Remove jekyll

This is a complete rewrite of my blog to a vastly simpler form. That is,
we only need HTML and a couple of build scripts to have something
equivalent, without all the dependencies that jekyll brings. I have also
taken the chance to remove a lot of stuff that was actually not needed.

Signed-off-by: Miquel Sabaté Solà <mikisabate@gmail.com>

2 files changed, 128 insertions, 0 deletions

diff --git a/scripts/build.sh b/scripts/build.sh
new file mode 100755
index 0000000..2985cfc
--- /dev/null
+++ b/scripts/build.sh
@@ -0,0 +1,10 @@
+#!/usr/bin/env sh
+
+ROOT_DIR="$(cd "$( dirname "${BASH_SOURCE[0]}" )/.." &> /dev/null && pwd )"
+
+rm -r $ROOT_DIR/build/*
+cp -r $ROOT_DIR/src/public/* $ROOT_DIR/build/
+find build/ -type f -name *.html -exec ruby scripts/render.rb {} \;
+
+cp -r $ROOT_DIR/images $ROOT_DIR/build/
+cp -r $ROOT_DIR/stylesheets $ROOT_DIR/build/
diff --git a/scripts/render.rb b/scripts/render.rb
new file mode 100755
index 0000000..f17c078
--- /dev/null
+++ b/scripts/render.rb
@@ -0,0 +1,118 @@
+#!/usr/bin/env ruby
+# frozen_string_literal: true
+
+require 'erb'
+
+if ARGV.size != 1
+  puts 'Yo! Just one argument which is the file path.'
+  exit 1
+end
+
+##
+# Utility methods.
+
+# Returns a string with the proper url for the given path. This is always needed
+# so the subpath is always applied if available.
+def url_for(path)
+  first = ARGV.first.split('/')[1]
+
+  default = if %w[en posts].include?(first)
+              '../'
+            else
+              ''
+            end
+
+  ENV.fetch('JO_SUBPATH', default) + path
+end
+
+# Returns a string with the path of the currently evaluated string but without
+# the first directory (e.g. so it's 'en/index.html' instead of
+# 'build/en/index.html').
+def current_path
+  ARGV.first.split('/')[1..].join('/')
+end
+
+# Returns a string with the url moved into the given language code.
+def url_set(lang)
+  path = if lang == 'ca'
+           current_path.split('/')[1..].join('/')
+         else
+           "#{lang}/#{current_path}"
+         end
+
+  url_for(path)
+end
+
+# Returns the language to be used for the current path.
+def current_lang
+  case current_path.split('/').first
+  when 'en'
+    'en'
+  else
+    'ca'
+  end
+end
+
+# Returns whether there is an english version of the page.
+def english_available?
+  !%w[posts/zelda.html].include?(current_path)
+end
+
+# Returns the class to be used for the <main> element.
+def main_class
+  if current_path.start_with?('posts/')
+    'post'
+  end
+end
+
+##
+# Hash with meta strings for each path.
+
+META = {
+  'cv.html' => {
+    title: 'Miquel Sabaté Solà - CV',
+    description: 'El meu Curriculum Vitae'
+  },
+  'license.html' => {
+    title: 'Miquel Sabaté Solà - Llicència',
+    description: 'Llicència d\'aquest espai web'
+  },
+  'en/index.html' => {
+    title: 'Miquel Sabaté Solà - Home page',
+    home: 'Home page',
+    locale: 'en_US',
+    locale_alternate: 'ca_ES'
+  },
+  'en/cv.html' => {
+    title: 'Miquel Sabaté Solà - CV',
+    description: 'My CV',
+    home: 'Home page',
+    locale: 'en_US',
+    locale_alternate: 'ca_ES'
+  },
+  'en/license.html' => {
+    title: 'Miquel Sabaté Solà - License',
+    description: 'License of this site',
+    home: 'Home page',
+    locale: 'en_US',
+    locale_alternate: 'ca_ES'
+  },
+  'posts/zelda.html' => {
+    title: 'Miquel Sabaté Solà - Sobre \'The Legend of Zelda\'',
+    description: 'Entrada sobre la saga de videojocs \'The Legend of Zelda\''
+  },
+}.fetch(current_path, {})
+
+##
+# Behold the complexity of my build!
+
+# Load the current path and evaluate it as an ERB file.
+template = ERB.new(IO.read(ARGV.first))
+inner_html = template.result(binding)
+
+# Load the layout file as an ERB file and apply all the bindings, which include
+# `inner_html` with the file that we previously evaluated.
+template = ERB.new(IO.read('src/templates/layout.html.erb'))
+
+# Re-write the file we were evaluating with the new data.
+File.write(ARGV.first, template.result(binding))