dj2 / ruby-rtf Goto Github PK

Hi,
I was using this gem for parsing rtf file
ruby version is 2.4.0 and rails is 'rails', '~> 5.1.4'when I execute doc = RubyRTF::Parser.new.parse(File.open(path).read)`

where "path" being equal to a rtf file in public folder -- Ruby native File loads file.
CompatibilityError in
ruby-rtf-0.0.3/lib/ruby-rtf/parser.rb:187:in handle_control'`

Any help or reference to any rtf parser without JAVA dependencies I know there are 2 such gems?

Try to parse attached file
test.zip

LibreOffice and all other known to me editors correctly opens it

But parsing it with ruby-rtf result are:

require_relative 'lib/ruby-rtf'

rtf = RubyRTF::Parser.new.parse(File.open('/home/lobashov/temp/test.rtf').read)
rtf.sections.each do |sec|
  puts sec[:text]
end

Outpus (ignoring warnings about unknown control):

�đėĔ

ĖāĠĖ

ĄāĔā




�āĕĖ


 

Tested on ruby-rtf v0.0.5 with ruby 2.7

I could not use the gem. After debugging the problem checked the the code below does not run as it should.

lib/ruby-rtf/parse.rb: 39

while (current_pos < len)
        char = src[current_pos]
        current_pos += 1

        case(char)
        when '\\' then
          name, val, current_pos = parse_control(src, current_pos)
          current_pos = handle_control(name, val, src, current_pos)

I found that src [current_pos] returns an integer, so it never enters the case with the options to char.

After modifying the line 39 for "case (char.chr)" worked properly.

I use ruby 1.8.7 on Windows 7. Would be some incompatibility?

Ran into this while parsing some RTFs in the wild, here's a minimal test case:

{\rtf1 \~}

$ RubyRTF::VERSION
=> "0.0.5"
$ RubyRTF::Parser.new.parse('{\rtf1 \~}')
NoMethodError: undefined method `[]' for nil:NilClass
from [...]/gems/ruby-rtf-0.0.5/lib/ruby-rtf/parser.rb:107:in `parse_control'

Seems like \~ should be a valid directive according to https://www.biblioscape.com/rtf15_spec.htm?

My workaround for now is just to replace instances of \~ with whitespace characters before passing them to RubyRTF::Parser.

RTF has the concept of hidden text. While the library has a flag for text that has been struckthrough, it doesn't seem to have anything to indicate that certain text is visible or not:

2.4.3 :269 > doc.sections[0]
 => {:text=>"HIDDEN TEXT", :modifiers=>{:justification=>:left, :left_indent=>0.0, :right_indent=>0.0, :font=>#<RubyRTF::Font:0x00007fdbdd024ef8 @family_command=:swiss, @name="Calibri", @alternate_name="", @non_tagged_name="", @panose="020f0502020204030204", @number=31506, @character_set=0, @pitch=:variable, @theme=:himinor>, :font_size=>12.0}}