Answered

Struggling to convert an HTML table into an array

(0) Share

Report

Posted on by CU16012125-0

Hello, I'm struggling to complete a flow that would convert email body table into an array of key-value.

I start off by capturing the email and converting HTML to Text. The email looks like this:

Dear customer, please see your orders below

Order #	Purchase Date	Return Date
AAA1	2/8/2020	5/8/2020
B21	2/8/2021	6/8/2021
C12	2/8/2022	3/8/2022
DDD1	2/8/2023	7/8/2023
EEF3	2/8/2024	8/8/2024

Please confirm.

First, I need to only keep data that is in the table. The only way I could think of is cutting out everything before the word "Order#" and after "Please", including the word. Is there a better way?

Then, by using delimiters, I convert the text into an array that looks like this: "Order #", "Purchase Date", "Return Date", "AAA1", "2/8/2020" and so on. How can I turn it into an array of key-value that would look like this:

[
{
    "Order": "AAA1",
    "EffDate": "2/8/2020",
    "ExpDate": "5/8/2021"
},
{
    "Order": "B21",
    "EffDate": "2/8/2021",
    "ExpDate": "6/8/2021"
}
]

and so on. Any help is appreciated!

Categories:

Building flows

I have the same question (0)

All responses (4)

Answers (1)

bscarlavai33 738 Super User 2025 Season 2 on at

Like (1)

Report

This should help - https://www.youtube.com/watch?v=IwRKWaTnl3U

Was this reply helpful? Yes No
Suggested answer

MrGott 101 on at

Like (1)

Report

Another way could look like this:

you could even go further and use the xpath function to get the values, but i think this is more intuitive

PS: The first line mus be filtered out. Normaly HTML uses th tag for the table headers, but yours uses td with a strong tag.

Was this reply helpful? Yes No
Chriddle 8,441 Super User 2025 Season 2 on at

Like (1)

Report
To get the table, use somethig like the following expression with your email body (with outputs('Compose') as your email body).
If there are multiple tables, use nthIndexOf.

slice( outputs('Compose'), indexOf( outputs('Compose'), '<table' ), add( indexOf( outputs('Compose'), '</table>' ), 8 ) )

If the HTML table is XML compliant (or you can make it so, e.g. with some string replacements), you can use the following:

(Assuming the HTML is in outputs('Compose'))

From:

skip( xpath( xml(outputs('Compose')), '//tbody/tr' ), 1 )

Map Order:

trim( xpath( item(), 'string(//td[1])' ) )

Map Purchase:

trim( xpath( item(), 'string(//td[2])' ) )

Map Return:

trim( xpath( item(), 'string(//td[3])' ) )

Was this reply helpful? Yes No
Verified answer

Chriddle 8,441 Super User 2025 Season 2 on at

Like (1)

Report
If your values don't contain spaces, you can do the following:

From:

chunk( split( replace( replace( body('Html_to_text'), 'Order # Purchase Date Return Date ', '' ), decodeUriComponent('%0A'), ' ' ), ' ' ), 3 )

Map Order:

item()[0]

Map Purchase:

item()[1]

Map Return:

item()[2]

Was this reply helpful? Yes No