mirror of
https://gitlab.com/news-flash/article_scraper.git
synced 2025-07-07 08:05:31 +02:00
no need for head
This commit is contained in:
parent
8c7cdacd26
commit
6116ba38ae
1 changed files with 0 additions and 16 deletions
|
@ -122,8 +122,6 @@ impl FullTextParser {
|
||||||
Node::new("article", None, &document).map_err(|()| FullTextParserError::Xml)?;
|
Node::new("article", None, &document).map_err(|()| FullTextParserError::Xml)?;
|
||||||
document.set_root_element(&root);
|
document.set_root_element(&root);
|
||||||
|
|
||||||
Self::generate_head(&mut root, &document)?;
|
|
||||||
|
|
||||||
for page_html in pages {
|
for page_html in pages {
|
||||||
self.parse_page(&mut article, &page_html, &mut root, config, global_config)?;
|
self.parse_page(&mut article, &page_html, &mut root, config, global_config)?;
|
||||||
}
|
}
|
||||||
|
@ -1091,20 +1089,6 @@ impl FullTextParser {
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
fn generate_head(root: &mut Node, document: &Document) -> Result<(), FullTextParserError> {
|
|
||||||
if let Ok(mut head_node) = Node::new("head", None, document) {
|
|
||||||
if let Ok(()) = root.add_prev_sibling(&mut head_node) {
|
|
||||||
if let Ok(mut meta) = head_node.new_child(None, "meta") {
|
|
||||||
if meta.set_property("charset", "utf-8").is_ok() {
|
|
||||||
return Ok(());
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
Err(FullTextParserError::Xml)
|
|
||||||
}
|
|
||||||
|
|
||||||
fn prevent_self_closing_tags(context: &Context) -> Result<(), FullTextParserError> {
|
fn prevent_self_closing_tags(context: &Context) -> Result<(), FullTextParserError> {
|
||||||
// search document for empty tags and add a empty text node as child
|
// search document for empty tags and add a empty text node as child
|
||||||
// this prevents libxml from self closing non void elements such as iframe
|
// this prevents libxml from self closing non void elements such as iframe
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue