Power Platform Community Forum Thread Details

Context:
I am building a web scraper in Power Automate Desktop (Attended). The flow iterates through a list of URLs, opens each in a browser instance, and attempts to extract details from a declaration page.

The Issue:
I keep getting the error Data extraction element not found when trying to extract simple text values (e.g., "Date", "Net Weight") that are visually present on the page.

Technical Details:

Target Site: Uses Vue.js / Nuxt.js (Dynamic DOM).
Page Structure: The data looks like a standard table.
Error: Microsoft.Flow.RPA.Desktop.Modules.SDK.Extended.Exceptions.InternalActionException: Data extraction element not found.

What I have established (The "Smoking Gun"):
I used PAD's UI Element picker to capture both the Label ("Date") and the Value ("2025-09-26"). PAD generated these Full XPaths, proving they are siblings in the same row:

Label Full XPath: /html/.../table/tbody/tr[1]/td[1]
Value Full XPath: /html/.../table/tbody/tr[1]/td[2]

What I have tried (and failed):

Default Selector: PAD generates a specific selector like td[Text="2025-09-26"]. This works for the first record but obviously fails on the next one where the date changes.
Relative XPath (Sibling): I tried //td[contains(., 'Date')]/following-sibling::td[1]. Result: Error "Element not found".
Relative XPath (Parent TR): I tried //tr[td[contains(., 'Date')]]/td[2]. Result: Error "Element not found".
Scoped XPath: I tried constraining it to the main container: //div[@id='main-content']//tr[td[contains(., 'Date')]]/td[2]. Result: Error "Element not found".
Wait Actions: I added a Wait for page content action before extraction to ensure the page is loaded (waiting for a reliable anchor element). The page is loaded.
Cache Clearing: I deleted the UI element and the Action, recreated them from scratch, and pasted the custom XPath into the "Text Editor" of the selector builder to ensure no old selectors were cached.

My Question:
Why does PAD fail to find the element using a valid relative XPath (//tr[td[contains(., 'Date')]]/td[2]) when the Full XPath clearly shows a standard tr/td structure?
Is there a specific quirk with Vue.js/Nuxt.js hydration in PAD's browser automation that makes these nodes "invisible" to XPath despite being visible in the DOM?

Any advice on a more robust selector strategy for this scenario would be appreciated.

Categories:

Building flows