handle Unicode characters properly

Ticket

+commit
 

Do not create these, it causes renaming problems associated with blank keys.

 

The issue is that in creating the unique keys identifying each card, we basically strip out everything that isn't ASCII alphanumeric.  What we really want to do, I think, is strip out just the non-alphanumeric ASCII characters.

 

Cards with no ascii alphanumeric characters will thus be given a blank key ("") which means two things:

  1. there can be only one such card on a Wagn
  2. some renaming (which stupidly uses a blank keys as an interim placeholder) breaks.

 

 

 

 

John's original "issue" text:

someone created 中文测试 (probably by New Card) which seemed to work, however it looks like any Chinese characters end up referring to the same card! E.g. 里是 or 邸报. Some testing needed to identify what characters trigger this vs. handle HTML entity characters properly.

 


unicode characters are no longer scraped from keys. I'm not sure there won't be some other unicode issues arising, but those should be ticketed more specifically.

  --Ethan McCutchen.....Fri Nov 05 11:47:22 -0700 2010


Do we need a ticket to migrate existing card keys for Unicode fix? I just had to rename http://grouppatternlanguage.org/wagn/Ens to http://grouppatternlanguage.org/wagn/Ensō - um, and then there's the issue that the URL isn't being recognized here...

  --John Abbe.....Thu Dec 16 18:02:00 -0800 2010


yeah, we probably need both of these tickets. The migration should be medium pri, I guess.

  --Ethan McCutchen.....Fri Dec 17 08:32:35 -0800 2010


Added the migration ticket, and recognize URLs with non-ASCII characters.

  --John Abbe.....Fri Dec 17 12:04:28 -0800 2010