From 0df3631b3d81f6704e6d1da83f65a4a6ce0cce44 Mon Sep 17 00:00:00 2001
From: areille <32739983+augustinreille@users.noreply.github.com>
Date: Fri, 28 Aug 2020 17:39:19 +0000
Subject: [PATCH] Add minify support (#1146)

* Doc add a missing arg to `get_taxonomy_url` (#1139)

This feature is already exist, but not in the doc yet

Related #766

* Add minify support

* Add documentation

* Code review

* Fix error in documentation

* Update minify-html to 0.3.6

* Move minify into write_content function

* Fix multiple calls to minify()

* Add test for minified output

* Fix breaking test

Co-authored-by: Ken <2770219+ken0x0a@users.noreply.github.com>
---
 Cargo.lock                                    | 12 ++++++
 components/config/src/config/mod.rs           |  3 ++
 components/site/Cargo.toml                    |  1 +
 components/site/src/lib.rs                    | 37 +++++++++++++++++--
 components/site/tests/site.rs                 | 12 ++++++
 .../getting-started/configuration.md          |  3 ++
 6 files changed, 64 insertions(+), 4 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 55044502..e3f53151 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1320,6 +1320,17 @@ dependencies = [
  "unicase",
 ]
 
+[[package]]
+name = "minify-html"
+version = "0.3.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b917ef1eb405ca3b1a489608f36392f0718ff4740bbc6f7d89edd4ec4c869fb7"
+dependencies = [
+ "aho-corasick",
+ "lazy_static",
+ "memchr",
+]
+
 [[package]]
 name = "miniz_oxide"
 version = "0.3.7"
@@ -2177,6 +2188,7 @@ dependencies = [
  "lazy_static",
  "library",
  "link_checker",
+ "minify-html",
  "rayon",
  "sass-rs",
  "search",
diff --git a/components/config/src/config/mod.rs b/components/config/src/config/mod.rs
index 0e715354..15f5f753 100644
--- a/components/config/src/config/mod.rs
+++ b/components/config/src/config/mod.rs
@@ -74,6 +74,8 @@ pub struct Config {
 
     /// Whether to compile the `sass` directory and output the css files into the static folder
     pub compile_sass: bool,
+    /// Whether to minify the html output
+    pub minify_html: bool,
     /// Whether to build the search index for the content
     pub build_search_index: bool,
     /// A list of file glob patterns to ignore when processing the content folder. Defaults to none.
@@ -320,6 +322,7 @@ impl Default for Config {
             hard_link_static: false,
             taxonomies: Vec::new(),
             compile_sass: false,
+            minify_html: false,
             mode: Mode::Build,
             build_search_index: false,
             ignored_content: Vec::new(),
diff --git a/components/site/Cargo.toml b/components/site/Cargo.toml
index 12373c84..60008852 100644
--- a/components/site/Cargo.toml
+++ b/components/site/Cargo.toml
@@ -8,6 +8,7 @@ include = ["src/**/*"]
 [dependencies]
 tera = "1"
 glob = "0.3"
+minify-html = "0.3.6"
 rayon = "1"
 serde = "1"
 serde_derive = "1"
diff --git a/components/site/src/lib.rs b/components/site/src/lib.rs
index cd8b6bff..61961b79 100644
--- a/components/site/src/lib.rs
+++ b/components/site/src/lib.rs
@@ -11,6 +11,7 @@ use std::sync::{Arc, Mutex, RwLock};
 
 use glob::glob;
 use lazy_static::lazy_static;
+use minify_html::{truncate, Cfg};
 use rayon::prelude::*;
 use tera::{Context, Tera};
 
@@ -445,6 +446,21 @@ impl Site {
         html
     }
 
+    /// Minifies html content
+    fn minify(&self, html: String) -> Result<String> {
+        let cfg = &Cfg { minify_js: false };
+        let mut input_bytes = html.as_bytes().to_vec();
+        match truncate(&mut input_bytes, cfg) {
+            Ok(_len) => match std::str::from_utf8(&mut input_bytes) {
+                Ok(result) => Ok(result.to_string()),
+                Err(err) => bail!("Failed to convert bytes to string : {}", err),
+            },
+            Err(minify_error) => {
+                bail!("Failed to truncate html at character {}:", minify_error.position);
+            }
+        }
+    }
+
     /// Copy the main `static` folder and the theme `static` folder if a theme is used
     pub fn copy_static_directories(&self) -> Result<()> {
         // The user files will overwrite the theme files
@@ -511,10 +527,19 @@ impl Site {
             create_directory(&current_path)?;
         }
 
+        let final_content = if !filename.ends_with("html") || !self.config.minify_html {
+            content
+        } else {
+            match self.minify(content) {
+                Ok(minified_content) => minified_content,
+                Err(error) => bail!(error),
+            }
+        };
+
         match self.build_mode {
             BuildMode::Disk => {
                 let end_path = current_path.join(filename);
-                create_file(&end_path, &content)?;
+                create_file(&end_path, &final_content)?;
             }
             BuildMode::Memory => {
                 let path = if filename != "index.html" {
@@ -527,7 +552,7 @@ impl Site {
                 }
                 .trim_end_matches('/')
                 .to_owned();
-                &SITE_CONTENT.write().unwrap().insert(path, content);
+                &SITE_CONTENT.write().unwrap().insert(path, final_content);
             }
         }
 
@@ -543,8 +568,12 @@ impl Site {
         let output = page.render_html(&self.tera, &self.config, &self.library.read().unwrap())?;
         let content = self.inject_livereload(output);
         let components: Vec<&str> = page.path.split('/').collect();
-        let current_path =
-            self.write_content(&components, "index.html", content, !page.assets.is_empty())?;
+        let current_path = self.write_content(
+            &components,
+            "index.html",
+            content,
+            !page.assets.is_empty(),
+        )?;
 
         // Copy any asset we found previously into the same directory as the index.html
         for asset in &page.assets {
diff --git a/components/site/tests/site.rs b/components/site/tests/site.rs
index bf9f3358..c7419d8e 100644
--- a/components/site/tests/site.rs
+++ b/components/site/tests/site.rs
@@ -712,6 +712,18 @@ fn can_build_site_custom_builtins_from_theme() {
     assert!(file_contains!(public, "404.html", "Oops"));
 }
 
+#[test]
+fn can_build_site_with_html_minified() {
+    let (_, _tmp_dir, public) = build_site_with_setup("test_site", |mut site| {
+        site.config.minify_html = true;
+        (site, true)
+    });
+
+    assert!(&public.exists());
+    assert!(file_exists!(public, "index.html"));
+    assert!(file_contains!(public, "index.html", "<!DOCTYPE html><html lang=en><head><meta charset=UTF-8>"));
+}
+
 #[test]
 fn can_ignore_markdown_content() {
     let (_, _tmp_dir, public) = build_site("test_site");
diff --git a/docs/content/documentation/getting-started/configuration.md b/docs/content/documentation/getting-started/configuration.md
index e029e1cb..e1096074 100644
--- a/docs/content/documentation/getting-started/configuration.md
+++ b/docs/content/documentation/getting-started/configuration.md
@@ -85,6 +85,9 @@ languages = []
 # When set to "true", the Sass files in the `sass` directory are compiled.
 compile_sass = false
 
+# When set to "true", the HTML output files will be minified
+minify_html = false
+
 # A list of glob patterns specifying asset files to ignore when the content
 # directory is processed. Defaults to none, which means that all asset files are
 # copied over to the `public` directory.