With 3.7+ aspirations of improving savefile interoperability between 32-bit
and 64-bit builds, as well as between platforms, it is better to not have
the underlying struct/array content be conditional.
This splits off some of the MAIL code into MAIL_STRUCTURES code. In theory,
since MAIL_STRUCTURES is unconditionally included, the macro could
just go away and leave that code unconditional, but this commit doesn't
go that far.
Dying in a shop while carrying partly eaten food would place that food
on the floor without marking it no_charge. But marking it that way
wouldn't have helped because as bones data gets saved, every object on
the level has its no_charge flag cleared. So 'no_charge' needs to be
explicitly set for partly eaten food in tended shops as a bones level
gets loaded.
Most of the shk.c diff is reformatting, but it does change the
get_pricing_units() routine to lie that the quantity is zero for
partly eaten food so that when multiplying with price it won't matter
whether the price has been forced to zero or been left non-zero.
Make some progress on a couple of next minor release checklist
items, hopefully without introducing too many new bugs. This
is just the initial commit, and work continues.
Checklist items:
Savefiles compatible between Windows versions, whether 64-bit
or 32-bit in little-endian field format.
Selection of file formats:
historical (structlevel saves),
lendian (little-endian, fieldlevel saves),
and just for proof-of-concept, ascii fieldlevel saves
(the ascii is huge! 10x bigger than little-endian).
For the fieldlevel save, all complex data structures recursively
get broken down until until it is one of the simple types that
can't be broken down any further, and that gets when it gets
written to the output file.
New files needed for this build:
hand-coded:
include/sfprocs.h
src/sfbase.c - really a dispatcher to one of the
output/input format routines.
src/sflendian.c - little-endian output writer/reader.
src/sfascii.c - ascii text output writer/reader.
auto-coded (generated):
include/sfproto.h
src/sfdata.c
This is just one approach. I'm sure there are countless others
and they have different pros and cons.
For producing the auto-coded files a utility called
universal-ctags, that is actively maintained and evolving,
was used to do all the heavy-lifting of parsing the
NetHack C sources to tabulate the data fields, and store
them in an intermediate file called util/nethack.tags
(not required for building NetHack if you already have a
generated include/sfproto.h and src/sfdata.c)
util/readtags (also not required for building NetHack
itself) will decipher the nethack.tags file and produce
the functions that can deal with the NetHack struct data
fields.
You can obtain the source for universal-ctags by cloning it
from here:
https://github.com/universal-ctags/ctags.git
The combination universal-ctags + util/readtags has been
tried and tested under both Windows and Linux, so it is
not tied to a particular platform.
Note: util/readtags will work only with universal-ctags
output, so other ctags are unlikely to work as-is.
Universal-ctags can be build from source very easily
under Linux, or under Windows using visual studio.
Being turned into green slime never drops hero's inventory so invent
objects shouldn't be subject to obj_not_held() handling.
obj_not_held() does apply to undead. Arising as a mummy or vampire
doesn't go through the trouble of dropping everything and picking it
back up, but there is a point in the die...arise sequence where the
hero is implicitly a corpse so nobody is holding his/her stuff.
Dead hero's map coordinates are set to <0,0> part way through bones
creation, then were being used to record grave location for overview
feedback with "final resting place for <dead hero>" if/when another
character got those bones and found the grave (actually, spotted the
location where first hero died regardless of whether a grave gets
placed there). Record dead hero's pre-<0,0> coordinates as intended.
Not previously noticed because in wizard mode the final resting place
becomes part of overview info as soon as bones are loaded rather than
waiting for the death location to be reached.
The followup message about the fix for #5056 was trapped by the spam
filter so didn't reach us for a while.
xlogfile has an extra field to track various achievements made during
the game it logs, two of which are fully exploring the gnomish mines
and fully exploring sokoban. Those are accomplished by finding the
special 'prize' item on the final level of their branch: luckstone
for mines and bag of holding or amulet of reflecition for sokoban.
3.6.0 had a bug where any item of the target type found anywhere in
the dungeon resulted in achieving the relevant goal. A post-3.6.1 fix
for that required that the item be found on the end level of the branch
and attempted to require that it an item explicitly placed there by the
special level loader, but the latter aspect had a bug which meant that
random items of the appropriate type placed on final level would count
as the prize. Chance of extra luckstones on mines' end is fairly high,
so potential for false completion of the achievement was also high.
The second complaint was that since the achievement was only recorded
if the special prize item was found on final level, then if a monster
took it to another level then the achievement became impossible. (Not
true, the player could take it back, drop it, and pick it up again, but
that is admittedly a pretty silly hoop to jump through.) On the other
hand, if a monster removed the item before the hero found it, then a
case could be made that the hero hadn't really fully explored the
level. However, this fix records the achievement no matter where the
hero picks up the item. The final level must be entered--otherwise no
monster could possibly acquire and transport the item--but it isn't
guaranteed to have been fully explored. Big deal....
The prize could also be acquired in bones data. Before the second
portion of this fix, that wouldn't have mattered. But now it does, so
clear the prize indicator when saving bones unless it happens to be the
same level where that item is created (impossible for sokoban, where no
bones are left; not sure offhand about mines' end). The former prize
stone or bag or amulet becomes an ordinary one of its type.
This can all be done in a much cleaner fashion once we give up on the
current save file compatability. Putting obj->o_id values into new
context.mines_prize and context.soko_prize, plus a hack to mkobj() to
not reuse those two values if the o_id counter ever wraps back to 0,
would cover most of the details. Adding an achievement tracking flag
to lev_comp's object handling for use by the special level loader
would cover most of the rest.
The BONES_POOLS implementation added an extra dot to the bones file
name (only when enabled) which would be a problem on some filesystems.
This changes the name from "bonD0.15.3" to "bon3D0.15" which avoids
the second dot and also fits within 8.3 characters. To enforce that,
the maximum value for BONES_POOLS is now 10 (yielding single-digit pool
numbers 0 through 9).
BONES_POOLS==1 will omit the pool number (that's not a change, just a
reminder), yielding "bonD0.15" and so on. Right now, BONES_POOLS==0
is equivalent to BONES_POOLS=1, but it could be changed someday to
mean that bones files shouldn't be used if we decide to support that.
The pool number as a suffix was being included in content validation,
so it wasn't possible to move "bonD0.15.3" to pool 2 by renaming it to
"bonD0.15.2". I'm not sure whether that was intentional, but it seems
overly draconian. "bon3D0.15" can be renamed to "bon2D0.15" and then
be loaded by a game assigned to pool 2. Also, pre-pool bones can be
retained by renaming to any valid pool and should still work.
The three letter filecode for quest bones has made the bonesid be
broken since 3.3.0 introduced it (the three letter code, not bones-id).
"QArc.2" for level 2 of the Archeologist quest was being written into
the bones file as "rc.2", but worked as intended because validation
when loading bones had the same mistake. This fixes it to use "QArc.2"
when saving and accept either "QArc.2" or "rc.2" when loading, so 3.6.0
bones files (and existing to-be-3.6.1 bones) will continue to work.
Add some new routines for dealing with fruit. I had hoped they would
let the existing fruit handling be simplified quite a bit, but the
improvement wasn't great. However, they're also groundwork for fixing
an old bug.
Avoid the possibility of a user-supplied name interfering with killer
reason truncation. A monster named ", while" that killed the hero
would result in "killed by <mon-type> called " being displayed on the
tombstone after stripping while-helpless reason to shorten the text.
Make a fix suggested during beta testing: you can read scrolls while
blind if you know the label, and you can write a scroll with a magic
marker while blind, but the result was flagged as description unknown
so you couldn't read the newly written scroll until regaining sight
or obtaining object identification. So change writing a previously
discovered scroll while blind to set dknown since a successful write
always yields the type of scroll requested. Getting lucky while
attempting to write an undiscovered scroll--which has to be done by
scroll's type name (for instance "food detection") rather than by its
label ("YUM YUM")--still leaves the description flagged as unknown
since hero hasn't seen the what sort of label the new scroll has.
Along the way I got side-tracked by the possibilty of writing a scroll
of mail. It's allowed and yielded the same result as finding such a
scroll in bones, or wishing for one: when read, it was junk mail from
Larn. Make one written via marker give different feedback since it
comes from creation of a stamped scroll without any stamps available.
Also, suppress an "argument not used" warning for readmail().
With DEBUG suppressed, I started getting
16 warning: empty body in an if-statement
and 2 warning: empty body in an else-statement
from gcc.
Using braces for an empty block instead of just ';' avoids the warning:
if (foo)
debugpline("foo");
is bad,
if (bar) {
debugpline("bar");
}
is good. ;-)
The changes to lint.h are just precautionary.
modified:
include/lint.h
src/attrib.c, bones.c, dbridge.c, dig.c, eat.c,
makemon.c, mkmaze.c, mon.c, sp_lev.c
Reformat some trailing &&, || operators followed by end-of-line comment,
missed by the earlier continuation formating.
An
#if 0
something {
#else
something_else {
#endif
construct in rhack(cmd.c) confused the automated reformatter, resulting
in some code from inside a function ending up in column 1.
I'll push a formatting guide at some point. There may still be
outstanding changes, but please feel free to resolve those as you arrive
a them.
To the best of my knowledge, there is no changes to the actual code
content, but the formatter does have the occasional bug. If you run into
an issue, please fix it!
Changes to be committed:
modified: include/extern.h
modified: src/bones.c
modified: src/do.c
modified: src/files.c
modified: src/music.c
modified: src/restore.c
modified: src/save.c
modified: sys/share/pcmain.c
modified: sys/share/pcsys.c
modified: sys/share/pcunix.c
In order to get level file locking correctly again post 3.4.3
with the newer compilers for windows, I had to funnel close()
calls to an intercepting routine.
I had two choices:
1. Surround every close() in at least 9 source files with messy:
#ifdef WIN32
nhclose(fd);
#else
close(fd);
#endif
OR
2. Replace every close() with nhclose() and
deal with the special code in the nhclose()
version for windows, while just calling
close() for other platforms (in files.c).
It is also possible, although not done in this commit,
to
#define nhclose(fd) close(fd)
in a header file for non-windows, rather than funnel
though a real nhclose() function in files.c.
* Replace variadic debugpline() with fixed argument debugpline0(str),
debugpline1(fmt,arg), and so on so that C99 support isn't required;
* showdebug() becomes a function rather than a macro and handles a
bit more;
* two debugpline() calls in light.c have been changed to impossible();
* DEBUGFILES macro (in sys.c) can substitute for SYSCF's DEBUGFILES
setting in !SYSCF configuration (I hope that's temporary).
Move debugging output into couple preprocessor defines, which
are no-op without DEBUG. To show debugging output from a
certain source files, use sysconf:
DEBUGFILES=dungeon.c questpgr.c
Also fix couple debug lines which did not compile.
This also includes fixes due to Derek Ray to depugpline to work better
on other platforms.
1) add graves to the dungeon features being tracked;
2) report on known bones (determined by seeing map spot(s) where previous
hero(es) died since there's no guarantee of graves or ghosts);
3) add automatic annotations for oracle, sokoban, bigroom, rogue level,
Ft.Ludios, castle, valley, and Moloch's sanctum. For bigroom and rogue
level you just need to visit that level, for the others you need to get
far enough along to learn something specific (oracle: her room, sokoban:
annotation is either "solved" or "unsolved" depending upon whether all
the holes and pits have been filled, fort and castle: see the drawbridge,
valley and sanctum: see inside the tended temple). Discovering the
relevant locations via magic mapping counts as "far enough along".
There should probably also be automatic annotations for Medusa and the
vibrating square but I'm not sure what criteria should be used for the
former or what phrasing to use for the latter. Demon lord/prince lairs fall
into similar category as Medusa.
TODO: add final #overview as an end of game disclosure option. (I was
planning this even before I saw that nitrohack has implemented it....)
