Unable to read the last line by using regexp

I used the following code to read text line by line. But the result excludes the last line. For example, if my text is:

void myfunc1() 
{
    // body;
}

by using the following code:

matchedText = regexp(myText,'(?<=\n.*?(\r)\n','match');

The matchedText is:

matchedText = {'void myfunc1()','{','','//body',''};

How can I get the last line of the text? -- '}' is missing.

Thanks.

[Guillaume]

I doubt your regular expression matches anything since it has unbalanced brackets. If the whole purpose is just to split text into lines, then it's overly complex anyway.

What are you trying to do?

[Adam Danz]

"I used the following code to read text line by line..."

Why not use fgetl()?

[dpb]

Or in memory, what about splitlines()?

[Walter Roberson]

The last line might possibly not have a \r\n -- the file might just end.

[Rui Zhang]

I am testing this function before using it.

The last line in my text file doesn't have \r\n as Walter said.

[Guillaume]

Again, the regexp that you've written in the question is invalid.

>> myText = sprintf('\nvoid myfunc1()\r\n{\r\n\t// body\r\n}')
myText =
    '
     void myfunc1()
     {
     	// body
     }'
>> matchedText = regexp(myText,'(?<=\n.*?(\r)\n','match')  %exact copy/paste of what you wrote in the question
matchedText =
  0×0 empty cell array

The regexp is missing a closing bracket. Maybe you meant (notice the ) after the first \n)

  >> matchedText = regexp(myText,'(?<=\n).*?(\r)\n','match')
matchedText =
  1×3 cell array
    {'void myfunc1()←↵'}    {'{←↵'}    {'→// body←↵'}

You'll notice that:

• I had to start the text with a \n because of the look-behind
• you capture all the line return\line feed

Again, there are much simpler ways to split text into lines.

[Rui Zhang]

Oh, I missed a ')' in my post. You are right.

My actual code in my file is:

%% read a line
searchPattern = '(?<=\n).*?(\r)\n';
matchedText = regexp(mytext,searchPattern,'match');

Thank you very much for your help!

Answer 1

[Guillaume]

As I said, your regular expression is invalid, and even corrected doesn't work very well. I also suspect that the (\r) was meant to be (\r)?

Anyway, if the goal is just to split lines:

>> myText = sprintf('void myfunc1()\r\n{\r\n\t// body\r\n}')  %demo data
myText =
    'void myfunc1()
     {
     	// body
     }'
 
>> matchedText = regexp(myText, '\r?\n', 'split')
matchedText =
  1×4 cell array
    {'void myfunc1()'}    {'{'}    {'→// body'}    {'}'}

[Walter Roberson]

Note that when you use split to split into lines that it is common to end up with an empty character vector as the last element. That shows up when the file does end in newline, because it splits at the newline and then there does not happen to be anything more in the string.