bridgetown-core 0.7.0 → 0.7.1

data/lib/bridgetown-core/tags/link.rb

@@ -0,0 +1,41 @@
+# frozen_string_literal: true
+
+module Bridgetown
+  module Tags
+    class Link < Liquid::Tag
+      include Bridgetown::Filters::URLFilters
+
+      class << self
+        def tag_name
+          name.split("::").last.downcase
+        end
+      end
+
+      def initialize(tag_name, relative_path, tokens)
+        super
+
+        @relative_path = relative_path.strip
+      end
+
+      def render(context)
+        @context = context
+        site = context.registers[:site]
+        relative_path = Liquid::Template.parse(@relative_path).render(context)
+
+        site.each_site_file do |item|
+          return relative_url(item) if item.relative_path == relative_path
+          # This takes care of the case for static files that have a leading /
+          return relative_url(item) if item.relative_path == "/#{relative_path}"
+        end
+
+        raise ArgumentError, <<~MSG
+          Could not find document '#{relative_path}' in tag '#{self.class.tag_name}'.
+
+          Make sure the document exists and the path is correct.
+        MSG
+      end
+    end
+  end
+end
+
+Liquid::Template.register_tag(Bridgetown::Tags::Link.tag_name, Bridgetown::Tags::Link)

data/lib/bridgetown-core/tags/post_url.rb

@@ -0,0 +1,107 @@
+# frozen_string_literal: true
+
+module Bridgetown
+  module Tags
+    class PostComparer
+      MATCHER = %r!^(.+/)*(\d+-\d+-\d+)-(.*)$!.freeze
+
+      attr_reader :path, :date, :slug, :name
+
+      def initialize(name)
+        @name = name
+
+        all, @path, @date, @slug = *name.sub(%r!^/!, "").match(MATCHER)
+        unless all
+          raise Bridgetown::Errors::InvalidPostNameError,
+                "'#{name}' does not contain valid date and/or title."
+        end
+
+        escaped_slug = Regexp.escape(slug)
+        @name_regex = %r!^_posts/#{path}#{date}-#{escaped_slug}\.[^.]+|
+          ^#{path}_posts/?#{date}-#{escaped_slug}\.[^.]+!x
+      end
+
+      def post_date
+        @post_date ||= Utils.parse_date(
+          date,
+          "'#{date}' does not contain valid date and/or title."
+        )
+      end
+
+      def ==(other)
+        other.relative_path.match(@name_regex)
+      end
+
+      def deprecated_equality(other)
+        slug == post_slug(other) &&
+          post_date.year  == other.date.year &&
+          post_date.month == other.date.month &&
+          post_date.day   == other.date.day
+      end
+
+      private
+
+      # Construct the directory-aware post slug for a Bridgetown::Post
+      #
+      # other - the Bridgetown::Post
+      #
+      # Returns the post slug with the subdirectory (relative to _posts)
+      def post_slug(other)
+        path = other.basename.split("/")[0...-1].join("/")
+        if path.nil? || path == ""
+          other.data["slug"]
+        else
+          path + "/" + other.data["slug"]
+        end
+      end
+    end
+
+    class PostUrl < Liquid::Tag
+      include Bridgetown::Filters::URLFilters
+
+      def initialize(tag_name, post, tokens)
+        super
+        @orig_post = post.strip
+        begin
+          @post = PostComparer.new(@orig_post)
+        rescue StandardError => e
+          raise Bridgetown::Errors::PostURLError, <<~MSG
+            Could not parse name of post "#{@orig_post}" in tag 'post_url'.
+             Make sure the post exists and the name is correct.
+             #{e.class}: #{e.message}
+          MSG
+        end
+      end
+
+      def render(context)
+        @context = context
+        site = context.registers[:site]
+
+        site.posts.docs.each do |document|
+          return relative_url(document) if @post == document
+        end
+
+        # New matching method did not match, fall back to old method
+        # with deprecation warning if this matches
+
+        site.posts.docs.each do |document|
+          next unless @post.deprecated_equality document
+
+          Bridgetown::Deprecator.deprecation_message "A call to "\
+            "'{% post_url #{@post.name} %}' did not match " \
+            "a post using the new matching method of checking name " \
+            "(path-date-slug) equality. Please make sure that you " \
+            "change this tag to match the post's name exactly."
+          return relative_url(document)
+        end
+
+        raise Bridgetown::Errors::PostURLError, <<~MSG
+          Could not find post "#{@orig_post}" in tag 'post_url'.
+          Make sure the post exists and the name is correct.
+        MSG
+      end
+    end
+  end
+end
+
+Liquid::Template.register_tag("post_url", Bridgetown::Tags::PostUrl)

data/lib/bridgetown-core/url.rb

@@ -0,0 +1,164 @@
+# frozen_string_literal: true
+
+# Public: Methods that generate a URL for a resource such as a Post or a Page.
+#
+# Examples
+#
+#   URL.new({
+#     :template => /:categories/:title.html",
+#     :placeholders => {:categories => "ruby", :title => "something"}
+#   }).to_s
+#
+module Bridgetown
+  class URL
+    # options - One of :permalink or :template must be supplied.
+    #           :template     - The String used as template for URL generation,
+    #                           for example "/:path/:basename:output_ext", where
+    #                           a placeholder is prefixed with a colon.
+    #           :placeholders - A hash containing the placeholders which will be
+    #                           replaced when used inside the template. E.g.
+    #                           { "year" => Time.now.strftime("%Y") } would replace
+    #                           the placeholder ":year" with the current year.
+    #           :permalink    - If supplied, no URL will be generated from the
+    #                           template. Instead, the given permalink will be
+    #                           used as URL.
+    def initialize(options)
+      @template     = options[:template]
+      @placeholders = options[:placeholders] || {}
+      @permalink    = options[:permalink]
+
+      if (@template || @permalink).nil?
+        raise ArgumentError, "One of :template or :permalink must be supplied."
+      end
+    end
+
+    # The generated relative URL of the resource
+    #
+    # Returns the String URL
+    def to_s
+      sanitize_url(generated_permalink || generated_url)
+    end
+
+    # Generates a URL from the permalink
+    #
+    # Returns the _unsanitized String URL
+    def generated_permalink
+      (@generated_permalink ||= generate_url(@permalink)) if @permalink
+    end
+
+    # Generates a URL from the template
+    #
+    # Returns the unsanitized String URL
+    def generated_url
+      @generated_url ||= generate_url(@template)
+    end
+
+    # Internal: Generate the URL by replacing all placeholders with their
+    # respective values in the given template
+    #
+    # Returns the unsanitized String URL
+    def generate_url(template)
+      if @placeholders.is_a? Drops::UrlDrop
+        generate_url_from_drop(template)
+      else
+        generate_url_from_hash(template)
+      end
+    end
+
+    def generate_url_from_hash(template)
+      @placeholders.inject(template) do |result, token|
+        break result if result.index(":").nil?
+
+        if token.last.nil?
+          # Remove leading "/" to avoid generating urls with `//`
+          result.gsub("/:#{token.first}", "")
+        else
+          result.gsub(":#{token.first}", self.class.escape_path(token.last))
+        end
+      end
+    end
+
+    # We include underscores in keys to allow for 'i_month' and so forth.
+    # This poses a problem for keys which are followed by an underscore
+    # but the underscore is not part of the key, e.g. '/:month_:day'.
+    # That should be :month and :day, but our key extraction regexp isn't
+    # smart enough to know that so we have to make it an explicit
+    # possibility.
+    def possible_keys(key)
+      if key.end_with?("_")
+        [key, key.chomp("_")]
+      else
+        [key]
+      end
+    end
+
+    def generate_url_from_drop(template)
+      template.gsub(%r!:([a-z_]+)!) do |match|
+        pool = possible_keys(match.sub(":", ""))
+
+        winner = pool.find { |key| @placeholders.key?(key) }
+        if winner.nil?
+          raise NoMethodError,
+                "The URL template doesn't have #{pool.join(" or ")} keys. "\
+                "Check your permalink template!"
+        end
+
+        value = @placeholders[winner]
+        value = "" if value.nil?
+        replacement = self.class.escape_path(value)
+
+        match.sub(":#{winner}", replacement)
+      end.squeeze("/")
+    end
+
+    # Returns a sanitized String URL, stripping "../../" and multiples of "/",
+    # as well as the beginning "/" so we can enforce and ensure it.
+
+    def sanitize_url(str)
+      "/#{str}".gsub("..", "/").gsub("./", "").squeeze("/")
+    end
+
+    # Escapes a path to be a valid URL path segment
+    #
+    # path - The path to be escaped.
+    #
+    # Examples:
+    #
+    #   URL.escape_path("/a b")
+    #   # => "/a%20b"
+    #
+    # Returns the escaped path.
+    def self.escape_path(path)
+      return path if path.empty? || %r!^[a-zA-Z0-9./-]+$!.match?(path)
+
+      # Because URI.escape doesn't escape "?", "[" and "]" by default,
+      # specify unsafe string (except unreserved, sub-delims, ":", "@" and "/").
+      #
+      # URI path segment is defined in RFC 3986 as follows:
+      #   segment       = *pchar
+      #   pchar         = unreserved / pct-encoded / sub-delims / ":" / "@"
+      #   unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~"
+      #   pct-encoded   = "%" HEXDIG HEXDIG
+      #   sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"
+      #                 / "*" / "+" / "," / ";" / "="
+      Addressable::URI.encode(path).encode("utf-8").sub("#", "%23")
+    end
+
+    # Unescapes a URL path segment
+    #
+    # path - The path to be unescaped.
+    #
+    # Examples:
+    #
+    #   URL.unescape_path("/a%20b")
+    #   # => "/a b"
+    #
+    # Returns the unescaped path.
+    def self.unescape_path(path)
+      path = path.encode("utf-8")
+      return path unless path.include?("%")
+
+      Addressable::URI.unencode(path)
+    end
+  end
+end

data/lib/bridgetown-core/utils.rb

@@ -0,0 +1,367 @@
+# frozen_string_literal: true
+
+module Bridgetown
+  module Utils
+    extend self
+    autoload :Ansi, "bridgetown-core/utils/ansi"
+    autoload :Exec, "bridgetown-core/utils/exec"
+    autoload :Internet, "bridgetown-core/utils/internet"
+    autoload :Platforms, "bridgetown-core/utils/platforms"
+    autoload :ThreadEvent, "bridgetown-core/utils/thread_event"
+    autoload :WinTZ, "bridgetown-core/utils/win_tz"
+
+    # Constants for use in #slugify
+    SLUGIFY_MODES = %w(raw default pretty simple ascii latin).freeze
+    SLUGIFY_RAW_REGEXP = Regexp.new('\\s+').freeze
+    SLUGIFY_DEFAULT_REGEXP = Regexp.new("[^[:alnum:]]+").freeze
+    SLUGIFY_PRETTY_REGEXP = Regexp.new("[^[:alnum:]._~!$&'()+,;=@]+").freeze
+    SLUGIFY_ASCII_REGEXP = Regexp.new("[^[A-Za-z0-9]]+").freeze
+
+    # Takes a slug and turns it into a simple title.
+    def titleize_slug(slug)
+      slug.split("-").map!(&:capitalize).join(" ")
+    end
+
+    # Non-destructive version of deep_merge_hashes! See that method.
+    #
+    # Returns the merged hashes.
+    def deep_merge_hashes(master_hash, other_hash)
+      deep_merge_hashes!(master_hash.dup, other_hash)
+    end
+
+    # Merges a master hash with another hash, recursively.
+    #
+    # master_hash - the "parent" hash whose values will be overridden
+    # other_hash  - the other hash whose values will be persisted after the merge
+    #
+    # This code was lovingly stolen from some random gem:
+    # http://gemjack.com/gems/tartan-0.1.1/classes/Hash.html
+    #
+    # Thanks to whoever made it.
+    def deep_merge_hashes!(target, overwrite)
+      merge_values(target, overwrite)
+      merge_default_proc(target, overwrite)
+      duplicate_frozen_values(target)
+
+      target
+    end
+
+    def mergable?(value)
+      value.is_a?(Hash) || value.is_a?(Drops::Drop)
+    end
+
+    def duplicable?(obj)
+      case obj
+      when nil, false, true, Symbol, Numeric
+        false
+      else
+        true
+      end
+    end
+
+    # Read array from the supplied hash favouring the singular key
+    # and then the plural key, and handling any nil entries.
+    #
+    # hash - the hash to read from
+    # singular_key - the singular key
+    # plural_key - the plural key
+    #
+    # Returns an array
+    def pluralized_array_from_hash(hash, singular_key, plural_key)
+      array = []
+      value = value_from_singular_key(hash, singular_key)
+      value ||= value_from_plural_key(hash, plural_key)
+
+      array << value
+      array.flatten!
+      array.compact!
+      array
+    end
+
+    def value_from_singular_key(hash, key)
+      hash[key] if hash.key?(key) || (hash.default_proc && hash[key])
+    end
+
+    def value_from_plural_key(hash, key)
+      if hash.key?(key) || (hash.default_proc && hash[key])
+        val = hash[key]
+        case val
+        when String
+          val.split
+        when Array
+          val.compact
+        end
+      end
+    end
+
+    def transform_keys(hash)
+      result = {}
+      hash.each_key do |key|
+        result[yield(key)] = hash[key]
+      end
+      result
+    end
+
+    # Apply #to_sym to all keys in the hash
+    #
+    # hash - the hash to which to apply this transformation
+    #
+    # Returns a new hash with symbolized keys
+    def symbolize_hash_keys(hash)
+      transform_keys(hash) { |key| key.to_sym rescue key }
+    end
+
+    # Apply #to_s to all keys in the Hash
+    #
+    # hash - the hash to which to apply this transformation
+    #
+    # Returns a new hash with stringified keys
+    def stringify_hash_keys(hash)
+      transform_keys(hash) { |key| key.to_s rescue key }
+    end
+
+    # Parse a date/time and throw an error if invalid
+    #
+    # input - the date/time to parse
+    # msg - (optional) the error message to show the user
+    #
+    # Returns the parsed date if successful, throws a FatalException
+    # if not
+    def parse_date(input, msg = "Input could not be parsed.")
+      Time.parse(input).localtime
+    rescue ArgumentError
+      raise Errors::InvalidDateError, "Invalid date '#{input}': #{msg}"
+    end
+
+    # Determines whether a given file has
+    #
+    # Returns true if the YAML front matter is present.
+    # rubocop: disable Naming/PredicateName
+    def has_yaml_header?(file)
+      File.open(file, "rb", &:readline).match? %r!\A---\s*\r?\n!
+    rescue EOFError
+      false
+    end
+
+    # Determine whether the given content string contains Liquid Tags or Vaiables
+    #
+    # Returns true is the string contains sequences of `{%` or `{{`
+    def has_liquid_construct?(content)
+      return false if content.nil? || content.empty?
+
+      content.include?("{%") || content.include?("{{")
+    end
+    # rubocop: enable Naming/PredicateName
+
+    # Slugify a filename or title.
+    #
+    # string - the filename or title to slugify
+    # mode - how string is slugified
+    # cased - whether to replace all uppercase letters with their
+    # lowercase counterparts
+    #
+    # When mode is "none", return the given string.
+    #
+    # When mode is "raw", return the given string,
+    # with every sequence of spaces characters replaced with a hyphen.
+    #
+    # When mode is "default", "simple", or nil, non-alphabetic characters are
+    # replaced with a hyphen too.
+    #
+    # When mode is "pretty", some non-alphabetic characters (._~!$&'()+,;=@)
+    # are not replaced with hyphen.
+    #
+    # When mode is "ascii", some everything else except ASCII characters
+    # a-z (lowercase), A-Z (uppercase) and 0-9 (numbers) are not replaced with hyphen.
+    #
+    # When mode is "latin", the input string is first preprocessed so that
+    # any letters with accents are replaced with the plain letter. Afterwards,
+    # it follows the "default" mode of operation.
+    #
+    # If cased is true, all uppercase letters in the result string are
+    # replaced with their lowercase counterparts.
+    #
+    # Examples:
+    #   slugify("The _config.yml file")
+    #   # => "the-config-yml-file"
+    #
+    #   slugify("The _config.yml file", "pretty")
+    #   # => "the-_config.yml-file"
+    #
+    #   slugify("The _config.yml file", "pretty", true)
+    #   # => "The-_config.yml file"
+    #
+    #   slugify("The _config.yml file", "ascii")
+    #   # => "the-config-yml-file"
+    #
+    #   slugify("The _config.yml file", "latin")
+    #   # => "the-config-yml-file"
+    #
+    # Returns the slugified string.
+    def slugify(string, mode: nil, cased: false)
+      mode ||= "default"
+      return nil if string.nil?
+
+      unless SLUGIFY_MODES.include?(mode)
+        return cased ? string : string.downcase
+      end
+
+      # Drop accent marks from latin characters. Everything else turns to ?
+      if mode == "latin"
+        I18n.config.available_locales = :en if I18n.config.available_locales.empty?
+        string = I18n.transliterate(string)
+      end
+
+      slug = replace_character_sequence_with_hyphen(string, :mode => mode)
+
+      # Remove leading/trailing hyphen
+      slug.gsub!(%r!^\-|\-$!i, "")
+
+      slug.downcase! unless cased
+      Bridgetown.logger.warn("Warning:", "Empty `slug` generated for '#{string}'.") if slug.empty?
+      slug
+    end
+
+    # Add an appropriate suffix to template so that it matches the specified
+    # permalink style.
+    #
+    # template - permalink template without trailing slash or file extension
+    # permalink_style - permalink style, either built-in or custom
+    #
+    # The returned permalink template will use the same ending style as
+    # specified in permalink_style.  For example, if permalink_style contains a
+    # trailing slash (or is :pretty, which indirectly has a trailing slash),
+    # then so will the returned template.  If permalink_style has a trailing
+    # ":output_ext" (or is :none, :date, or :ordinal) then so will the returned
+    # template.  Otherwise, template will be returned without modification.
+    #
+    # Examples:
+    #   add_permalink_suffix("/:basename", :pretty)
+    #   # => "/:basename/"
+    #
+    #   add_permalink_suffix("/:basename", :date)
+    #   # => "/:basename:output_ext"
+    #
+    #   add_permalink_suffix("/:basename", "/:year/:month/:title/")
+    #   # => "/:basename/"
+    #
+    #   add_permalink_suffix("/:basename", "/:year/:month/:title")
+    #   # => "/:basename"
+    #
+    # Returns the updated permalink template
+    def add_permalink_suffix(template, permalink_style)
+      template = template.dup
+
+      case permalink_style
+      when :pretty, :simple
+        template << "/"
+      when :date, :ordinal, :none
+        template << ":output_ext"
+      else
+        template << "/" if permalink_style.to_s.end_with?("/")
+        template << ":output_ext" if permalink_style.to_s.end_with?(":output_ext")
+      end
+
+      template
+    end
+
+    # Work the same way as Dir.glob but seperating the input into two parts
+    # ('dir' + '/' + 'pattern') to make sure the first part('dir') does not act
+    # as a pattern.
+    #
+    # For example, Dir.glob("path[/*") always returns an empty array,
+    # because the method fails to find the closing pattern to '[' which is ']'
+    #
+    # Examples:
+    #   safe_glob("path[", "*")
+    #   # => ["path[/file1", "path[/file2"]
+    #
+    #   safe_glob("path", "*", File::FNM_DOTMATCH)
+    #   # => ["path/.", "path/..", "path/file1"]
+    #
+    #   safe_glob("path", ["**", "*"])
+    #   # => ["path[/file1", "path[/folder/file2"]
+    #
+    # dir      - the dir where glob will be executed under
+    #           (the dir will be included to each result)
+    # patterns - the patterns (or the pattern) which will be applied under the dir
+    # flags    - the flags which will be applied to the pattern
+    #
+    # Returns matched pathes
+    def safe_glob(dir, patterns, flags = 0)
+      return [] unless Dir.exist?(dir)
+
+      pattern = File.join(Array(patterns))
+      return [dir] if pattern.empty?
+
+      Dir.chdir(dir) do
+        Dir.glob(pattern, flags).map { |f| File.join(dir, f) }
+      end
+    end
+
+    # Returns merged option hash for File.read of self.site (if exists)
+    # and a given param
+    def merged_file_read_opts(site, opts)
+      merged = (site ? site.file_read_opts : {}).merge(opts)
+      if merged[:encoding] && !merged[:encoding].start_with?("bom|")
+        merged[:encoding] = "bom|#{merged[:encoding]}"
+      end
+      if merged["encoding"] && !merged["encoding"].start_with?("bom|")
+        merged["encoding"] = "bom|#{merged["encoding"]}"
+      end
+      merged
+    end
+
+    private
+
+    def merge_values(target, overwrite)
+      target.merge!(overwrite) do |_key, old_val, new_val|
+        if new_val.nil?
+          old_val
+        elsif mergable?(old_val) && mergable?(new_val)
+          deep_merge_hashes(old_val, new_val)
+        else
+          new_val
+        end
+      end
+    end
+
+    def merge_default_proc(target, overwrite)
+      if target.is_a?(Hash) && overwrite.is_a?(Hash) && target.default_proc.nil?
+        target.default_proc = overwrite.default_proc
+      end
+    end
+
+    def duplicate_frozen_values(target)
+      target.each do |key, val|
+        target[key] = val.dup if val.frozen? && duplicable?(val)
+      end
+    end
+
+    # Replace each character sequence with a hyphen.
+    #
+    # See Utils#slugify for a description of the character sequence specified
+    # by each mode.
+    def replace_character_sequence_with_hyphen(string, mode: "default")
+      replaceable_char =
+        case mode
+        when "raw"
+          SLUGIFY_RAW_REGEXP
+        when "pretty"
+          # "._~!$&'()+,;=@" is human readable (not URI-escaped) in URL
+          # and is allowed in both extN and NTFS.
+          SLUGIFY_PRETTY_REGEXP
+        when "ascii"
+          # For web servers not being able to handle Unicode, the safe
+          # method is to ditch anything else but latin letters and numeric
+          # digits.
+          SLUGIFY_ASCII_REGEXP
+        else
+          SLUGIFY_DEFAULT_REGEXP
+        end
+
+      # Strip according to the mode
+      string.gsub(replaceable_char, "-")
+    end
+  end
+end