Changing letters to other letters with regexprep

6 visualizaciones (últimos 30 días)
Ryoma Kawakami
Ryoma Kawakami el 28 de Sept. de 2018
Comentada: Walter Roberson el 28 de Sept. de 2018
I'm trying to change letters in a string to different ones with one line of code. I simplified my code down to the following:
regexprep('ab',{'a','b'},{'b','c'});
Ideally, this would output 'bc'. Currently, regexprep runs through the string twice and ends up outputting 'cc'.
Is there any way to prevent this from happening? Is it possible with regexprep? Also, I would prefer to keep the letters as strings, as I'm planning on using the 'preservecase' condition.

Respuesta aceptada

Stephen23
Stephen23 el 28 de Sept. de 2018
Editada: Stephen23 el 28 de Sept. de 2018
You don't need regexprep, using ismember and indexing is simpler:
>> str = 'a':'z'
str = abcdefghijklmnopqrstuvwxyz
>> old = 'bcd';
>> new = 'abc';
>> [X,Y] = ismember(str,old);
>> str(X) = new(Y(X))
str = aabcefghijklmnopqrstuvwxyz
Note that this works efficiently for any characters, not just a limited subset. In order to use regexprep you would probably need to use a dynamic replacement expression.
  1 comentario
Walter Roberson
Walter Roberson el 28 de Sept. de 2018
Well, any characters representable in the first 65536 code points ;-) If you have codepoints beyond that, such as Linear B or CJK Unified Ideograms like ? then you would need more work.

Iniciar sesión para comentar.

Más respuestas (2)

OCDER
OCDER el 28 de Sept. de 2018
Note that regexprep will replace things Sequentially.
regexprep('ab',{'a','b'},{'b','c'});
ab -> bb -> cc
To fix, reverse order like this:
regexprep('ab',{'b', 'a'},{'c', 'b'});
ab -> ac -> bc
  2 comentarios
Ryoma Kawakami
Ryoma Kawakami el 28 de Sept. de 2018
I apologize for not mentioning what I exactly planned to do after this.
I want regexprep (or something similar) to work for any shift in letters (such as 'bcd' to 'abc'), which means that reversing the order of evaluation wouldn't work.
OCDER
OCDER el 28 de Sept. de 2018
replace('a b c', {'a', 'b', 'c'}, {'c', 'a', 'b'})

Iniciar sesión para comentar.


Walter Roberson
Walter Roberson el 28 de Sept. de 2018
old = 'bcd';
new = 'abc';
lookup = char(0:255);
lookup(old+1) = new;
str = 'a':'z'
str = lookup(str+1)
You do not need to add 1 if you are willing to omit the possibility of char(0) being in the string.
  2 comentarios
Stephen23
Stephen23 el 28 de Sept. de 2018
Editada: Stephen23 el 28 de Sept. de 2018
Interesting. This will not work for characters >255 (or whatever limit you pick).
Walter Roberson
Walter Roberson el 28 de Sept. de 2018
Correct. You can extend it to max(str) easily, though.

Iniciar sesión para comentar.

Categorías

Más información sobre Characters and Strings en Help Center y File Exchange.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by