directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Emmanuel Lecharny <elecha...@gmail.com>
Subject Re: Questions about the versions in the repository
Date Sat, 29 Jul 2006 22:22:58 GMT
Jörg Henne a écrit :

> Ok, thanks for the overview. Since I am focused on getting something 
> production-ish ready, I'll stick with 1.0-trunks for now.
> I've had some problems getting 
> org.apache.directory.shared.ldap.ldif.LdifReaderTest.java 
> <mailbox:///home/elecharny/.thunderbird/ztdegx62.default/Mail/Local%20Folders/Inbox.sbd/Apache.sbd/Directory.sbd/Dev?number=12739957&header=quotebody&part=1.2&filename=LdifReaderTest.java>

> to succeed:
> - the generated test files use the absolute path "/tmp/" as a prefix. 
> This is fine on *nix, but fails on Windows, of course.

Yeah, this has been fixed in 1.1 trunk, and also in 
1.0-optimization-tunk, which will be merged back to 1.0-trunk soon 
(tonite Paris time ?)

> - the file:// URLs have the same Problem, of course.

Same response.

> - the UTF-8 used in the file doesn't seem to be recognized out of the 
> box. This causes problems with three tests. I don't know whether this 
> is a problem with maven, my setup, or Windows. Anyway: I converted two 
> of them to use \uxxxx escapes, however there is still 
> testLdifParserRFC2849Sample4() which I don't readily see how to 
> convert, since eclipse doesn't display some characters (and I do have 
> CJK fonts installed!). Maybe I'll start another try with vi tomorrow.

Ok, this is something I have to explain. I had to set those chars in the 
file using an encoding. Java files use Unicode, but only when compiled. 
Until then, you need to store files which use an encoding scheme, and 
explain the compiler which encoding those java files have been created 
with. ATM, all the java files are using UTF-8 encoding. If you svn up 
the project and try to compile it on windows, there is a good luck that 
you are using ISO-8859-1 or CP-152 encoding, which is very different 
from UTF-8 encoding (especially when trying to use chines or japanese 
chars :). What you may do is :
- change the default encoding on your computer (if you build it under 
CygWin, I think changing LANG variable to en_??:UTF-8 should be enough)
- fix maven property file to include the |${maven.compiler.encoding} 
property.| I have no idea how to do that :(
- fix the test files to use unicode escaped sequences for all the 
special chars. Painfull :(

Sorry about the choice I have made...

Emmanuel

Mime
View raw message