Extracting URL from email body

[pedrocantor]

Hey everyone, 

I've been trying to solve this myself by viewing the topics/answers that everyone's so helpfully contributed to. There have been a few methods I've looked at but what looks closest/least convoluted is 

Solved: how to automate outlook mail download links - Power Platform Community (microsoft.com)

Reflected in my flows below I tried researching whether I need to convert html to text then search the text to split out the download string. But can't I just have an expression that searches the email for a download link in html without needing to convert to text? 

This is the subject I'm referring to: 

Solved: Extract URL from Email body - Power Platform Community (microsoft.com)

At this point in time I just want to resolve this step and have a crack at the upload to Sharepoint myself (can't learn otherwise!) 

I've had a crack with other methods but I just need some directions, as I get a few steps into a solution before realising perhaps it isn't what I can use

The below seems to be the most promising but I can't see the expression/variable set in the image

Solved: how to automate outlook mail download links - Power Platform Community (microsoft.com)

I have had a few goes at it, such as converting the html to text, then using substring/index of to look for a string (the hyperlink) but it keeps returning 0 or null (different flow actions, I've got a few non-working flows!)

From the below I don't know if I'm taking the instructions too literally. It says search text so i said 'http' but I'm not sure if I need to tell automate 'http' is a string

Got this to output 0 so a little bit of an improvement 😛 

Substring attempt: 

End result (of additional flows I'll add on) is download report from link to SPO, then something reads the spreadsheet, finds the UPN, looks at the manager column, then populates Azure AD

I just need to get the first part sorted!

[v-yueyun-msft]

Hi , @pedrocantor 

According to your screenshot , you need to use string() function to convert the body to string first before using the IndexOf() function.

string(outputs('html'))

If this reply can help you , you can click mark this reply as solution (Accept solution) which can help more people, thanks in advance! 

Best Regards,

Yueyun Zhang

[AlexEncodian]

@pedrocantor 

Encodian has since released a utility action (one that consumes only 0.05 credits per operation) to extract URLs

https://support.encodian.com/hc/en-gb/articles/11056297407261-Utility-Extract-URL-s-from-Text

See this example solution:

https://www.encodian.com/blog/extract-urls-from-text-and-documents-with-power-automate

[pedrocantor]

Thanks for the reply @v-yueyun-msft !

That's a good tip for the progression of this. I don't know if I've intuited your solution correctly, I've kind of just replaced the substring part with string and kept the indexof:

When researching the need to convert the body of email from HTML to text, THEN from text to string, it's opened up a few other challenges

Other articles I've read say to step through the email using a loop function to convert each line in the body (array) to a string:

Solved: Convert array to string - Power Platform Community (microsoft.com)

Another video suggests splitting the text in the array using a delimiter. 

It relies on the email being a fixed format (which is what I'm working with) so this will bypass the need to convert the array to string, and I can use delimiters to parse out the http link I need

Power Automate - Extract Text Values from an Email Body (youtube.com)

Ideally, initializing the body of text from object to string so I can then use IndexOf would be ideal but a few actions to split out the text might not be so bad either

Thanks again for reading

[v-yueyun-msft]

Hi , @pedrocantor 

Thanks for your response! According to your screenshot , you still do not convert the object to string. You can try to use this code first to test in your side.

Best Regards,

Yueyun Zhang

[pedrocantor]

Thank you for doing that for me! 🙂 

This is what I've got now

I used

string(outputs('Html_to_text'),
indexOf(string(outputs('Html_to_text'),'http')))

when I got the below error, I just used

indexOf(string(outputs('Html_to_text'),'http')))

Same error unfortunately

Any additional insight would be appreciated as always 🙂 

[v-yueyun-msft]

Hi , @pedrocantor 

Thanks for your response! Your ')' seems to put to the wrong position.

You can try to use this code:

indexOf(string(outputs('Html_to_text')) ,'http' )

Best Regards,

Yueyun Zhang

[pedrocantor]

Great that worked, thank you!

Though I'm getting it output as: 

Instead I was expecting this 

What it looks like is IndexOf is returning the number of the column/row HTTPs starts at?

[v-yueyun-msft · Accepted solution]

Hi , @pedrocantor 

Yes , you need to use the substring() function to extract the url in your side.

substring(string(outputs('Html_to_text')) ,    indexOf(string(outputs('Html_to_text')) ,'http' )    , < your url length>)

Best Regards,

Yueyun Zhang

[pedrocantor]

Thats great, thank you for doing that!

My version of your expression: 

It didn't return the HTTP string I was after but that's because HTTP appears in the Output headers

I did it with length 20

"Set-Cookie": "ARRAffinity=51c5964be65011b1cad3526b8255ea473f9a641923fda7092968c1277815672b;Path=/;HttpOnly;Secure;Domain=conversionservice-ae.azconn-ae-001.p.azurewebsites.net,ARRAffinitySameSite=51c5964be65011b1cad3526b8255ea473f9a641923fda7092968c1277815672b;Path=/;HttpOnly;SameSite=None;Secure;Domain=conversionservice-ae.azconn-ae-001.p.azurewebsites.net",

So I have some fine tuning to do but you've helped so much

Hopefully if there is someone searching for the same solution, our discussion/documentation/screenshots will assist them