aboutsummaryrefslogtreecommitdiff
path: root/v_windows/v/examples/web_crawler/web_crawler.v
diff options
context:
space:
mode:
Diffstat (limited to 'v_windows/v/examples/web_crawler/web_crawler.v')
-rw-r--r--v_windows/v/examples/web_crawler/web_crawler.v24
1 files changed, 24 insertions, 0 deletions
diff --git a/v_windows/v/examples/web_crawler/web_crawler.v b/v_windows/v/examples/web_crawler/web_crawler.v
new file mode 100644
index 0000000..00d4dfa
--- /dev/null
+++ b/v_windows/v/examples/web_crawler/web_crawler.v
@@ -0,0 +1,24 @@
+import net.http
+import net.html
+
+fn main() {
+ // http.fetch() sends an HTTP request to the URL with the given method and configurations.
+ config := http.FetchConfig{
+ user_agent: 'Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:88.0) Gecko/20100101 Firefox/88.0'
+ }
+ resp := http.fetch(http.FetchConfig{ ...config, url: 'https://tuicool.com' }) or {
+ println('failed to fetch data from the server')
+ return
+ }
+ // html.parse() parses and returns the DOM from the given text.
+ mut doc := html.parse(resp.text)
+ // html.DocumentObjectModel.get_tag_by_attribute_value() retrieves all the tags in the document that has the given attribute name and value.
+ tags := doc.get_tag_by_attribute_value('class', 'list_article_item')
+ for tag in tags {
+ href := tag.children[0].attributes['href'] or { panic('key not found') }
+ title := tag.children[0].attributes['title'] or { panic('key not found') }
+ println('href: $href')
+ println('title: $title')
+ println('')
+ }
+}