Jabref: Filepath saved on Windows does not open on Linux

[JoKalliauer]

JabRef 4.0
Linux 4.10.0-37-generic amd64 
Java 1.8.0_144

Steps to reproduce:

1. create an idential folderstructure on Windows and on Linux (f.e. use sync-systems like: dropbox, onedrive, googledrive, owncloud)
2. JabRef on Windows&Linux must have the same "Main file directory" (f.e. Linux: /home/jkalliau/ownCloud and Windows: C:/Users/jkalliau/ownCloud)
3. add a PDF-File to an article in JabRef to a bibtex-file on Windows
4. try to open this PDF-File in JabRef on Linux

The problem is that Windows saves:

File = {:..\\..\\Literature\\DNA\\Weitere\\Bai_Martin_Scheres_Dietz_2012_PNAS.pdf:PDF},
file = {:PhDJK\\Literature\\MD\\Amber17.pdf:PDF},

But Linux saves as:

file = {:PhDJK/Literature/MD/Amber17.pdf:PDF},

With JabRef 3.8.2 neither Windows nor Linux had problems using different File-Structures, but now with JabRef 4 on Linux it does (JabRef can't find the file). :-(

On Windows with JabRef4.0 it still works also with Linux-Filepathes.

This bib-file has developed over some time with JabRef 3.x and it always worked to open with JabRef 3.8.2 on Windows and on Linux, also both added file-paths in a different form.

I would stay with JabRef 3.8.2, but due to bug #2879 I prefer JabRef 4.

Otherwise I might have to use sed to correct the path-mistakes.

bib-file with filepath saved on Windows

```
% Encoding: UTF-8

@Article{bai2012cryo,
Title = {{Cryo-EM} structure of a {3D} {DNA-origami} object},
Author = {Bai, Xiao-chen and Martin, Thomas~G and Scheres, Sjors~HW and Dietz, Hendrik},
Journal = {Proceedings of the National Academy of Sciences},
Year = {2012},

Month = dec,
Number = {49},
Pages = {20012--20017},
Volume = {109},

Adsnote = {Provided by the SAO/NASA Astrophysics Data System},
Adsurl = {http://adsabs.harvard.edu/abs/2012PNAS..10920012B },
Doi = {10.1073/pnas.1215713109},
File = {:..\..\Literature\DNA\Weitere\Bai_Martin_Scheres_Dietz_2012_PNAS.pdf:PDF},
Groups = {DNA},
Keywords = {rank2},
Publisher = {National Acad Sciences}
}

@Article{case2017amber,
author = {Case, David A and Walker, Ross C and Cheatham III, Thomas E and Simmerling, Carlos LI and Roitberg, Adrian and Merz, Kenneth M and Luo, Ray and Darden, Tom A and Wang, Junmei and Duke, Robert E and Daniel R. Roe and Scott LeGrand and Jason Swails and Andreas W. Götz and David Cerutti and Tyler Luchko and Gérald Monard and Celeste Sagui and Feng Pan and Charles Lin and Daniel Mermelstein and Pengfei Li and Alexey Onufriev and Saeed Izadi and Romain M. Wolf and Xiongwu Wu and Holger Gohlke and Nadine Homeyer and Ruxi Qi and Wesley M. Botello-Smith and Li Xiao and D’Artagnan Greene and Tim Giese and Taisung Lee and Darrin York and Jian Liu and Hai Nguyen and Igor Omelyan and Andriy Kovalenko and Peter A. Kollman and Zhang, W and others},
title = {Amber 2017 Reference Manual (Covers Amber16 and AmberTools17)},
journal = {University of California, San Francisco},
year = {2017},
month = may,
day = {29},
file = {:PhDJK\Literature\MD\Amber17.pdf:PDF},
url = {http://ambermd.org/doc12/Amber17.pdf},
}

@Comment{jabref-meta: databaseType:bibtex;}

@Comment{jabref-meta: grouping:
0 AllEntriesGroup:;
1 StaticGroup:DNA\;0\;1\;\;\;\;;
}
```

[lenhard]

@Siedlerchr You have recently modified the file path serialization, right? Can you comment on this?

[Siedlerchr]

I will look into this, first guess is that the relative path is passed as
String to the open method

Am 17.10.2017 2:24 nachm. schrieb "Jörg Lenhard" [e-mail]:

@Siedlerchr https://github.com/siedlerchr You have recently modified
the file path serialization, right? Can you comment on this?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/JabRef/jabref/issues/3311#issuecomment-337214329, or mute
the thread
https://github.com/notifications/unsubscribe-auth/AATi5DMB-ywN5Xe5hFN5nLL69DggN50dks5stJxjgaJpZM4P6kZ8
.

[lenhard]

@JoKalliauer We think we have a fix for this. Can you please try out the version that includes the fix, which is available here: http://builds.jabref.org/linuxlink/ and let us know if it works for you.

[JoKalliauer]

JabRef 4.1-dev--snapshot--2017-11-04--linuxlink--4ffdabf5c
Linux 4.10.0-38-generic amd64 
Java 1.8.0_151

JabRef 4.1-dev--snapshot--2017-11-04--linuxlink--4ffdabf5c
Windows 7 6.1 amd64 
Java 1.8.0_152



md5-e1e51c1d6dafe98a1fb1c9d4250c2b3b



sed -ri 's/  (f|F)ile[ ]+= \{:([-[:alnum:] _.\/\;\:]+)\\\\([-[:alnum:] _.\\\/\;\:\(\)]+).([[:alpha:]]{3}):(MHT|PDF|Text|JPG-Bild|Word)\},/  \1ile = \{:\2\/\3.\4:\5\},/' filename.bib



md5-e801c04fb47e851648fc7153172c7a45



Warning: no message translation for "Show console output (only necessary when the launcher is used)" for locale de
Opening: /home/jkalliau/ownCloud/Master/BP2/literature.bib
Opening: /home/jkalliau/ownCloud/Master/jk/JK2EuroC/a4sample.bib
Opening: /home/jkalliau/ownCloud/PhDJK/LaTeX/Style/LiteratureJK.bib
Warning: no message translation for "JabRef cannot access the file because it is being used by another process." for locale de
Warning: no message translation for "JabRef cannot access the file because it is being used by another process." for locale de
Warning: no message translation for "JabRef cannot access the file because it is being used by another process." for locale de
Warning: no message translation for "JabRef cannot access the file because it is being used by another process." for locale de
Warning: no message translation for "JabRef cannot access the file because it is being used by another process." for locale de
Warning: no message translation for "JabRef cannot access the file because it is being used by another process." for locale de

[Siedlerchr]

Thanks for your feedback!
You could run a cleanup operation to rename the files, that should then set the files links correctly.

Regarding the translations:
They will usually be added before the next release, (at least the German, to avoid double work during dev).
The changes are now in the master. So you can switch to that to stay on top with actual changes

[JoKalliauer]

@lenhard: Please remove the waiting-for-feedback-label, or is there feedback missing?

[lenhard]

@JoKalliauer Nope, I've removed the label.