0000472: Sorting Unicode characters

(0001051)
Kry (manager)
2005-06-06 19:10

How are you, hum, suposed to sort wide chars?

(0001052)
benjavalero (reporter)
2005-06-08 21:03

Well... at least in Spanish and German dictionaries, a vowel with an accent is treated as a simple vowel to get sorted, so Benjamín and Benjamin would be sorted in the same place.

I hope I have explained... my English is very poor :(

Benja

(0001053)
Kry (manager)
2005-06-08 22:40

Problem is, even if they look alike, the chars have a completely different identifier. Dunno if there's any way to map them, but I'm sure I don't know it.

(0001054)
GonoszTopi (administrator)
2005-06-09 10:02

It is possible to do, but doesn't worth the effort. The unicode character table (available from The Unicode Consortium) contains the composite map for each (composite) character, so it would reveal that Á = A + ´, so it can be sorted as "A"

However, it was a good solution only if we were talking of one single language. I don't think it would be a good idea to include a several megabytes database in aMule, just to fix sorting.

So, technically the issue is closed. Reopen it when cheap sorting solutions found for unicode.

Issue History
Date Modified	Username	Field	Change
2005-06-06 11:31	benjavalero	New Issue
2005-06-06 19:10	Kry	Note Added: 0001051
2005-06-08 21:03	benjavalero	Note Added: 0001052
2005-06-08 22:40	Kry	Note Added: 0001053
2005-06-09 10:02	GonoszTopi	Status	new => resolved
2005-06-09 10:02	GonoszTopi	Resolution	open => won't fix
2005-06-09 10:02	GonoszTopi	Assigned To	=> GonoszTopi
2005-06-09 10:02	GonoszTopi	Note Added: 0001054

Notes
(0001051) Kry (manager) 2005-06-06 19:10	How are you, hum, suposed to sort wide chars?

(0001052) benjavalero (reporter) 2005-06-08 21:03	Well... at least in Spanish and German dictionaries, a vowel with an accent is treated as a simple vowel to get sorted, so Benjamín and Benjamin would be sorted in the same place. I hope I have explained... my English is very poor :( Benja

(0001053) Kry (manager) 2005-06-08 22:40	Problem is, even if they look alike, the chars have a completely different identifier. Dunno if there's any way to map them, but I'm sure I don't know it.

(0001054) GonoszTopi (administrator) 2005-06-09 10:02	It is possible to do, but doesn't worth the effort. The unicode character table (available from The Unicode Consortium) contains the composite map for each (composite) character, so it would reveal that Á = A + ´, so it can be sorted as "A" However, it was a good solution only if we were talking of one single language. I don't think it would be a good idea to include a several megabytes database in aMule, just to fix sorting. So, technically the issue is closed. Reopen it when cheap sorting solutions found for unicode.

Relationships