igf/igf2018-talks/convert.go

package main

import (
	"fmt"
	"log"
	"os"
	"strings"

	"golang.org/x/net/html"
)

func main() {
	fi, err := os.Open(os.Args[1])
	if err != nil {
		log.Fatal(err)
	}

	doc, err := html.Parse(fi)
	if err != nil {
		log.Fatal(err)
	}

	content := findContent(doc)

	fmt.Println(extractText(content))
}

func findContent(n *html.Node) *html.Node {
	for _, attr := range n.Attr {
		if attr.Key == "id" && attr.Val == "content" {
			return n
		}
	}
	for c := n.FirstChild; c != nil; c = c.NextSibling {
		if found := findContent(c); found != nil {
			return found
		}
	}

	return nil
}

func clean(data string) string {
	data = strings.TrimSpace(data)
	return data
}

func extractText(n *html.Node) string {
	text := ""

	if n.Type == html.TextNode {
		return clean(n.Data)
	}

	if n.Type == html.ElementNode && n.Data == "br" {
		return "\n"
	}

	for c := n.FirstChild; c != nil; c = c.NextSibling {
		extr := extractText(c)
		if extr != "" {
			text += extr
		}
	}

	if n.Type == html.ElementNode &&
		(n.Data == "p" || n.Data == "div" || n.Data == "li") &&
		!strings.HasSuffix(text, "\n") {
		text += "\n"
	}

	return text
}

func toMap(content string) map[string]interface{} {
	return nil
}