deba 0.6.0 → 0.7.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: f39bcdf2a7a1d34172bb6c68196532c119c13fdb
-  data.tar.gz: e911b81e44f25984e658abddabed6c9af80d895b
+  metadata.gz: 2c8a544deca7b66f05de8209b957f2ce5c07f1e2
+  data.tar.gz: 3ec0e8bb25b23ad2145867dc130ea9a9f475f64b
 SHA512:
-  metadata.gz: b1f78c6a8cdde9ee7f7608bac295a262bbe7631c147ad6f43f541d02a0841de46615370cc7b48de20e150236ff1cb124384a80c4cd1ffe52e520bf3336929ba5
-  data.tar.gz: 5fad6708011e66e11825e32e3edd01508dcc4e8cfe6eed2719c86abb08a2ccb1dc2f2c1d057a9494e8cce42d94d9cba16ee668f051b28bd24253f65967b3e596
+  metadata.gz: cdd042c6214229de58365925f74b435e14cc9f54f9ca40304189354706c7182a805ccf971bf5a8e59820164faafa6d7549f1d26f7b6fefcf77fb5f19f0c18018
+  data.tar.gz: 3bdfc3fcd212b0dacd8228c04c391324a67ec619c853b74255c063036a4748b4249abcb2e0052837c09428a42cd5fcbd7adb192278db760c1dd3ff491d79ef94

data/exe/deba

@@ -0,0 +1,6 @@
+#!/usr/bin/env ruby
+
+require "bundler/setup"
+require "deba"
+
+print Deba.extract(File.read(ARGV.first))

data/lib/deba.rb

@@ -1,7 +1,7 @@
 require "nokogiri"
 
 module Deba
-  VERSION = "0.6.0"
+  VERSION = "0.7.0"
 end
 
 require "deba/utils"
@@ -9,6 +9,7 @@ require "deba/stringifier"
 require "deba/document"
 require "deba/break"
 require "deba/heading"
+require "deba/list_item"
 require "deba/paragraph"
 require "deba/text_runner"
 require "deba/extractor"

data/lib/deba/document.rb

@@ -10,6 +10,6 @@ class Deba::Document
   end
 
   def to_s
-    @blocks.map { |block| block.to_s }.join.strip
+    @blocks.map { |block| block.to_s }.join.chomp("\n")
   end
 end

data/lib/deba/extractor.rb

@@ -1,7 +1,7 @@
 class Deba::Extractor
   HEADING_TAGS = %w(h1 h2 h3 h4 h5 h6)
   BLOCK_INITIATING_TAGS = %w(article aside body blockquote dd dt header li nav ol p pre section td th ul)
-  ENHANCERS = { %w(b strong) => "*", %(i em) => "_" }
+  ENHANCERS = { %w(b strong) => "*", %w(i em) => "_" }
 
   attr_reader :blocks
 
@@ -59,6 +59,16 @@ class Deba::Extractor
       return
     end
 
+    if node_name == 'li'
+      last_item = node.xpath('count(following-sibling::li)').to_i == 0
+      index = node.xpath('boolean(ancestor::ol)') ? (node.xpath('count(preceding-sibling::li)').to_i + 1) : nil
+      @text_run.break(Deba::ListItem, last_item, index)
+      node.children.each { |n| process(n) }
+      @text_run.break(Deba::Paragraph)
+
+      return
+    end
+
     #These tags terminate the current paragraph, if present, and start a new paragraph
     if BLOCK_INITIATING_TAGS.include?(node_name)
       @text_run.break(Deba::Paragraph)

data/lib/deba/list_item.rb

@@ -0,0 +1,19 @@
+class Deba::ListItem
+  attr_reader :segments
+
+  def initialize(segments, last, index)
+    @segments = segments
+    @last = last
+    @index = index
+  end
+
+  def to_s
+    prefix = if @index.nil?
+      "* "
+    else
+      "#{@index}. "
+    end
+
+    "#{prefix}#{Deba::Stringifier.new(@segments).stringify}\n#{"\n" if @last}"
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: deba
 version: !ruby/object:Gem::Version
-  version: 0.6.0
+  version: 0.7.0
 platform: ruby
 authors:
 - Brenton "B-Train" Fletcher
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2017-01-30 00:00:00.000000000 Z
+date: 2017-02-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -56,7 +56,8 @@ description: Deba takes a HTML document or fragment and extracts the text conten
   into a plaintext format that is a strict subset of markdown.
 email:
 - i@bloople.net
-executables: []
+executables:
+- deba
 extensions: []
 extra_rdoc_files: []
 files:
@@ -68,11 +69,13 @@ files:
 - bin/console
 - bin/setup
 - deba.gemspec
+- exe/deba
 - lib/deba.rb
 - lib/deba/break.rb
 - lib/deba/document.rb
 - lib/deba/extractor.rb
 - lib/deba/heading.rb
+- lib/deba/list_item.rb
 - lib/deba/paragraph.rb
 - lib/deba/stringifier.rb
 - lib/deba/text_runner.rb