Character encoding issue in abstract with connector

alhb

The character encoding issues have been reported in several posts, but I would appreciate your clarification on how to retrieve data depending on the channel used: add item by connector / by DOI / by .ris file.
For example, with this article :
Schneider, G. (2010). Par la musique, accompagner et soutenir l'enfant né prématurément, sa famille et les professionnels qui les accueillent, du CAMSP au centre néonatal. Contraste, 33(2), 103-116. doi:10.3917/cont.033.0103,
accessed from the French database Cairn, I wonder about the data of the abstract.
- adding by Zotero connector: character encoding issue
- adding by .ris file: character encoding OK
(- adding by DOI : no abstract ; because optional element in Crossref ?)

Could the character encoding issue (adding by connector) be related to the proxy? Or to the translator? Or nothing to do with that?
I'm not sure my question is well formulated...

Thank you for your help.

dstillman

You'd have to say exactly how you're saving when you're seeing the issue so we can try to reproduce it.

alhb

Thank you.
Here is the text retrieved for the abstract field :
Adding item via my browser (Firefox 76.0.1) with Zotero connector :
RésuméLes rares évaluations du dispositif français de protection de l’enfance montrent que les résultats concernant l’état affectif et intellectuel des enfants placés sont mauvais. Une des raisons de cet échec est l’absence de dispositif d’écoute individuelle de l’enfant. Les causes de cette lacune sont décrites, ainsi que les caractéristiques d’une écoute adaptée : continuité, durée, discrétion, supervision. Seul un tel travail fait contrepoids à ce que ressent un enfant confronté à des parents très inadéquats : idéalisation, collage, séduction, terreur. Un exemple détaillé illustre ces principes.

VS.
adding item by .ris :
RésuméLes rares évaluations du dispositif français de protection de l’enfance montrent que les résultats concernant l’état affectif et intellectuel des enfants placés sont mauvais. Une des raisons de cet échec est l’absence de dispositif d’écoute individuelle de l’enfant. Les causes de cette lacune sont décrites, ainsi que les caractéristiques d’une écoute adaptée : continuité, durée, discrétion, supervision. Seul un tel travail fait contrepoids à ce que ressent un enfant confronté à des parents très inadéquats : idéalisation, collage, séduction, terreur. Un exemple détaillé illustre ces principes.

alhb

Damn ! encoding issue was automatically corrected when I posted the comment

dstillman

Again, you have to tell us exactly how you're saving. Just saying you're saving from the Connector isn't sufficient. We open a browser. We want to reproduce this. How do we do it?

dstillman

https://www.zotero.org/support/reporting_problems#provide_steps_to_reproduce

alhb

Ok, sorry.
1. I open the Cairn base from my library site (https://bibliotheque.univ-paris-diderot.fr/) after logging in :
-> https://www-cairn-info.rproxy.sc.univ-paris-diderot.fr/
2. I run a search (by the way, the results page is not recognized by the connector ; web page vs. folder icon)
3. Then I open a result (recognized as an article) : https://www-cairn-info.rproxy.sc.univ-paris-diderot.fr/revue-de-psychotherapie-psychanalytique-de-groupe-2001-2-page-99.htm
4. Then I save the item in Zotero by the connector.

You will find a screenshot of encoding issue here :
https://www.dropbox.com/s/2i7g5bxr2ovq416/encoding_issue_abstract.docx?dl=0

dstillman

OK, I can reproduce the problem with that item. The page you listed in your first post doesn't include an abstract, so (at least for me) the save button doesn't include an abstract at all, and, as you said, if you use the RIS export it's correct.

This second page is just serving bad data in the citation_abstract meta field, and you would need to report that to them. If you view the page source, you can see that the og:description field has correct HTML entities, but in the citation_abstract meta field they're encoded twice, so è becomes é. (They're also including an invalid, encoded <titre> element that doesn't belong there.)

Zotero is importing it correctly given the data they're providing.

alhb

Thank you for your answer and your explanations. I will therefore contact them to point out these anomalies.