p209327/nlnieuws
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update

Browse Source
This commit is contained in:
Peter Kleiweg
2026-06-05 16:05:46 +02:00
parent efa301cc4a
commit a76fa21584
12 changed files with 645 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

387
Volkskrant/cmd/volkskrant/volkskrant.go Normal file
View File

@@ -0,0 +1,387 @@
package main
import (
e "codeberg.org/pebbe/errors"
"github.com/jbowtie/gokogiri"
u "git.web.rug.nl/p209327/nlnieuws/internal/util"
//"encoding/json"
"encoding/xml"
"fmt"
//"html"
"io"
"net/http"
"net/url"
"os"
"strings"
"time"
)
type Rss struct {
XMLName xml.Name `xml:"rss"`
Items []ItemT `xml:"channel>item"`
}
type ItemT struct {
PubDate string `xml:"pubDate"`
UnixTime int64 `xml:"unixTime"`
Guid string `xml:"guid"`
Link string `xml:"link"`
Title string `xml:"title"`
Data []byte `xml:",innerxml"`
}
/*
type GraphT struct {
Graph []map[string]any `json:"@graph"`
}
*/
var (
p = e.PanicErr
w = e.WarnErr
agent = "AhrefsBot/7.0"
)
func exists(filename string) bool {
_, err := os.Stat(filename)
return err == nil
}
func fileDate(filename string) string {
b, err := os.ReadFile(filename)
if err != nil {
return ""
}
s := string(b)
i1 := strings.Index(s, "<unixTime>") + 10
i2 := strings.Index(s, "</unixTime>")
if i2 < i1 {
return ""
}
return s[i1:i2]
}
func main() {
defer func() {
if e.Panicked {
_ = recover()
os.Exit(1)
}
}()
myLock := "/net/corpora/nlnieuws/Volkskrant/lock"
u.MkLock(myLock)
defer func() {
_ = os.Remove(myLock)
}()
req, err := http.NewRequest("GET", "https://www.volkskrant.nl/rss.xml", 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/Volkskrant/%d/%02d/%02d", t.Year(), int(t.Month()), t.Day())
if exists(dirname + "/lock") {
continue
}
basename := item.Guid
filename := dirname + "/" + url.PathEscape(basename)
ts := fmt.Sprintf("%d", t.Unix())
needUpdate := fileDate(filename+".xml") != ts
p(os.MkdirAll(dirname, 0777))
func() {
var ok bool
defer func() {
if e.Panicked {
fmt.Fprintln(os.Stderr, "----", filename)
fmt.Fprintln(os.Stderr, "----", item.Link)
}
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))
ok = doArticle(filename, item.Link, item.Title, t, needUpdate)
}()
}
}
func doArticle(filename string, url string, title string, timestamp time.Time, needUpdate bool) (ok bool) {
if exists(filename + ".skip") {
return true
}
if needUpdate {
_ = os.Remove(filename + ".err")
_ = os.Remove(filename + ".html")
// _ = os.Remove(filename + ".json")
_ = os.Remove(filename + ".txt")
} else {
if exists(filename + ".txt") {
return true
}
}
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())
body = u.HtmlFix(body)
doc, err := gokogiri.ParseHtml(body)
p(err)
/*
s := string(body)
ok = true
i1 := strings.Index(s, `<script type="application/ld+json"`)
if i1 < 0 {
ok = false
} else {
i1 += strings.Index(s[i1:], `>`) + 1
i2 := i1 + strings.Index(s[i1:], `</script>`)
if i2 < i1 {
ok = false
} else {
s = html.UnescapeString(s[i1:i2])
}
}
if !ok {
_ = w(fmt.Errorf("script jsonld not found: %s", url))
fp, err := os.Create(filename + ".err")
p(err)
p(fmt.Fprintf(fp, "script jsonld not found: %s\n", url))
p(fp.Close())
p(os.Chtimes(filename+".err", timestamp, timestamp))
fp, err = os.Create(filename + ".html")
p(err)
p(fp.Write(body))
p(fp.Close())
p(os.Chtimes(filename+".html", timestamp, timestamp))
return false
}
var graph GraphT
p(json.Unmarshal([]byte(s), &graph))
for _, g := range graph.Graph {
t := g["@type"]
switch v := t.(type) {
case string:
if v == "NewsArticle" {
b, err := json.Marshal(g)
p(err)
s = string(b)
}
}
}
fp, err := os.Create(filename + ".json")
p(err)
p(fp.WriteString(s))
p(fp.Close())
p(os.Chtimes(filename+".json", timestamp, timestamp))
*/
root := doc.Root()
articles, err := root.Search(`//article[@id="article-content"]`)
p(err)
if len(articles) == 0 {
_ = w(fmt.Errorf("empty: %s", url))
fp, err := os.Create(filename + ".err")
p(err)
p(fmt.Fprintf(fp, "empty: %s\n", url))
p(fp.Close())
p(os.Chtimes(filename+".err", timestamp, timestamp))
fp, err = os.Create(filename + ".html")
p(err)
p(fp.Write(body))
p(fp.Close())
p(os.Chtimes(filename+".html", timestamp, timestamp))
return false
}
article := articles[0]
live, err := article.Search(`.//*[@data-test-id="live-blog-label"]`)
p(err)
if len(live) > 0 {
fp, err := os.Create(filename + ".skip")
p(fp.WriteString("liveblog\n"))
p(err)
p(os.Chtimes(filename+".skip", timestamp, timestamp))
return true
}
headers, err := article.Search(`.//header`)
p(err)
if len(headers) == 0 {
_ = w(fmt.Errorf("no header: %s", url))
fp, err := os.Create(filename + ".err")
p(err)
p(fmt.Fprintf(fp, "no elements: %s\n", url))
p(fp.Close())
p(os.Chtimes(filename+".err", timestamp, timestamp))
fp, err = os.Create(filename + ".html")
p(err)
p(fp.Write(body))
p(fp.Close())
p(os.Chtimes(filename+".html", timestamp, timestamp))
return false
}
header := headers[0]
isOpinie := false
tags := make([]string, 0)
ell, err := header.Search(`.//*[@data-test-id="article-label"]`)
p(err)
if len(ell) == 0 {
_ = w(fmt.Errorf("no labels: %s", url))
}
for _, el := range ell {
s := strings.TrimSpace(el.Content())
if s != "" && s != "Nieuws" {
tags = append(tags, s)
}
if strings.ToLower(s) == "opinie" {
isOpinie = true
}
}
pars := make([]string, 0)
found := false
ell, err = header.Search(`.//*[@data-test-id="header-intro"]`)
p(err)
for _, el := range ell {
s := strings.TrimSpace(el.Content())
if s != "" {
pars = append(pars, s)
found = true
}
}
if !found && !isOpinie {
_ = w(fmt.Errorf("no intro: %s", url))
}
specials, err := article.Search(`.//section//aside | .//section//figure | .//section//b`)
p(err)
for _, special := range specials {
special.Remove()
}
ell, err = article.Search(`.//section//*[@data-article-element-index]`)
p(err)
if len(ell) == 0 {
_ = w(fmt.Errorf("no elements: %s", url))
fp, err := os.Create(filename + ".err")
p(err)
p(fmt.Fprintf(fp, "no elements: %s\n", url))
p(fp.Close())
p(os.Chtimes(filename+".err", timestamp, timestamp))
fp, err = os.Create(filename + ".html")
p(err)
p(fp.Write(body))
p(fp.Close())
p(os.Chtimes(filename+".html", timestamp, timestamp))
return false
}
found = false
for _, el := range ell {
s := strings.TrimSpace(el.Content())
if s != "" {
pars = append(pars, s)
found = true
}
}
if !found {
_ = w(fmt.Errorf("no text, skipping: %s", url))
fp, err := os.Create(filename + ".skip")
p(fp.WriteString(url + "\n"))
p(err)
p(os.Chtimes(filename+".skip", timestamp, timestamp))
fp, err = os.Create(filename + ".html")
p(err)
p(fp.Write(body))
p(fp.Close())
p(os.Chtimes(filename+".html", timestamp, timestamp))
return true
}
fp, err := os.Create(filename + ".txt")
p(err)
if len(tags) == 0 {
p(fmt.Fprintln(fp, "##META text tag ="))
} else {
for _, tag := range tags {
p(fmt.Fprintf(fp, "##META text tag = %s\n", u.FixSpace(tag)))
}
}
p(fp.WriteString(u.AddEnd(u.FixSpace(title))))
for _, par := range pars {
p(fp.WriteString(u.AddEnd(u.FixSpace(par))))
}
p(fp.Close())
p(os.Chtimes(filename+".txt", timestamp, timestamp))
return true
}