add link_checker settings for external_level and internal_level (#1848)

* add external_level and internal_level

* remove unnecessary debug derive on LinkDef

* clarify doc comment about link check levels

* simplify link checker logging

* add missing warn prefix

* simplify link level logging, remove "Level" from linklevel variants

* remove link level config from test site

* switch back to using bail! from get_link_domain

* move console's deps to libs

* remove unnecessary reference

* calling console::error/warn directly

* emit one error, or one warning, per link checker run

* various link checker level changes

* add docs about link checker levels

* remove accidentally committed test site

* remove completed TODO

Author: mwcz

Cargo.lock

@@ -22,9 +22,7 @@ time = "0.3"
 name = "zola"
 
 [dependencies]
-atty = "0.2.11"
 clap = { version = "3", features = ["derive"] }
-termcolor = "1.0.4"
 # Below is for the serve cmd
 hyper = { version = "0.14.1", default-features = false, features = ["runtime", "server", "http2", "http1"] }
 tokio = { version = "1.0.1", default-features = false, features = ["rt", "fs", "time"] }
@@ -39,6 +37,7 @@ mime_guess = "2.0"
 
 site = { path = "components/site" }
 errors = { path = "components/errors" }
+console = { path = "components/console" }
 utils = { path = "components/utils" }
 libs = { path = "components/libs" }
 

Cargo.toml

@@ -1,10 +1,28 @@
 use serde::{Deserialize, Serialize};
 
+#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
+pub enum LinkCheckerLevel {
+    #[serde(rename = "error")]
+    Error,
+    #[serde(rename = "warn")]
+    Warn,
+}
+
+impl Default for LinkCheckerLevel {
+    fn default() -> Self {
+        Self::Error
+    }
+}
+
 #[derive(Clone, Debug, Default, PartialEq, Eq, Serialize, Deserialize)]
 #[serde(default)]
 pub struct LinkChecker {
     /// Skip link checking for these URL prefixes
     pub skip_prefixes: Vec<String>,
     /// Skip anchor checking for these URL prefixes
     pub skip_anchor_prefixes: Vec<String>,
+    /// Emit either "error" or "warn" for broken internal links (including anchor links).
+    pub internal_level: LinkCheckerLevel,
+    /// Emit either "error" or "warn" for broken external links (including anchor links).
+    pub external_level: LinkCheckerLevel,
 }

components/config/src/config/link_checker.rs

@@ -5,8 +5,8 @@ mod theme;
 use std::path::Path;
 
 pub use crate::config::{
-    languages::LanguageOptions, link_checker::LinkChecker, search::Search, slugify::Slugify,
-    taxonomies::TaxonomyConfig, Config,
+    languages::LanguageOptions, link_checker::LinkChecker, link_checker::LinkCheckerLevel,
+    search::Search, slugify::Slugify, taxonomies::TaxonomyConfig, Config,
 };
 use errors::Result;
 

components/config/src/lib.rs

@@ -0,0 +1,8 @@
+[package]
+name = "console"
+version = "0.1.0"
+edition = "2021"
+
+[dependencies]
+errors = { path = "../errors" }
+libs = { path = "../libs" }

components/console/Cargo.toml

@@ -0,0 +1,57 @@
+use std::env;
+use std::io::Write;
+
+use libs::atty;
+use libs::once_cell::sync::Lazy;
+use libs::termcolor::{Color, ColorChoice, ColorSpec, StandardStream, WriteColor};
+
+/// Termcolor color choice.
+/// We do not rely on ColorChoice::Auto behavior
+/// as the check is already performed by has_color.
+static COLOR_CHOICE: Lazy<ColorChoice> =
+    Lazy::new(|| if has_color() { ColorChoice::Always } else { ColorChoice::Never });
+
+pub fn info(message: &str) {
+    colorize(message, ColorSpec::new().set_bold(true), StandardStream::stdout(*COLOR_CHOICE));
+}
+
+pub fn warn(message: &str) {
+    colorize(
+        &format!("{}{}", "Warning: ", message),
+        ColorSpec::new().set_bold(true).set_fg(Some(Color::Yellow)),
+        StandardStream::stdout(*COLOR_CHOICE),
+    );
+}
+
+pub fn success(message: &str) {
+    colorize(
+        message,
+        ColorSpec::new().set_bold(true).set_fg(Some(Color::Green)),
+        StandardStream::stdout(*COLOR_CHOICE),
+    );
+}
+
+pub fn error(message: &str) {
+    colorize(
+        &format!("{}{}", "Error: ", message),
+        ColorSpec::new().set_bold(true).set_fg(Some(Color::Red)),
+        StandardStream::stderr(*COLOR_CHOICE),
+    );
+}
+
+/// Print a colorized message to stdout
+fn colorize(message: &str, color: &ColorSpec, mut stream: StandardStream) {
+    stream.set_color(color).unwrap();
+    write!(stream, "{}", message).unwrap();
+    stream.set_color(&ColorSpec::new()).unwrap();
+    writeln!(stream).unwrap();
+}
+
+/// Check whether to output colors
+fn has_color() -> bool {
+    let use_colors = env::var("CLICOLOR").unwrap_or_else(|_| "1".to_string()) != "0"
+        && env::var("NO_COLOR").is_err();
+    let force_colors = env::var("CLICOLOR_FORCE").unwrap_or_else(|_| "0".to_string()) != "0";
+
+    force_colors || use_colors && atty::is(atty::Stream::Stdout)
+}

components/console/src/lib.rs

@@ -6,6 +6,7 @@ edition = "2021"
 [dependencies]
 ahash = "0.7.6"
 ammonia = "3"
+atty = "0.2.11"
 base64 = "0.13"
 csv = "1"
 elasticlunr-rs = {version = "2", default-features = false, features = ["da", "no", "de", "du", "es", "fi", "fr", "it", "pt", "ro", "ru", "sv", "tr"] }
@@ -34,6 +35,7 @@ slug = "0.1"
 svg_metadata = "0.4"
 syntect = "5"
 tera = { version = "1", features = ["preserve_order"] }
+termcolor = "1.0.4"
 time = "0.3"
 toml = "0.5"
 unic-langid = "0.9"

components/libs/Cargo.toml

@@ -6,6 +6,7 @@
 
 pub use ahash;
 pub use ammonia;
+pub use atty;
 pub use base64;
 pub use csv;
 pub use elasticlunr;
@@ -34,6 +35,7 @@ pub use slug;
 pub use svg_metadata;
 pub use syntect;
 pub use tera;
+pub use termcolor;
 pub use time;
 pub use toml;
 pub use unic_langid;

components/libs/src/lib.rs

@@ -11,6 +11,7 @@ pest_derive = "2"
 errors = { path = "../errors" }
 utils = { path = "../utils" }
 config = { path = "../config" }
+console = { path = "../console" }
 libs = { path = "../libs" }
 
 [dev-dependencies]

components/markdown/Cargo.toml

@@ -1,12 +1,13 @@
 use std::fmt::Write;
 
+use errors::bail;
 use libs::gh_emoji::Replacer as EmojiReplacer;
 use libs::once_cell::sync::Lazy;
 use libs::pulldown_cmark as cmark;
 use libs::tera;
 
 use crate::context::RenderContext;
-use errors::{anyhow, Context, Error, Result};
+use errors::{Context, Error, Result};
 use libs::pulldown_cmark::escape::escape_html;
 use utils::site::resolve_internal_link;
 use utils::slugs::slugify_anchors;
@@ -139,7 +140,18 @@ fn fix_link(
                 resolved.permalink
             }
             Err(_) => {
-                return Err(anyhow!("Relative link {} not found.", link));
+                let msg = format!(
+                    "Broken relative link `{}` in {}",
+                    link,
+                    context.current_page_path.unwrap_or("unknown"),
+                );
+                match context.config.link_checker.internal_level {
+                    config::LinkCheckerLevel::Error => bail!(msg),
+                    config::LinkCheckerLevel::Warn => {
+                        console::warn(&msg);
+                        link.to_string()
+                    }
+                }
             }
         }
     } else if is_external_link(link) {

components/markdown/src/markdown.rs

@@ -10,6 +10,7 @@ serde = { version = "1.0", features = ["derive"] }
 
 errors = { path = "../errors" }
 config = { path = "../config" }
+console = { path = "../console" }
 utils = { path = "../utils" }
 templates = { path = "../templates" }
 search = { path = "../search" }

components/site/Cargo.toml

@@ -295,10 +295,45 @@ impl Site {
         tpls::register_tera_global_fns(self);
 
         // Needs to be done after rendering markdown as we only get the anchors at that point
-        link_checking::check_internal_links_with_anchors(self)?;
+        let internal_link_messages = link_checking::check_internal_links_with_anchors(self);
+
+        // log any broken internal links and error out if needed
+        if let Err(messages) = internal_link_messages {
+            let messages: Vec<String> = messages
+                .iter()
+                .enumerate()
+                .map(|(i, msg)| format!("  {}. {}", i + 1, msg))
+                .collect();
+            let msg = format!(
+                "Found {} broken internal anchor link(s)\n{}",
+                messages.len(),
+                messages.join("\n")
+            );
+            match self.config.link_checker.internal_level {
+                config::LinkCheckerLevel::Warn => console::warn(&msg),
+                config::LinkCheckerLevel::Error => return Err(anyhow!(msg.clone())),
+            }
+        }
 
+        // check external links, log the results, and error out if needed
         if self.config.is_in_check_mode() {
-            link_checking::check_external_links(self)?;
+            let external_link_messages = link_checking::check_external_links(self);
+            if let Err(messages) = external_link_messages {
+                let messages: Vec<String> = messages
+                    .iter()
+                    .enumerate()
+                    .map(|(i, msg)| format!("  {}. {}", i + 1, msg))
+                    .collect();
+                let msg = format!(
+                    "Found {} broken external link(s)\n{}",
+                    messages.len(),
+                    messages.join("\n")
+                );
+                match self.config.link_checker.external_level {
+                    config::LinkCheckerLevel::Warn => console::warn(&msg),
+                    config::LinkCheckerLevel::Error => return Err(anyhow!(msg.clone())),
+                }
+            }
         }
 
         Ok(())