Broken ThesesFR translator

mjthoraval · August 3, 2023

I am trying to import the following thesis:https://www.theses.fr/2014PA066627
On that page it uses the ThesesFR translator. But it can only download the metadata and cannot get the PDF file. Some metadata is put in a note, without any formatting. It would be good to add some line breaks between the different parts.

If I try to import from https://www.theses.fr/2014PA066627# (with a "#" at the end), the ThesesFR translator returns an error, and switches to the Embedded Metadata translator. In that case, it is able to download the PDF file properly: https://www.theses.fr/2014PA066627.pdf

Debug ID: D1216339559

poettli · August 18, 2023

This is weird. I tried to import this item, out of curiosity. It was somehow successful (I got both the metadata and the PDF file), but thrown an error. I restarted both Firefox and Zotero, to check again, and now I can see the behavior described above, i.e., it (definitively?) switched to Embedded Metadata translator (while still showing the icon of the ThesesFR translator). And will use this translator with any entry of theses.fr.

mjthoraval · September 11, 2025

I still cannot get the PDF file from the ThesesFr translator.
Is it possible to fix it?

https://www.theses.fr/2014PA066627
Zotero Desktop 8.0-beta.8+4ae1ea1da (64-bit) - Debug ID: D1785963687
Zotero Connector 5.0.182beta1 - Debug ID: D530935253
Windows 10
Firefox 142.0.1 (64-bit)

Other link tested:
https://theses.fr/2024UPSLS024

AbeJellinek · September 11, 2025

OK, made some updates. The "#" URL thing should be fixed, and we now get PDFs for most theses, though some (e.g., https://theses.fr/2014PA066627) won't work because the PDF link doesn't go straight to the PDF. I can look into improving that further if there's a relatively small set of sites that we need to handle.

Your Zotero Connector should auto-update within a few minutes, or you can update translators manually from the Advanced pane of the Zotero settings.

mjthoraval · September 11, 2025

Thank you very much!

Here are a few examples for which it is still not working:
https://theses.fr/2024UPSLM051
https://theses.fr/2005REIMS006
https://theses.fr/2014PA066166
https://theses.fr/2000PA066381
https://theses.fr/2014PA066627

HAL seems to be the main target website to find the PDF file. It is fine in those cases to import directly from HAL instead of Theses.fr.

Just this one has a direct link which is not in first position, which breaks the import of the PDF file:
https://theses.fr/2005REIMS006