enumark 0.1.3 → 0.1.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 24753e0838c8e355e1ddeec38ab241488f2cf71197dd4fe136dccb62a377b9dc
-  data.tar.gz: 5f5659f7fd18b5d21ac5f63a367cb778537d296f21ff4bcd90beda20667a4be2
+  metadata.gz: b85bfbf07990e9f2a38461056ade65766019e3b322c7f4f64778d8627a1c6561
+  data.tar.gz: 5c42a625adbfd33c04d8d5cec939fb7fc91355e127dfe6181328433a7c7f7dc6
 SHA512:
-  metadata.gz: 9e72d815fde4bd327f34018d65dfd1407bca74b05cd119daf451e1a407644ee53a0e6c4ec205d50e923b6a43badc7d1a0c17d229669edae2d32f1292b81451e8
-  data.tar.gz: c2b795969b331004b1332100ba0827fcf0b41a511c7f358c345b4453db53db008a5fa607f73568d32dc82e374188c69f236f50e6a4f1eb90ad272eda0420f17b
+  metadata.gz: ed84f93d9a0b17acce86b08cc50e64517c023c46e85e29661471088afafb54c749bd3e20da4d42aecd6f91ecb1cbe201ce8d2e95ee50aafdfd729a440f33ed34
+  data.tar.gz: 9b679c8511689282e29c8c9b602d33510088d7ef00d1cba6bf88e60871d49c1ac02995de895dccff4a3385a5ae45729ecd2d85a353838c74ea1e096d7c922e62

data/README.md

@@ -53,6 +53,17 @@ enum.each_category do |cate|
 end
 ```
 
+Explore trends of your dump files:
+
+```ruby
+dir = Enumark::Dir.new('/path/to/directory_with_bookmark_dump_files_more_than_one')
+
+dir.added # select items in last file but not in second to last
+dir.deleted # reject items in last file
+dir.uniq # union all items
+dir.static # select items appear in all files
+```
+
 ## Development
 
 After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake test` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment.

data/lib/enumark.rb

@@ -1,111 +1,18 @@
 # frozen_string_literal: true
 
 require_relative "enumark/version"
+require 'enumark/item'
+require 'enumark/category'
+require 'enumark/grouping'
+require 'enumark/dir'
 
 class Enumark
   include Enumerable
 
-  CATEGORY_START = /^\s.*<DT><H3/
-  CATEGORY_END = /^\s.*<\/DL><p>/
-  CATEGORY_NAME = /ADD_DATE="(.*?)".*LAST_MODIFIED="(.*?)".*>(.*)<\/H3/
-
-  ITEM_PREFIX = /^\s.*<DT><A/
-  ITEM_NAME = /HREF="(.*?)".*ADD_DATE="(.*?)".*>(.*)<\/A>/
-
-  class Category
-    attr_reader :name
-    alias_method :inspect, :name
-    alias_method :to_s, :name
-
-    def initialize(line)
-      m = line.match(CATEGORY_NAME)
-      @add_date = m[1]
-      @last_mod = m[2]
-      @name = m[3]
-    end
-  end
-
-  class Item
-    attr_reader :name, :href, :categories
-
-    USELESS_SHARP = /\#.*$/
-
-    def initialize(line, categories)
-      m = line.match(ITEM_NAME)
-      @href = m[1].gsub(USELESS_SHARP, '')
-      @add_date = m[2]
-      @name = m[3]
-      @categories = categories
-    end
-
-    def inspect
-      @inspect ||= "#{categories_str}> #{name}"
-    end
-
-    def categories_str
-      @categories_str ||= "/#{categories.join('/')}"
-    end
-
-    def to_s
-      inspect
-    end
-
-    def host
-      @host ||= (URI.parse(href).host rescue 'unknown')
-    end
-  end
-
-  class Hostname
-    attr_reader :name, :items
-
-    def initialize(name)
-      @name = name
-      @items = []
-    end
-
-    def add(item)
-      @items << item
-    end
-
-    def inspect
-      @name
-    end
-
-    def to
-      inspect
-    end
-  end
-
-  class Grouping
-    Group = Struct.new(:name, :items)
-
-    def initialize(enumark, key, &post)
-      @lock = Mutex.new
-      @collection = nil
-
-      @enumark = enumark
-      @key = key
-      @post = post
-    end
-
-    def each(&block)
-      unless @collection
-        @lock.synchronize do
-          @collection = @enumark.group_by(&@key)
-          @collection = @post.call(@collection) if @post
-          @collection = @collection.map{ |k, items| Group.new(k, items) }
-        end
-      end
-
-      @collection.each(&block)
-    end
-  end
-
-  def initialize(file)
+  def initialize(file, items: nil)
     @file = file
     @lock = Mutex.new
-    @read = false
-    @items = []
+    @items = items
 
     @hosts = Grouping.new(self, :host)
     @dup_titles = Grouping.new(self, :name){ |groups| groups.select{ |_, items| items.count > 1 } }
@@ -115,6 +22,7 @@ class Enumark
 
   def each(&block)
     read_all_lines
+    sort_by_add_date!
     @items.each(&block)
   end
 
@@ -134,32 +42,50 @@ class Enumark
     @cates.each(&block)
   end
 
+  [:+ ,:-, :&, :|].each do |op|
+    class_eval <<-EOM
+      def #{op}(another)
+        new_items = self.to_a #{op} another.to_a
+        Enumark.new(nil, items: new_items)
+      end
+    EOM
+  end
+
   private
 
   def read_all_lines
-    return if @read
+    return if @items
 
     @lock.synchronize do
-      next if @read
-
-      _read_all_lines
-      @read = true
+      _read_all_lines unless @items
     end
   end
 
   def _read_all_lines
     categories = []
+    @items = []
 
     File.new(@file).each do |line|
       case line
-      when ITEM_PREFIX
+      when Item::PREFIX
         item = Item.new(line, categories.dup)
         @items.push(item)
-      when CATEGORY_START
+      when Category::START
         categories.push(Category.new(line))
-      when CATEGORY_END
+      when Category::ENDIND
         categories.pop
       end
     end
   end
+
+  def sort_by_add_date!
+    return if @sorted
+
+    @lock.synchronize do
+      next if @sorted
+
+      @items.sort!{ |i1, i2| i2.add_date <=> i1.add_date }
+      @sorted = true
+    end
+  end
 end

data/lib/enumark/category.rb

@@ -0,0 +1,21 @@
+# frozen_string_literal: true
+
+class Enumark
+  class Category
+
+    START = /^\s.*<DT><H3/
+    ENDIND = /^\s.*<\/DL><p>/
+    PATTERN = /ADD_DATE="(.*?)".*LAST_MODIFIED="(.*?)".*>(.*)<\/H3/
+
+    attr_reader :name
+    alias_method :inspect, :name
+    alias_method :to_s, :name
+
+    def initialize(line)
+      m = line.match(PATTERN)
+      @add_date = m[1]
+      @last_mod = m[2]
+      @name = m[3]
+    end
+  end
+end

data/lib/enumark/dir.rb

@@ -0,0 +1,26 @@
+# frozen_string_literal: true
+
+class Enumark
+  class Dir
+    def initialize(dir)
+      @enumarks = ::Dir.glob(dir).map{ |f| ::Enumark.new(f) }
+      raise 'Not enough to process' if @enumarks.count <= 1
+    end
+
+    def added
+      @added ||= (@enumarks[-1] - @enumarks[-2])
+    end
+
+    def deleted
+      @deleted ||= @enumarks[0..-2].reverse_each.reduce(&:|) - @enumarks[-1]
+    end
+
+    def uniq
+      @uniq ||= @enumarks.reverse_each.reduce(&:|)
+    end
+
+    def static
+      @static ||= @enumarks.reverse_each.reduce(&:&)
+    end
+  end
+end

data/lib/enumark/grouping.rb

@@ -0,0 +1,29 @@
+# frozen_string_literal: true
+
+class Enumark
+
+  class Grouping
+    Group = Struct.new(:name, :items)
+
+    def initialize(enumark, key, &post)
+      @lock = Mutex.new
+      @collection = nil
+
+      @enumark = enumark
+      @key = key
+      @post = post
+    end
+
+    def each(&block)
+      unless @collection
+        @lock.synchronize do
+          @collection = @enumark.group_by(&@key)
+          @collection = @post.call(@collection) if @post
+          @collection = @collection.map{ |k, items| Group.new(k, items) }
+        end
+      end
+
+      @collection.each(&block)
+    end
+  end
+end

data/lib/enumark/item.rb

@@ -0,0 +1,44 @@
+# frozen_string_literal: true
+
+class Enumark
+  class Item
+
+    PREFIX = /^\s.*<DT><A/
+    PATTERN = /HREF="(.*?)".*ADD_DATE="(.*?)".*>(.*)<\/A>/
+    USELESS_SHARP = /\#.*$/
+
+    attr_reader :name, :href, :add_date, :categories
+
+    def initialize(line, categories)
+      m = line.match(PATTERN)
+      @href = m[1].gsub(USELESS_SHARP, '')
+      @add_date = Time.at(m[2].to_i)
+      @name = m[3]
+      @categories = categories
+    end
+
+    def inspect
+      @inspect ||= "#{add_date.strftime('%F %T')} #{categories_str}> #{name}"
+    end
+
+    def categories_str
+      @categories_str ||= "/#{categories.join('/')}"
+    end
+
+    def to_s
+      inspect
+    end
+
+    def hash
+      href.hash
+    end
+
+    def eql?(another)
+      href.eql?(another.href)
+    end
+
+    def host
+      @host ||= (URI.parse(href).host rescue 'unknown')
+    end
+  end
+end

data/lib/enumark/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 class Enumark
-  VERSION = "0.1.3"
+  VERSION = "0.1.4"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: enumark
 version: !ruby/object:Gem::Version
-  version: 0.1.3
+  version: 0.1.4
 platform: ruby
 authors:
 - ken
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2021-02-14 00:00:00.000000000 Z
+date: 2021-02-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: pry
@@ -41,6 +41,10 @@ files:
 - bin/setup
 - enumark.gemspec
 - lib/enumark.rb
+- lib/enumark/category.rb
+- lib/enumark/dir.rb
+- lib/enumark/grouping.rb
+- lib/enumark/item.rb
 - lib/enumark/version.rb
 homepage: https://github.com/turnon/enumark
 licenses: