Parse a SRT file with jQuery Javascript

Question

I'm trying to parse .srt but I get an internal error and I can't figure out what is it.

Here is my code:

    var subtitles;
    jQuery.get('SB_LKRG-eng.srt', function(data) {
        //alert(data);
        function strip(s) {
            return s.replace(/^\s+|\s+$/g,"");
        }
        srt = data.replace(/\r\n|\r|\n/g, '\n');
        //alert(srt);
        srt = strip(srt);
        //alert(srt);
        var srt_ = srt.split('\n\n');
        alert(srt_);
        var cont = 0;
        for(s in srt_) {
            st = srt_[s].split('\n');
            alert(st);
            if(st.length >=2) {
              n = st[0];
              i = strip(st[1].split(' --> ')[0]);
              o = strip(st[1].split(' --> ')[1]);
              t = st[2];

              if(st.length > 2) {
                for(j=3; j<st.length;j++)
                  t += '\n'+st[j];
              }

            subtitles[cont].number = n;
            subtitles[cont].start = i;
            subtitles[cont].end = o;
            subtitles[cont].text = t;
            //alert(subtitles[cont].start);
            }
            cont++;
        }

    });

I can extract the first 4 subtitles and then the code stops and breaks exception: TypeError, I can't understand why... Here a sample of the subtitles file:

1
00:00:01,000 --> 00:00:04,000
Descargados de www.AllSubs.org

2
00:00:49,581 --> 00:00:52,049
Bueno, tienes que escapar, tengo que ir a jugar

3
00:00:52,084 --> 00:00:55,178
Tengo que encontrar un día que está lleno de nada más que sol

4
00:00:55,220 --> 00:00:57,552
Crucero por la calle, moviéndose al compás

5
00:00:57,589 --> 00:01:00,683
Todos los que conoces está teniendo nada más que diversión

6
00:01:00,726 --> 00:01:03,251
Deja todo detrás de ti

7
00:01:03,295 --> 00:01:06,128
Siente esas palmeras soplan

8
00:01:06,165 --> 00:01:09,157
La gente en el norte no puede encontrar

9
00:01:09,201 --> 00:01:11,829
Están fuera de palear la nieve

10
00:01:11,870 --> 00:01:14,998
El tiempo para moverse, pero no seas lento

11
00:01:15,040 --> 00:01:17,941
En sus marcas, prepárate para ir

Part of the code is from: http://v2v.cc/~j/jquery.srt/jquery.srt.js

Can anyone help me?

Thank you

Peluso

Could you post the content of .srt file (or its part)? – Sicken

@Sicken I have posted a sample of the .srt file, thanks! – Peluso

@LGSon I have edited the post because this function is not important for my problem, thanks!! – Peluso

Shouldn't the cont++; be inside the for loop? – Odiliaodille

@LGSon yes it is inside – Peluso

Sorry, I ment inside the if(st.length >=2) { statement? – Odiliaodille

@LGSon thanks for replying. I think cont++ has to be outside if(st.length) because I need the same count for every iteration, because subtitles is a tuple-array, where inside each position there are 4 fields, and I need the same count number to write down each field of one position, am I right? – Peluso

Use push method instead of series of assignments – Sicken

And don't use for..in with array. It's only intended for objects, not arrays. – Sicken

Now, when cont++ is outside the first if(st.length >=2) {, you always count, but you only store in subtitles[cont] when the if(st.length >=2) is true. ... Note though, that I don't mean the inner if(st.length > 2) { – Odiliaodille

Answer 1

var PF_SRT = function() {
  //SRT format
  var pattern = /(\d+)\n([\d:,]+)\s+-{2}\>\s+([\d:,]+)\n([\s\S]*?(?=\n{2}|$))/gm;
  var _regExp;

  var init = function() {
    _regExp = new RegExp(pattern);
  };
  var parse = function(f) {
    if (typeof(f) != "string")
      throw "Sorry, Parser accept string only.";

    var result = [];
    if (f == null)
      return _subtitles;

    f = f.replace(/\r\n|\r|\n/g, '\n')

    while ((matches = pattern.exec(f)) != null) {
      result.push(toLineObj(matches));
    }
    return result;
  }
  var toLineObj = function(group) {
    return {
      line: group[1],
      startTime: group[2],
      endTime: group[3],
      text: group[4]
    };
  }
  init();
  return {
    parse: parse
  }
}();

　

  jQuery.get('demo.srt')
   .done(function(text) {
       try {
         //Array with {line, startTime, endTime, text}
         var result = PF_SRT.parse(text);
       } catch (e) {
         //handle parsing error
       }
  });

Demo

https://jsfiddle.net/5v7wz4bq/

Reena

When I run your code in this fiddle demo, jsfiddle.net/876sscf7/4, it errors out in the middle of the first iteration. Could you explain why your demo doesn't ? ... Btw, when I add my fix for the inner split, then it works jsfiddle.net/876sscf7/5 – Odiliaodille

Your fiddle makes st[1].split(' --> ')[1] empty variable, because script on line2 should perform a global match, find all matches rather than stopping after the first match. jsfiddle.net/21kd95cj – Reena

Wow .... I see that now ... one of my typical quickfix error .. thanks for your time – Odiliaodille

your code is very good. but on some .srt files I see some rows only have time without text and your code stop on this row! can you help me? jsfiddle.net/21kd95cj/4 on row 8 thanks – Accordingly

@Reena Thank you So Much ❤️❤️❤️ – Accordingly

@Reena I have problem, when some rows have more than 2 sentences, this code detect only first sentence. For example jsfiddle.net/5v7wz4bq/3 on row 3 we have 3 sentences and on result, show only first sentence. can you help me? – Accordingly

@Reena I think, I solved it. Please check this: jsfiddle.net/5v7wz4bq/5 I only change /(\d+)\n([\d:,]+)\s+-{2}\>\s+([\d:,]+)\n([\s\S]*?(?=\n{2}|=\n{2}))/gm Thank you – Accordingly

Answer 2

Here is one problem:

o = strip(st[1].split(' --> ')[1]);

At this line, when there isn't any ' --> ' to split, the returned length of the array is 1, which errors when you ask for array item 2.

And here is another:

subtitles[cont].number = n;
....

Neither is the subtitles declared, nor its properties .number, ... etc.

Update

Here is a sample that works (switched the jQuery "read srt file" part for the data)

var data = document.getElementById("data").innerHTML;
data = data.replace(/>/g,">");

function strip(s) {
    return s.replace(/^\s+|\s+$/g,"");
}
srt = data.replace(/\r\n|\r|\n/g, '\n');
srt = strip(srt);
var srt_ = srt.split('\n\n');
var cont = 0;
var subtitles = [];

for(s in srt_) {
    st = srt_[s].split('\n');
    if(st.length >=2) {

        var st2 = st[1].split(' --> ');
        var t = st[2];

        if(st.length > 2) {
            for(j=3; j < st.length;j++)
                t += '\n'+st[j];
        }
        
        subtitles[cont] = { number : st[0],
                            start : st2[0],
                            end : st2[1],
                            text : t
                          }
        
        console.log(subtitles[cont].number + ": " + subtitles[cont].text);
        document.body.innerHTML += subtitles[cont].number + ": " + subtitles[cont].text + "<br>";
        cont++;
    }
}

<div id="data" style="display:none">1
00:00:01,000 --> 00:00:04,000
Descargados de www.AllSubs.org

2
00:00:49,581 --> 00:00:52,049
Bueno, tienes que escapar, tengo que ir a jugar

3
00:00:52,084 --> 00:00:55,178
Tengo que encontrar un día que está lleno de nada más que sol

4
00:00:55,220 --> 00:00:57,552
Crucero por la calle, moviéndose al compás

5
00:00:57,589 --> 00:01:00,683
Todos los que conoces está teniendo nada más que diversión

6
00:01:00,726 --> 00:01:03,251
Deja todo detrás de ti

7
00:01:03,295 --> 00:01:06,128
Siente esas palmeras soplan

8
00:01:06,165 --> 00:01:09,157
La gente en el norte no puede encontrar

9
00:01:09,201 --> 00:01:11,829
Están fuera de palear la nieve

10
00:01:11,870 --> 00:01:14,998
El tiempo para moverse, pero no seas lento

11
00:01:15,040 --> 00:01:17,941
En sus marcas, prepárate para ir
</div>

Odiliaodille

Answer 3

It is better to use the following regex to cover them if the number of lines of text in each section increases

/(\d+)\n([\d:,]+)\s+-{2}\>\s+([\d:,]+)\n([\s\S]*?(?=\n{2}|$))/g

View the output on the console

let subtitle = document.getElementById('subtitle').value;
console.log(_subtitle(subtitle));

function _subtitle(text) {

        let Subtitle = text;
        let Pattern = /(\d+)\n([\d:,]+)\s+-{2}\>\s+([\d:,]+)\n([\s\S]*?(?=\n{2}|$))/g;
        let _regExp = new RegExp(Pattern);
        let result = [];

        if (typeof (text) != "string") throw "Sorry, Parser accept string only.";
        if (Subtitle === null) return Subtitle;

        let Parse = Subtitle.replace(/\r\n|\r|\n/g, '\n');
        let Matches;

        while ((Matches = Pattern.exec(Parse)) != null) {

result.push({
                Line: Matches[1],
                Start: Matches[2],
                End: Matches[3],
                Text: Matches[4],
            })

        }
        
        return result;

    }

<textarea id="subtitle">1
00:00:00,000 --> 00:00:00,600
Hi my friends

2
00:00:00,610 --> 00:00:01,050
In the first line, everything works properly
But there is a problem in the second line that I could not solve :(

3
00:00:01,080 --> 00:00:03,080
But then everything is in order and good

4
00:00:03,280 --> 00:00:05,280
You do me a great favor by helping me. Thankful</textarea>

Village