The macros defining unit stat development in tech are somewhat
inconvenient to use. Define more convenient functions, and hide away
the macros near the function definitions.
If gcc's preprocessor chokes, it leaves an empty dependency file
behind, and doesn't touch the object file. If an old object file
exists, and is newer than the .c file, make will then consider the
object file up-to-date. This can lead to nasty version errors.
delete_old_news() and init_nreport() tested for non-zero nws_when,
which is breaks for news at the epoch. Not likely to happen, but
still wrong.
ncache() tested nws_uid, which breaks for the first entry in the news
file. This made ncache() overlook that entry in the cache, and create
a new news item instead of incrementing nws_ntm. Was always broken
that way.
Fix by testing nws_vrb instead.
The call of ef_ensure_space() serves no purpose: the blank records it
adds are ignored on read, and overwritten (not updated) by new news.
Get rid of it.
Only treaty extension initialized ef_type and uid properly. None of
them zeroed unused members and holes in the struct. comm and trade
extension called ef_extend(), which had no effect, so remove that.
New struct emptypedstr to avoid depending on empobj.h there.
Remove now superfluous manual initializations elsewhere.
This doesn't fix any missing initializations.
Make sure all members of unit empobj_storage share uid in addition to
ef_type.
Add matching uid member to struct gamestr, struct natstr and struct
sctstr, and set them.
Swap struct empobj members uid and own to make that easier, and update
struct comstr, struct lndstr, struct lonstr, struct loststr, struct
nwsstr, struct nukstr, struct plnstr, struct realmstr, struct shpstr,
struct trdstr, struct trtstr accordingly.
Note that the uid isn't properly set for struct nwsstr, struct lonstr,
struct trdstr, struct comstr and struct loststr. To be fixed.
The get_FOOp() macros are generally avoided outside the update,
because direct access to the sector cache needs synchronization to be
safe. unit_map() didn't access the cache directly until it was
converted from get_ship() & friends to get_empobjp() in commit
fec9878c. Switching to get_empobj() reverts the change to direct
access while keeping the simplification.
There's no technical reason for rejecting sector access by id. It's
unusual, but not wrong.
Also remove the superfluous test for EF_BAD; ef_ptr() covers that.
put_empobj() used struct empobj member uid, which is valid only for
units. Existing users pass only units, fortunately. Fix by making it
take type and uid parameters.
This led to a bogus message when an interactive explore moved onto a
bridge and got prompted, the bridge was destroyed, and the player
stopped the explore "on the water".
get_empobj_chr() and emp_obj_chr_name() access struct sctstr member
sct_type through struct empobj member type. This is technically
non-portable, because the two differ in signedness. It was also
undocumented. Fix by making sct_type signed. sct_newtype as well,
for consistency.
map_char() uses unsigned char for a sector type argument. Change that
to int. Matches how this is done elsewhere.
Use ef_ensure_space() in getobj(). This also makes sure objects are
properly initialized before undumping writes to them.
Clean up how sentinels are appended: instead of keeping its slot
reserved while undumping, keep it in the table, and strip it off when
done.
ef_extend() extended the file bypassing the cache, which screws up the
cache if it's EFF_MEM. It fixed that by closing and reopening the
table. Cheesy, and worked only for file-backed tables.
Rewrite ef_extend() to remap the cache properly for EFF_MEM. While
there, simplify the !EFF_MEM case: steal a cache slot instead of
allocating a buffer.
Factor cache mapping out of ef_open() and ef_extend() into new
ef_realloc_cache().
Read-only was a bit of a misnomer: you could write to the table by
obtaining a pointer into it from ef_ptr(), you just couldn't write to
the backing file.
Semantic changes:
* ef_flush() is now allowed when the table is file-backed or privately
mapped. Before, it had to be file-backed. Flushing a privately
mapped table does nothing, just like flushing a read-only table did.
* ef_write() is now allowed when the table is file-backed or fully
cached. Before, it had to be file-backed and not read-only.
Writing to a privately mapped file-backed table doesn't write to the
file.
* ef_extend() is not implemented for privately mapped tables, just
like it wasn't implemented for read-only tables.
