View Issue Details

Jump to NotesJump to History
IDProjectCategoryView StatusDate SubmittedLast Update
0002739unrealircdpublic2006-01-05 07:172007-04-27 04:06
ReporterTrocotronic Assigned To 
PrioritynormalSeverityminorReproducibilityalways
Status closedResolutionwont fix 
PlatformAMD K6 32bitsOSWindows XP ProfessionalOS VersionSP2
Product Version3.2.4 
Summary0002739: Badwords does not support charsets
DescriptionI'll explain this with a simple example.
I'm spanish and we have the 'ñ/Ñ' letter (241/209 chars). As you can see, they are not standar. So, if you ban a word like "españa" you are able to use "espaÑa" or "ESPAÑA" or whatever word has Ñ as uppercase.
In other words, there are some problems with lower/uppercase foreign letters (ñ/Ñ, ç/Ç, á/Á, é/É, etc.).

(In our language there are a lot of badwords containing 'ñ')
TagsNo tags attached.
3rd party modules

Relationships

related to 0003101 closed Suggestion about cyrillic nicks 
related to 0002882 acknowledged Configurable CASEMAPPING (Lower/Uppercase with charsets) 

Activities

vonitsanet

2006-01-05 08:39

reporter   ~0010923

This is not a bug. it works like this.
i dont know how possible is to add this check for every language..

Stealth

2006-01-05 09:10

reporter   ~0010925

Use regex in the badword block.

Trocotronic

2006-01-22 15:38

reporter   ~0011018

spamfilter {
    regex "coño";
    target channel;
    reason "No palabrotas";
    action kill;
};

If I use coño works fine. I'm killed. But if I use COÑO, I'm not.

JasonTik

2006-01-22 19:16

reporter   ~0011027

Last edited: 2006-01-22 19:17

 In spamfilter, which is regex, you can use [ñÑ] which will match either.

In badwords, you can use that, and it will change from glob matching to regex matching, however, then * and ? will no longer work properly, and should be replaced with '[^ ]*' and '[^ ]' respectively. Also any regular expressions special characters will need to be escaped, but no normal letters qualify for that, so it shouldnt be too big of an issue.

aquanight

2006-01-23 21:36

reporter   ~0011045

http://www.vulnscan.org/UnrealIRCd/unreal32docs.html#feature_nickchars

I think it's stated pretty clearly there that casemapping is still only for US-ASCII (iow: a-z == A-Z and that's it).

stskeeps

2007-04-27 04:06

reporter   ~0013786

Too early version - please resubmit for 3.2.6/3.3

Issue History

Date Modified Username Field Change
2006-01-05 07:17 Trocotronic New Issue
2006-01-05 08:39 vonitsanet Note Added: 0010923
2006-01-05 09:10 Stealth Note Added: 0010925
2006-01-22 15:38 Trocotronic Note Added: 0011018
2006-01-22 19:16 JasonTik Note Added: 0011027
2006-01-22 19:17 JasonTik Note Edited: 0011027
2006-01-23 21:36 aquanight Note Added: 0011045
2006-11-01 07:33 syzop Relationship added related to 0003101
2006-11-01 07:33 syzop Relationship added related to 0002882
2007-04-27 04:06 stskeeps Status new => closed
2007-04-27 04:06 stskeeps Note Added: 0013786
2007-04-27 04:06 stskeeps Resolution open => wont fix