Module: OCR

Extended by:
OCR
Included in:
OCR
Defined in:
lib/ocr.rb

Overview

helper for OCR’ing single digits that were screen captured

Constant Summary collapse

GOCR = 
"gocr"
CACHE = 
{}
CACHE_FILE = 
File.expand_path('~/.sensible-cinema-ocr.marshal')

Instance Method Summary collapse

Instance Method Details

#clear_cache! ⇒ Object 



97
98
99
100
 
# File 'lib/ocr.rb', line 97

def clear_cache!
  CACHE.clear
  File.delete CACHE_FILE if File.exist?(CACHE_FILE)
end

#identify_digit(memory_bitmap, options = {}) ⇒ Object

options are :might_be_colon, :should_invert



43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
 
# File 'lib/ocr.rb', line 43

def identify_digit memory_bitmap, options = {}
  require 'mini_magick' # here because of installation woe, but actually not a big slowdown

  if CACHE.has_key?(memory_bitmap)
    return CACHE[memory_bitmap] unless (defined?($OCR_NO_CACHE) && $OCR_NO_CACHE)
  else
    puts 'cache miss' if $DEBUG && $VERBOSE
  end
  
  if options[:might_be_colon]
    # do special processing <sigh>
    total = (memory_bitmap.scan(/\x00{5}+/)).length
    if total >= 3 # really should be 4 for VLC
      # it had some darkness...therefore have been a colon!
      CACHE[memory_bitmap] = ":"
      return ":"
    end
  end
  image = MiniMagick::Image.read(memory_bitmap)
  # any operation on image is expensive, requires convert.exe in path...
  if options[:should_invert] 
    # hulu wants negate
    # but doesn't want sharpen, for whatever reason...
    # mogrify calls it negate...
    image.negate 
  end

  image.format(:pnm)
  # I think it's VLC full screen that wants sharpening...
  image.sharpen(2) if options[:sharpen] # hulu does *not* want sharpen, though I haven't checked it too closely...

  previous = nil
  p options if $DEBUG
  raise 'you must pass in OCR levels in the player description' unless options[:levels]
  for level in options[:levels]
    command = "#{GOCR} -l #{level} #{image.path} 2>NUL"
    a = `#{command}`
    if a =~ /[0-9]/
      # it might be funky like "_1_\n"
      a.strip!
      a.gsub!('_', '')
      a = a.to_i
      return CACHE[memory_bitmap] = a
    end
  end
  # cache failures here, for VLC's hour clock' sake
  CACHE[memory_bitmap] = nil
  nil
end

#load_from_ocr_seed ⇒ Object 



114
115
116
117
118
119
120
121
 
# File 'lib/ocr.rb', line 114

def load_from_ocr_seed
  for file in Dir[__DIR__ + "/ocr_seed/*.bmp"]
    file =~ /(\d+)\.bmp/i
    digit = $1.to_i
    raise unless digit < 10
    CACHE[File.binread(file)] = digit
  end
end

#serialize_cache_to_disk ⇒ Object 



104
105
106
 
# File 'lib/ocr.rb', line 104

def serialize_cache_to_disk
  File.binwrite(CACHE_FILE, Marshal.dump(CACHE))
end

#unserialize_cache_from_disk ⇒ Object 



108
109
110
111
112
 
# File 'lib/ocr.rb', line 108

def unserialize_cache_from_disk  
  if File.exist? CACHE_FILE
    CACHE.merge!(Marshal.load(File.binread(CACHE_FILE)))
  end
end

#version ⇒ Object 



93
94
95
 
# File 'lib/ocr.rb', line 93

def version
  `#{GOCR} -h 2>&1`
end