Performance / Some documents take hours to convert

[hornschorsch]

Hello, some documents, for example todays document 00315692_2024.xml take hours to be converted into html. It seems as if the software is running in an endless loop because it generates thousands of lines like this:

<section title="block040115">5.1.12&nbsp;<span class="label">Bedingungen für die Auftragsvergabe</span><section title="block04011503"><span class="label">Bedingungen für die Einreichung</span><span class="text">:</span><section title="block0401150301"><span class="label">Elektronische Einreichung</span><span class="text">:â<80><88></span><span class="dynamic-label">Erforderlich</span></section>

After two hours it had generated 23.000 lines like that and i killed the process...

Is this a bug in the software or in the document?

In general it would be great if the performance of this tool could be enhanced or if the bulk download files could be provided in html. This would be one conversion run for you, while at the moment lots of bulk download users have to convert the xml to html on their own.

[bertrand-lorentz]

Which version are you using ?

With the current develop branch, I was able to generate the HTML from the XML notice downloaded from the TED website:

$ time java -jar target/eforms-notice-viewer-0.11.0-SNAPSHOT-app.jar de ../../Downloads/2024-OJS103-00315692.xml
Created HTML file: build\output-html\16-de.html

real    3m43.391s
user    0m0.000s
sys     0m0.125s

It does take some time (almost 4 minutes), I guess because the notice has 158 lots. But it does work in the end.

Please note that, as indicated in the README, the eForms Notice Viewer is a sample application:

The eForms Notice Viewer is intended to demonstrate the use of concepts and resources available in the eForms SDK. You can use it as an inspiration or a starting point for your own applications.

So we haven't looked into performance improvements for it. We would be happy to accept contributions in this area.

Please also note that we offer an online service to convert XML notices to HTML or PDF.
It is the "visualisation API" described at https://docs.ted.europa.eu/api/index.html
I was able to convert this notice to HTML, and got a response in less than 6 seconds.

In addition, you can download the notice in HTML from the TED website: https://ted.europa.eu/en/simap/developers-corner-for-reusers#download-notices-various-formats

For your suggestion to include the HTML files in the bulk download, please contact the TED HelpDesk: https://ted.europa.eu/en/contact

[hornschorsch]

Which version are you using ?

I was using 0.10.0 but even with the current 0.11.0-SNAPSHOT the problem occured. Finally it seemd to be a caching issue, after calling the programm with -f it converted the file in 3 minutes.

Please note that, as indicated in the README, the eForms Notice Viewer is a sample application:
So we haven't looked into performance improvements for it. We would be happy to accept contributions in this area.

I know, but i am no java programmer and i dont know the eforms sdk at all. We are a small agency and have no capacities to employ a java programmer.

Please also note that we offer an online service to convert XML notices to HTML or PDF.
It is the "visualisation API" described at https://docs.ted.europa.eu/api/index.html
I was able to convert this notice to HTML, and got a response in less than 6 seconds.

So can´t you provide the software that is used there to the public?

In addition, you can download the notice in HTML from the TED website: https://ted.europa.eu/en/simap/developers-corner-for-reusers#download-notices-various-formats

I am afraid that this download page and also the API you mentioned earlier are rate limited?

Are these HTML-Files generated on the fly if you download them or are they lying on the server already. In the latter case it would be very easy for you / the EU, to pack them into an archive and provide a link...

Anyway, thanks for your reply!