SEUM and dynamic content

SeUM does not follow hard coded links, but instead does a fuzzy match based on certain learned elements collected as part of the initial recording. While it's true that if all reference to these elements are missing from the page the transaction will fail. This is normally the desired behavior. It seems like you have a particular use case whereby you don't care where the link takes you, provided it returns results and you're able to click through on the first returned result. Is that correct? Are you using or aware of any product or solution that provides the functionality you describe?

Also when I open the saved XML file, I see very limited detail including only hard links, so is there more to each script that just what is saved in the XML file? Is that additional information generated on the first run?

The hard URL is saved as part of the recording, but this is only one of many components used in the fuzzy match. SeUM was designed from the ground up to work with dynamic content. The link location in the DOM tree for example, is also saved in the recording. All of these elements are then used to make a fuzzy match of what link to follow.

Are you currently experiencing issues with the way SeUM is working, or is this more of a preemptive question?  

Hello,

as aLTeReGo wrote, we use multiple criteria to find element on the page. However you can try one thing.

If your search results page is not changing except results table, you can try to force element search based only on specific location of element on the page.

• Export your recording into XML file and open it. 
• Find action that is trying to click on first result.
• In <action> tag for this action find <element> tag.
• In <element> tag delete <innerHtml> and <attributes> tags with their content. Leave only <element> itself with it's attributes (especially "location" attribute is required).
• Save XML and try to play it back in recorder.

This operation removes some values used for element search and search should be performed only based on element location on the page. If first result is always on the same location in HTML DOM tree, it should be able to always click on first element.

However if you have some dynamic content that can change DOM structure of the page and shift first result to some different DOM location, this will still not work. In this case it would need some user-defined search logic that would be designed for specific page. It could be a good feature.

Please let us know if it works for you.

Hello,

as aLTeReGo wrote, we use multiple criteria to find element on the page. However you can try one thing.

If your search results page is not changing except results table, you can try to force element search based only on specific location of element on the page.

• Export your recording into XML file and open it. 
• Find action that is trying to click on first result.
• In <action> tag for this action find <element> tag.
• In <element> tag delete <innerHtml> and <attributes> tags with their content. Leave only <element> itself with it's attributes (especially "location" attribute is required).
• Save XML and try to play it back in recorder.

This operation removes some values used for element search and search should be performed only based on element location on the page. If first result is always on the same location in HTML DOM tree, it should be able to always click on first element.

However if you have some dynamic content that can change DOM structure of the page and shift first result to some different DOM location, this will still not work. In this case it would need some user-defined search logic that would be designed for specific page. It could be a good feature.

Please let us know if it works for you.