summaryrefslogtreecommitdiff
path: root/lex-fetch/de
diff options
context:
space:
mode:
authorMartin Fischer <martin@push-f.com>2025-04-08 19:25:36 +0200
committerMartin Fischer <martin@push-f.com>2025-04-14 07:04:45 +0200
commite29d27533725819ec3f6d05a27048d3d2627b53e (patch)
tree5afba50408b25179edb4ea6445acfe1d3e051488 /lex-fetch/de
parent96236c9d80cea2d6ba83591a7d08a8cc096fd8d3 (diff)
refactor: port fetchers to Go
* Austria: upgraded to RIS API v2.6 because v2.5 has been turned off
Diffstat (limited to 'lex-fetch/de')
-rw-r--r--lex-fetch/de/de.go66
1 files changed, 66 insertions, 0 deletions
diff --git a/lex-fetch/de/de.go b/lex-fetch/de/de.go
new file mode 100644
index 0000000..2c0a033
--- /dev/null
+++ b/lex-fetch/de/de.go
@@ -0,0 +1,66 @@
+package de
+
+import (
+ "fmt"
+ "log/slog"
+ "net/http"
+ "net/url"
+ "strings"
+
+ "github.com/antchfx/htmlquery"
+ "golang.org/x/text/encoding/charmap"
+ "golang.org/x/text/transform"
+ "push-f.com/lex-surf/internal/lex"
+ "push-f.com/lex-surf/lex-fetch/progress"
+)
+
+type Fetcher struct{}
+
+var pages = []rune("ABCDEFGHIJKLMNOPQRSTUVWYZ123456789")
+
+func (s *Fetcher) Fetch(log *slog.Logger, client *http.Client, progress *progress.Reporter) ([]lex.Law, error) {
+ progress.Total = len(pages)
+ var laws []lex.Law
+ for i := range len(pages) {
+ resp, err := client.Get(fmt.Sprintf("https://www.gesetze-im-internet.de/Teilliste_%c.html", pages[i]))
+ if err != nil {
+ return nil, fmt.Errorf("failed to fetch page: %w", err)
+ }
+ if resp.StatusCode != 200 {
+ return nil, fmt.Errorf("unexpected status code %d", resp.StatusCode)
+ }
+ reader := transform.NewReader(resp.Body, charmap.ISO8859_1.NewDecoder())
+ doc, err := htmlquery.Parse(reader)
+ if err != nil {
+ return nil, fmt.Errorf("failed to HTML parse: %w", err)
+ }
+ div := htmlquery.FindOne(doc, "//div[@id='paddingLR12']")
+ if div == nil {
+ return nil, fmt.Errorf("didn't find expected HTML div")
+ }
+ for child := range div.ChildNodes() {
+ if child.FirstChild != nil {
+ href := htmlquery.SelectAttr(child.FirstChild, "href")
+ redir, ok := strings.CutPrefix(href, "./")
+ if !ok {
+ return nil, fmt.Errorf("expected href to start with ./ but found %s", href)
+ }
+ redir, ok = strings.CutSuffix(redir, "/index.html")
+ if !ok {
+ return nil, fmt.Errorf("expected href to end with /index.html but found %s", href)
+ }
+
+ hrefUrl, _ := url.Parse(href)
+
+ laws = append(laws, lex.Law{
+ URL: resp.Request.URL.ResolveReference(hrefUrl).String(),
+ Title: strings.TrimSpace(child.FirstChild.NextSibling.NextSibling.Data),
+ Redir: redir,
+ Abbr: redir,
+ })
+ }
+ }
+ progress.ReportProgress(i + 1)
+ }
+ return laws, nil
+}