markdownlyze 0.0.2

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: 2893ad36d725df82ee075003813643374a0a6b612ba5f31595f4f3b889952ea1
+  data.tar.gz: 730467d9ad5ca5ed36bfb79e6a583e275a53d599da4cc1eedc9583775bdd7f03
+SHA512:
+  metadata.gz: b2999d43f01fe126efbd00ed2c86b258bbcf1726c76c27228aca2d45376cbace5b5ec2d2b8cb854617932413896867992777a993bfa49794e4485d2e283a2b01
+  data.tar.gz: 1c159cf78113b05268a0bd56db22678ba8eeaffcf00a55723a50ef329a37e2af2f3cc3f584e60beb1e1a8a63e497822588b7517cb2f4b3a789c4970d65f96228

data/lib/markdownlyze/element_name_matcher.rb

@@ -0,0 +1,39 @@
+module Markdownlyze
+  class ElementNameMatcher
+    MATCHERS = {
+      /^# .*/ => :h1,
+      /^## .*/ => :h2,
+      /^### .*/ => :h3,
+      /^#### .*/ => :h4,
+      /\!\[.*\]\(\.\/images\/.*\)/ => :image,
+      /```[a-z]*/ => :code_block,
+      /^\> .*/ => :quote,
+      [/^1\. .*/, lambda { |next_line| next_line.to_s.match?(/^2\. .*/) }] => :ol,
+      [/^(\*|\-) .*/, lambda { |next_line| next_line.to_s.match?(/^(\*|\-) .*/) }] => :ul,
+      [/^|.* \| .*$/, lambda { |next_line| next_line.to_s.count('|') > 1 }] => :table
+    }.freeze
+
+    class << self
+      def call(current_line:, next_line:)
+        return :blank_line if current_line.nil? || current_line.strip.empty?
+
+        MATCHERS.each do |pattern, element|
+          return element if matching?(current_line, next_line, pattern)
+        end
+
+        :paragraph
+      end
+
+      private
+
+      def matching?(current_line, next_line, pattern)
+        if pattern.is_a?(Regexp)
+          current_line.match?(pattern)
+        else
+          regex_pattern, lambda_pattern = pattern
+          current_line.match?(regex_pattern) && lambda_pattern.call(next_line)
+        end
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/base.rb

@@ -0,0 +1,40 @@
+module Markdownlyze
+  module Elements
+    class Base
+      attr_reader :skip_until
+
+      def initialize(line:, index:, lines:)
+        @line = line
+        @index = index
+        @lines = lines
+        @skip_until = nil
+      end
+
+      def options = {}
+
+      private
+
+      def to_html(value)
+        # avoid case when 1. something is created as one element ol list
+        if value.to_s.match?(/^\d+\. .*/)
+          captures = value.match(/^(?<start_number>\d+)\. (?<raw_value>.*)/).named_captures
+          parsed_value = markdown.render(captures.fetch('raw_value')).gsub(/\<(|\/)p\>/, '').gsub("\n", '').gsub("&#39;", "'")
+          element_number = captures.fetch('start_number')
+
+          "#{element_number}. #{parsed_value}"
+        elsif value.to_s.strip == '>'
+          ''
+        else
+          markdown.render(value).gsub(/\<(|\/)p\>/, '').gsub("\n", '').gsub("&#39;", "'")
+        end
+      end
+
+      def markdown
+        @markdown ||= begin
+          renderer = Redcarpet::Render::HTML.new(render_options = {})
+          Redcarpet::Markdown.new(renderer, extensions = {})
+        end
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/blank_line.rb

@@ -0,0 +1,7 @@
+module Markdownlyze
+  module Elements
+    class BlankLine < ::Markdownlyze::Elements::Base
+      def value = nil
+    end
+  end
+end

data/lib/markdownlyze/elements/code_block.rb

@@ -0,0 +1,26 @@
+module Markdownlyze
+  module Elements
+    class CodeBlock < ::Markdownlyze::Elements::Base
+      def value
+        @lines[@index..-1].each_with_index do |subline, subindex|
+          if subline.to_s.strip == '```'
+            @skip_until = @index + subindex + 1
+            break
+          end
+        end
+
+        @lines[(@index + 1)..(@skip_until - 2)].join("\n")
+      end
+
+      def options
+        { language: language }
+      end
+
+      private
+
+      def language
+        (@line.match(/```(?<language>[a-z].*)/) || { language: 'bash' })[:language]
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/h1.rb

@@ -0,0 +1,9 @@
+module Markdownlyze
+  module Elements
+    class H1 < ::Markdownlyze::Elements::Base
+      def value
+        @line.gsub(/^# /, '')
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/h2.rb

@@ -0,0 +1,9 @@
+module Markdownlyze
+  module Elements
+    class H2 < ::Markdownlyze::Elements::Base
+      def value
+        @line.gsub(/^## /, '')
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/h3.rb

@@ -0,0 +1,9 @@
+module Markdownlyze
+  module Elements
+    class H3 < ::Markdownlyze::Elements::Base
+      def value
+        @line.gsub(/^### /, '')
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/h4.rb

@@ -0,0 +1,9 @@
+module Markdownlyze
+  module Elements
+    class H4 < ::Markdownlyze::Elements::Base
+      def value
+        @line.gsub(/^#### /, '')
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/image.rb

@@ -0,0 +1,9 @@
+module Markdownlyze
+  module Elements
+    class Image < ::Markdownlyze::Elements::Base
+      def value
+        @line.match(/\!\[.*\]\((?<file_name>.*)\)/).named_captures.fetch('file_name')
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/ol.rb

@@ -0,0 +1,18 @@
+module Markdownlyze
+  module Elements
+    class Ol < ::Markdownlyze::Elements::Base
+      def value
+        @lines[@index..-1].each_with_index do |subline, subindex|
+          if !subline.match?(/^\d+\. .*/)
+            @skip_until = @index + subindex
+            break
+          end
+        end
+
+        @skip_until = @lines.size if @skip_until.nil?
+
+        @lines[@index..(@skip_until - 1)].map { |line| to_html(line.gsub(/^\d+\. /, '')) }
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/paragraph.rb

@@ -0,0 +1,7 @@
+module Markdownlyze
+  module Elements
+    class Paragraph < ::Markdownlyze::Elements::Base
+      def value = to_html(@line)
+    end
+  end
+end

data/lib/markdownlyze/elements/quote.rb

@@ -0,0 +1,18 @@
+module Markdownlyze
+  module Elements
+    class Quote < ::Markdownlyze::Elements::Base
+      def value
+        @lines[@index..-1].each_with_index do |subline, subindex|
+          if !subline.match?(/^\> .*/) && subline.to_s.strip != '>'
+            @skip_until = @index + subindex
+            break
+          end
+        end
+
+        @skip_until = @lines.size if @skip_until.nil?
+
+        @lines[@index..(@skip_until - 1)].map { |line| to_html(line.gsub(/^\> /, '')) }
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/table.rb

@@ -0,0 +1,26 @@
+module Markdownlyze
+  module Elements
+    class Table < ::Markdownlyze::Elements::Base
+      def value
+        @lines[@index..-1].each_with_index do |subline, subindex|
+          if subline.count('|') < 2
+            @skip_until = @index + subindex
+            break
+          end
+        end
+
+        table_lines = @lines[@index..(skip_until.to_i - 1)]
+        headers = table_lines.shift
+        table_lines.shift
+        raw_table = [headers] | table_lines
+        rows = raw_table.map { |row| row.split("|").map(&:strip)[1..-1].map { |value| to_html(value) } }
+
+        if @skip_until.to_i.zero?
+          @skip_until = @index + rows.size + 1
+        end
+
+        rows
+      end
+    end
+  end
+end

data/lib/markdownlyze/elements/ul.rb

@@ -0,0 +1,18 @@
+module Markdownlyze
+  module Elements
+    class Ul < ::Markdownlyze::Elements::Base
+      def value
+        @lines[@index..-1].each_with_index do |subline, subindex|
+          if !subline.match?(/^(\*|\-) .*/)
+            @skip_until = @index + subindex
+            break
+          end
+        end
+
+        @skip_until = @lines.size if @skip_until.nil?
+
+        @lines[@index..(@skip_until - 1)].map { |line| to_html(line.gsub(/^(\*|\-) /, '')) }
+      end
+    end
+  end
+end

data/lib/markdownlyze/version.rb

@@ -0,0 +1,9 @@
+module Markdownlyze
+  module Version
+    module_function
+
+    def to_s
+      "0.0.2"
+    end
+  end
+end

data/lib/markdownlyze.rb

@@ -0,0 +1,54 @@
+require 'redcarpet'
+
+require_relative 'markdownlyze/element_name_matcher'
+
+require_relative 'markdownlyze/elements/base'
+require_relative 'markdownlyze/elements/h1'
+require_relative 'markdownlyze/elements/h2'
+require_relative 'markdownlyze/elements/h3'
+require_relative 'markdownlyze/elements/h4'
+require_relative 'markdownlyze/elements/image'
+require_relative 'markdownlyze/elements/blank_line'
+require_relative 'markdownlyze/elements/paragraph'
+require_relative 'markdownlyze/elements/code_block'
+require_relative 'markdownlyze/elements/table'
+require_relative 'markdownlyze/elements/ol'
+require_relative 'markdownlyze/elements/ul'
+require_relative 'markdownlyze/elements/quote'
+
+module Markdownlyze
+  class << self
+    def parse(content)
+      markdown_lines = content.to_s.split("\n")
+      skip_until = nil
+
+      elements = markdown_lines.map.with_index do |line, index|
+        unless skip_until.nil?
+          next if index < skip_until
+        end
+
+        element_name = ::Markdownlyze::ElementNameMatcher.call(
+          current_line: line,
+          next_line: markdown_lines[index + 1]
+        )
+
+        element_class = ::Markdownlyze::Elements.const_get(element_name.to_s.split('_').map(&:capitalize).join)
+        element = element_class.new(
+          line: line,
+          index: index,
+          lines: markdown_lines
+        )
+
+        { element: element_name, value: element.value }.merge(element.options).tap do
+          skip_until = element.skip_until
+        end
+      end
+
+      elements.compact
+    end
+
+    def parse_file(file_path)
+      parse(File.read(file_path))
+    end
+  end
+end

metadata

@@ -0,0 +1,99 @@
+--- !ruby/object:Gem::Specification
+name: markdownlyze
+version: !ruby/object:Gem::Version
+  version: 0.0.2
+platform: ruby
+authors:
+- Paweł Dąbrowski
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2024-08-19 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: pry
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: redcarpet
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.6.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.6.0
+description: Turn markdown documents into easily parseable format
+email: contact@paweldabrowski.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/markdownlyze.rb
+- lib/markdownlyze/element_name_matcher.rb
+- lib/markdownlyze/elements/base.rb
+- lib/markdownlyze/elements/blank_line.rb
+- lib/markdownlyze/elements/code_block.rb
+- lib/markdownlyze/elements/h1.rb
+- lib/markdownlyze/elements/h2.rb
+- lib/markdownlyze/elements/h3.rb
+- lib/markdownlyze/elements/h4.rb
+- lib/markdownlyze/elements/image.rb
+- lib/markdownlyze/elements/ol.rb
+- lib/markdownlyze/elements/paragraph.rb
+- lib/markdownlyze/elements/quote.rb
+- lib/markdownlyze/elements/table.rb
+- lib/markdownlyze/elements/ul.rb
+- lib/markdownlyze/version.rb
+homepage:
+licenses: []
+metadata: {}
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.4.10
+signing_key:
+specification_version: 4
+summary: Turn markdown documents into easily parseable format
+test_files: []