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

fmt

Browse source
This commit is contained in:
Jan Lukas Gernert 2022-10-07 09:32:39 +02:00
parent 69659da983
commit 7b205e8e27
1 changed files with 5 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

12
src/lib.rs
View file

@ -168,8 +168,7 @@ impl ArticleScraper {
xpath_single_page_link xpath_single_page_link
); );
if let Some(single_page_url) = Util::find_page_url(&xpath_ctx, xpath_single_page_link) if let Some(single_page_url) = Util::find_page_url(&xpath_ctx, xpath_single_page_link) {
{
// parse again with single page url // parse again with single page url
debug!("Single page link found '{}'", single_page_url); debug!("Single page link found '{}'", single_page_url);
@ -574,15 +573,14 @@ impl ArticleScraper {
// strip elements using Readability.com and Instapaper.com ignore class names // strip elements using Readability.com and Instapaper.com ignore class names
// .entry-unrelated and .instapaper_ignore // .entry-unrelated and .instapaper_ignore
// See http://blog.instapaper.com/post/730281947 // See http://blog.instapaper.com/post/730281947
let _ = Util::strip_node(context,
"//*[contains(@class,' entry-unrelated ') or contains(@class,' instapaper_ignore ')]");
// strip elements that contain style="display: none;"
let _ = Util::strip_node( let _ = Util::strip_node(
context, context,
"//*[contains(@style,'display:none')]", "//*[contains(@class,' entry-unrelated ') or contains(@class,' instapaper_ignore ')]",
); );
// strip elements that contain style="display: none;"
let _ = Util::strip_node(context, "//*[contains(@style,'display:none')]");
// strip all comments // strip all comments
let _ = Util::strip_node(context, "//comment()"); let _ = Util::strip_node(context, "//comment()");