html2tex 0.1.4 → 0.1.5

data/lib/html2tex/basic_processor.rb

@@ -23,7 +23,7 @@ class HTML2TeX
     def next_element
       if scanner.eos?
         nil
-      elsif s = scanner.scan(/<[^>]+>/)
+      elsif s = scanner.scan(/<!--|<[^>]+>/)
         tag(s[/<([^>\s]+)/, 1].downcase)
       elsif scanner.scan(/\s+/)
         whitespace
@@ -79,11 +79,25 @@ class HTML2TeX
         "\\end{itemize}\n"
       when "li"
         "\\item "
-      when "/li"
+      when "/li", "/dd"
         "\n"
+      when "dl"
+        "\\begin{description}\n"
+      when "/dl"
+        "\\end{description}\n"
+      when "dt"
+        "\\item["
+      when "/dt"
+        "]"
+      when "dd"
+        " "
       when /^h\d/
         @squash_next = true
         HeadingProcessor.new(scanner, t, @options).to_tex + "\n\n"
+      when "script"
+        NullProcessor.new(scanner, "</script>", @options).to_tex
+      when "!--"
+        NullProcessor.new(scanner, "-->", @options).to_tex
       else
         ""
       end
@@ -96,3 +110,4 @@ class HTML2TeX
 end
 
 require "html2tex/heading_processor"
+require "html2tex/null_processor"

data/lib/html2tex/null_processor.rb

@@ -0,0 +1,16 @@
+class HTML2TeX
+  class NullProcessor
+    attr_reader :scanner, :options
+
+    def initialize(scanner, end_on, options)
+      @scanner = scanner
+      @end_on  = end_on
+      @options = options
+    end
+
+    def to_tex(buffer="")
+      scanner.scan(%r{.*?#{Regexp.escape(@end_on)}}im)
+      ""
+    end
+  end
+end

data/lib/html2tex/version.rb

@@ -1,3 +1,3 @@
 class HTML2TeX
-  VERSION = "0.1.4"
+  VERSION = "0.1.5"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: html2tex
 version: !ruby/object:Gem::Version 
-  version: 0.1.4
+  version: 0.1.5
 platform: ruby
 authors: 
 - Paul Battley
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-05-16 00:00:00 +01:00
+date: 2010-05-23 00:00:00 +01:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -58,6 +58,7 @@ files:
 - lib/html2tex/basic_processor.rb
 - lib/html2tex/cli.rb
 - lib/html2tex/preamble_processor.rb
+- lib/html2tex/null_processor.rb
 - lib/html2tex/heading_processor.rb
 - lib/html2tex/tex.rb
 - README.md