[physfs] My shot at PHYSFS_stat
indy at driftsolutions.com
Mon Feb 15 15:41:20 EST 2010
Looks good, just one or 2 things to mention in the PHYSFS_stat function:
At the top, I originally had it as "int retval = -1;" not "int retval = 0;" that way it defaults to failure in case any of the other parts fail or the file doesn't exist.
On the note "/* !!! FIXME: what should this be set to if we fail completely? */": basically it wouldn't matter what is in the PHYSFS_Stat struct if the function fails, none of the values should be used unless PHYSFS_stat() returns 0.
On this part:
retval = 1; /* !!! FIXME: What's the right value? */"
I don't remember exactly why I did that since I made it so long ago, but I assume I had a reason. I think it may have been for debugging purposes so it would return 1 if the mount point existed instead of -1 or something like that.
Monday, February 15, 2010, 2:12:14 PM, you wrote:
>> Both this work and Indy's have merit. I'm likely going to merge pieces
>> of both and add you both to the credits.
RCG> Ok, this is in now, changeset 1054:57f4af811ffb.
RCG> I cleaned up some things, but the biggest change is that I got rid of
RCG> getFileInfo()...I suppose this is meant to be an optimization for
RCG> targets that could get, say, file timestamps easily but not file sizes
RCG> without uncompressing the whole file...but it doesn't apply to any
RCG> current archivers, and I'd rather this use the simpler stat() interface.
RCG> It makes the code much cleaner.
RCG> Anyhow, you'll have to merge when you pull from my repo again, since
RCG> even things that didn't get a specific change probably had tab
RCG> characters removed, etc.
RCG> physfs mailing list
RCG> physfs at icculus.org
mailto:indy at driftsolutions.com
P.S. This is what part of the alphabet would look like if Q & R were eliminated.
More information about the physfs