Regex to handle variability in ID setting

[hellyars]

I need one regex to rule them all.

\w{2}\d{3} works but it captures too much.  Using the first example, it captures both PE015 (the desired target) and LM250.  The challenge is that sometimes the ID is at the start of the line with a space, with a colon, or at the end in parentheses, etc.

RAW_TEXT	ID
PE015 - LM2500 GAS DIGITAL FUEL CONTROL	PE015
GB003 - Twisted Rudder	GB003
AASP WM018	WM018
GW055: SPEED LOGS	GW055
CJ061 Titanium	CJ061
Equipment Upgrades (DY006)	DY006

[DataNath]

Does this do the trick?

Expression: (^\w{2}\d{3}|\w{2}\d{3}$|\(\w{2}\d{3}\))

Workbook attached.

[MarqueeCrew]

\b[a-zA-Z]{2}\d{3}\b

regex_replace([text],".*(\b[a-zA-Z]{2}\d{3}\b).*",'$1')

 try it out and let me know

 cheers

[Qiu]

@hellyars 
I can not really be sure but at least it works on the date set you procided.

.*?([A-Z]{2}\d{3})\W*.*