Power Automate: Remove Characters From A String

A Power Automate flow can unexpectedly fail to due invalid characters in a file name. To solve this problem we can obtain the file name and use Power Automate to remove characters from a string. I will show you two approaches to do this: defining a list of non valid characters and removing them from a filename and creating a list of valid characters and keeping them in the file name.

Table of Contents
• Introduction: The File Name With Special Characters

#1 Remove Invalid Characters Method
• Remove Characters From A String In Power Automate
• Flow Expressions To Remove Characters From A String
• Run The Flow To Remove Characters From A String
• Understanding How The Flow To Remove Characters Works

#2 Keep Only Valid Characters Method
• Keep Only Valid Characters In A Text String
• Flow Expressions To Keep Only Valid Characters In A Text String
• Run The Flow To Keep Only Valid Characters In A String
• Explaining The Flow To Keep Only Valid Characters In A String

Introduction: The File Name With Special Characters

A member of the Technology department uses a Power Automate flow to ensure filenames have only valid characters.

The flow removes any special characters from the filename.

#1 Remove Invalid Characters Method

Remove Characters From A String In Power Automate

Open Power Automate and create a new instant flow to remove the non valid characters from a text string. Add all of the actions shown the image below to the flow. It will take the following steps:

• Compose a string of the non valid characters to be removed from the text string
• Pass in the string value to be cleaned
• Check each character in the string value one-by-one
• Filter and remove any non-valid characters
  
  

Flow Expressions To Remove Characters From A String

Copy and paste these text values and flow expressions into the appropriate actions to quickly build the flow.

Compose: Non Valid Characters – Inputs field

~`!@#$%^&*()+|{}[]<>/\?;:.,

Compose: Text String To Validate – Inputs field

data$batch!-2024.03.02

Select: Array Of Text String Chars To Validate – From field

range(0,length(outputs('Compose:_Text_String_To_Validate')))

Select: Array Of Text String Chars To Validate – Map field – Index value

indexOf(outputs('Compose:_Non_Valid_Characters'), substring(outputs('Compose:_Text_String_To_Validate'),item(),1))

Filter array – Index Equals -1 (choose a value field)

item()['Index']

Select: Array Of Text String Chars To Validate – Map field – Char value

substring(outputs('Compose:_Text_String_To_Validate'),item(),1)

Compose: Valid Text String – Inputs field

join(body('Select:_Array_Of_Valid_Characters'), '')

Run The Flow To Remove Characters From A String

After the flow is created give it a test to run ensure it works as expected. The text string in our example data$batch!-2024.03.02 is changed to databatch-20240302.

Understanding How The Flow To Remove Characters Works

The flow successfully removes non valid characters from a string but how does it work? The flow action Select: Array Of Text String Chars To Validate builds an array of characters in the text string paired with an index number. An index of -1 means the character was not found in the string of non valid characters. Indexes with a value not equals to -1 have a non valid character.

Then we filter the array to show only valid characters in the action Filter array: Index Equals -1.

The filtered array is transformed into a simple array of characters in Select: Array Of Valid Characters which are re-joined into a text string in the action Compose: Valid Text String.

#2 Keep Only Valid Characters Method

Keep Only Valid Characters In A Text String

Compiling a list all of the non valid characters which can be found in a string is difficult. It is simpler to approach the problem by defining a list of valid characters instead. We will create another flow to illustrate this method.

Open Power Automate and create a new instant flow to remove the non valid characters from a text string. Add all of the actions shown the image below to the flow

Flow Expressions To Keep Only Valid Characters In A Text String

Copy and paste these text values and flow expressions into the appropriate actions to quickly build the flow.

Compose: Valid Characters – Inputs field

ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789_-

Compose: Text String To Validate – Inputs field

data$batch!-2024.03.02

Select: Array Of Text String Chars To Validate – From field

range(0,length(outputs('Compose:_Text_String_To_Validate')))

Select: Array Of Text String Chars To Validate – Map field – Index value

indexOf(outputs('Compose:_Valid_Characters'), substring(outputs('Compose:_Text_String_To_Validate'),item(),1))

Filter array: Index Does Not Equal -1 (choose a value field)

item()['Index']

Select: Array Of Text String Chars To Validate – Map field – Char value

substring(outputs('Compose:_Text_String_To_Validate'),item(),1)

Compose: Valid Text String – Inputs field

join(body('Select:_Array_Of_Valid_Characters'), '')

Run The Flow To Keep Only Valid Characters In A String

Now that the flow is complete perform a test to run ensure it works properly. Once again, the text string in our example data$batch!-2024.03.02 is changed to databatch-20240302.

Explaining The Flow To Keep Only Valid Characters In A String

Just as before, the flow action Select: Array Of Text String Chars To Validate builds an array of characters in the text string paired with an index number. An index of -1 means the character was not found in the string of non valid characters. Indexes with a value not equals to -1 have a non valid character.

But this time we filter the array differently to show only characters that do not equal -1 in the action Filter array: Index Does Not Equal -1.

Finally we, reconstruct the simple array of characters in Select: Array Of Valid Characters are once again joined into a text string in the action Compose: Valid Text String.

Questions?

If you have any questions or feedback about Power Automate: Remove Characters From A String please leave a message in the comments section below. You can post using your email address and are not required to create an account to join the discussion.