Hey folks,

            Is it possible to have a regex that will parse out groups of characters from the following example. The '|' characters are part of th e text, if anyone's familiar with Fitnesse they'll recognise the syntax


| This | a | is | b | the | c | full | d | method | e | name |

basically the line is delimited using the '|' character and every odd token i.e. This, is, the, full, method, name makes up the whole method name. Thisisthefullmethodname.

Is it possible to write a regular expression that will parse out these values? Even to provide a Java pattern that will allow me to check each odd group?

I've tried the following but to no avail.

(\|[^\|]*?)*

Thanks,
Mark.

Well, there's a trivial non-regex way to do this in Java.

String text = "|This|a|is|b|demo|"
String[] chunks = text.split("\\|");
StringBuffer result = new StringBuffer(text.length());
for ( int c = 1; c < chunks.length(); c+=2 ) // chunk 0 has no content
{
    result.append( chunks[c] );
}
String answer = result.toString();

No?

(Actually, there is a regex in there...the argument to split.)

But if you have some real reason to do this all via a single regex...Ehhh...I don't see one, but it's a fun challenge.

Hey Old Pedant,
         Exactly pal. I realise it could be done fairly handily
with java but one regex to rule them all??? Now yer talking.
If ye know of any decent discussions of grouping and backreferences
that would be great.

Thanks,
Mark.