Do you have examples that could become test cases? On Sat, Feb 13, 2010 at 2:32 AM, Villu Ruusmann <[email protected]>wrote:
> now I see approximately 10 > % of my "text extraction" tests failing. The problem is related to > incorrect text decoding (eg. there is gibberish like "b?fi??" instead > of text). > -- Ted Dunning, CTO DeepDyve
