wrong line number information on some source files

Semih Özmen January 20, 2013

Hi,

there are some files in our repository that are generating some warning logs in wrapper.log file & does not show the source code correctly in source view of fisheye.

* file is 5503 lines long (checked with RAW view of fisheye and with local version. last revision of file)

* in source view of fisheye, it tries to show 5571 lines. after line number 5505 they are all empty and the lines before 5505 are correct

* in wrapper.log file of fisheye server, tones of warnings written, for every time of viewing the source file with fisheye

...

INFO | jvm 1 | 2013/01/21 12:13:37 | 2013-01-21 12:13:37,038 WARN - unknown line number 5562 only know about 5505 in SVN:Comosys_3.0_Plugins:mechanical/trunk/elements/TankDetailedDrawingDlg.cpp

INFO | jvm 1 | 2013/01/21 12:13:37 | 2013-01-21 12:13:37,038 WARN - unknown line number 5562 only know about 5505 in SVN:Comosys_3.0_Plugins:mechanical/trunk/elements/TankDetailedDrawingDlg.cpp

INFO | jvm 1 | 2013/01/21 12:13:37 | 2013-01-21 12:13:37,038 WARN - unknown line number 5563 only know about 5505 in SVN:Comosys_3.0_Plugins:mechanical/trunk/elements/TankDetailedDrawingDlg.cpp

...

* in the meanwhile wrapper.log file is getting bigger & bigger

* there is not only one file atm, i've found 2 files but it's such a big repo. not investigated all files

what's going on and how can i figure it out ? thanks in advance

1 answer

0 votes
Foong
Atlassian Team
Atlassian Team members are employees working across the company in a wide variety of roles.
January 22, 2013

Hi Eyyub,

This is related to the file encoding. Modifying it to be UTF-8 fix the problem.

Semih Özmen January 23, 2013

it's not possible to change my source codes to UTF-8. i've changed the repository charset to windows 1252 & 1254 nothing changed.

not all the compilers support UTF-8 with BOM and UTF-8 Without BOM should cover ANSI charset. if you treat an ANSI file as it's UTF-8 without BOM there should be no problem.

this make me to think that it's not about UTF-8. it's just a symptom

Foong
Atlassian Team
Atlassian Team members are employees working across the company in a wide variety of roles.
February 19, 2013

found out to be caused by mixed-encoding repositories

feature request https://jira.atlassian.com/browse/FE-4477

workaround: using '-Dfile.encoding=ISO-8859-1'

Suggest an answer

Log in or Sign up to answer