Using Regex to Find String by ASCII Number

[KieranRidge]

Hello,

I have an example of a problem I'd like to solve which will help me solve similar problems in the future. Say I have a record labeled:

<FormulaField expression="[Metric D]/[Metric B]" field="Calc1" size="8" type="Double" /> <FormulaField expression="([Metric C]-[Metric B])/[metric B]" field="Calc2" size="8" type="Double" /> <FormulaField expression="([Metric A]*[Metric A])/([Metric B]-[Metric C])" field="Calc3" size="8" type="Double" />

which has so many non-word characters. However I'm looking to count only the number of total appearances of the following non-word characters, ignoring anything else:

(

)

*

+

,

-

In the ASCII Table (https://www.cs.cmu.edu/~pattis/15-1XX/common/handouts/ascii.html) these values are 40-45. How could I write in the Regex code that I only want to count the characters in this ASCII grouping? This post here (https://community.alteryx.com/t5/Alteryx-Designer-Discussions/Remove-ALL-ascii-Characters/td-p/35381...) regards hexagonal grouping which I can't really finagle to work in this case.

File attached (a text input of just the value above).

Thanks,

Kieran

[Claje · Accepted solution]

Hi,

I think the following should work:

REGEX_CountMatches([Text],'[\x28-\x2d]')

I tested this and a manual method and found 9 matches in each.

The key was escaping the hexadecimal value with the "\" character.

Hope this helps! 

EDIT:

Notably, you can also do this (look for a range of characters in ascii) without any hexadecimal values.  The following expression works as well.

REGEX_CountMatches([Text],'[(--]')

[CharlieS]

How about the following expression that can be placed in a Formula tool:

length(REGEX_Replace([Field1],"[^\(\)\*\+\,\-]",""))

The RegEx there will replace any character not in the set with nothing, effectively removing all but the characters of interest. Once that is complete, use the length( function to count the characters.

[MarqueeCrew · Accepted solution]

@KieranRidge ,

I was going to say:  

REGEX_CountMatches([Text],'[\x28-\x2d]')

But @Claje stole my thunder.

Cheers,

Mark 

[danilang]

Hi @KieranRidge 

There are multiple places(here's one) where you can find ASCII charts to help you map the decimal range (40-45) to the corresponding hexadecimal range (\x28-\x2d), that @Claje (and now @MarqueeCrew) used in their regex formula

Dan

[KieranRidge]

Thanks, everybody and @Claje has earned the Accept Solution due to being the quickest and giving two different answers that fit the criteria.

[MarqueeCrew · Accepted solution]

A bunch of folks jumped in to help save you. It's ok to skip me, but please do consider recognizing all who took the time to suggest workable solutions for you.

Cheers,

Mark

[KieranRidge]

Apologies - didn't realize I could assign multiple Accept Solutions. Fixing now.

[MarqueeCrew]

Thanks! Alteryx is always evolving and someone a year from now might just post an easier answer.

Cheers,

Mark