Showing results for 
Search instead for 
Did you mean: 

Support for regex - either in conditions or as an action with capture groups exposed

It would be really good to have support for regular expressions.


This could be surfaced in the UI ideally in both conditions and actions.


A conditional regex would just test whether a string matched against a regex.

An action regex would expose capture groups.

Status: Under Review

This is a great suggestion - we are trying to identify a way to safely run regex in Flow.

Regular Visitor

I have processes where I need to remove timestamps and reformat email attachments before storing them into sharepoint, regex would really help here.

Not applicable

Agreed - 

And to clarify - if I understand correctly, the regex patterns supported by office 365 transport rules do not apply to flow, correct?

New Member


Advocate II

Any updates on regex? I really need it. For instance i have an email adress like "Ramazan_BIYIK <>"  I want to take email section between '<' and '>' . I can't take it with json parse or html conversion so i need regex.

Frequent Visitor

I would put every vote into this.

Regular Visitor

I completely support this, regex would make Flow a whole load more powerful and useful.


Ask  your self... who would create a modern scripting/automation/taskflow system and not include regex support? 

Memorable Member

I want to use the Regular Expression by Microsoft Flow,

for match a pattern, extract a parts to the string.



- Get a product code from a string, like [0-9]

- Get a digits from a string, like \d{3}



Resolver I

Oh yes please! I currently do string manipulation -extract text from sentences- using dozens of variables and complex mixes of substring() and indexOf() functions...


This could be replaced by one call with a corresponding regex and capturing groups.

Kudo Collector

...or to provide an easy way to remove spaces between words.

Super User



Would be really good to have Regex to use within Microsoft Flow - For example I'm trying to extract the profit centre from a file name (anywhere in the name) and then save that to a SP library of the found profit centre name... and when you have lists that comprise thousands of entries for codes in format (\d{2})-(\d{3})-(\d{3}) you don't want to (can't?) be doing if/else matches for that volume of entries!