New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

References numbers in PDF abstracts missing #3203

Closed
uhuaha opened this Issue Jan 11, 2018 · 7 comments

Comments

Projects
None yet
2 participants
@uhuaha

uhuaha commented Jan 11, 2018

When converting abstracts (single abstract and book of abstracts) to a PDF the references numbers go missing. For example:

"This is just an example first sentence [1]. In the second example sentence they showed that ... [2] is ..."

[1] ....
... <= This line should start with "[2]" but this number is missing and all following numbers as well if there were more than two references. "[1]" is always present though.

We have Indico version 2.0a1. The output of "root@indico:~# tex --version" yields "TeX 3.14159265 (TeX Live 2015/Debian)". Is there a possibility to get the references numbers in the abstracts? Thank you very much for your help.

@ThiefMaster

This comment has been minimized.

Member

ThiefMaster commented Jan 11, 2018

Could you please provide a non-truncated example to reproduce the issue?

Anyway, guessing from your example: Abstracts use Markdown, so something like [1]: foo declares foo as the location for the link [1] which would usually be written as [something][1] (which would then result in "something" linking to the URL "foo".

@uhuaha

This comment has been minimized.

uhuaha commented Jan 12, 2018

Thanks for your quick answer. Here is a more elaborate case (I copied one of the abstracts from the management area in the "edit contribution" window):

Small clusters in the gas phase are ideal model systems for investigating the fundamental aspects of complex reactions [1,2]. The good control over cluster size, composition, and charge state allows distinguishing the effects of these parameters on the studied reactions. In the high-vacuum conditions of a gas-phase experiment there are no contaminating agents that could affect the reaction mechanism. In addition, for small-sized clusters, ranging from two to a few tens of atoms, direct comparison between experiment and quantum-chemical calculations is possible.
In this talk, I will discuss two examples in which infrared multiple photon dissociation spectroscopy (IRMPD) is used to investigate the interaction of metal clusters with small molecules. First, the adsorption of CO on Mo and Nb doped cationic platinum clusters is studied with the goal to enhance understanding about the CO poisoning effect of Pt catalyst nanoparticles in proton exchange membrane fuel cells. Reactivity measurement in a low-pressure collision cell showed significant reduction in the reactivity of the doped Pt clusters compared to bare Pt clusters, which is attributed to electron transfer from the highly coordinated dopants to the Pt atoms and the concomitant lower CO binding energies [3]. Stretching frequencies of the adsorbed CO molecules reflect dopant-induced charge redistribution within the clusters [4]. 
Second, I will present a study of hydrogen adsorption on the V and Rh doped cationic aluminum clusters. IRMPD spectra, which provide a fingerprint of the hydrogen binding geometry, prove that H$_2$ dissociates upon adsorption for the V doped clusters. Orbital analysis shows that the activation barriers are due to an unfavorable overlap between cluster and hydrogen orbitals [5]. For Rh doped clusters, depending on the cluster sizes, either the H$_2$ binds dissociatively or it adsorbs molecularly in Kubas complexes. This size dependence is not due to kinetic impediment of the hydrogenation reaction by an activation barrier, but to binding energy differences [6].

[1] D. Justes, R. Mitrić, N. Moore, V. Bonačić-Koutecký, A. Castleman, Jr., J. Am. Chem. Soc. 125, 6289 (2003).
[2] S. M. Lang, T. Bernhardt, R. Barnett, U. Landman, Angew. Chem. Int. Ed. 49, 980 (2010)
[3] P. Ferrari, L.M. Molina, V.E. Kaydashev, J.A. Alonso, P. Lievens, E. Janssens, Angew. Chem. Int. Ed. 55, 11059 (2016)
[4] P. Ferrari, J. Vanbuel, N.M. Tam, M.T. Nguyen, S. Gewinner, W. Schöllkopf, A. Fielicke, E. Janssens, Chem. Eur. J. 23, 4120 (2017)
[5] J. Vanbuel, E.M. Fernández, P. Ferrari, S. Gewinner, W. Schöllkopf, L.C. Balbás, A. Fielicke, E. Janssens, Chem. Eur. J. 23, 15638 (2017)
[6] J. Vanbuel, M.Y. Jia, P. Ferrari, S. Gewinner, W. Schöllkopf, M.T. Nguyen, A. Fielicke, E. Janssens, Top. Catal. (2017). https://doi.org/10.1007/s11244-017-0878-x

So, here everything is fine but when I create a PDF from this all numbers [x] starting from [2] in the listed references are missing. [1] is always present though. It seems to me that pdflatex may not function correctly. I recently installed latex/texlive as demonstrated in https://docs.getindico.io/en/latest/installation/latex/. Do you guys have an explanation for the observed behaviour? Thanks so much for your help!

@ThiefMaster

This comment has been minimized.

Member

ThiefMaster commented Jan 12, 2018

OK, I can reproduce this locally:

This output from xelatex seems related (you can find this in /opt/indico/tmp/indixo-texgen-*/ as well)

! Illegal unit of measure (pt inserted).
<to be read again>
                   }
l.122 [2]
          S. M. Lang, T. Bernhardt, R. Barnett, U. Landman, Angew. Chem. Int...

! Illegal unit of measure (pt inserted).
<to be read again>
                   }
l.123 [3]
          P. Ferrari, L.M. Molina, V.E. Kaydashev, J.A. Alonso, P. Lievens, ...

! Illegal unit of measure (pt inserted).
<to be read again>
                   }
l.124 [4]
          P. Ferrari, J. Vanbuel, N.M. Tam, M.T. Nguyen, S. Gewinner, W. Sch...

! Illegal unit of measure (pt inserted).
<to be read again>
                   }
l.125 [5]
          J. Vanbuel, E.M. Fernández, P. Ferrari, S. Gewinner, W. Schöllko...

! Illegal unit of measure (pt inserted).
<to be read again>
                   }
l.126 [6]
          J. Vanbuel, M.Y. Jia, P. Ferrari, S. Gewinner, W. Schöllkopf, M.T...

I wonder if we need to escape anything in latex when square brackets are involved...

@ThiefMaster ThiefMaster added the bug label Jan 12, 2018

@ThiefMaster

This comment has been minimized.

Member

ThiefMaster commented Jan 12, 2018

Minimal abstract content to reproduce it:

Hello world

[1] foo
[2] bar

shows up as

[1] foo
bar

The full generated LaTeX source of this example is attached: bug.tex

@ThiefMaster

This comment has been minimized.

Member

ThiefMaster commented Jan 12, 2018

This sounds like the problem: https://tex.stackexchange.com/a/49105/1651

@ThiefMaster ThiefMaster self-assigned this Jan 12, 2018

@ThiefMaster

This comment has been minimized.

Member

ThiefMaster commented Jan 12, 2018

This will be fixed in 2.0.1; but since it's a rather minor issue we'll wait if more issues show up in 2.0 before doing the first bugfix release (you were a bit unlucky, we released 2.0 just earlier today, otherwise it'd have been included here).

Anyway, it's a very small patch so I'd suggest you to update to 2.0, and then apply the patch manually. To do so, simply run th is command as the indico user. Since it also includes a changelog updated which cannot be applied, you will have to press Enter twice so it skips that part of the patch:

curl -s https://github.com/indico/indico/commit/0839c48ee13f935433bf79de7d07c41c837c9dc7.diff | patch -d ~/.venv/lib/python2.7/site-packages/ -p1

After applying the patch, use touch ~/web/indico.wsgi to reload the server.

@uhuaha

This comment has been minimized.

uhuaha commented Jan 12, 2018

Thank you very much, @ThiefMaster! Upgrading to the latest version and then applying the patch did the trick. Awesome support!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment