Open gnott opened 5 years ago
Thanks @gnott for getting a head start on this!
I noticed in one article, 00003, there were some extra
tags that pass through into the abstract
At eLife we don't allow references in the abstract as the abstract should be a stand alone piece as it's delivered to other places, such as PubMed, without full text. This is standard in STM Publishing so it may be that we never encountered it as an issue in the library before.
It would be good to do some work to remove <xref>
tags during processing.
Shall we add that as a task in the Crossref eLife repo?
Plan sounds good, thanks!
<abstract>
<p>In this paper, the costs and benefits associated with postponing retirement are
simulated in a standard simulation model for Belgium, using the approach of
<xref ref-type="bibr" rid="bib18">Stock and Wise (1990)</xref>. Unlike
earlier microsimulation-based applications of this approach, such as Gruber and
Wise (<xref ref-type="bibr" rid="bib9">1999</xref>, <xref ref-type="bibr"
rid="bib10">2004</xref>), this model does not take a representative sample
as the point of departure, but simulates the costs and benefits of postponing
retirement for four fictitious employees, representing male and female white-
and blue-collar workers. While confirming conclusions drawn by other authors,
this model allows for the separation of specific retirement schemes, and of the
effect of different fiscal regimes for those retired and working. It is shown
that differences between retirement schemes show up in differences in
replacement rates and by whether or not the retirement benefit is a function of
career length. Furthermore, advantageous fiscal regulations for the retired have
a strong impact on the implicit costs of postponing retirement.</p>
</abstract>
I converted this to an issue so I can add some results of my initial test.
I tried converting four IJM XML in the Crossref generation library, with very minimal configuration variables, which are also unknown or incomplete at this time:
The results are fairly good. All four output files are valid when I test on the Crossref validation tool.
I noticed in one article,
00003
, there were some extra<xref>
tags that pass through into the abstract. I could check why these end up in the Crossref deposit abstract, which can probably be removed by the library.I didn't check all the XML files, but I could convert them all when we have actual URLs, and these could probably be uploaded via the Crossref web GUI fairly quickly.
@Melissa37 these are results from my initial test, looks fairly good and the basics are supported.