Java string normalization

Question

Is there a Java library to normalize a string by removing spaces/special characters, lowercase all letters, for example: S-cube Abc' Inc. to scubeabcinc?

Answer 1

There is java.text.Normalizer. Java holds text in Unicode, and é can be written as one Unicode symbol, code point, or as two, an e and a zero-width '. Unicode normalisation is very important, for dictionaries, file names. The Normalizer can be used to decompose into letters and accents (diacritical marks), and with a regex replaceAll remove all accents.

Character has Unicode support giving Unicode names to code points, classifying code points as letters, digits, several scripts etcetera.

There is Collate, Locale oriented, that creates specific keys for words, for ordering, as Comparator. In one locale the order could be AaBbCcĉD.. and in another ABC...abc and such. Locale specifies toUpperCase. For instance in Turkish there is a letter i-without-dot Iı and i-with-dot İi.

And then there is your use-case: a reduction. There is for instance the soundex algorithm (third party) for sound-alike representation. Regex can remove interpunction etcetera with String.replaceAll.

Answer 2

No need for a library other than String, String.replaceAll and String.toLowerCase does what you're looking for:

  String s = "S-cube Abc' Inc.";
  s = s.replaceAll("[^a-zA-Z]", "").toLowerCase();

Answer 3

No library is needed. Just use regex and String#toLowerCase:

String s = "S-cube Abc' Inc.";
s = s.replaceAll("[^a-zA-Z]", "");
s = s.toLowerCase();
System.out.println(s);