Retrieve the comments from an hacker news post

package inner

import (

	"strconv"

	"strings"

	"github.com/gocolly/colly/v2"

)

type comment struct {

	Author  string `selector:"a.hnuser" json:"author"`

	URL     string `selector:".age a[href]" attr:"href" json:"url"`

	Comment string `selector:".comment" json:"comment"`

	Replies []*comment `json:"replies"`

	depth   int

}

func main(itemID string) (interface{}, error) {

	comments := make([]*comment, 0)

	// Instantiate default collector

	c := colly.NewCollector()

	// Extract comment

	c.OnHTML(".comment-tree tr.athing", func(e *colly.HTMLElement) {

		width, err := strconv.Atoi(e.ChildAttr("td.ind img", "width"))

		if err != nil {

			return

		}

		// hackernews uses 40px spacers to indent comment replies,

		// so we have to divide the width with it to get the depth

		// of the comment

		depth := width / 40

		c := &comment{

			Replies: make([]*comment, 0),

			depth:   depth,

		}

		e.Unmarshal(c)

		c.Comment = strings.TrimSpace(c.Comment[:len(c.Comment)-5])

		if depth == 0 {

			comments = append(comments, c)

			return

		}

		parent := comments[len(comments)-1]

		// append comment to its parent

		for i := 0; i < depth-1; i++ {

			parent = parent.Replies[len(parent.Replies)-1]

		}

		parent.Replies = append(parent.Replies, c)

	})

	c.Visit("https://news.ycombinator.com/item?id=" + itemID)

	return comments, nil

}