Will there be cases where there are multiple gmail addresses in a given notes field? Or will it be a single gmail address in a given notes field?

a single email address in a field of text

Find @ and parse left to the space and right to the space.

exactly! I’ve tried so many ways to write the formula but nothing works

Like this?

Exactly like that. Bill. you even pulled the emails that were not “gmail” perfect

Great - this is ugly because Airtable has so few tools to work with. It’s like starting a fire with wet hay and a couple of non-ferrous rocks in a 40mph crosswind.

But, it can be done despite the lack of simple string handling tools such as Split().

The formula to do this with Split is -

{Notes}.Split('@')t0].Split(' ')t1].Split(' ')t0]

Without Split() -

CONCATENATE(

    REPLACE(LEFT({Notes}, FIND("@", {Notes})),1,FIND("~",SUBSTITUTE(LEFT({Notes}, FIND("@", {Notes}))," ","~",LEN(LEFT({Notes}, FIND("@", {Notes}))) - LEN(SUBSTITUTE(LEFT({Notes}, FIND("@", {Notes}))," ","")))),""), 

    LEFT(RIGHT({Notes}, LEN({Notes}) - FIND("@", {Notes})), FIND(" ", RIGHT({Notes}, LEN({Notes}) - FIND("@", {Notes}))) - 1)

)

/sigh]

Don’t ask how I know how to do this. Just enjoy…

CONCATENATE(

    REPLACE(LEFT({Notes}, FIND("@", {Notes})),1,FIND("~",SUBSTITUTE(LEFT({Notes}, FIND("@", {Notes}))," ","~",LEN(LEFT({Notes}, FIND("@", {Notes}))) - LEN(SUBSTITUTE(LEFT({Notes}, FIND("@", {Notes}))," ","")))),""), 

    LEFT(RIGHT({Notes}, LEN({Notes}) - FIND("@", {Notes})), FIND(" ", RIGHT({Notes}, LEN({Notes}) - FIND("@", {Notes}))) - 1)

)

Thank you very much! I will enjoy and try to reverse engineer it to try to find out how you pull it off.

Do so at risk to your sanity. :winking_face:

Opps Sorry for this but it seems not to be pulling in the “gmail.com” of the email addresses… Am i missing something?

Or is it just easier to add that part in a different field?

This is one of the Notes texts I tested with:

vdjh bill.french@gmail.com hjfv jhd vjh

This is the outcome from analyzing that row:

bill.french@gmail.com

It appears that it is performing the correct entity extraction at least in terms of this collection of tests.

Here’s another way to tackle it:

TRIM(MID(SUBSTITUTE({Test string}, " ", REPT(" ", 30)), FIND("@", SUBSTITUTE({Test string}, " ", REPT(" ", 30))) - 30, 60))

TRIM(MID(SUBSTITUTE({Test string}, " ", REPT(" ", 30)), FIND("@", SUBSTITUTE({Test string}, " ", REPT(" ", 30))) - 30, 60))

@Justin_Barrett,

That looks far simpler. What is the purpose of the 30 and 60 values in the formula?

I’m replacing each space with 30 spaces, grabbing a chunk of the resulting larger string that begins 30 characters before the @—which is guaranteed to be somewhere the mega-space block just before the email address—and is 60 characters long—which is mostly guaranteed to be in the mega-space block after the address—then trimming the extra spaces off the ends.