On Sat, Sep 30, 2006 at 12:49:41PM -0700, Debian Bug Tracking System wrote: > In the current version in sid I can't find the occurrence of nnn > as written in your patch, so this stuff was fixed upstream. > > I'm interested in your semi automatic tools, are they publicly > available? If so, where?
See http://www.mail-archive.com/debian-i18n%40lists.debian.org/msg04809.html. Please note that also many manual work was involved. Most trivial grep calls show hundreds of errors if applied to a sufficient large data set. German contains only 4 8-bit characters (ä,ö,ü,ß). Searching for these UTF-8 encoded characters in latin1 texts and the opposite, searching for doubled words or spaces in front of ",", ".", "!", "?", give many matches ... Once you start this process you will notice many other errors in surrounding lines. Search for these errors globally and you will find more errors during your attempts to fix these, ... There are also lists of common typos available (for German e.g. for wikipedia editors). Just start with one method and you will never stop ... But at least you can fix many important errors. Once the translators got informed they will (hopefully) learn from the mistakes and avoid these in the future. Jens
