Hi to all.
I have a problem like this:
I have to navigate within a web tool.
Collect elements in the table.
How can I collect such an item?
For Middleware… i neet do collect “message server”…
For Change Logs… “use change logs” and “Include change…” …
Does anyone know how to do?
Thanks for any suggestions
Aaron
Hello @AaronMark ,
You can start with this control:
You can also look up Anchor Base elements that will serve as reference to your check boxes.
For example, in this case the “Use Change Logs” text is the Anchor element for its checkbox that sits to the left of the text anchor.
The other option is to get an attribute of the checkbox element by using the Get Attribute activity. Usually, a selected Check box will have an attribute named checked. It may also have a name that will be related to its function (in this case Use Change Logs)
Thanks
I had tried this approach, but it becomes very complex to manage it.
Basically the activity as a whole, wants to collect the data present on a WEB page, build a table, and copy it on a word document.
The page that contains these elements is not constant. (number of rows, cell names, varies from type to type).
Can someone suggest me an alternative and adaptable way to collect all the data in the table?
At the moment I’m collecting element by element.
“Find Element” associated with an anchor base by importing each element into the database.
I was wondering if there was a faster method.
Perhaps taking advantage of an image scraping? However, it is an activity that I do not know how to use correctly and I do not know if it is for me.
I need help!
Hi!
What do you mean by collect item?
Do you wants to use Check/Uncheck the item or you wants to extract the Message server Text?
Regards,
NaNi
I have to collect the header of the cells, and the text of its content, where the cells are.
In the boxes where a setting is selected, I need to collect the text of that setting.
In my first example, I should create a table like this:
Name || JDBC…
Connection Type || Uplinked
Database Type || Oracle
Configuration Type || Standard
For the second table, something similar.
MIDDLEWARE || Message Server
Change Logs || Use Change logs
Change Logs || Include Change logs…
Or as an alternative, create a static table,
where list, as column A: Message Server,
and in column B: if it is selected, I write YES.
Perhaps this second solution is easier to do.
I don’t know how to determine if it is checked or not.
Are these just 2 examples that you have posted here? Or are your forms of only 2 different types?
I’m asking because if your form keeps changing into infinite formats then that is not a typical repeatable/robotic scenario.
If it’s just two different formats, then you could try the following options:
Screen-scraping
The challenge will be the check boxes/radio buttons. Each type of OCR engine will have it’s challenges. But you have to test it to see how they would turn up using different OCR engines and choose the best one.
Document Understanding
You could try using this technology to create templates out of your form and have the DU framework extract the elements based on which format of the form it is looking at
Third option (I must admit this is a bit nutty!)
Read the HTML code of the form between the <form> and </form> tags (most likely) . A well-formed html is a kind of XML and you may try to use RegEx to look for form field names and if a match is found you can try to use the regex to extract the data between the tags.
Now, why would you want to get the HTML code is the question? The answer to that is you don’t have to struggle with the uncertainties of Screen scraping or DU. The HTML code is definitive and if something exists in your form, it would exist in a formatted state.
Thanks
Hi Andy
There are theoretically only two scenarios.
I do not exclude that there are others, but not infinite.
Maybe 3 or 4. (I’m working in a pre-production environment for now)
Everything you have listed, I think is quite complex for my computer skills.
I have developed more basic bots to date. (click on objects, import data on db, perform calculations, create reports …)
This has a slightly higher difficulty level.
If I may ask. Have you ever made something like this?
Do you have any WF that you can forward to me to understand if I can learn something from an already made xaml?
…and last question, but not least, of the three options, would you ask the Third then?
Ok, then it’s good to start with a narrow field first before diving into a broader solution. Which leads to your other point:
Before you make that assessment, I would suggest you give at least one of the solutions a try.
Especially the third one (unless, someone else on the forum chimes in and presents an easier solution, that would be welcomed very much)
Even if you can’t use RegEx, you should be able to use basic string functions (and using these functions is even easier in StudioX) to extract information from the html code.
At least it’s a start. But if you prefer to stick to the point-and-click solution, then your approach of collection information element-by-element is the easiest, although laborious.
If you could post the html code of the form (by right clicking on the page, view page code) and post your samples here as text files, maybe someone can help you come up with a solution that you can use to work with other form templates.
thanks
