I'm trying to compile a LaTex file online with sharelatex. It works fine, but on top of my generated references the first seems to be broken. It only shows " [1] ." . All other 25 references seem to be fine. It is also awkward that the [1] reference doesn't even appear in the document anywhere.
I'm not sure, but I think this warning might be connected, since it is the only one I get:
/usr/local/texlive/2017/texmf-dist/tex/latex/biblatex/blx-bibtex.def Package biblatex Warning: Using fall-back BibTeX(8) backend: functionality may be reduced/unavailable.
The bibilography is generated with \printbibliography
I found a solution after looking on the right forum:
https://tex.stackexchange.com/questions/365060/warning-using-fall-back-bibtex8-backendbiblatex-functionality-may-be-redu
I had to use
\usepackage[backend=bibtex,style=numeric,sortcites,natbib=true,sorting=none]{biblatex}
instead of
%\usepackage[backend=bibtex, style=numeric]{biblatex}
I still don't really get why it works now. The warning is still displayed, but the problem with the first reference is gone. When i try to use biber with \usepackage[backend=biber, style=numeric]{biblatex} it simply doesn't output any table of references.
edit:
It seems what I thought fixed it here didn't, but just hid the broken reference in between fine ones. I made a mistake in the .bib file, one closing bracket } to much. I didn't notice it before because I didn't know how to assassinate the broken reference with its counterpart in the .bib file.
Related
I have the following reference identified in file A:
.. _my-label:
and I reference it in file B :
this is a reference to file A :ref:`my-label`
This generates a cross-reference as expected when outputting HTML. However, when outputting LaTeX, it does not and I have the classical warning:
LaTeX Warning: Hyper reference `my-label:my-label' on page XX undefined on input line YY.
Is there a LaTeX trick like double compilation or something similar that I am not doing correctly?
I encountered the same issue. HTML compiled without errors for me, but LaTeX compilation did throw the hyperref errors you described. It seems to me that, for some obscure reason, Sphinx does not create the labels that hyperref tries to reference.
I came up with the following solution: since I do not know how to include the missing labels, I will just make it so that LaTeX does not look for them anymore. In detail, I am doing this by overwriting the \hyperref command.
I included the following code in my conf.py file:
latex_elements = {
'preamble': r'''
\renewcommand{\hyperref}[2][]{#2}
'''
}
This includes the \renewcommand{... in the preamble of the LaTeX document created by Sphinx. It will overwrite the \hyperref command so that it won't try to insert a link, but just print the link text.
Obviously, with this solution, the reference that caused the errors will not appear as hyperlinks in your PDF document, but at least it is compiling without errors.
Note
What I described worked perfectly fine for my use case, however, it is described in the Hyperref manual that the \hyperref command can be invoked in two different ways (\hyperref{URL}{category}{name}{text} and \hyperref[label]{text}). I am only overwriting the second one, as that seems to be the one that Sphinx is using for cross references. However, not accounting for the first one when overwriting the command might lead to issues in some cases.
I have the MvxBindingAttributes.xml in my UI.Droid/Resources/Values but everytime I try to use the bindings in my layouts I get
Warning attribute is not declared.
No resource identifier found for attribute 'MvxBind' in package
'DroidUI'
so for some reason it cannot find those attributes. I have tried to close/reopen/clean/rebuild with no luck, tried removing the layouts and then building (Hoping it would add them at build) and then try to add them again, but nothing seems to work
Have you checked that the MvxBindingAttributes.xml file is marked in properties as an AndroidResource?
VisualStudio likes marking the xml files as content or none instead - especially if you've added the MvxBindingAttributes.xml file as a link.
As an update to this, you can also now use a special 'auto' xml scheme:
xmlns:local="http://schemas.android.com/apk/res-auto"
In the project properties, add a Android Manifest file and add a package name that contains at least two periods, for example:
com.DroidUI
In the axml file, type the package name instead of the namespace:
xmlns:local="http://schemas.android.com/apk/res/com.DroidUI
xmlns:local="http://schemas.android.com/apk/res-auto
Fixed it for me.
First, thanks for the answer. I had the same problem.
But in my case, attempting to use the suggestions above failed. or at least that is what i thought.
When i used local:MvxItemTemplate=#layout/My_Manifest while i had the My_Manifest.axml file correctly placed in Layout folder, and correct xml schemas, it still had problems.
But, when i changed My_Manifest to my_manifest while leaving the actual file name as My_Manifest.axml, everything was working.
just wanted to share
In my case for some reason the MvxBindingAttributes.xml file existed, but hadn't been included in the project. Right-clicking on MvxBindingAttributes.xml and choosing "Include in Project" sorted the problem for me.
Just in case: You might also have rubbish (in my case, it were dots) in some of your XML. But the error pointed "Maybe there's something wrong with the auto-generated XML lol" so it took me half an hour to notice it... Sigh!
When I generate a .pdf file from a .tex file using pdflatex, only the "Contents" title is shown with no actual TOC. If I run pdflatex my.tex once more, it generates the TOC just fine. I can reproduce this simply by removing the .toc file. What I think is happening is that my .toc file is being generated too late -- so how can I make the TOC work first time? Should I be generating the TOC beforehand without using pdflatex?
This is normal. LaTeX document need several compilations to reach a stable state. Use rubber -d my to compile the right number of times (rubber comes as a package on many linux distros).
I might be mistaken, but I think, that this is the default behaviour. I assume, you also won't find correct cross references (footnotes, end notes, literature) after the first run of pdflatex.
The point is, that LaTeX needs the extra rounds to resolve the references pointing inside the document, to get numbering and page numbers right.
I experienced the same problem with the editor Latexian. What solved the problem was changing the preferences. I changed "Number of typesetter runs at end" to 3, instead of the default 1. Then I added the "Refresh" button to the toolbar and tried refreshing and it worked.
I added a line "\cite{test}" as a test to my working Latex document. When I compiled the bibtex "!bibtex name_of_my_file, I got the expected error:
Warning--I didn't find a database entry for "test"
Then, I removed the line and compiled the bibtex again, hoping to have a working Latex file again. However, the same error occurs, even with a fresh shell. I cannot understand the behaviour. What is the logic? How can I get my Latex document working again?
[Updated Info]
The problem dissapeared as unexpectedly as it emerged. I have no idea why but it works now. Do you know the reason for the odd behaviour?
I think you are tripping over the multi-pass nature of LaTex plus Bibtex. If you look at Step 3 in this discussion, you'll see the following:
The first run (through latex)
generates an auxiliary file,
paper.aux, containing information
about citations (and other types of
references), the bibliography style
used, and the name of the bibtex
database. The second run (through
bibtex) uses the information in the
auxiliary file, along with the data
contained in the bibtex database, to
create a file paper.bbl. This file
contains a thebibliography environment
with \bibitem entries formatted
according to the bibliography style
specified.
So, what I think is happening is that your name_of_my_file.aux file still contains your placeholder \cite{test}. If you remove the auxiliary file, you should be able to start over with:
latex name_of_my_file
bibtex name_of_my_file
latex name_of_my_file
latex name_of_my_file
[Update based on additional info]: The problem was that you had a .aux file with your \cite{} still embedded. The second time that you ran latex, you overrode the old file with the new. That's why the complete set of steps includes an initial latex call, a bibtex call and two follow-up latex calls. Think of it as a multi-pass compiler and it might be more intuitive.
You could have a look at latexmk, which will take care of the fix point compilation for you.
Anyway, you should be able to build the document (pdflatex blah.tex), even if you're missing a bibliography item. The corresponding references will just appear as question marks in the PDF.
Rerun latex to regenerate the aux file.
Have a look at this discussion for pointers to a bit more information. Basically, you may have taken your citation out of the .tex file, but it still exists in one of the derived files (aux, bbl, whatever...)
Check if your bib file has the extension .bib and not .tex.
If it is .tex, just change it to .bib and that should do it.
Once I changed it accidentally to tex, by adding some references, and saving it with the "save as" option, without specifying the bib extension. That's how it can happen all of a sudden.
delete all your .aux and temporal files, re run with latex and then bibtex and then latex twice.
In LaTeX, the \includeonly statement exists to selectively add or exclude sub-documents.
One quirk in the whole process is that when sections are excluded, LaTeX decides to insert a page-break. The article at To have no pagebreak after \include in LaTeX recommends using newclude to avoid this issue. This package defines \include* which provides a work-around by omitting the implied \clearpage.
I am occasionally seeing the behaviour where if I comment out lines of my \includeonly statement the sub-document still gets included. Has anyone else seen this behaviour?
For whatever reason, everything is now working as it should.
Just in case the order of the \usepackage directives is important, in the working version \usepackage{newclude} is the first command after \documentclass{article}.
It's not a bug, it's a feature. Really. Does \input do what you need, rather than \include?