While testing the Web interactions I found some difficulties. Could be a misinterpretation or a bug in the use of the extension.
If I use the Web Interaction in a Loop, there is not a way to end the Interaction. If I connect the interaction the loop fails.
In this case I have a list of files to download in the same page. It didn’t make sense to me, start the Interaction navigate to the page, send the click, en the interaction multiple times. (Maybe is the only way)
If I don’t run the Web interaction End, the process never ends until starvation.
If I use a headless browser
Question: Which is the correct way to use the web interactions in a loop?
Question: How to change the download directory in the advance browser setting if possible.
Question: Is it possible to change the behavior of opening the pdf file on the page to “download”?
BUG or I need to adjust the column results?
Content Retriever - Works great with static paths.
With relative paths I’m getting these results in the link column, and the process fails. (I will try to test try-catch to avoid keep running the web interaction)
The calculation of the base url is: get left of the first find of /
href=“”**./**encuestas/ensanutnl2022/informes.php
"<?xml version=""1.0"" encoding=""UTF-8""?>
<a class=""on-iframe"" href=""./encuestas/ensanutnl2022/informes.php"">
Informes
</a>"
https://ensanut.insp.mx./encuestas/ensanutnl2022/informes.php
Not a problem
To download the file I got the links in the page
Content Retriever returns
"<?xml version=""1.0"" encoding=""UTF-8""?>
<a href=""../../encuestas/ensanutnl2022/doctos/informes/NvoLeon22_Ensanut.pdf"" target=""_blank"">
<img alt=""InformeFinal2022"" src=""../../encuestas/ensanutnl2022/img/NvoLeon22_Ensanut.png"">
</img>
</a>"
This Link fails: …/…/
https://ensanut.insp.mx.../../encuestas/ensanutnl2022/doctos/informes/NvoLeon22_Ensanut.pdf
Based on the page position, I guess the link should be based on the active page
https://ensanut.insp.mx./encuestas/ensanutnl2022/../../encuestas/ensanutnl2022/doctos/informes/NvoLeon22_Ensanut.pdf
Which the server transform to:
https://ensanut.insp.mx./encuestas/ensanutnl2022/doctos/informes/NvoLeon22_Ensanut.pdf
By now I’m killing the Tasks and adjusting the paths. It is easier than download the files manually.
Thanks