how to use jsl to test if a string contains only A~Z and 0~9?

[#1]

how to use jsl to test if a string contains only A~Z and 0~9?

my data shall  contains only A~Z and 0~9 but when 1 or few bit flip then there is not printable characters.

many thanks in advance!

[#2 · Accepted solution]

You can use regular expressions to check this.  Create a new column that contains the following formula:

If (Regex(:Name("2nd"), "^[a-zA-Z0-9]+$") == :Name("2nd"), "Yes", "No")

The new column will have the value "Yes" if it is purely alphanumeric, "No" otherwise.

[#3]

thanks for your reply, PMroz.

What ^ and +$ for?

how can I make it return 1 or 0 so that I could use it together with other comparison algorithm?

[#4 · Accepted solution]

The ^ means the start of the string, the + means to match at least one character, and the $ means the end of the string.

To return a 1 or 0 simply change Yes to 1 and No to 0:

If (Regex(:Name("2nd"), "^[a-zA-Z0-9]+$") == :Name("2nd"), 1, 0)

There are a ton of websites out there that explain regular expressions btw.

[#5]

Thank you@pmroz for the instructions. Very useful with some particularly dirty data.

If anyone else needs a handy example try the following:

If(
	Regex( :"2nd"n, "^[A-Z]+$" ) == :"2nd"n, "Uppercase",
	Regex( :"2nd"n, "^[a-z]+$" ) == :"2nd"n, "Lowercase",
	Regex( :"2nd"n, "^[0-9]+$" ) == :"2nd"n, "Numbers",
	Regex( :"2nd"n, "^[A-Z0-9]+$" ) == :"2nd"n, "Uppercase and numbers",
	Regex( :"2nd"n, "^[a-z0-9]+$" ) == :"2nd"n, "lowercase and numbers",
	Regex( :"2nd"n, "^[A-Za-z]+$" ) == :"2nd"n, "No numbers",
	Regex( :"2nd"n, "^[A-Za-z0-9]+$" ) == :"2nd"n, "All characters",
	Regex( :"2nd"n, "^[A-Za-z0-9]+$" ) != :"2nd"n,
		"includes non alphanumeric characters"
)

[#6]

Hi 

I wonder if regex can be used in text? 

Text = " today I m at FL12, tomorrow I will be at FL102"

 I have rows of text that I want to find out if each row contains FL + anydigit then yes, or print the FL+digit. In this case, it will be FL12 and FL103.

If nor Regex, can you suggest other function?

[#7]

If you only need to know found or not found, it is fairly simple with one regex, if you also need the matches you need to loop over the string while substituting the found matches. Also if you need to test your Regex I suggest regex101

Names Default To Here(1);

original_txt1 = " today I m at FL12, tomorrow I will be at FL 103";
original_txt2 = " today I m at FLb12, tomorrow I will be at FL a123";

// if you only need to know if there is a match combine regex with !IsMissing()
found1 = !Is Missing(Regex(original_txt1, "FL\s?\d+"));
found2 = !Is Missing(Regex(original_txt2, "FL\s?\d+"));
show(found1, found2);

// if you also need the matches loop over the string
matches = {};
While(!Is Missing(rgx_match = Regex(original_txt1, "FL\s?\d+")),
	Insert Into(matches, rgx_match);
	Substitute Into(original_txt1, rgx_match, "");
);
show(matches);

To my knowledge JMP cannot perform "global" regex search with Regex Match() and I have made wish list item for it Add flag to Regex Match() to find all non-overlapping occurances of pattern ). You could most likely use JMP's own Pattern Matching to do tasks like finding multiple matches, but I haven't bothered to learn it. Here is good post about thatPattern Matching 

[#8]

Thanks! The regex101 helps to understand the code you shared.