Removal of unwanted text in Gmail

[Confused61]

Simple need...

What I want; From using ARRAY option of expected 'starting text', search email text and SLICE expected but unwanted text for removal as identified in the array.

eg "On Sun", "On Mon" etc or similar in the array. I want to check if the text CONTAINS any text starting with what's in the ARRAY and SLICE it for removal from the body.

I've tried a few formula options but the only thing I've found to work is to laboriously identify & check the text for each individual 'starting text', slicing and replacing each time.

Surely there's a shorter, quicker method using ARRAY?

My knowledge of Apply To Each functions is as good as non-existent but I have tried to adapt such a method I've got working for me elsewhere but with no success.

Any help much appreciated!!

[VictorIvanidze]

Could you please show the text  before conversion and the text after conversion?

[Confused61]

OK. You receive an email and it may have been forwarded or it's a reply.

You will receive an email message, responding to your original message, or perhaps it's a forwarded message, with your original message at the bottom. Or even a message received which has a consistent footer appended after the message text.

The email message, will have the original message being responded to (or footer on a new message), appended at the bottom. The response message, will have the original appended message at the bottom and it will start with a date, either starting with the day eg, "On Tue" or possibly even the month first, eg "Mar 28th", or similar.

So you might have a reply with the original message following as follows. Here, the appended original message shown starting with a date in red text;

"

Dear You,

Text text text text text text text text.

Regards,

Me.

On Tue, Mar 28, 2023 at 12:43 PM BST, Jack Black <jblack@gmail.com> wrote:

Hi. The advice is clearly of little use, given the email address I've sent my query to hasn't generated a response as yet.
So, 'admin' is the place to go...

blah, blah, blah.

"

I can easily slice away all the text from "On Tue", using indexOf , leaving just the reply message text, which is what I want.

However, the day (or month) can obviously change, so "On Tue" can be any other day of the week. Ditto months-first date of "Mar 28th" being one of 12 months.

My ARRAY will therefore need to include all the possibilities, ie. 7 days of the week and 12 months of the year. Or any other constant which may occur.

Therefore, I might be able to use for example; createArray('On Sun', 'On Mon', 'On Tue', 'Jan','Feb','Mar', etc etc etc).

I'm looking for a solution which will "search for" any of the array of possibilities in the whole message text (possibly using expression contains), indexing from that point and so enabling all text thereafter to be removed, leaving just the principal message text.

Hope this is clear.

Cheers,

Pete

[VictorIvanidze · Accepted solution]

Thanks for detailed explanation.

Have a look at HTML text of the message. I guess there is unique div tag just before the red On. If yes use it to split the HTML text. Something like

first(split(variables('MessageHTMLtext'),variables('BeforeOnTag')))

[Confused61]

Brilliantly, simple! Why didn't I think of looking at the HTML before conversion to Plain text?🙄.

However, the formula has also saved me a deal of time "getting there". So many thanks for this experienced help and wisdom 😊👌

[VictorIvanidze]

Thanks for you kind words. I have to note that if everybody can describe their problem as clear as you, life become a bit easier 😉

By the way, this formula is better:

concat(first(split(variables('MessageHTMLtext'),variables('BeforeOnTag'))),'</body></html>')

[Confused61]

Your welcome. I'm very grateful, being only a little familiar now with PA after just a few weeks with simple Flows. The main obstacles for me have been 1., to understand the range of possibilities in the combined use of the 'dynamics' and 'expressions' and 2., to identify the constants within texts.

Whilst I've achieved "simple" flows that work, they are over-complex but whilst as in this example, there are better, shorter, more efficient, simpler ways of achieving the same result, sometimes it seems, there's no short-cut. But I don't have an in-depth knowledge of all the uses of the expressions, just reading & learning what appear/s to be the right one/s. EG, I've not 'noticed' the expression and use of 'First' before, so that's been a real help.

I'll have a look at your improved formula but I'm hoping that the HTML approach will deliver the constant/s I need to strip out the unwanted text. As message HTML can vary in composition, the 'tags' can vary and not always be in the place needed if a given tag is adopted as the constant for formula use. Therefore, the 'uniqueness' in positional terms is not guaranteed. Nevertheless, if I only have to focus on a couple of tags, it's still going to save a lot of repetitive 'stages' in the flow.

Can't thank you enough in this respect!! 😊👍

[Confused61]

Hi again. I've hit a snag. A message from an iPhone doesn't show any html tags so the tag-finding formula hasn't worked.

Are you aware of this being the case for ANY message issued from an iPhone, or even other AppleMac devices?

Many thanks for any clues...

[VictorIvanidze]

Hi.

What's content type of such message?

Are you sure it's just a plain text message?

[Confused61]

Well I'm not sure of the plain text question. Seems to suggest somehow that it's not. Odd thing happened though, as the flow was "activity suspended" status this morning for no apparent reason to me. Anyhow, turned back on and worked but this particular message I've had to include a step to remove text from "Sent from".

Any ideas? Cheers.

[VictorIvanidze]

Looks like you have to check if the body is HTML or just a plain text and handle it differently.

For the plain text body you can use -- as a divider.

[Confused61]

Yep, it's rather unexpected though. Must be something to do with how iPhones or the like format text in messages, at least🤔👍

[Confused61]

Hi again. Unfortunately the fab first(split(etc formula, which will be of great use to me, isn't going to help as I was expecting using HTML text body. The variety of HTML in different messages doesn't deliver the key textual position constants I need. So, I'm looking to having to revert to multiple steps to identify and remove what I don't want.

Is there any way I can utilise the first(split(etc formula with an array of the constants I do have or can expect?

Many thanks for any thoughts on this.

Pete