switch html-py-ever example from kuchiki to scraper

examples/html-py-ever/Cargo.toml

@@ -5,9 +5,8 @@ authors = ["konstin <konstin@mailbox.org>"]
 edition = "2021"
 
 [dependencies]
-kuchiki = "0.8.0"
+scraper = "0.22"
 pyo3 = "0.27"
-tendril = "0.4.3"
 
 [lib]
 name = "html_py_ever"

examples/html-py-ever/README.md

@@ -1,6 +1,6 @@
 # html-py-ever
 
-Demoing how to use [html5ever](https://github.com/servo/html5ever) through [kuchiki](https://github.com/kuchiki-rs/kuchiki) to speed up html parsing and css-selecting.
+Demoing how to use [html5ever](https://github.com/servo/html5ever) through [scraper](https://github.com/rust-scraper/scraper) to speed up html parsing and css-selecting.
 
 ## Usage
 

examples/html-py-ever/noxfile.py

@@ -12,3 +12,12 @@ def test(session: nox.Session):
     session.install("--no-build-isolation", ".")
     # Test Python package
     session.run("pytest", *session.posargs)
+
+
+@nox.session()
+def bench(session: nox.Session):
+    session.install(SETUPTOOLS_RUST, "pytest", "pytest-benchmark", "beautifulsoup4")
+    # Ensure build uses version of setuptools-rust under development
+    session.install("--no-build-isolation", ".")
+    # Test Python package
+    session.run("pytest", "--benchmark-enable", *session.posargs)

examples/html-py-ever/pyproject.toml

@@ -36,3 +36,6 @@ target = "html_py_ever.html_py_ever"
 # ^-- The last part of the name (e.g. "html_py_ever") has to match lib.name in Cargo.toml,
 #     but you can add a prefix to nest it inside of a Python package.
 # See reference for RustExtension in https://setuptools-rust.readthedocs.io/en/latest/reference.html
+#
+[tool.pytest.ini_options]
+addopts = "--benchmark-disable"

examples/html-py-ever/rust/lib.rs

@@ -3,37 +3,43 @@ use pyo3::prelude::*;
 #[pymodule]
 mod html_py_ever {
     use pyo3::prelude::*;
+    use scraper::{Html, Selector};
+    use std::fs;
     use std::io::Read;
     use std::path::Path;
-    use tendril::stream::TendrilSink;
 
     /// A parsed html document
     #[pyclass(unsendable)]
     struct Document {
-        node: kuchiki::NodeRef,
+        html: Html,
     }
 
     #[pymethods]
     impl Document {
         /// Returns the selected elements as strings
-        fn select(&self, selector: &str) -> Vec<String> {
-            self.node
-                .select(selector)
-                .unwrap()
-                .map(|css_match| css_match.text_contents())
-                .collect()
+        fn select(&self, selector: &str) -> PyResult<Vec<String>> {
+            let selector = Selector::parse(selector)
+                .map_err(|e| PyErr::new::<pyo3::exceptions::PyValueError, _>(format!("{e:?}")))?;
+            Ok(self
+                .html
+                .select(&selector)
+                .map(|element| element.html())
+                .collect())
         }
     }
 
     impl Document {
         fn from_reader(reader: &mut impl Read) -> PyResult<Document> {
-            let node = kuchiki::parse_html().from_utf8().read_from(reader)?;
-            Ok(Document { node })
+            let mut html_string = String::new();
+            reader.read_to_string(&mut html_string)?;
+            let html = Html::parse_document(&html_string);
+            Ok(Document { html })
         }
 
         fn from_file(path: &Path) -> PyResult<Document> {
-            let node = kuchiki::parse_html().from_utf8().from_file(path)?;
-            Ok(Document { node })
+            let html_string = fs::read_to_string(path)?;
+            let html = Html::parse_document(&html_string);
+            Ok(Document { html })
         }
     }
 

examples/html-py-ever/rust/main.rs

@@ -1,11 +1,11 @@
 //! Pure rust version for comparing with python based calls
 
-use kuchiki;
+use scraper::{Html, Selector};
 
 use std::env;
+use std::fs;
 use std::path::PathBuf;
 use std::time::Instant;
-use tendril::stream::TendrilSink;
 
 fn main() {
     let path = PathBuf::from(
@@ -15,13 +15,15 @@ fn main() {
     );
 
     let now = Instant::now();
-    let document = kuchiki::parse_html().from_utf8().from_file(&path).unwrap();
+    let html_string = fs::read_to_string(&path).unwrap();
+    let document = Html::parse_document(&html_string);
     println!("{:?}", now.elapsed());
+
     let now2 = Instant::now();
+    let selector = Selector::parse("a[href]").unwrap();
     let links: Vec<String> = document
-        .select("a[href]")
-        .unwrap()
-        .map(|css_match| css_match.text_contents())
+        .select(&selector)
+        .map(|element| element.text().collect())
         .collect();
     println!("{} {:?}", links.len(), now2.elapsed());
 }

examples/html-py-ever/tests/test_selector.py

@@ -1,12 +1,11 @@
 #!/usr/bin/env python
-from glob import glob
 import os
+from glob import glob
 
 import html_py_ever
 import pytest
 from bs4 import BeautifulSoup
 
-
 HTML_FILES = glob(os.path.join(os.path.dirname(__file__), "*.html"))