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

clean js-links & add new test

Browse source
This commit is contained in:
Jan Lukas Gernert 2023-03-26 11:31:59 +02:00
parent da12fcdab6
commit 873e081c33
7 changed files with 564 additions and 492 deletions
Download patch file Download diff file Expand all files Collapse all files

936
resources/tests/readability/clean-links/expected.html
View file

File diff suppressed because it is too large Load diff

2
resources/tests/readability/hukumusume/expected.html
View file

@ -107,7 +107,7 @@
おしまい おしまい
</p> </p>
<p> <p>
<a href="javascript:history.back();" onmouseout="MM_swapImgRestore()" onmouseover="MM_swapImage('Image10','','../../../gazou/pc_gazou/all/top_bana/back_logo_b.gif',1)" target="_blank"><img src="http://fakehost/gazou/pc_gazou/all/top_bana/back_logo_r.gif" alt="前のページへ戻る" name="Image10" width="175" height="32" id="Image10"></a><br> <span><img src="http://fakehost/gazou/pc_gazou/all/top_bana/back_logo_r.gif" alt="前のページへ戻る" name="Image10" width="175" height="32" id="Image10"></span><br>
<br> <br>
<br> <br>
<br> <br>

5
resources/tests/readability/js-link-replacement/expected.html Normal file
View file

@ -0,0 +1,5 @@
<article><DIV id="readability-page-1"><span>
<p>abc</p>
<p>def</p>
ghi
</span></DIV></article>

14
resources/tests/readability/js-link-replacement/source.html Normal file
View file

@ -0,0 +1,14 @@
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8"/>
<title>Replace javascript: links</title>
</head>
<body>
<a href="javascript:">
<p>abc</p>
<p>def</p>
ghi
</a>
</body>
</html>

65
src/full_text_parser/mod.rs
View file

@ -17,7 +17,7 @@ use crate::util::Util;
use encoding_rs::Encoding; use encoding_rs::Encoding;
use fingerprints::Fingerprints; use fingerprints::Fingerprints;
use libxml::parser::Parser; use libxml::parser::Parser;
use libxml::tree::{Document, Node}; use libxml::tree::{Document, Node, NodeType};
use libxml::xpath::Context; use libxml::xpath::Context;
use reqwest::header::HeaderMap; use reqwest::header::HeaderMap;
use reqwest::{Client, Url}; use reqwest::{Client, Url};
@ -180,7 +180,7 @@ impl FullTextParser {
if article.thumbnail_url.is_none() { if article.thumbnail_url.is_none() {
Self::check_for_thumbnail(&xpath_ctx, article); Self::check_for_thumbnail(&xpath_ctx, article);
} }
Self::prep_content(&xpath_ctx, config, global_config, &article.url); Self::prep_content(&xpath_ctx, config, global_config, &article.url, &document);
let found_body = Self::extract_body(&xpath_ctx, root, config, global_config)?; let found_body = Self::extract_body(&xpath_ctx, root, config, global_config)?;
if !found_body { if !found_body {
@ -198,7 +198,7 @@ impl FullTextParser {
let html = Self::download(&url, client, headers).await?; let html = Self::download(&url, client, headers).await?;
document = Self::parse_html(&html, config, global_config)?; document = Self::parse_html(&html, config, global_config)?;
xpath_ctx = Self::get_xpath_ctx(&document)?; xpath_ctx = Self::get_xpath_ctx(&document)?;
Self::prep_content(&xpath_ctx, config, global_config, &url); Self::prep_content(&xpath_ctx, config, global_config, &url, &document);
let found_body = Self::extract_body(&xpath_ctx, root, config, global_config)?; let found_body = Self::extract_body(&xpath_ctx, root, config, global_config)?;
if !found_body { if !found_body {
@ -262,7 +262,7 @@ impl FullTextParser {
let xpath_ctx = Self::get_xpath_ctx(&document)?; let xpath_ctx = Self::get_xpath_ctx(&document)?;
metadata::extract(&xpath_ctx, config, Some(global_config), article); metadata::extract(&xpath_ctx, config, Some(global_config), article);
Self::check_for_thumbnail(&xpath_ctx, article); Self::check_for_thumbnail(&xpath_ctx, article);
Self::prep_content(&xpath_ctx, config, global_config, url); Self::prep_content(&xpath_ctx, config, global_config, url, &document);
Self::extract_body(&xpath_ctx, root, config, global_config)?; Self::extract_body(&xpath_ctx, root, config, global_config)?;
Ok(()) Ok(())
@ -505,6 +505,7 @@ impl FullTextParser {
xpath: &str, xpath: &str,
attribute: &str, attribute: &str,
article_url: &url::Url, article_url: &url::Url,
document: &Document,
) -> Result<(), FullTextParserError> { ) -> Result<(), FullTextParserError> {
let node_vec = Util::evaluate_xpath(context, xpath, false)?; let node_vec = Util::evaluate_xpath(context, xpath, false)?;
for mut node in node_vec { for mut node in node_vec {
@ -514,26 +515,54 @@ impl FullTextParser {
.err() .err()
.map(|err| err == url::ParseError::RelativeUrlWithoutBase) .map(|err| err == url::ParseError::RelativeUrlWithoutBase)
.unwrap_or(false); .unwrap_or(false);
let is_javascript = trimmed_url.contains("javascript:");
let completed_url = if is_relative_url { if is_relative_url {
article_url.join(trimmed_url)? let completed_url = match article_url.join(trimmed_url) {
Ok(joined_url) => joined_url,
Err(_) => continue,
};
_ = node.set_attribute(attribute, completed_url.as_str());
} else if is_javascript {
// if the link only contains simple text content, it can be converted to a text node
let mut child_nodes = node.get_child_nodes();
let child_count = child_nodes.len();
let first_child_is_text = child_nodes
.first()
.and_then(|n| n.get_type())
.map(|t| t == NodeType::TextNode)
.unwrap_or(false);
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")
} else {
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);
}
container
};
_ = parent.replace_child_node(new_node, node);
}
} else if let Ok(parsed_url) = Url::parse(trimmed_url) {
_ = node.set_attribute(attribute, parsed_url.as_str());
} else { } else {
Url::parse(trimmed_url)? _ = node.set_attribute(attribute, trimmed_url);
}; };
node.set_attribute(attribute, completed_url.as_str())
.map_err(|_| FullTextParserError::Scrape)?;
} }
} }
Ok(()) Ok(())
} }
fn fix_urls(context: &Context, url: &Url) { fn fix_urls(context: &Context, url: &Url, document: &Document) {
_ = Self::repair_urls(context, "//img", "src", url); _ = Self::repair_urls(context, "//img", "src", url, document);
_ = Self::repair_urls(context, "//a", "src", url); _ = Self::repair_urls(context, "//a", "src", url, document);
_ = Self::repair_urls(context, "//a", "href", url); _ = Self::repair_urls(context, "//a", "href", url, document);
_ = Self::repair_urls(context, "//object", "data", url); _ = Self::repair_urls(context, "//object", "data", url, document);
_ = Self::repair_urls(context, "//iframe", "src", url); _ = Self::repair_urls(context, "//iframe", "src", url, document);
} }
fn prep_content( fn prep_content(
@ -541,6 +570,7 @@ impl FullTextParser {
config: Option<&ConfigEntry>, config: Option<&ConfigEntry>,
global_config: &ConfigEntry, global_config: &ConfigEntry,
url: &Url, url: &Url,
document: &Document,
) { ) {
// replace H1 with H2 as H1 should be only title that is displayed separately // replace H1 with H2 as H1 should be only title that is displayed separately
if let Ok(h1_nodes) = Util::evaluate_xpath(context, "//h1", false) { if let Ok(h1_nodes) = Util::evaluate_xpath(context, "//h1", false) {
@ -642,7 +672,7 @@ impl FullTextParser {
_ = Util::strip_node(context, "//link"); _ = Util::strip_node(context, "//link");
_ = Util::strip_node(context, "//aside"); _ = Util::strip_node(context, "//aside");
Self::fix_urls(context, url); Self::fix_urls(context, url, document);
} }
/** /**
@ -858,7 +888,6 @@ impl FullTextParser {
pub(crate) fn post_process_document(document: &Document) -> Result<(), FullTextParserError> { pub(crate) fn post_process_document(document: &Document) -> Result<(), FullTextParserError> {
if let Some(mut root) = document.get_root_element() { if let Some(mut root) = document.get_root_element() {
Self::simplify_nested_elements(&mut root)?; Self::simplify_nested_elements(&mut root)?;
Self::clean_attributes(&mut root)?; Self::clean_attributes(&mut root)?;
Self::remove_single_cell_tables(&mut root); Self::remove_single_cell_tables(&mut root);
Self::remove_extra_p_and_div(&mut root); Self::remove_extra_p_and_div(&mut root);

27
src/full_text_parser/readability/mod.rs
View file

@ -293,13 +293,32 @@ impl Readability {
let mut top_candidate = top_candidates.first().cloned().unwrap_or_else(|| { let mut top_candidate = top_candidates.first().cloned().unwrap_or_else(|| {
// If we still have no top candidate, just use the body as a last resort. // If we still have no top candidate, just use the body as a last resort.
// We also have to copy the body node so it is something we can modify. // We also have to copy the body node so it is something we can modify.
let mut rt = document.get_root_element().expect("doc should have root"); let mut root = document.get_root_element().expect("doc should have root");
Self::initialize_node(&mut rt, &state).expect("init should not fail"); if let Some(body) = root
.get_child_elements()
.into_iter()
.find(|n| n.get_name().to_uppercase() == "BODY")
{
root = body;
}
let mut new_top_candidate =
Node::new("DIV", None, &document).expect("can't create new node");
for mut child in root.get_child_elements().drain(..) {
child.unlink();
new_top_candidate.add_child(&mut child).unwrap();
}
root.add_child(&mut new_top_candidate).unwrap();
Self::initialize_node(&mut new_top_candidate, &state)
.expect("init should not fail");
needed_to_create_top_candidate = true; needed_to_create_top_candidate = true;
rt new_top_candidate
}); });
//Util::serialize_node(&top_candidate, "top_candidate.html"); // Util::serialize_node(&top_candidate, "top_candidate.html");
let mut alternative_candidate_ancestors = Vec::new(); let mut alternative_candidate_ancestors = Vec::new();
// Find a better top candidate node if it contains (at least three) nodes which belong to `topCandidates` array // Find a better top candidate node if it contains (at least three) nodes which belong to `topCandidates` array

7
src/full_text_parser/readability/tests.rs
View file

@ -21,7 +21,7 @@ async fn run_test(name: &str) {
let document = crate::FullTextParser::parse_html(&html, None, &empty_config).unwrap(); let document = crate::FullTextParser::parse_html(&html, None, &empty_config).unwrap();
let xpath_ctx = crate::FullTextParser::get_xpath_ctx(&document).unwrap(); let xpath_ctx = crate::FullTextParser::get_xpath_ctx(&document).unwrap();
crate::FullTextParser::prep_content(&xpath_ctx, None, &empty_config, &url); crate::FullTextParser::prep_content(&xpath_ctx, None, &empty_config, &url, &document);
let mut article = Article { let mut article = Article {
title: None, title: None,
author: None, author: None,
@ -252,6 +252,11 @@ async fn ietf_1() {
run_test("ietf-1").await run_test("ietf-1").await
} }
#[tokio::test]
async fn js_link_replacement() {
run_test("js-link-replacement").await
}
#[tokio::test] #[tokio::test]
async fn webmd_1() { async fn webmd_1() {
run_test("webmd-1").await run_test("webmd-1").await