misc-mirrors/article_scraper
1
0
Fork 0
mirror of https://gitlab.com/news-flash/article_scraper.git synced 2025-07-08 16:40:00 +02:00
Code Releases Activity

fix strip unlikely table-child & add 2 new tests

Browse source
This commit is contained in:
Jan Lukas Gernert 2023-03-26 11:54:13 +02:00
parent 873e081c33
commit 2189f527d7
8 changed files with 2579 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

6
src/full_text_parser/mod.rs
View file

@ -535,9 +535,11 @@ impl FullTextParser {
if let Some(mut parent) = node.get_parent() {
let new_node = if child_count == 1 && first_child_is_text {
let link_content = node.get_content();
Node::new_text(&link_content, document).expect("Failed to create new text node")
Node::new_text(&link_content, document)
.expect("Failed to create new text node")
} else {
let mut container = Node::new("span", None, document).expect("Failed to create new span container node");
let mut container = Node::new("span", None, document)
.expect("Failed to create new span container node");
for mut child in child_nodes.drain(..) {
child.unlink();
_ = container.add_child(&mut child);