Class: SmTranscript::TtmlReader

Inherits:
Object
  • Object
show all
Defined in:
lib/sm_transcript/ttml_reader.rb

Constant Summary collapse

BASE_DATE =

these constants help calculate timed text start, end, and duration times. The choice of BASE_DATE is completely arbitrary. It simply allows us to remove the day, month and year from the calculation.

"1/1/1970"
BASE_TIME = 
Time.parse("#{BASE_DATE} 0:0:0")

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(src_file) ⇒ TtmlReader 



26
27
28
29
30
31
 
# File 'lib/sm_transcript/ttml_reader.rb', line 26

def initialize(src_file)
  @metadata = {}
  @words = []
  parse_metadata(src_file)
  parse_words(src_file)
end

Instance Attribute Details

#metadata ⇒ Object (readonly)

Returns the value of attribute metadata.



12
13
14
 
# File 'lib/sm_transcript/ttml_reader.rb', line 12

def metadata
  @metadata
end

#words ⇒ Object (readonly)

Returns the value of attribute words.



13
14
15
 
# File 'lib/sm_transcript/ttml_reader.rb', line 13

def words
  @words
end

Class Method Details

.from_file(file_name) ⇒ Object 



21
22
23
24
 
# File 'lib/sm_transcript/ttml_reader.rb', line 21

def self.from_file(file_name)
  # p File.expand_path(file_name)
  new(File.open(file_name))
end

Instance Method Details

#get_millisecs(time_val) ⇒ Object

convert a rfc2822 formated string representing hours, minutes, and seconds into an integer representing the equivalent number of milliseconds on an undefined date. Time.parse() returns seconds.



108
109
110
111
112
113
114
 
# File 'lib/sm_transcript/ttml_reader.rb', line 108

def get_millisecs(time_val)
  t = time_val.to_s
  if t.match(/\d+ms/).nil?
    t = (Time.parse("#{BASE_DATE} #{t}") - BASE_TIME) * 1000
  end
  t.to_i
end

#parse_metadata(src_file) ⇒ Object 



33
34
35
 
# File 'lib/sm_transcript/ttml_reader.rb', line 33

def parse_metadata(src_file)
  # not currently parsing metadata from ttml files
end

#parse_words(src_file) ⇒ Object 



37
38
39
40
41
42
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
92
93
94
95
96
97
98
99
100
101
102
 
# File 'lib/sm_transcript/ttml_reader.rb', line 37

def parse_words(src_file)
  doc = Nokogiri::XML(open(src_file.path))
  # each word/phrase is the content of a <p> element
  doc.css('p').each do |e|
    start_time = get_millisecs(e.attr("begin"))
    duration   = get_millisecs(e.attr("dur"))
    end_time   = get_millisecs(e.attr("end"))

    duration = end_time - start_time unless duration.to_i > 0
    end_time = start_time + duration unless end_time.to_i > 0

    # replace tab characters with space
    phrase = e.inner_html.gsub(/\t/, " ")
    
    # replace <br></br> or <br/> with %
    phrase = phrase.gsub(/(<\/?br ?\/?>)+/, "%")

    # replace newline characters with %
    phrase = phrase.gsub(/\n/, "%")

    # replace series of % characters with single %
    phrase = phrase.gsub(/%+/, "%")

    # remove leading % character
    phrase = phrase.gsub(/^\s*%\s*/, "")

    # remove trailing % characters
    phrase = phrase.gsub(/\s*%\s*$/, "")

    # remove more trailing % characters
    phrase = phrase.gsub(/%\s*$/, "")

    # Split phrase at delimiter to create smaller phrases.  Why? Some
    # ttml files have phrases too long for useful karoke.  Fortunately
    # they're broken by newlines and <br/> elements.  So we take advantage
    # of this to split the phrases into smaller chunks.
    arr = phrase.split(/%/)

    # invariant: arr.length > 0
    if arr.length == 0
      next
    end
    
    if arr.length == 1
      # no break point so pass it through
      @words << SmTranscript::Word.new(start_time, end_time, duration, phrase)
    else
      # we don't know the actual duration, so make each equal.
      phrase_duration = duration / arr.length
      seg_start_time = start_time
      arr.each do |seg|
        end_time = seg_start_time + phrase_duration
        @words << SmTranscript::Word.new(seg_start_time, end_time, phrase_duration, seg)
        seg_start_time = seg_start_time + phrase_duration
      end
    end

    # p "#{start_time}, #{duration}, #{end_time}"
    # p e.content
    # p e.content.scan(/<br\/>/).length
    # this regex selects any HTML tags (the gsub removes it)
    # phrase = e.content.gsub(/<\/?[^>]*>/, "")

    # @words << SmTranscript::Word.new(start_time, end_time, duration, phrase)
  end
end