Ruby/Rails text parse into objects

Question

I'm trying to create objects from each recurring set in the text below (an .srt subtitle file):

1
00:02:12,446 --> 00:02:14,406
The Hovitos are near.

2
00:02:15,740 --> 00:02:18,076
The poison is still fresh,
three days.

3
00:02:18,076 --> 00:02:19,744
They're following us.

For example, I could take the three or four lines and assign them to attributes of the new object. So for the first set, I could have Sentence.create(number: 1, time_marker: '00:02:12', content: "The Hovitos are near.")

Start with script.each_line, and what other general structure might put me on the right track? I'm having a hard time with this and any help would be fantastic!

Edit

Some of the messy unfinished code I have so far is below. It actually works (I think). Would you have taken a totally different route? I don't have any experience with this.

number = nil
time_marker = nil
content = []

script = script.strip
script.each_line do |line|
  line = line.strip
  if line =~ /^\d+$/
    number = line.to_i
  elsif line =~ /-->/
    time_marker = line[0..7]
  elsif line =~ /^\b\D/
    content << line
  else
    if content.size > 1
      content = content.join("\n") 
    else
      content = content[0]
    end

    Sentence.create(movie: @movie, number: number, 
      time_marker: time_marker, content: content)
    content = []
  end
end

Mischa · Accepted Answer · 2014-04-12 06:44:32Z

1

Here is a way you can do it:

File.read('subtitles.srt').split(/^\s*$/).each do |entry| # Read in the entire text and split on empty lines
  sentence = entry.strip.split("\n")
  number = sentence[0] # First element after empty line is 'number'
  time_marker =  sentence[1][0..7] # Second element is 'time_marker'
  content = sentence[2..-1].join("\n") # Everything after that is 'content'
end

edited Apr 12, 2014 at 6:44

answered Apr 12, 2014 at 4:30

Mischa

43.4k9 gold badges101 silver badges111 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

Stoic Over a year ago

Assuming fixed width of time_marker is wrong, time can be greater than an hour. But, I like the split on whitespace lines. Somehow, didnt occur to me :)

Mischa Over a year ago

@Stoic, in theory, yes, it's wrong, but I don't know any movies that last longer than 99 hours ;-)

Stoic Over a year ago

(offtopic) :P try watching a Bollywood movie.

aymm Over a year ago

lol, 99 hours? I like both your suggestions, @Stoic and Mischa. Many thanks. The whitespace splitting idea is awesome.

Stoic · Accepted Answer · 2014-04-12 04:32:19Z

Assume that subtitles are in the following variable:

subtitles = %q{1
00:02:12,446 --> 00:02:14,406
The Hovitos are near.

2
00:02:15,740 --> 00:02:18,076
The poison is still fresh,
three days.

3
00:02:18,076 --> 00:02:19,744
They're following us.}

Then, you can do this:

def split_subs subtitles
  grouped, splitted = [], []
  subtitles.split("\n").push("\n").each do |sub|
    if sub.strip.empty?
      splitted.push({
        number: grouped[0],
        time_marker: grouped[1].split(",").first,
        content: grouped[2..-1].join(" ")
      })
      grouped = []
    else
      grouped.push sub.strip
    end
  end
  splitted
end

puts split_subs(subtitles)

# output:
# ➲ ruby 23025546.rb                                  [10:00:07] ▸▸▸▸▸▸▸▸▸▸
# {:number=>"1", :time_marker=>"00:02:12", :content=>"The Hovitos are near."}
# {:number=>"2", :time_marker=>"00:02:15", :content=>"The poison is still fresh, three days."}
# {:number=>"3", :time_marker=>"00:02:18", :content=>"They're following us."}

Collectives™ on Stack Overflow

Ruby/Rails text parse into objects

2 Answers 2

4 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

4 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related