E-Mail Text Parse Returning Unexpected Results

[DataFun]

I have created a flow that is triggered by an incoming e-mail that meets a set of criteria.  Part of my flow is to create a response e-mail that contains some of the text in the original e-mail.  To do this, I convert the HTML to text, and then I attempt to split the text into one line per row.  When I try to create my output strings, I am not returning the info I would like to return.  Below is a section on my flow.

• Html to Text is performed on the Body of the incoming e-mail
• New Line is a Compose that simply creates a new line, so the output in the subsequent step will go to a new line
• Spit Text splits the text into individual lines:  split(outputs('Html_to_text')?['body'], outputs('New_Line'))
• Filter Out Blank Lines simply removes lines with no text: length(trim(item())) where the result is > 0.

Here is an example of a section of text from the incoming e-mail:

Event Details

Name:  HRA 1:1

Begin:  03. Apr 2023 (Monday)

End:  03. Apr 2023 (Monday)

The data after converting html to text and removing the blank lines looks like the following (the url path is removed for this example):

Line 11:  "Event Details Name: HRA 1:1",

Line 12:  "[https://<url path>] Begin: 03. Apr",

Line 13:  "2023 (Monday) End: 03. Apr 2023 (Monday)",

When I convert the HTML to text, I would expect this info to be on their own lines, however, that is not the case.  In spite of this, I would expect first(split(body('Filter_Out_Blank_Lines')[11], 'Name:  ')) to capture the HRA 1:1 after Name:  .  My expression returns:

[https://<url path>] Begin: 03. Apr

Why is my data not splitting like I would expect it to? In the example above I would expect “Event Details” to be on one line and “Name:  HRA 1:1” to be on the next.

When I try to capture the information after “Name:  “, why is it returning “[https://<url path>] Begin: 03. Apr” instead of “HRA 1:1”?  My line number reference is correct as I tested the expression with 10 & 12 to confirm.

This same issue happens with other text I search for.  It appears to me the issue is with how the text appears after converting HTML to text.

[Nived_Nambiar]

Hi @DataFun 

Could you change the expression highlighted below 

with this : 

split(outputs('Html_to_text')?['body'], decodeUriComponent('%0A'))

Hope it helps 🙂

[VictorIvanidze]

Don't use "HTML to Text" action. Parse the HTML itself.

[DataFun]

I got the same result using the code you suggested.

[VictorIvanidze]

Could you show HTML text before conversion?

[Nived_Nambiar]

Hi @DataFun 

Before converting the email body to html trying this split expression using <br> as delimiter

split({Email_body}, '<br>')

Specify dynamic content of email body replacing {Email_body} 

Hope it helps 🙂

[DataFun]

Here is the html from the incoming e-mail:

<p>&nbsp;</p>

<p>[cid:logo]</p>

<p>[cid:mail-banner-signup]</p>

<p><br />
THANK YOU FOR YOUR REGISTRATION</p>

<p>Dear Registrant,</p>

<p>You successfully registered for the event HRA 1:1<br />
[https://******]. If you have<br />
any further event related questions feel free to contact the event owner.</p>

<p>You have registered for an 1:1 consultation with an HR Advisor and will receive<br />
a calendar invitation with the meeting details. We look forward to connecting<br />
with you!</p>

<p>Event Details</p>

<p>Name:</p>

<p>HRA 1:1 [https://*****]</p>

<p>Begin:</p>

<p>05. Apr 2023 (Wednesday)</p>

<p>End:</p>

<p>05. Apr 2023 (Wednesday)</p>

<p>The following details have been entered:</p>

<p>Q: Please confirm you are a people manager</p>

<p>A: Yes</p>

<p>Q: What topic would you like to discuss during the meeting?</p>

<p>A: General Consultation (local policies, change management, culture &amp; inclusion,<br />
data and org planning, compensation topics)</p>

<p>Q: What specific questions do you have?</p>

<p>A: - Answer #3</p>

<p>&nbsp;</p>

<p>Best Regards,<br />
Your Support Team</p>

<p>&nbsp;</p>

<p>&nbsp;</p>

<p>&nbsp;</p>

<p>&nbsp;</p>

[DataFun]

I think I understand why I am not getting the info as I expect.  The incoming e-mail uses a table template, so what appears to be on the same line in the e-mail gets output to two separate lines when run through html to text.  I assume html-to-text is putting the first column on one line and the second column on the next line for each row in the table.  I copied the text of the message and pasted with no formatting into an e-mail and sent it through my flow and it parsed as I would have expected.  

The e-mail template has some lines with a single column and others with two columns.  Is there a way I can join the columns coming out of the e-mail table template into a single line?

[Nived_Nambiar]

Hi @DataFun 

Does it appears in email with table format, looking at html, i think it is coming as paragraph wise 

[DataFun]

As mentioned above, some lines appear to be a single column, however, the highlighted info below is the info I am trying to retrieve and it appears to be in separate columns.  I can select the first column for all three rows without the second column be selected.  This indicates to me that this information is formatted as two separate columns.  If I hit Reply and select this section of the e-mail, the Table Design tab appears on the Ribbon.  

Here's how the e-mail looks if I add borders.  Ideally, I would like to be able to just pull the Event Details and the responses to the questions.  If that doesn't work, I would be happy if I could remove the first two rows that have images in them so I am left without the lines:

[cid:co-logo]

[cid:mail-banner-signup]

[Nived_Nambiar]

Hi @DataFun 

So to confirm finally what you need is the values highlighted in below image 

and the html structure of that is like below

[cid:logo]

[cid:mail-banner-signup]

THANK YOU FOR YOUR REGISTRATION

Dear Registrant,

You successfully registered for the event HRA 1:1
[https://******]. If you have
any further event related questions feel free to contact the event owner.

You have registered for an 1:1 consultation with an HR Advisor and will receive
a calendar invitation with the meeting details. We look forward to connecting
with you!

Event Details

Name:

HRA 1:1 [https://*****]

Begin:

05. Apr 2023 (Wednesday)

End:

05. Apr 2023 (Wednesday)

The following details have been entered:

Q: Please confirm you are a people manager

A: Yes

Q: What topic would you like to discuss during the meeting?

A: General Consultation (local policies, change management, culture & inclusion,
data and org planning, compensation topics)

Q: What specific questions do you have?

A: - Answer #3

Best Regards,
Your Support Team

Right ?

[DataFun]

Correct.  I wouldn't need the text Name:, Begin:, or End:, but it is fine if I do get it.  Thank you for your help on this.  I am new to Power Automate and I have been trying my best to resolve this on my own.  

[Nived_Nambiar]

One more question, would the questions and field names would be always fixed?

[DataFun]

Yes, the questions are fixed.  Only the answers change.  From time to time we will change the questions, but I can cross that bridge later, if needed.  As for the Event Details, the Name, Begin, and End are fixed and the response details vary from message to message.