Split a word with regexp in matlab; startIndex for 'split'?

Question

My aim is to generate the phonetic transcription for any word according to a set of rules.

First, I want to split words into their syllables. For example, I want an algorithm to find 'ch' in a word and then separate it like shown below:

Input: 'aachbutcher'
Output: 'a' 'a' 'ch' 'b' 'u' 't' 'ch' 'e' 'r'

I have come so far:

check=regexp('aachbutcher','ch');

if (isempty(check{1,1})==0)          % Returns 0, when 'ch' was found.

   [match split startIndex endIndex] = regexp('aachbutcher','ch','match','split')

   %Now I split the 'aa', 'but' and 'er' into single characters:
   for i = 1:length(split)
       SingleLetters{i} = regexp(split{1,i},'.','match');
   end

end

My problem is: How do I put the cells together, such that they are formatted like the desired output? I only have the starting indexes for the match parts ('ch') but not for the split parts ('aa', 'but','er').

Any ideas?

Take the diff of `startIndex` to get the lengths? – Floris Apr 22 '14 at 21:22 — Floris, Apr 22 '14 at 21:22

Daniel · Accepted Answer · 2014-04-23T11:28:25.930

0

You don't need to work with the indices or length. Simple logic: Process first element from match, then first from split, then second from match etc....

[match,split,startIndex,endIndex] = regexp('aachbutcher','ch','match','split');

%Now I split the 'aa', 'but' and 'er' into single characters:
SingleLetters=regexp(split{1,1},'.','match');

for i = 2:length(split)
   SingleLetters=[SingleLetters,match{i-1},regexp(split{1,i},'.','match')];
end

edited Apr 23 '14 at 11:28

answered Apr 22 '14 at 21:59

Daniel

36,610
3
36
69

worked perfectly, thanks! can you correct the typo in the third line (the 'i'). I thought that method would not work when the word starts with 'ch', but luckily split{1,1} is an empty entry in that case. – jonen Apr 23 '14 at 11:05

score 0 · Answer 2 · answered Apr 22 '14 at 22:04

So, you know the length of 'ch', it's 2. You know where you found it from regex, as those indices are stored in startIndex. I'm assuming (Please, correct me if I'm wrong) that you want to split all other letters of the word into single-letter cells, like in your output above. So, you can just use the startIndex data to construct your output, using conditionals, like this:

check=regexp('aachbutcher','ch');

if (isempty(check{1,1})==0)          % Returns 0, when 'ch' was found.

    [match split startIndex endIndex] = regexp('aachbutcher','ch','match','split')

    %Now I split the 'aa', 'but' and 'er' into single characters:
    for i = 1:length(split)
       SingleLetters{i} = regexp(split{1,i},'.','match');
    end

end

j = 0;
for i = 1 : length('aachbutcher')
    if (i ~= startIndex(1)) && (i ~= startIndex(2)) 
        j = j +1;
        output{end+1} = SingleLetters{j};
    else
        i = i + 1;    
        output{end+1} = 'ch';
    end
end

I don't have MATLAB right now, so I can't test it. I hope it works for you! If not, let me know and I'll take anther shot at it.

Split a word with regexp in matlab; startIndex for 'split'?

2 Answers2