Donnerstag, 18. April 2024
latex4
diegolego, 22:15h
One more experience. I am including various chapters with \include{}. Each chapter should have its own bibliography. Therefore, I am using biber with
> \refsection
before the first section and
> \printbibliography
> \endrefsection
at the end.
.bib-files are included via \addbibresource{} in the main file.
The tricky part is that people leave many more bib-entries in their files than they cite. I don't know how biber deals with double entries and if it takes the first or last occurence. In any case, did it happen, that I had to fix the bib-entry of another chapter than the one using it. This makes sense since \addbibresource{} simply pools the bib-files and they are not associated to a specific \include{}.
> \refsection
before the first section and
> \printbibliography
> \endrefsection
at the end.
.bib-files are included via \addbibresource{} in the main file.
The tricky part is that people leave many more bib-entries in their files than they cite. I don't know how biber deals with double entries and if it takes the first or last occurence. In any case, did it happen, that I had to fix the bib-entry of another chapter than the one using it. This makes sense since \addbibresource{} simply pools the bib-files and they are not associated to a specific \include{}.
... link
Mittwoch, 17. April 2024
latex3
diegolego, 09:25h
An error - this time it goes on my account - is the following. I got the message
> File ended while scanning use of \newlabel.
and online people said it is caused by missing closing bracket. I tried so many things and it was strange because the error was a bit erratic. I even counted the frequency of characters
> grep -o . filename | sort | uniq -c
with the expectation that the number of { should be the same as the count of } - although no sufficient proof.
I then removed most of the text and kept adding paragraphs until the error occurred. However, this procedure led to inconsistent findings.
At the end I went back to the original source. Since I included the .tex file with \include{}, I had to comment out the \begin{document} and \end{document}. Precisely, the 2nd I did not do. This means there was an additional \end{document}.
> File ended while scanning use of \newlabel.
and online people said it is caused by missing closing bracket. I tried so many things and it was strange because the error was a bit erratic. I even counted the frequency of characters
> grep -o . filename | sort | uniq -c
with the expectation that the number of { should be the same as the count of } - although no sufficient proof.
I then removed most of the text and kept adding paragraphs until the error occurred. However, this procedure led to inconsistent findings.
At the end I went back to the original source. Since I included the .tex file with \include{}, I had to comment out the \begin{document} and \end{document}. Precisely, the 2nd I did not do. This means there was an additional \end{document}.
... link
Dienstag, 16. April 2024
latex2
diegolego, 00:05h
Another problem that I encountered with LaTeX was encoding. People provided messy .bib files and I copied the necessary entries into a single file. After removing illegal characters. I got the following message.
> ERROR - Data file '... .bib' cannot be read in encoding 'utf8': Can't decode ill-formed UTF-8 octet sequence at /usr/lib/perl5/vendor_perl/5.26.1/File/Slurper.pm line 59.
Searching for the error message brought me to this explanation. Apparently, it means that my .tex and .bib files have different encodings. On this website I found that the encoding can be checked with "file -i" and it can be changed with
> iconv -f iso-8859-1 -t utf-8 bib1.bib -o bib2.bib
which solved the problem.
Again, overleaf ignored any errors. I guess the idea is not to frustrate latex beginners, but the consequence is that they generate sloppy latex-sources.
Update: mendely creates the worst bib-files
> ERROR - Data file '... .bib' cannot be read in encoding 'utf8': Can't decode ill-formed UTF-8 octet sequence at /usr/lib/perl5/vendor_perl/5.26.1/File/Slurper.pm line 59.
Searching for the error message brought me to this explanation. Apparently, it means that my .tex and .bib files have different encodings. On this website I found that the encoding can be checked with "file -i" and it can be changed with
> iconv -f iso-8859-1 -t utf-8 bib1.bib -o bib2.bib
which solved the problem.
Again, overleaf ignored any errors. I guess the idea is not to frustrate latex beginners, but the consequence is that they generate sloppy latex-sources.
Update: mendely creates the worst bib-files
... link
Montag, 15. April 2024
{\'\i}
diegolego, 20:35h
Sometimes LaTeX is an asshole. This time I had a very reluctant error in my source. I kept getting this message:
> Package inputenc Error: Unicode character ́ (U+0301)
It means there is an illegal character. It had to be either the .tex file or the .bib one. I found these tips how to locate it.
With this command I found a few illegal characters
> cat filename | grep -n -P "[^|a-zA-Z\{\}\s%\./\-:;,0-9@=\\\\\"'\(\)_~\$\!&\`\?+#\^<>\[\]\*]"
but the same error message continued to inhebit the compilation.
With a bit more of experimenting I was sure it had to be the bibliography - but I simply could not find any remaining invalid character.
It took me a while until I checked the .bbl file. It contained many special characters which were introduced by biber (essentially diacritic/accents) - but there were also two instances of unicode 0301. The person from whom I received the source used
> \'\i instead of \'i
to generate an accent on an i. I could not find it because the code itself generated the illegal character.
To make things worse, overleaf seems to filter out such issues. The person writing the source compiled on overleaf - but locally it did not work.
Update (17.4.2024): It seems like people copy bibtex-entries from google-scholar and it creates these \'\i which in my oppinion are not good latex. Could some please clarify.
> Package inputenc Error: Unicode character ́ (U+0301)
It means there is an illegal character. It had to be either the .tex file or the .bib one. I found these tips how to locate it.
With this command I found a few illegal characters
> cat filename | grep -n -P "[^|a-zA-Z\{\}\s%\./\-:;,0-9@=\\\\\"'\(\)_~\$\!&\`\?+#\^<>\[\]\*]"
but the same error message continued to inhebit the compilation.
With a bit more of experimenting I was sure it had to be the bibliography - but I simply could not find any remaining invalid character.
It took me a while until I checked the .bbl file. It contained many special characters which were introduced by biber (essentially diacritic/accents) - but there were also two instances of unicode 0301. The person from whom I received the source used
> \'\i instead of \'i
to generate an accent on an i. I could not find it because the code itself generated the illegal character.
To make things worse, overleaf seems to filter out such issues. The person writing the source compiled on overleaf - but locally it did not work.
Update (17.4.2024): It seems like people copy bibtex-entries from google-scholar and it creates these \'\i which in my oppinion are not good latex. Could some please clarify.
... link
Freitag, 22. März 2024
The Cormorant
diegolego, 14:47h
... link
Dienstag, 5. März 2024
System
diegolego, 06:56h
Serj Tankian, der Sänger von System of a Down, hat ein Buch geschrieben namens Down with the System - A Memoir (of Sorts).
... link
Montag, 12. Februar 2024
Dauerwerbesendung
diegolego, 08:31h
... link
Sonntag, 21. Januar 2024
Conway
diegolego, 21:01h
... link
Freitag, 15. Dezember 2023
Hinterhof
diegolego, 09:42h
... link
Donnerstag, 14. Dezember 2023
besser spät als nie
diegolego, 09:17h
adventofcode.com
Tobi ist mit großem Einsatz dabei - vielleicht postet er ja mal seine Lösungen.
Tobi ist mit großem Einsatz dabei - vielleicht postet er ja mal seine Lösungen.
... link
... nächste Seite
