nlnieuws/Amsterdam/amsterdam.go

package main

import (
	e "codeberg.org/pebbe/errors"
	"github.com/jbowtie/gokogiri"

	"encoding/xml"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"os"
	"strings"
	"time"
)

type Rss struct {
	XMLName xml.Name `xml:"rss"`
	Items   []ItemT  `xml:"channel>item"`
}

type ItemT struct {
	Title    string `xml:"title"`
	PubDate  string `xml:"pubDate"`
	UnixTime int64  `xml:"unixTime"`
	Guid     string `xml:"guid"`
	Link     string `xml:"link"`
	Data     []byte `xml:",innerxml"`
}

var (
	p     = e.PanicErr
	agent = "AhrefsBot/7.0"
	// agent = "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/145.0.0.0 Safari/537.36"
)

func exists(filename string) bool {
	_, err := os.Stat(filename)
	return err == nil
}

func main() {
	defer func() {
		if e.Panicked {
			_ = recover()
			os.Exit(1)
		}
	}()

	req, err := http.NewRequest("GET", "https://www.amsterdam.nl/nieuws/nieuwsoverzicht/?rss=true", nil)
	p(err)
	req.Header.Set("User-Agent", agent)

	client := &http.Client{}
	resp, err := client.Do(req)
	p(err)
	body, err := io.ReadAll(resp.Body)
	p(err)
	p(resp.Body.Close())

	var rss Rss
	p(xml.Unmarshal(body, &rss))

	if len(rss.Items) == 0 {
		p(fmt.Errorf("len(rss.Items) == 0"))
	}

	for _, item := range rss.Items {
		t, err := time.Parse(time.RFC1123Z, item.PubDate)
		if err != nil {
			t, err = time.Parse(time.RFC1123, item.PubDate)
		}
		p(err)
		dirname := fmt.Sprintf("/net/corpora/nlnieuws/amsterdam/%d/%02d", t.Year(), int(t.Month()))
		if exists(dirname + "/lock") {
			continue
		}
		filename := dirname + "/" + url.PathEscape(strings.TrimPrefix(item.Guid, "https://www.amsterdam.nl/nieuws/"))
		p(os.MkdirAll(dirname, 0777))
		func() {
			var ok bool
			defer func() {
				if e.Panicked {
					fmt.Fprintln(os.Stderr, "----", filename, "----")
				}
				if !ok {
					_ = os.Remove(filename + ".xml")
				}
			}()
			fp, err := os.Create(filename + ".xml")
			p(err)
			p(fp.WriteString("<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<item>\n"))
			p(fmt.Fprintf(fp, "<unixTime>%d</unixTime>", t.Unix()))
			p(fp.Write(item.Data))
			p(fp.WriteString("</item>\n"))
			p(fp.Close())
			p(os.Chtimes(filename+".xml", t, t))
			doArticle(filename, item.Link, item.Title, t)
			ok = true
		}()
	}
}

func doArticle(filename string, url string, title string, timestamp time.Time) {
	if exists(filename + ".txt") {
		return
	}
	time.Sleep(2 * time.Second)

	req, err := http.NewRequest("GET", url, nil)
	p(err)
	req.Header.Set("User-Agent", agent)

	client := &http.Client{}
	resp, err := client.Do(req)
	p(err)
	body, err := io.ReadAll(resp.Body)
	p(err)
	p(resp.Body.Close())

	doc, err := gokogiri.ParseHtml(body)
	p(err)

	root := doc.Root()

	fp, err := os.Create(filename + ".txt")
	p(err)

	p(fp.WriteString(addEnd(title)))

	count := 0

	pp, err := root.Search(`//div[@id="zone_intro"]//div[contains(@class, "inleiding")]/p`)
	p(err)
	for _, p1 := range pp {
		p(fp.WriteString(addEnd(p1.Content())))
		count++
	}

	ell, err := root.Search(`//div[@id="zone_content"]//div[contains(@class, "tekst")]/child::*`)
	p(err)
	for _, el := range ell {
		if n := el.Name(); n == "p" || n == "h3" {
			p(fp.WriteString(addEnd(el.Content())))
			count++
		}
	}

	p(fp.Close())

	p(os.Chtimes(filename+".txt", timestamp, timestamp))

	if count == 0 {
		fp, err := os.Create(filename + ".debug.html")
		p(err)
		p(fp.Write(body))
		p(fp.Close())
		p(os.Chtimes(filename+".debug.html", timestamp, timestamp))
	}
}

func addEnd(s string) string {
	s = strings.TrimSpace(s)
	n := len(s)
	if n == 0 {
		return ""
	}
	if n > 0 {
		if strings.ContainsAny(s[n-1:], ".!?") {
			return s + "\n"
		}
	}
	if n > 1 {
		s2 := s[n-2:]
		if s2 == `."` || s2 == `!"` || s2 == `?"` || s2 == `.'` || s2 == `!'` || s2 == `?'` {
			return s + "\n"
		}
	}
	return s + ".\n"
}