手把手教你用Go打造带可视化的网络爬虫

package main
​
import (
    "fmt"
    "log"
    "os"
    "strconv"
    "strings"
​
    "github.com/gocolly/colly"
    "github.com/go-echarts/go-echarts/v2/charts"
    "github.com/go-echarts/go-echarts/v2/opts"
)
​
// 定义数据结构
type Product struct {
    Name  string
    Price float64
    Stars float64
}
​
func main() {
    // 初始化爬虫
    c := colly.NewCollector(
        colly.AllowedDomains("books.toscrape.com"),
    )
​
    var products []Product
​
    // 查找产品信息
    c.OnHTML(".product_pod", func(e *colly.HTMLElement) {
        // 提取书名
        title := e.ChildAttr(".image_container img", "alt")
        
        // 提取价格
        priceStr := e.ChildText(".price_color")
        priceStr = strings.Replace(priceStr, "£", "", 1)
        price, err := strconv.ParseFloat(priceStr, 64)
        if err != nil {
            log.Printf("价格转换错误: %v", err)
            return
        }
        
        // 提取评分
        starClass := e.ChildAttr(".star-rating", "class")
        stars := parseStars(starClass)
        
        products = append(products, Product{
            Name:  title,
            Price: price,
            Stars: stars,
        })
    })
​
    // 处理下一页
    c.OnHTML(".next a", func(e *colly.HTMLElement) {
        nextPage := e.Request.AbsoluteURL(e.Attr("href"))
        c.Visit(nextPage)
    })
​
    // 设置爬虫错误处理
    c.OnError(func(r *colly.Response, err error) {
        log.Println("请求错误:", err)
    })
​
    // 开始爬取
    fmt.Println("开始爬取数据...")
    err := c.Visit("http://books.toscrape.com/")
    if err != nil {
        log.Fatal(err)
    }
​
    fmt.Printf("共爬取到 %d 条数据\n", len(products))
    
    // 生成可视化图表
    generateChart(products)
}
​
// 解析星级评分
func parseStars(starClass string) float64 {
    parts := strings.Split(starClass, " ")
    if len(parts) < 2 {
        return 0
    }
    
    switch parts[1] {
    case "One":
        return 1
    case "Two":
        return 2
    case "Three":
        return 3
    case "Four":
        return 4
    case "Five":
        return 5
    default:
        return 0
    }
}
​
// 生成可视化图表
func generateChart(products []Product) {
    // 准备数据
    var names []string
    var prices []opts.BarData
    var stars []opts.ScatterData
    
    for _, p := range products {
        names = append(names, p.Name)
        prices = append(prices, opts.BarData{Value: p.Price})
        stars = append(stars, opts.ScatterData{Value: p.Stars})
    }
    
    // 创建柱状图显示价格
    bar := charts.NewBar()
    bar.SetGlobalOptions(
        charts.WithTitleOpts(opts.Title{
            Title: "图书价格分布",
        }),
        charts.WithXAxisOpts(opts.XAxis{
            Name: "图书名称",
            AxisLabel: &opts.AxisLabel{Show: true, Interval: "0", Rotate: 45},
        }),
        charts.WithYAxisOpts(opts.YAxis{
            Name: "价格(£)",
        }),
        charts.WithInitializationOpts(opts.Initialization{
            Width:  "1200px",
            Height: "600px",
        }),
    )
    
    bar.SetXAxis(names).
        AddSeries("价格", prices)
    
    // 创建散点图显示评分
    scatter := charts.NewScatter()
    scatter.SetGlobalOptions(
        charts.WithTitleOpts(opts.Title{
            Title: "图书评分分布",
        }),
        charts.WithXAxisOpts(opts.XAxis{
            Name: "图书名称",
            AxisLabel: &opts.AxisLabel{Show: true, Interval: "0", Rotate: 45},
        }),
        charts.WithYAxisOpts(opts.YAxis{
            Name: "评分",
        }),
        charts.WithInitializationOpts(opts.Initialization{
            Width:  "1200px",
            Height: "600px",
        }),
    )
    
    scatter.SetXAxis(names).
        AddSeries("评分", stars)
    
    // 创建HTML文件
    f, err := os.Create("book_analysis.html")
    if err != nil {
        log.Fatal(err)
    }
    defer f.Close()
    
    // 渲染图表到HTML
    page := charts.NewPage()
    page.AddCharts(bar, scatter)
    page.Render(f)
    
    fmt.Println("可视化图表已生成到 book_analysis.html")
}

go mod init crawler-viz
go get github.com/gocolly/colly
go get github.com/go-echarts/go-echarts/v2

go run main.go