> Since ap_bucket_split_any() is what we agree on, then let's move on and > figure out > what the best approach is to improving it to handle sockets and pipes better. > It > really is too bad that the read() function ignores the len parameter. If it > didn't, > then split_any() could just tell it how much to read, and this wouldn't be as > much > of an issue. Alternately, split_any() could just keep track of how much was > read
That doesn't work. What happens if I pass 1024 to the read function, but there is only 512 bytes on the socket/pipe? > and loop through an arbitrary number of reads, incrementing until the totals > matched. That's the least intrusive approach. I'll implement it if people > agree. > Thoughts? split_any needs to make sure it is splitting at the correct location. However, now we are talking about a different function, which is a brigade split based on offset. If that is what we are after, then the name is incorrect. Ryan _______________________________________________________________________________ Ryan Bloom [EMAIL PROTECTED] 406 29th St. San Francisco, CA 94131 -------------------------------------------------------------------------------
