Check command global (#683)

* Add check subcommand

* Add some brief documentation for the check subcommand

* Start working on parallel link checks

* Check all external links in Site

* Return *all* dead links in site

Cargo.lock

@@ -2162,6 +2162,7 @@ dependencies = [
  "glob 0.2.11 (registry+https://github.com/rust-lang/crates.io-index)",
  "imageproc 0.1.0",
  "library 0.1.0",
+ "link_checker 0.1.0",
  "rayon 1.0.3 (registry+https://github.com/rust-lang/crates.io-index)",
  "sass-rs 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "search 0.1.0",

components/library/src/content/page.rs

@@ -76,6 +76,8 @@ pub struct Page {
     pub lang: String,
     /// Contains all the translated version of that page
     pub translations: Vec<Key>,
+    /// Contains the external links that need to be checked
+    pub external_links: Vec<String>,
 }
 
 impl Page {
@@ -104,6 +106,7 @@ impl Page {
             reading_time: None,
             lang: String::new(),
             translations: Vec::new(),
+            external_links: Vec::new(),
         }
     }
 
@@ -263,6 +266,7 @@ impl Page {
         self.summary = res.summary_len.map(|l| res.body[0..l].to_owned());
         self.content = res.body;
         self.toc = res.toc;
+        self.external_links = res.external_links;
 
         Ok(())
     }
@@ -339,6 +343,7 @@ impl Default for Page {
             reading_time: None,
             lang: String::new(),
             translations: Vec::new(),
+            external_links: Vec::new(),
         }
     }
 }

components/library/src/content/section.rs

@@ -56,6 +56,8 @@ pub struct Section {
     pub lang: String,
     /// Contains all the translated version of that section
     pub translations: Vec<Key>,
+    /// Contains the external links that need to be checked
+    pub external_links: Vec<String>,
 }
 
 impl Section {
@@ -85,6 +87,7 @@ impl Section {
             reading_time: None,
             lang: String::new(),
             translations: Vec::new(),
+            external_links: Vec::new(),
         }
     }
 
@@ -189,6 +192,8 @@ impl Section {
         })?;
         self.content = res.body;
         self.toc = res.toc;
+        self.external_links = res.external_links;
+
         Ok(())
     }
 
@@ -255,6 +260,7 @@ impl Default for Section {
             word_count: None,
             lang: String::new(),
             translations: Vec::new(),
+            external_links: Vec::new(),
         }
     }
 }

components/rendering/src/markdown.rs

@@ -9,7 +9,6 @@ use config::highlighting::{get_highlighter, SYNTAX_SET, THEME_SET};
 use context::RenderContext;
 use errors::{Error, Result};
 use front_matter::InsertAnchor;
-use link_checker::check_url;
 use table_of_contents::{make_table_of_contents, Header};
 use utils::site::resolve_internal_link;
 use utils::vec::InsertMany;
@@ -25,6 +24,7 @@ pub struct Rendered {
     pub body: String,
     pub summary_len: Option<usize>,
     pub toc: Vec<Header>,
+    pub external_links: Vec<String>,
 }
 
 // tracks a header in a slice of pulldown-cmark events
@@ -66,7 +66,7 @@ fn is_colocated_asset_link(link: &str) -> bool {
         && !link.starts_with("mailto:")
 }
 
-fn fix_link(link_type: LinkType, link: &str, context: &RenderContext) -> Result<String> {
+fn fix_link(link_type: LinkType, link: &str, context: &RenderContext, external_links: &mut Vec<String>) -> Result<String> {
     if link_type == LinkType::Email {
         return Ok(link.to_string());
     }
@@ -83,17 +83,10 @@ fn fix_link(link_type: LinkType, link: &str, context: &RenderContext) -> Result<
         }
     } else if is_colocated_asset_link(&link) {
         format!("{}{}", context.current_page_permalink, link)
-    } else if context.config.check_external_links
-        && !link.starts_with('#')
-        && !link.starts_with("mailto:")
-    {
-        let res = check_url(&link);
-        if res.is_valid() {
-            link.to_string()
-        } else {
-            return Err(format!("Link {} is not valid: {}", link, res.message()).into());
-        }
     } else {
+        if !link.starts_with('#') && !link.starts_with("mailto:") {
+            external_links.push(link.to_owned());
+        }
         link.to_string()
     };
     Ok(result)
@@ -142,6 +135,7 @@ pub fn markdown_to_html(content: &str, context: &RenderContext) -> Result<Render
 
     let mut inserted_anchors: Vec<String> = vec![];
     let mut headers: Vec<Header> = vec![];
+    let mut external_links = Vec::new();
 
     let mut opts = Options::empty();
     let mut has_summary = false;
@@ -207,7 +201,7 @@ pub fn markdown_to_html(content: &str, context: &RenderContext) -> Result<Render
                         Event::Start(Tag::Image(link_type, src, title))
                     }
                     Event::Start(Tag::Link(link_type, link, title)) => {
-                        let fixed_link = match fix_link(link_type, &link, context) {
+                        let fixed_link = match fix_link(link_type, &link, context, &mut external_links) {
                             Ok(fixed_link) => fixed_link,
                             Err(err) => {
                                 error = Some(err);
@@ -302,6 +296,7 @@ pub fn markdown_to_html(content: &str, context: &RenderContext) -> Result<Render
             summary_len: if has_summary { html.find(CONTINUE_READING) } else { None },
             body: html,
             toc: make_table_of_contents(headers),
+            external_links: external_links,
         })
     }
 }

components/rendering/tests/markdown.rs

@@ -697,10 +697,9 @@ Some text
 }
 
 #[test]
-fn can_validate_valid_external_links() {
+fn correctly_captures_external_links() {
     let permalinks_ctx = HashMap::new();
-    let mut config = Config::default();
-    config.check_external_links = true;
+    let config = Config::default();
     let context = RenderContext::new(
         &ZOLA_TERA,
         &config,
@@ -708,58 +707,14 @@ fn can_validate_valid_external_links() {
         &permalinks_ctx,
         InsertAnchor::None,
     );
-    let res = render_content("[a link](http://google.com)", &context).unwrap();
-    assert_eq!(res.body, "<p><a href=\"http://google.com\">a link</a></p>\n");
-}
-
-#[test]
-fn can_show_error_message_for_invalid_external_links() {
-    let permalinks_ctx = HashMap::new();
-    let mut config = Config::default();
-    config.check_external_links = true;
-    let context = RenderContext::new(
-        &ZOLA_TERA,
-        &config,
-        "https://vincent.is/about/",
-        &permalinks_ctx,
-        InsertAnchor::None,
-    );
-    let res = render_content("[a link](http://google.comy)", &context);
-    assert!(res.is_err());
-    let err = res.unwrap_err();
-    assert!(format!("{}", err).contains("Link http://google.comy is not valid"));
-}
-
-#[test]
-fn doesnt_try_to_validate_email_links_mailto() {
-    let permalinks_ctx = HashMap::new();
-    let mut config = Config::default();
-    config.check_external_links = true;
-    let context = RenderContext::new(
-        &ZOLA_TERA,
-        &config,
-        "https://vincent.is/about/",
-        &permalinks_ctx,
-        InsertAnchor::None,
-    );
-    let res = render_content("Email: [foo@bar.baz](mailto:foo@bar.baz)", &context).unwrap();
-    assert_eq!(res.body, "<p>Email: <a href=\"mailto:foo@bar.baz\">foo@bar.baz</a></p>\n");
-}
-
-#[test]
-fn doesnt_try_to_validate_email_links_angled_brackets() {
-    let permalinks_ctx = HashMap::new();
-    let mut config = Config::default();
-    config.check_external_links = true;
-    let context = RenderContext::new(
-        &ZOLA_TERA,
-        &config,
-        "https://vincent.is/about/",
-        &permalinks_ctx,
-        InsertAnchor::None,
-    );
-    let res = render_content("Email: <foo@bar.baz>", &context).unwrap();
-    assert_eq!(res.body, "<p>Email: <a href=\"mailto:foo@bar.baz\">foo@bar.baz</a></p>\n");
+    let content = "
+[a link](http://google.com)
+[a link](http://google.comy)
+Email: [foo@bar.baz](mailto:foo@bar.baz)
+Email: <foo@bar.baz>
+    ";
+    let res = render_content(content, &context).unwrap();
+    assert_eq!(res.external_links, &["http://google.com".to_owned(), "http://google.comy".to_owned()]);
 }
 
 #[test]

components/site/Cargo.toml

@@ -19,6 +19,7 @@ front_matter = { path = "../front_matter" }
 search = { path = "../search" }
 imageproc = { path = "../imageproc" }
 library = { path = "../library" }
+link_checker = { path = "../link_checker" }
 
 [dev-dependencies]
 tempfile = "3"

components/site/src/lib.rs

@@ -15,6 +15,7 @@ extern crate library;
 extern crate search;
 extern crate templates;
 extern crate utils;
+extern crate link_checker;
 
 #[cfg(test)]
 extern crate tempfile;
@@ -33,7 +34,7 @@ use sass_rs::{compile_file, Options as SassOptions, OutputStyle};
 use tera::{Context, Tera};
 
 use config::{get_config, Config};
-use errors::{Error, Result};
+use errors::{Error, ErrorKind, Result};
 use front_matter::InsertAnchor;
 use library::{
     find_taxonomies, sort_actual_pages_by_date, Library, Page, Paginator, Section, Taxonomy,
@@ -42,6 +43,7 @@ use templates::{global_fns, render_redirect_template, ZOLA_TERA};
 use utils::fs::{copy_directory, create_directory, create_file, ensure_directory_exists};
 use utils::net::get_available_port;
 use utils::templates::{render_template, rewrite_theme_paths};
+use link_checker::check_url;
 
 #[derive(Debug)]
 pub struct Site {
@@ -243,9 +245,64 @@ impl Site {
         self.render_markdown()?;
         self.register_tera_global_fns();
 
+        if self.config.check_external_links {
+            self.check_external_links()?;
+        }
+
         Ok(())
     }
 
+    pub fn check_external_links(&self) -> Result<()> {
+        let library = self.library.write().expect("Get lock for check_external_links");
+        let page_links = library.pages()
+            .values()
+            .map(|p| {
+                let path = &p.file.path;
+                p.external_links.iter().map(move |l| (path.clone(), l))
+            })
+            .flatten();
+        let section_links = library.sections()
+            .values()
+            .map(|p| {
+                let path = &p.file.path;
+                p.external_links.iter().map(move |l| (path.clone(), l))
+            })
+            .flatten();
+        let all_links = page_links.chain(section_links).collect::<Vec<_>>();
+
+        // create thread pool with lots of threads so we can fetch
+        // (almost) all pages simultaneously
+        let threads = std::cmp::min(all_links.len(), 32);
+        let pool = rayon::ThreadPoolBuilder::new().num_threads(threads).build().map_err(|e| Error {
+            kind: ErrorKind::Msg(e.to_string().into()),
+            source: None,
+        })?;
+
+        let errors: Vec<_> = pool.install(|| {
+            all_links.par_iter().filter_map(|(path, link)| {
+                let res = check_url(link);
+                if res.is_valid() {
+                    None
+                } else {
+                    Some((path, res))
+                }
+            }).collect()
+        });
+
+        if errors.is_empty() {
+            Ok(())
+        } else {
+            let msg = errors.into_iter()
+                .map(|(path, check_res)| format!("Dead link in {:?}: {:?}", path, check_res))
+                .collect::<Vec<_>>()
+                .join("\n");
+            Err(Error {
+                kind: ErrorKind::Msg(msg.into()),
+                source: None,
+            })
+        }
+    }
+
     /// Insert a default index section for each language if necessary so we don't need to create
     /// a _index.md to render the index page at the root of the site
     pub fn create_default_index_sections(&mut self) -> Result<()> {

docs/content/documentation/getting-started/cli-usage.md

@@ -83,6 +83,11 @@ You can also point to another config file than `config.toml` like so - the posit
 $ zola --config config.staging.toml serve
 ```
 
+### check
+
+The check subcommand will try to build all pages just like the build command would, but without writing any of the
+results to disk. Additionally, it always checks external links regardless of the site configuration.
+
 ## Colored output
 
 Any of the three commands will emit colored output if your terminal supports it.

src/cli.rs

@@ -67,5 +67,7 @@ pub fn build_cli() -> App<'static, 'static> {
                         .takes_value(false)
                         .help("Do not start a server, just re-build project on changes")
                 ]),
+            SubCommand::with_name("check")
+                .about("Try building the project without rendering it. Checks links")
         ])
 }

src/cmd/check.rs

@@ -0,0 +1,28 @@
+use std::env;
+use std::path::PathBuf;
+
+use errors::Result;
+use site::Site;
+
+use console;
+
+pub fn check(
+    config_file: &str,
+    base_path: Option<&str>,
+    base_url: Option<&str>,
+) -> Result<()> {
+    let bp = base_path.map(PathBuf::from).unwrap_or(env::current_dir().unwrap());
+    let mut site = Site::new(bp, config_file)?;
+    // Force the checking of external links
+    site.config.check_external_links = true;
+    // Disable syntax highlighting since the results won't be used
+    // and this operation can be expensive.
+    site.config.highlight_code = false;
+    if let Some(b) = base_url {
+        site.set_base_url(b.to_string());
+    }
+    site.load()?;
+    console::notify_site_size(&site);
+    console::warn_about_ignored_pages(&site);
+    Ok(())
+}

src/cmd/mod.rs

@@ -1,7 +1,9 @@
 mod build;
 mod init;
 mod serve;
+mod check;
 
 pub use self::build::build;
 pub use self::init::create_new_project;
 pub use self::serve::serve;
+pub use self::check::check;

src/main.rs

@@ -89,6 +89,21 @@ fn main() {
                 }
             };
         }
+        ("check", Some(matches)) => {
+            console::info("Checking site...");
+            let start = Instant::now();
+            match cmd::check(
+                config_file,
+                matches.value_of("base_path"),
+                matches.value_of("base_url"),
+            ) {
+                Ok(()) => console::report_elapsed_time(start),
+                Err(e) => {
+                    console::unravel_errors("Failed to check the site", &e);
+                    ::std::process::exit(1);
+                }
+            };
+        }
         _ => unreachable!(),
     }
 }