"%(E00" (or "text%(E00" ) works on my system. 
Not sure that tells us exactly what type it is using, but it does get you out of the current jam. 
Regards, 

Iain

Hi Iain,

thanks for the quick reply.

Yes this works for this kind of search.

I will have a try, if % is a kind of escape character also for other searches.

Kind regards

Norbert

Just a thought but, if % is an escape character(it seems it is), then searches with literal % will create false results.  I tried a test and got the wrong components back.

There is also the problem that included procs in components can not work with the tool above, as there is no DICT relation between components and included procs.  Also, Precompiler Directives may or may not give unexpeceted results.

Personally, I would only trust in listings.

I suspect you have to escape the escape character. :) 
It will find code in included procs and point you to the included procs, if you change those, you need to use the Cross Reference option to determine which components are affected. 
I'd counter that searching the listings does not tell you the origin of the affected code.... 

Iain

I would still much prefer to use exported XML source code rather than anything else.

Regex, multi line searches. Basically the full potential of a tool like notepad ++

Me too. I do these things directly in git folder with VisualStudioCode (or SublimeText or Notepad++...)

Nowadays I seem more like a forensic than a developer. And archaeologist, when I look for into the git log.

We do have such entries in the (developer) ASN-File
[SETTINGS]
$LISTING_LEVEL=2
$SYMBOL_TABLE =3

[FILES]
*lst       \\\\server\\ufs\\application\\x_lst\\*.lst
*sym    \\\\server\\ufs\\application\\x_sym\\*.sym
*.cmi    \\\\server\\ufs\\application\\x_cmi\\*.cmi

So everytime you compile something, the listing and other things goes to the three folders
In this folders you can search by any tool (that can handle text files )  :-)

Ingo