perf(client): fix blocking and async mixture

.github/workflows/bench.yml

@@ -0,0 +1,42 @@
+name: Benches
+on:
+  push:
+    branches: [master]
+  pull_request:
+    branches: [master]
+
+env:
+  CARGO_TERM_COLOR: always
+  RUST_LOG: "off"
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+      - uses: actions/cache@v2
+        id: cache
+        with:
+          path: |
+            ~/.cargo/bin/
+            ~/.cargo/registry/index/
+            ~/.cargo/registry/cache/
+            ~/.cargo/git/db/
+            target/
+          key: ${{ runner.os }}-cargo-${{ hashFiles('**/Cargo.lock') }}
+      - name: Install Benchmark Dependencies
+        run: |
+          # install node-crawler
+          npm install -g crawler
+          # install go and deps
+          go mod init example.com/spider
+          cat go.mod
+          echo "github.com/gocolly/colly/v2 v2.1.0" >> ./go.mod
+          cat go.mod
+
+          go mod tidy
+
+          # install the local cli latest
+          cd ./spider_cli && cargo install  --path . && cd ../
+      - name: Run Benchmarks
+        run: cargo bench

.github/workflows/rust.yml

@@ -13,21 +13,17 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v3
+      - uses: actions/cache@v2
+        id: cache
+        with:
+          path: |
+            ~/.cargo/bin/
+            ~/.cargo/registry/index/
+            ~/.cargo/registry/cache/
+            ~/.cargo/git/db/
+            target/
+          key: ${{ runner.os }}-cargo-${{ hashFiles('**/Cargo.lock') }}
       - name: Build
         run: cargo build --verbose
       - name: Run tests
         run: cargo test --verbose --all-features
-      - name: Install Benchmark Dependencies
-        run: |
-          sudo apt update && sudo apt-get install build-essential
-          # install node-crawler
-          npm install crawler -g
-          # install the local cli latest
-          cd ./spider_cli && cargo install  --path . && cd ../
-          # install go and deps
-          echo "module github.com/x/y" >> go.mod
-          echo "go 1.14" >> go.mod
-          echo "require github.com/gocolly/colly/v2 v2.1.0" >> go.mod
-          go mod tidy
-      - name: Run Benchmarks
-        run: cargo bench

benches/.gitignore

@@ -4,4 +4,5 @@ node_modules
 node-crawler.js
 go-crolly.go
 go.mod
-go.sum
+go.sum
+output.txt

benches/Cargo.toml

@@ -5,7 +5,7 @@ publish = false
 edition = "2021"
 
 [dependencies]
-spider = { version = "1.5.5", path = "../spider" }
+spider = { version = "1.6.0", path = "../spider" }
 criterion = "0.3"
 
 [[bench]]

benches/crawl.rs

@@ -10,7 +10,7 @@ pub fn bench_speed(c: &mut Criterion) {
     let go_crawl_script = go_crolly::gen_crawl();
     let mut group = c.benchmark_group("crawl-speed");
 
-    group.sample_size(10).measurement_time(Duration::new(85, 0) + Duration::from_millis(500));
+    group.sample_size(10).measurement_time(Duration::new(180, 0) + Duration::from_millis(500));
     group.bench_function("Rust[spider]: with crawl 10 times", |b| b.iter(||Command::new("spider")
         .args(["--delay", "0", "--domain", "https://rsseau.fr", "crawl"])
         .output()

benches/go_crolly.rs

@@ -3,7 +3,7 @@ use std::io::{BufWriter, Write};
 
 pub fn crawl_stub() -> String {
     r#"
-    package main
+    package spider
 
     import (
         "fmt"

examples/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "spider_examples"
-version = "1.5.5"
+version = "1.6.0"
 authors = ["madeindjs <contact@rousseau-alexandre.fr>", "j-mendez <jeff@a11ywatch.com>"]
 description = "Multithreaded web crawler written in Rust."
 repository = "https://github.com/madeindjs/spider"
@@ -15,7 +15,7 @@ publish = false
 maintenance = { status = "as-is" }
 
 [dependencies.spider]
-version = "1.5.5"
+version = "1.6.0"
 path = "../spider"
 default-features = false
 

spider/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "spider"
-version = "1.5.5"
+version = "1.6.0"
 authors = ["madeindjs <contact@rousseau-alexandre.fr>", "j-mendez <jeff@a11ywatch.com>"]
 description = "Multithreaded web crawler written in Rust."
 repository = "https://github.com/madeindjs/spider"
@@ -15,7 +15,7 @@ edition = "2018"
 maintenance = { status = "as-is" }
 
 [dependencies]
-reqwest = { version = "0.11.10" }
+reqwest = { version = "0.11.10", features = ["blocking"] }
 scraper = "0.12"
 robotparser-fork = "0.10.5"
 url = "2.2"

spider/src/page.rs

@@ -91,7 +91,7 @@ fn parse_links() {
         .unwrap();
 
     let link_result = "https://choosealicense.com/";
-    let html = fetch_page_html(&link_result, &client).unwrap();
+    let html = fetch_page_html(&link_result, &client);
     let page: Page = Page::new(&link_result, &html);
 
     assert!(
@@ -111,7 +111,7 @@ fn test_abs_path() {
         .build()
         .unwrap();
     let link_result = "https://choosealicense.com/";
-    let html = fetch_page_html(&link_result, &client).unwrap();
+    let html = fetch_page_html(&link_result, &client);
     let page: Page = Page::new(&link_result, &html);
 
     assert_eq!(

spider/src/utils.rs

@@ -1,8 +1,18 @@
-pub use crate::reqwest::{Client, Error};
+pub use crate::reqwest::blocking::{Client};
+use reqwest::StatusCode;
 
-#[tokio::main]
-pub async fn fetch_page_html(url: &str, client: &Client) -> Result<String, Error> {
-    let body = client.get(url).send().await?.text().await?;
+pub fn fetch_page_html(url: &str, client: &Client) -> String {
+    let mut body = String::new();
 
-    Ok(body)
+    // silence errors for top level logging
+    match client.get(url).send() {
+        Ok(res) if res.status() == StatusCode::OK => match res.text() {
+            Ok(text) => body = text,
+            Err(_) => {},
+        },
+        Ok(_) => (),
+        Err(_) => {}
+    }
+
+    body
 }

spider/src/website.rs

@@ -39,8 +39,6 @@ pub struct Website<'a> {
     pub on_link_find_callback: fn(String) -> String,
     /// Robot.txt parser holder
     robot_file_parser: RobotFileParser<'a>,
-    // fetch client
-    client: Client,
     // ignore holding page in memory, pages will always be empty
     pub page_store_ignore: bool,
 }
@@ -50,14 +48,13 @@ impl<'a> Website<'a> {
     pub fn new(domain: &str) -> Self {
         Self {
             configuration: Configuration::new(),
-            domain: domain.to_string(),
-            links: HashSet::from([format!("{}/", domain)]),
             links_visited: HashSet::new(),
             pages: Vec::new(),
             robot_file_parser: RobotFileParser::new(&format!("{}/robots.txt", domain)), // TODO: lazy establish
+            links: HashSet::from([format!("{}/", domain)]),
             on_link_find_callback: |s| s,
-            client: Client::new(),
-            page_store_ignore: false
+            page_store_ignore: false,
+            domain: domain.to_owned(),
         }
     }
 
@@ -85,14 +82,13 @@ impl<'a> Website<'a> {
     }
 
     /// configure http client
-    pub fn configure_http_client(&mut self, user_agent: Option<String>) {
+    fn configure_http_client(&mut self, user_agent: Option<String>) -> Client {
         let mut headers = header::HeaderMap::new();
         headers.insert(CONNECTION, header::HeaderValue::from_static("keep-alive"));
 
-        self.client = Client::builder()
+        Client::builder()
             .default_headers(headers)
             .user_agent(user_agent.unwrap_or(self.configuration.user_agent.to_string()))
-            .pool_max_idle_per_host(0)
             .build()
             .expect("Failed building client.")
     }
@@ -108,7 +104,7 @@ impl<'a> Website<'a> {
     /// Start to crawl website
     pub fn crawl(&mut self) {
         self.configure_robots_parser();
-        self.configure_http_client(None);
+        let client = self.configure_http_client(None);
         let delay = self.get_delay();
         let on_link_find_callback = self.on_link_find_callback;
         let pool = self.create_thread_pool();
@@ -122,16 +118,15 @@ impl<'a> Website<'a> {
                     continue;
                 }
                 self.log(&format!("- fetch {}", &link));
-                self.links_visited.insert(link.to_string());
-
-                let thread_link = link.to_string();
+                self.links_visited.insert(String::from(link));
 
+                let link = link.clone();
                 let tx = tx.clone();
-                let cx = self.client.clone();
+                let cx = client.clone();
 
                 pool.spawn(move || {
-                    let link_result = on_link_find_callback(thread_link);
-                    let html = fetch_page_html(&link_result, &cx).unwrap_or_default();
+                    let link_result = on_link_find_callback(link);
+                    let html = fetch_page_html(&link_result, &cx);
                     let page = Page::new(&link_result, &html);
                     let links = page.links();
 
@@ -146,7 +141,7 @@ impl<'a> Website<'a> {
             rx.into_iter().for_each(|page| {
                 let (page, links) = page;
                 self.log(&format!("- parse {}", page.get_url()));
-
+                
                 new_links.extend(links);
 
                 if !self.page_store_ignore {
@@ -156,11 +151,9 @@ impl<'a> Website<'a> {
                 if self.configuration.delay > 0 {
                     thread::sleep(delay);
                 }
-
             });
 
             self.links = new_links;
-
         }
     }
 

spider_cli/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "spider_cli"
-version = "1.5.5"
+version = "1.6.0"
 authors = ["madeindjs <contact@rousseau-alexandre.fr>", "j-mendez <jeff@a11ywatch.com>"]
 description = "Multithreaded web crawler written in Rust."
 repository = "https://github.com/madeindjs/spider"
@@ -23,7 +23,7 @@ quote = "1.0.18"
 failure_derive = "0.1.8"
 
 [dependencies.spider]
-version = "1.5.5"
+version = "1.6.0"
 path = "../spider"
 default-features = false