding/hex"
	"encoding/json"
	"fmt"
	"io"
	"io/ioutil"
	"net/http"
	"net/url"
	"os"
	"regexp"
	"strings"
	"sync"
	"sync/atomic"
	"time"
)

type Record struct {
	URL  string `json:"url"`
	Text string `json:"text"`
	TS   string `json:"ts"`
}

var (
	visited   = make(map[string]bool)
	visitedMu sync.RWMutex
	queue     = make(chan string, 1000000)
	stats     struct {
		pages, bytes, errors int64
	}
	client = &http.Client{
		Timeout: 5 * time.Second,
		Transport: &http.Transport{
			MaxIdleConns:        1000,
			MaxIdleConnsPerHost: 50,
			IdleConnTimeout:     30 * time.Second,
		},
	}
	linkRe   = regexp.MustCompile(`href=["']([^"']+)["']`)
	scriptRe = regexp.MustCompile(`(?is)<script[^>]*>.*?</script>|<style[^>]*>.*?</style>`)
	tagRe    = regexp.MustCompile(`<[^>]+>`)
	spaceRe  = regexp.MustCompile(`\s+`)
)

func hash(s string) string {
	h := md5.Sum([]byte(s))
	return hex.EncodeToString(h[:8])
}

func extractText(html string) string {
	text := scriptRe.ReplaceAllString(html, " ")
	text = tagRe.ReplaceAllString(text, " ")
	text = spaceRe.ReplaceAllString(text, " ")
	text = strings.TrimSpace(text)
	if len(text) > 50000 {
		text = text[:50000]
	}
	return text
}

func extractLinks(html, baseURL string) []string {
	base, err := url.Parse(baseURL)
	if err != nil {
		return nil
	}
	var links []string
	matches := linkRe.FindAllStringSubmatch(html, 100)
	for _, m := range matches {
		if len(m) > 1 {
			href := m[1]
			if u, err := url.Parse(href); err == nil {
				resolved := base.ResolveReference(u)
				if resolved.Scheme == "http" || resolved.Scheme == "https" {
					links = append(links, resolved.String())
				}
			}
		}
	}
	return links
}

func worker(id int, output chan<- Record) {
	for urlStr := range queue {
		h := hash(urlStr)
		visitedMu.RLock()
		seen := visited[h]
		visitedMu.RUnlock()
		if seen {
			continue
		}
		visitedMu.Lock()
		visited[h] = true
		visitedMu.Unlock()

		resp, err := client.Get(urlStr)
		if err != nil {
			atomic.AddInt64(&stats.errors, 1)
			continue
		}

		ct := resp.Header.Get("Content-Type")
		if resp.StatusCode != 200 || !strings.Contains(ct, "text/html") {
			resp.Body.Close()
			atomic.AddInt64(&stats.errors, 1)
			continue
		}

		body, _ := ioutil.ReadAll(io.LimitReader(resp.Body, 1<<20))
		resp.Body.Close()
		html := string(body)

		text := extractText(html)
		if len(text) > 200 {
			output <- Record{URL: urlStr, Text: text, TS: time.Now().UTC().Format("2006-01-02T15:04:05")}
			atomic.AddInt64(&stats.pages, 1)
			atomic.AddInt64(&stats.bytes, int64(len(text)))
		}

		for _, link := range extractLinks(html, urlStr) {
			select {
			case queue <- link:
			default:
			}
		}
	}
}

func main() {
	seeds := []string{
		"https://en.wikipedia.org/wiki/Main_Page",
		"https://en.wikipedia.org/wiki/Special:Random",
		"https://news.ycombinator.com/",
		"https://www.reddit.com/r/all/",
		"https://www.reddit.com/r/programming/",
		"https://stackoverflow.com/questions",
		"https://medium.com/",
		"https://dev.to/",
	}

	for _, s := range seeds {
		queue <- s
	}

	os.MkdirAll("/workspace/go_crawl", 0755)
	outPath := fmt.Sprintf("/workspace/go_crawl/crawl_%s.jsonl", time.Now().Format("20060102_150405"))
	f, _ := os.Create(outPath)
	defer f.Close()
	w := bufio.NewWriter(f)
	defer w.Flush()

	output := make(chan Record, 10000)

	go func() {
		for rec := range output {
			data, _ := json.Marshal(rec)
			w.Write(data)
			w.WriteByte('\n')
		}
	}()

	start := time.Now()
	go func() {
		for {
			time.Sleep(10 * time.Second)
			elapsed := time.Since(start).Minutes()
			if elapsed > 0 {
				p := atomic.LoadInt64(&stats.pages)
				b := atomic.LoadInt64(&stats.bytes)
				e := atomic.LoadInt64(&stats.errors)
				fmt.Printf("[%s] %d pgs | %.0fMB | %.1fMB/min | Q:%d | Err:%d\n",
					time.Now().Format("15:04:05"), p, float64(b)/1024/1024, float64(b)/1024/1024/elapsed, len(queue), e)
			}
		}
	}()

	fmt.Printf("=== GO STDLIB CRAWLER ===\nWorkers: 1000\nOutput: %s\n\n", outPath)

	var wg sync.WaitGroup
	for i := 0; i < 1000; i++ {
		wg.Add(1)
		go func(id int) {
			defer wg.Done()
			worker(id, output)
		}(i)
	}
	wg.Wait()
	close(output)
}