Answered

Error while extracting HTML table with missing column header

(0) Share

Report

Posted on by Ipsifendus

I'm getting an error in a PAD flow that's using an "Extract data from web page" action to examine the contents of an HTML table that looks like this:

☐	Field1	Field2	Field3	Field4	Field5	Field6
☐	123456	abcde	dsf	adf	sdfaf	dfaf
☐	369121	fghij	dsfad	afa	dsfadf	dafdsf
☐	102030	klmno	sdfds	adfaf	dsfaf	dsfasdf

I'll paste the full error below, but my reading of it is that PAD doesn't like the fact the 1st and 7th columns have no column names. My problem is that I need to flow to identify a particular row in the table based on the "Field" values, and then place a checkmark in the corresponding checkbox.

Has anybody come across this? Are there any workarounds I can use to sidestep this problem?

Error text:

Correlation Id: e7bcfd46-f52a-4387-80fe-aa6fc2f011ba

Column names can't contain null or empty strings.

Parameter name: columnNames: Microsoft.Flow.RPA.Desktop.Modules.SDK.Extended.Exceptions.InternalActionException: Failed to extract data (web page error while extracting data). ---> System.ArgumentException: Column names can't contain null or empty strings.

Parameter name: columnNames

at Microsoft.Flow.RPA.Desktop.UIAutomation.WebAutomation.Runtime.Actions.LocalWebAutomationRuntime.ValidateColumnNames(IReadOnlyList`1 columnNames)

at Microsoft.Flow.RPA.Desktop.UIAutomation.WebAutomation.Runtime.Actions.LocalWebAutomationRuntime.ExtractEntireHtmlTableFromWebPage(WebBrowserInfo webBrowserInfo, WebExtractionSelector tableSelector, IEnumerable`1 columnNames, WebPagerParameters pagerParameters, WebActionRetryParameters retryParameters, Nullable`1 timeout)

--- End of stack trace from previous location where exception was thrown ---

at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()

at Microsoft.Flow.RPA.Desktop.UIAutomation.WebAutomation.Runtime.WebAutomationRuntimeRouter.Invoke(MethodInfo targetMethod, Object[] args)

--- End of stack trace from previous location where exception was thrown ---

at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()

at System.Reflection.DispatchProxyGenerator.Invoke(Object[] args)

at generatedProxy_3.ExtractEntireHtmlTableFromWebPage(WebBrowserInfo , WebExtractionSelector , IEnumerable`1 , WebPagerParameters , WebActionRetryParameters , Nullable`1 )

--- End of remote exception stack trace ---

at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()

at Microsoft.Flow.RPA.Desktop.UIAutomation.Shared.Rpc.RpcDispatchProxy`1.GetRemoteResult

Categories:

Power Automate Desktop

I have the same question (0)

All responses (5)

Answers (1)

Sort by

Deenuji_Loganathan_ 6,250 Moderator on at

Like (0)

Report
Copy link

Link copied!

@Ipsifendus

Is it possible to share the website URL to check from our end and assist you in appropriate ways?

Thanks,
Deenuji Loganathan 👩‍💻
Automation Evangelist 🤖
Follow me on LinkedIn 👥

-------------------------------------------------------------------------------------------------------------
If I've helped solve your query, kindly mark my response as the solution ✔ and give it a thumbs up!👍 Your feedback supports future seekers 🚀. If you'd like to appreciate me, please write a LinkedIn recommendation 🙏

Was this reply helpful? Yes No
Ipsifendus 43 on at

Like (0)

Report
Copy link

Link copied!

Not easily; it's a web-based application rather than a public facing website. The product is a LIMS system called LabVantage. Much of the page content appears to be dynamically generated by javascript. I'm attaching the results of using the browser's "View Source" function on the page in question.

pagesource.docx

Was this reply helpful? Yes No
Agnius Bartninkas Most Valuable Professional on at

Like (0)

Report
Copy link

Link copied!

It seems like you're using the "Entire HTML table" approach for extracting the data.

Capture each column you want individually and build the table, instead of extracting the entire HTML table with one click. Your columns will then get some default column names assigned. You can then either leave them as they are, or set your own custom names. But you will not be allowed to leave any of them blank. When they're not blank, it should work fine.

Was this reply helpful? Yes No
Deenuji_Loganathan_ 6,250 Moderator on at

Like (0)

Report
Copy link

Link copied!
@Ipsifendus
You can try above @Agnius approach and see if it works for you in case if not, you can consider below suggestions.

The above provided source code not helping much here, but I can offer some suggestions to help you solve it.

Since you're already retrieving table values from the columns named field1 to field6, you should have an understanding of the number of rows in your table. Try inspecting the checkboxes to determine their IDs. For instance, if the first checkbox's ID is "customcheck1" and the second one's is "customcheck2," you can make educated guesses about the IDs for the rest of the checkboxes.

Next, introduce a counter starting from one. Iterate through your data table and perform checkbox selection within the loop. Capture the checkbox you want to select within the loop and, once captured, adjust the checkbox ID from being static to dynamic by incorporating the counter variable into the checkbox ID.

Step1: Extract data from the webpage (Retrieve the entire HTML table).

Step2: Use browser inspection tools to identify the checkbox IDs.

Step3: Capture the checkboxes using the "set checkbox on webpage" action.
Step 4: Modify the checkbox ID from static to dynamic within the checkbox UI element.

Follow these steps to dynamically select checkboxes based on your data table. If you encounter difficulties with these actions, you can resort to "get details of web page" to retrieve the page source code and use regex functionality to extract the required element IDs.

Full flow screenshot:

Code:

WebAutomation.LaunchEdge.AttachToEdgeByTitle TabTitle: $'''Error while extracting HTML table with missing col... - Power Platform Community''' AttachTimeout: 5 BrowserInstance=> Browser # Method1: Using Data Extraction WebAutomation.ExtractData.ExtractHtmlTable BrowserInstance: Browser Control: $'''html > body > div:eq(1) > div:eq(3) > div > div > div > div:eq(0) > div:eq(2) > div:eq(0) > div > div:eq(2) > div > div > div > div > div:eq(0) > div:eq(1) > div > div:eq(1) > div > div > div:eq(2) > div > table''' ExtractionParameters: {[$'''Value #1''', $'''Value #2''', $'''Value #3''', $'''Value #4''', $'''Value #5''', $'''Value #6''', $'''Value #7''', $'''Value #8'''], [$'''''', $'''''', $'''''', $'''''', $'''''', $'''''', $'''''', $''''''] } PostProcessData: False TimeoutInSeconds: 60 ExtractedData=> DataFromWebPage SET TotalRows TO DataFromWebPage.RowsCount SET Counter TO 1 LOOP FOREACH DataRow IN DataFromWebPage WebAutomation.SetCheckboxState.SetCheckboxState BrowserInstance: Browser Control: appmask['Web Page \'h ... 598#M38574\'']['Input checkbox \'autoSubscribeToThread\''] State: WebAutomation.CheckboxState.Checked WaitForPageToLoadTimeout: 60 Variables.IncreaseVariable Value: Counter IncrementValue: 1 END # Method2: Get source code of your web page and parse your data using regex WebAutomation.GetDetailsOfWebPage BrowserInstance: Browser DetailsType: WebAutomation.DetailsType.Source Details=> WebPageProperty

Thanks,
Deenuji Loganathan 👩‍💻
Automation Evangelist 🤖
Follow me on LinkedIn 👥

-------------------------------------------------------------------------------------------------------------
If I've helped solve your query, kindly mark my response as the solution ✔ and give it a thumbs up!👍 Your feedback supports future seekers 🚀. If you'd like to appreciate me, please write a LinkedIn recommendation 🙏

Was this reply helpful? Yes No
Verified answer

Ipsifendus 43 on at

Like (1)

Report
Copy link

Link copied!

Wanted to make a note of how this was fixed...I determined that although two columns in the table lacked names in the header row, PAD was assigning the first unnamed column a default name, but not for the second unnamed column in the 6th position. Upon re-extracting the table, I discovered I could type a column name in to the blank space, and that fixed the error.

Was this reply helpful? Yes No