misc-mirrors/article_scraper
1
0
Fork 0
mirror of https://gitlab.com/news-flash/article_scraper.git synced 2025-07-07 16:15:32 +02:00
Code Releases Activity

tmp: dont strip scripts

Browse source
This commit is contained in:
Jan Lukas Gernert 2020-01-27 16:36:32 +01:00
parent 8247defe54
commit 90d45ea6b3
1 changed files with 20 additions and 21 deletions
Download patch file Download diff file Expand all files Collapse all files

41
src/lib.rs
View file

@ -115,10 +115,10 @@ impl ArticleScraper {
return Err(error)
}
if let Err(error) = ArticleScraper::eliminate_noscript_tag(&context) {
error!("Eliminating <noscript> tag failed - {}", error);
return Err(error)
}
// if let Err(error) = ArticleScraper::eliminate_noscript_tag(&context) {
// error!("Eliminating <noscript> tag failed - {}", error);
// return Err(error)
// }
if download_images {
if let Err(error) = self.image_downloader.download_images_from_context(&context).await {
@ -572,7 +572,7 @@ impl ArticleScraper {
let _ = ArticleScraper::strip_node(&context, &String::from("//*[contains(@style,'display:none')]"));
// strip all scripts
let _ = ArticleScraper::strip_node(&context, &String::from("//script"));
//let _ = ArticleScraper::strip_node(&context, &String::from("//script"));
// strip all comments
let _ = ArticleScraper::strip_node(&context, &String::from("//comment()"));
@ -706,25 +706,24 @@ impl ArticleScraper {
Ok(())
}
fn eliminate_noscript_tag(context: &Context) -> Result<(), ScraperError> {
// fn eliminate_noscript_tag(context: &Context) -> Result<(), ScraperError> {
// let xpath = "//noscript";
// let node_vec = Self::evaluate_xpath(context, xpath, false)?;
let xpath = "//noscript";
let node_vec = Self::evaluate_xpath(context, xpath, false)?;
// for mut node in node_vec {
// if let Some(mut parent) = node.get_parent() {
// node.unlink();
// let children = node.get_child_nodes();
// for mut child in children {
// child.unlink();
// let _ = parent.add_child(&mut child);
// }
// }
for mut node in node_vec {
if let Some(mut parent) = node.get_parent() {
node.unlink();
let children = node.get_child_nodes();
for mut child in children {
child.unlink();
let _ = parent.add_child(&mut child);
}
}
// }
}
Ok(())
}
// Ok(())
// }
}