zlib is used when available and falls back on gzio.
This allows performance improvements as our CRC32 and gzio
implementations are slower than base zlib.
As zlib is available when libpng is present, this is sensible to
benefit from it.
Just walking around the castle fixing bugs as I find them, e. g.:
- make decompression routine safer (this could write a couple of invalid byte in at least one case)
- also support compressed data header without actual data (required in the kitchen scene when killing the evil cook)
- fix some graphics/palette related glitches
This in fact slightly changes the priority order of added archives. Formerly,
all archives in SearchMan were preferred to the customly added ones in
ArchiveMan. All standard paths (i.e. path and extrapath) will be still be
searched before the custom ones (which are all priority 0 right now) but system
specific paths will be searched after (due to their priority being -1). Since
system specific paths shouldn't contain any game data files this should
hopefully be harmless.
This wasn't tested for games with CAB archives.
Now all of the Archive API is implemented to avoid possibly inconsistent
behavior when an ArchiveMan object is accessed via a class type higher in the
class hierachy.
Formerly the hasFile method was non-const, thus did not overwrite the virtual
hasFile method of SearchSet. Since all accesses to this function have been
done through a ArchiveMan typed variable, the missing overwrite should not
have caused any trouble.
It looks like this was forgotten in a6ec4f70da.
Playing with cabinet datafiles, this failed to find the "Charisma.smk"
file in the cabinets when in the Recreation (TV) room on Cygnus Alpha
and this caused an engine abort after the GUI dialog warning of the
"missing" video file. This was due to animation.cpp code using
Common::file::exists() instead of going via the ArchiveMan.
However,a hasFile() method implementation was also required to implement
fallback to decompressed (movie) files if the file requested is not in
the cabinet or the cabinet has been externally decompressed to files.
Thanks to fuzzie for the hasFile() patch.
Also, removed noisy warning which this correction triggers repeatedly in
installshield_cab.cpp hasFile(). This looks like leftover from
debugging, so not critical.
