Using if and else in regular expression

Firstly, in regexp, ?() is not a conditional. ? matches the character (group) to the left of it 0 or 1 times and () starts a capture group with nothing in it?... no conditionals I'm afraid :) The closest might be (a|b) which matches either a or b...

The regexp is a little difficult to read, so

^((\+\d{1,3}(-| )?\(?\d\)?(-| )?\d{1,3})|(\(?\d{2,3}\)?))(-| )?(\d{1,4})(-| )?(\d{6})(( x| ext)\d{1,5}){0,1}$

Try regexper.com, type in the regexp and it will draw you a state diagram...

Using some tabbing to break up the expression:

^(
    (\+\d{1,3}(-| )?
    \(?\d\)?(-| )?
    \d{1,3})
  |(
    \(?\d{2,3}\)?
   )
)

(-| )?(\d{1,4})
(-| )?(\d{6})
(
    ( x| ext)\d{1,5}
){0,1}$

(Note makes some spaces hard to read but we'll go through that by referencing the original)

^ matches the start of a line

The next group is ((\+\d{1,3}(-| )?\(?\d\)?(-| )?\d{1,3})|(\(?\d{2,3}\)?))

This has two parts: (X|Y), where X=(\+\d{1,3}(-| )?\(?\d\)?(-| )?\d{1,3}) and Y=(\(?\d{2,3}\)?). This will match either X or Y...

Breaking down X=(\+\d{1,3}(-| )?\(?\d\)?(-| )?\d{1,3}):

1. The outer () are a capture, so strip these...

2. \+ matches a literal plus sign. Note that it has to be escaped with the \ because + is a meta character meaning "match one or more of the previous".

3. \d{1,3} matches any decimal digit eiter 1, 2 or 3 times but no more or less

4. (-| )? matches either - or (space) zero or one times. The ? is wht specifies zero or one times.

5. \(?\d\) matches a literal '(' (notice the escape) zero or one times. Then a decimal digit, then another literal )

6. (-| )? we've seen before (matches either - or (space) zero or one times. The ? is wht specifies zero or one times.)

7. \d{1,3} we've also seen before (matches any decimal digit eiter 1, 2 or 3 times but no more or less)

So we can say that X matches (and captures - that's wat the outer () is doing) any string that starts with a plus, has 1 to 3 digits then possibly a space or a hyphen, a digit inside brackets, possibly another space or hyphen and then another 1 to 3 digits. This is captured as the first capture group... phew!

Breaking down Y=(\(?\d{2,3}\)?):

1. The outer () are a capture so string these...

2. \(? matches a literal ( zero or one times.

3. \d{2,3} matches any digit two or three times

4. \)? matches a literal ) zero or one times

So we can say that Y matches an two or three digit number, possibly surrounded by brackets. This is captures as the first capture group. Jeez!

Now we have X and Y we can see what the first chunk of the regexp matches (brain melting!).

The first chunk, call it CHUNK1 matches and captures either

1. any string that starts with a plus, has 1 to 3 digits then possibly a space or a hyphen, a digit inside brackets, possibly another space or hyphen and then another 1 to 3 digits OR

2. any two or three digit number, possibly surrounded by brackets

Continuing... (-| )? we've seen before (matches either - or (space) zero or one times. The ? is wht specifies zero or one times.)

(\d{1,4}) matches a string of digit characters that is 1,2,3 or 4 digits in length. This forms the second capture group.

(-| )? we've seen before (matches either - or (space) zero or one times. The ? is wht specifies zero or one times.)

(\d{6}) matches a string of exactly 6 digits

So here you are matching a string with a possible space or hypen, 1 to 4 numbers, another possible space or hyphen and then 6 numbers. Call this chunk2

So far we have matched any string consistiing of chunk1 followed immediately by chunk2...

This concludes the main bit of the phone number, the rest appears to handle extensions...

The next bit is (( x| ext)\d{1,5}){0,1}. Lets break this down a little.

1. The surrounding brackets are the capture group.

2. ( x| ext) matches either of the two literal strings ' x' or ' ext' - note the beginning space.

3. \d{1,5} matches any digit 1,2,3,4 or 5 times.

4. {0,1} matches the capture group zero or one times... i.e. the phone number does not need to have an extension

Finally $ matches the end of line.

Hopefully this has broken down the string well enough for you to work through :)