Regex Parse from the "right"

[bmalak_98]

Hello. I'm parsing some strings into component parts using the Alteryx RegEx node. I'd like to use a single formula to find the "parsing" characters from the "end of string." I could build a formula with if...then...else, reverse, and regex functions, but think this could be done with node.

In the last two rows of the table below, there are two instances of one of the parsing characters, "X".  I'd like to use the "last" instance of that value, or "from the right."  The character string lengths in front of or behind the parsing character could be variant lengths, so specifying the number of characters to look for is not a preferred solution.  

My current formula is

(-|X|D|R)(.*)

String	RegExOut1	RegExOut2
195/75D14	D	14
7-14.5	-	14.5
7.00-14	-	14
7.00-14	-	14
7.50-15	-	15
7.50-17	-	17
10.5/80-18	-	18
10/10.5-20	-	20
1050/50R32	R	32
10X61X5	X	61X5
10X9X100	X	9X100

But would like to get the portion that follows the "last" instance of the "X" character.  

Desired
RegExOut1	RegExOut2
D	14
-	14.5
-	14
-	14
-	15
-	17
-	18
-	20
R	32
X	5
X	100

These attempts return NULLs.

Returns NULLs:	\>(-|X|D|R)(.*)
Returns NULLs:	(-|\>X|D|R)(.*)

Any ideas? Thank you.

[Pang_Hee_Choy · Accepted solution]

if the second part always is number. you can try to use [\d\.] (any character is number or dot) and $ (refer to the end of the string)

(-|X|D|R)([\d.]+)$

[cmcclellan · Accepted solution]

This will give you 3 fields, but does what you want to do 🙂 

(-|X|D|R)(?!.*(-|X|D|R))(.*)$

[Jonathan-Sherman · Accepted solution]

Hi @bmalak_98,

You could use something along these lines:

.*?([a-z])?([\d\.]+)$

The dollar sign anchors the regex to the end of the string so you're searching for the last iteration of a character a-z (which could or may not be there) followed by the last number (including decimal places)

I've attached my workflow for you to download if needed!

Kind regards,

Jonathan

[bmalak_98]

Thank you, @Pang_Hee_Choy.

That works well.  

[bmalak_98]

Thank you, @cmcclellan.

I can use that as a solution.  

[bmalak_98]

Thank you, @Jonathan-Sherman 

That works great. 

I've incorporated bits and pieces from all the responses so far and have settled on the 

(-|X|D|R)([\d.]+)$

[Pang_Hee_Choy · Accepted solution]

@bmalak_98 

just in case the number is more than 1000 and have thousand separator (,).

you may add a comma in the [\d.]

(-|X|D|R)([\d.,]+)$

[bmalak_98]

Thank you for the update.