Merge docrails

2 files changed, 63 insertions, 0 deletions

diff --git a/railties/guides/rails_guides/generator.rb b/railties/guides/rails_guides/generator.rb
index 6c0d9f3c3b..f93282db2e 100644
--- a/railties/guides/rails_guides/generator.rb
+++ b/railties/guides/rails_guides/generator.rb
@@ -57,6 +57,7 @@ module RailsGuides
 
           result = view.render(:layout => 'layout', :text => textile(body))
           f.write result
+          warn_about_broken_links(result) if ENV.key?("WARN_BROKEN_LINKS")
         end
       end
     end
@@ -134,5 +135,38 @@ module RailsGuides
         code_blocks[$1.to_i]
       end
     end
+
+    def warn_about_broken_links(html)
+      anchors = extract_anchors(html)
+      check_fragment_identifiers(html, anchors)
+    end
+    
+    def extract_anchors(html)
+      # Textile generates headers with IDs computed from titles.
+      anchors = Set.new
+      html.scan(/<h\d\s+id="([^"]+)/).flatten.each do |anchor|
+        if anchors.member?(anchor)
+          puts "*** DUPLICATE HEADER ID: #{anchor}, please consider rewording" if ENV.key?("WARN_DUPLICATE_HEADERS")
+        else
+          anchors << anchor
+        end
+      end
+
+      # Also, footnotes are rendered as paragraphs this way.
+      anchors += Set.new(html.scan(/<p\s+class="footnote"\s+id="([^"]+)/).flatten)
+      return anchors
+    end
+    
+    def check_fragment_identifiers(html, anchors)
+      html.scan(/<a\s+href="#([^"]+)/).flatten.each do |fragment_identifier|
+        next if fragment_identifier == 'mainCol' # in layout, jumps to some DIV
+        unless anchors.member?(fragment_identifier)
+          guess = anchors.min { |a, b|
+            Levenshtein.distance(fragment_identifier, a) <=> Levenshtein.distance(fragment_identifier, b)
+          }
+          puts "*** BROKEN LINK: ##{fragment_identifier}, perhaps you meant ##{guess}."
+        end
+      end
+    end
   end
 end
diff --git a/railties/guides/rails_guides/levenshtein.rb b/railties/guides/rails_guides/levenshtein.rb
new file mode 100644
index 0000000000..4010b61e26
--- /dev/null
+++ b/railties/guides/rails_guides/levenshtein.rb
@@ -0,0 +1,29 @@
+module Levenshtein
+  # Based on the pseudocode in http://en.wikipedia.org/wiki/Levenshtein_distance.
+  def self.distance(s1, s2)
+    s = s1.unpack('U*')
+    t = s2.unpack('U*')
+    m = s.length
+    n = t.length
+
+    # matrix initialization
+    d = []
+    0.upto(m) { |i| d << [i] }
+    0.upto(n) { |j| d[0][j] = j }
+
+    # distance computation
+    1.upto(m) do |i|
+      1.upto(n) do |j|
+        cost = s[i] == t[j] ? 0 : 1
+        d[i][j] = [
+          d[i-1][j] + 1,      # deletion
+          d[i][j-1] + 1,      # insertion
+          d[i-1][j-1] + cost, # substitution
+        ].min
+      end
+    end
+
+    # all done
+    return d[m][n]
+  end
+end