Difference: JaysJournal (37 vs. 38)

Line: 1 to 1
	May 2010 archive 06/01/10
Line: 149 to 149
	More converting Look at fix mentioned above.
Added:
> >	06/10/10 Bahhhhh spent all day converting strings to c-strings. I've decided to convert only things having to do with sequences and qualities to c-strings. To make it easier to see, I put up some typedefs: #define NGSA_READLEN 128 typedef char SequenceBase; typedef char QualityBase; typedef SequenceBase SequenceString[NGSA_READLEN+1]; typedef QualityBase QualityString[NGSA_READLEN+1]; I'm about 75% done, I think. I finished the I/O classes and the drivers, as well as half the mappers. Just need to convert the rest, and then start on the pair-end matcher, and some other stuff. Also optimized some of the lower-level algorithms (such as ngsa::ReverseComplement) to make better use of `char` arrays. I hope the end result is a speed up, or at least doesn't result in a speed-down! Some parts still aren't very efficient, because there are other methods that still take in C++ strings, where I end up having to do a conversion (which takes O(n) time). I think `strlen` is also O(n), versus O(1) on `std::string::size` (I think that's what these are), since `std::string` can just update the size every time the string grows/shrinks, so I should make sure to avoid using `strlen` too often. To do: More conversion...