kemitix/paperoni
Archived
1
0
Fork 0
Code Issues 4 Pull requests 3 Projects Releases Packages Wiki Activity

Factor out text extraction into extractor module

Browse source
This commit is contained in:
Kenneth Gitere 2020-05-01 16:17:59 +03:00
parent 4527fb07d9
commit b5336e078d
2 changed files with 189 additions and 89 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

181
src/extractor.rs Normal file
View file

@ -0,0 +1,181 @@
use std::fs::File;
use kuchiki::{traits::*, NodeRef};
pub struct Extractor {
pub root_node: NodeRef,
}
impl Extractor {
/// Create a new instance of an HTML extractor given an HTML string
pub fn from_html(html_str: &str) -> Self {
Extractor {
root_node: kuchiki::parse_html().one(html_str),
}
}
/// Extract the value of an attribute
fn extract_attr_val<T: Fn(&str) -> U, U>(
&self,
css_selector: &str,
attr_target: &str,
mapper: T,
) -> Option<U> {
self.root_node
.select_first(css_selector)
.ok()
.and_then(|data| data.attributes.borrow().get(attr_target).map(mapper))
}
fn extract_inner_text(&self, css_selector: &str) -> Option<String> {
let node_ref = self.root_node.select_first(css_selector).ok()?;
extract_text_from_node(node_ref.as_node())
}
pub fn extract_content(&self) {
// Extract the useful parts of the head section
let author: Option<String> =
self.extract_attr_val("meta[name='author']", "content", |author| {
author.to_string()
});
let description =
self.extract_attr_val("meta[name='description']", "content", |description| {
description.to_string()
});
let tags = self.extract_attr_val("meta[name='keywords']", "content", |tags| {
tags.split(",")
.map(|tag| tag.trim().to_string())
.collect::<Vec<String>>()
});
let title = self.extract_inner_text("title").unwrap_or("".to_string());
let lang = self
.extract_attr_val("html", "lang", |lang| lang.to_string())
.unwrap_or("en".to_string());
let meta_attrs = MetaAttr::new(author, description, lang, tags, title);
dbg!(meta_attrs);
let article_ref = self.root_node.select_first("article").unwrap();
let mut out_file = File::create("out.html").expect("Can't make file");
for node_ref in article_ref.as_node().descendants() {
match node_ref.data() {
kuchiki::NodeData::Element(..) | kuchiki::NodeData::Text(..) => (),
_ => node_ref.detach(),
}
}
println!("Saving to file");
for node_ref in article_ref.as_node().children() {
match node_ref.data() {
kuchiki::NodeData::Element(_) => {
node_ref
.serialize(&mut out_file)
.expect("Serialization failed");
}
_ => (),
}
}
}
}
fn extract_text_from_node(node: &NodeRef) -> Option<String> {
node.first_child()
.map(|child_ref| child_ref.text_contents())
}
#[derive(Debug)]
pub struct MetaAttr {
author: Option<String>,
description: Option<String>,
language: String,
tags: Option<Vec<String>>,
title: String,
}
impl MetaAttr {
pub fn new(
author: Option<String>,
description: Option<String>,
language: String,
tags: Option<Vec<String>>,
title: String,
) -> Self {
MetaAttr {
author,
description,
language,
tags,
title,
}
}
}
#[cfg(test)]
mod test {
use super::*;
const TEST_HTML: &'static str = r#"
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="description" content="A sample document">
<meta name="keywords" content="test,Rust">
<meta name="author" content="Paperoni">
<title>Testing Paperoni</title>
</head>
<body>
<header>
<!-- Unimportant information -->
<h1>Testing Paperoni</h1>
</header>
<article>
<h1>Starting out</h1>
<p>Some Lorem Ipsum text here</p>
<p>Observe this picture</p>
<img src="./img.jpg" alt="Random image">
</article>
<footer>
<p>Made in HTML</p>
</footer>
</body>
</html>
"#;
#[test]
fn test_extract_attr_val() {
let extractor = Extractor::from_html(TEST_HTML);
let ext_author =
extractor.extract_attr_val("meta[name='author']", "content", |val| val.to_string());
assert!(ext_author.is_some());
assert_eq!("Paperoni", &ext_author.unwrap());
let ext_author =
extractor.extract_attr_val("meta[name='invalid-name']", "content", |val| {
val.to_string()
});
assert!(ext_author.is_none());
let lang_attr = extractor.extract_attr_val("html", "lang", |lang| lang.to_string());
assert!(lang_attr.is_some());
assert_eq!("en".to_string(), lang_attr.unwrap());
}
#[test]
fn test_extract_inner_text() {
let extractor = Extractor::from_html(TEST_HTML);
let title_text = extractor.extract_inner_text("title");
assert!(title_text.is_some());
assert_eq!("Testing Paperoni".to_string(), title_text.unwrap());
let title_text = extractor.extract_inner_text("titln");
assert!(title_text.is_none());
}
#[test]
fn test_extract_text() {
let extractor = Extractor::from_html(TEST_HTML);
let h1_node = extractor.root_node.select_first("h1").unwrap();
let h1_text = extract_text_from_node(h1_node.as_node());
assert!(h1_text.is_some());
assert_eq!("Testing Paperoni".to_string(), h1_text.unwrap());
}
}

97
src/main.rs
View file

@ -1,6 +1,8 @@
use async_std::task; use async_std::task;
use kuchiki::traits::*;
mod extractor;
use extractor::Extractor;
fn main() { fn main() {
task::block_on(async { task::block_on(async {
let urls = vec![ let urls = vec![
@ -9,10 +11,12 @@ fn main() {
"https://saveandrun.com/posts/2020-01-08-working-with-git.html", "https://saveandrun.com/posts/2020-01-08-working-with-git.html",
"https://blog.hipstermojo.xyz/posts/redis-orm-preface/", "https://blog.hipstermojo.xyz/posts/redis-orm-preface/",
"https://vuejsdevelopers.com/2020/03/31/vue-js-form-composition-api/?utm_campaign=xl5&utm_medium=article&utm_source=vuejsnews#adding-validators", "https://vuejsdevelopers.com/2020/03/31/vue-js-form-composition-api/?utm_campaign=xl5&utm_medium=article&utm_source=vuejsnews#adding-validators",
"https://medium.com/typeforms-engineering-blog/the-beginners-guide-to-oauth-dancing-4b8f3666de10" "https://medium.com/typeforms-engineering-blog/the-beginners-guide-to-oauth-dancing-4b8f3666de10",
"https://dev.to/steelwolf180/full-stack-development-in-django-3768"
]; ];
let html = fetch_url(urls[0]).await; let html = fetch_url(urls[4]).await;
extract_content(html); let extractor = Extractor::from_html(&html);
extractor.extract_content();
}); });
} }
@ -25,88 +29,3 @@ async fn fetch_url(url: &str) -> String {
.await .await
.expect("Unable to fetch URL") .expect("Unable to fetch URL")
} }
fn extract_content(html_str: String) {
let document = kuchiki::parse_html().one(html_str);
let author: Option<String> =
document
.select_first("meta[name='author']")
.ok()
.and_then(|data| {
data.attributes
.borrow()
.get("content")
.map(|name| name.to_string())
});
let description = document
.select_first("meta[name='description']")
.ok()
.and_then(|data| {
data.attributes
.borrow()
.get("content")
.map(|description| description.to_string())
});
let tags = document
.select_first("meta[name='keywords']")
.ok()
.and_then(|data| {
data.attributes.borrow().get("content").map(|tags| {
tags.split(",")
.map(|tag_str| tag_str.trim().to_string())
.collect::<Vec<String>>()
})
});
let title = if let Some(title_node) = document.select_first("title").ok() {
title_node
.as_node()
.first_child()
.and_then(|text_node| {
text_node
.as_text()
.map(|text_ref| text_ref.borrow().to_string())
})
.unwrap_or("".to_string())
} else {
"".to_string()
};
let lang = document
.select_first("html")
.ok()
.and_then(|data| {
data.attributes
.borrow()
.get("lang")
.map(|val| val.to_string())
})
.unwrap_or("en".to_string());
let meta_attrs = MetaAttr::new(author, description, lang, tags, title);
dbg!(meta_attrs);
}
#[derive(Debug)]
struct MetaAttr {
author: Option<String>,
description: Option<String>,
language: String,
tags: Option<Vec<String>>,
title: String,
}
impl MetaAttr {
pub fn new(
author: Option<String>,
description: Option<String>,
language: String,
tags: Option<Vec<String>>,
title: String,
) -> Self {
MetaAttr {
author,
description,
language,
tags,
title,
}
}
}