Using role:!wizard to limit which roles would be candidates for
random selection didn't work as I expected. It required a separate
option setting for role to exclude. This implements how I thought
it worked:
|OPTIONS=role:!ranger !samurai !wizard
will exclude multiple roles with a space-separated list in a single
option setting. It also adds support for
|OPTIONS=!role:ranger samurai wizard
to do the same thing. (OPTIONS=!role:!ranger isn't allowed.)
I thought 'OPTIONS=role:barbarian caveman knight' could be used to
limit random selection to those choices, but that doesn't work and
I haven't attempted to implement it.
This also renames the 'align' option to 'alignment'. That made the
truncation to 'align' become ambiguous, so it got added back as an
alias for the full name.
Guidebook.tex is lagging; I'm burned out.
Move the new stuff from PR #946 out of xname() into new routine
armor_simple_name().
Noticed while testing: tweak 'call object type' so that it doesn't
list instances of pre-discovered armor as likely candidates since
assigning a name to such wasn't showing up in the discoveries list.
Add "silver shield" as wishing synonym for "polished silver shield".
This ended up combining several unrelated changes.
Add missing 'fixes' entry for curses-specific item in New Features.
When answering "Shall I pick ... for you? [ynaq]", accept \m as well
as \n and space for choosing the default of 'y', same as normal
ynaq() would. Also add '*' to '@' as not-shown potential answers;
they force 'random'.
When tty tore down any of the menus, things were reasonable if they
were short enough for corner windows, but tall ones that switch to
full screen weren't fully erased. The parts of those outside of the
map window stayed behind when the tall menu was closed and cleared.
Mainly affects picking the "~ - reset filtering" choice but also
affected the role menu on 24 line tty screens. (Didn't affect curses
because it tracks and refreshes its base window when some overlaying
window goes away.)
The role menu used 25 lines so required a second page for the case
of a 24 line screen on tty. Dealing with that is a bit ugly but it
wasn't an issue when this form of role selection was tty-only (because
the info about choices made so far was displayed on the base window
rather than in an extra menu line back then) so I added a hack for it.
If the role menu will take one more line than the screen height, the
separator between 'random' (below 'Wizard') and 'pick race first' gets
squeezed out. If the menu needs two more lines (doesn't happen now,
except by changing screen size to 23 lines for testing), a second line
gets squeezed out. (Not attempted for curses because it wouldn't
help. 'windowborders' and one or two extra separators it adds make
menus taller. I doubt if many players use curses on 24-line screens
but if they do, they'll be using something new rather than going from
something that used to fit on one page with 3.6.x.)
Move the tty role/race/&c selection from wintty.c to role.c and remove
its references to BASE_WINDOW. Have curses call the same routine now
so that the player has the option to choose role, race, gender, and
alignment in any order and to confirm or override random settings
prior to starting play. Also if you went through "who are you?" then
final confirmation includes an extra menu choice to rename the hero.
It still has the quirk of sometimes remembering some of the previous
aspects when you re-pick a new value for some aspect which already
been selected.
The menus pop up on top of the copyright screen and that looks a bit
strange. I don't think core code has any way to erase that base
window without erasing the entire screen so to fix the strangeness
the window ports would need to do that before calling the selection
routine. I didn't do that because the very first prompt, "Shall I
pick ... for you? [ynaq]" shows up in that window rather than in a
popup over it, and having it be all by itself on an otherwise blank
screen seemed to be even stranger.
X11 and Qt both have more sophisticated selection routines so I
haven't tried to switch either of them to use this. They both use a
fancy role-selection-specific menu with all the aspects present at
once so this wouldn't fit without more work than I care to tackle.
Don't require the list of generic usernames in sysconf to need to be
ordered to guard against false substring matches. If the list was
"nethacker nethack" and the tentative character name was "nethack",
it wouldn't be recognized as generic. The old code forced the list
to be "nethack nethacker" for the matching to work correctly because
it only checked the first matching substring. Either order works now.
It also failed to recognize a generic name if the player used
|nethack -u nethack-samurai-human-male-lawful
because it checked for generic names before stripping off the role
aspects. Now that will at least recognize the name as generic and
prompt with "who are you?", but the role/race/&c info gets discarded.
The unlink call wasn't operating the same on a makedefs built
on mingw, and dat/mdXXXX files were being left behind post-build.
Provide an alternative way of doing the temporary files if
MD_USE_TMPFILE_S is defined during the compile of makedefs.c
This should fix the problem of polymorphing or stone-to-fleshing a
pile of multiple boulders and having some underneath ones which get
changed resist and not get changed, producing a pile with one or more
non-boulders above one or more boulders. If that situation arises,
re-stack the pile so that boulders are moved to the top.
This also revises zapping up or down while hiding under something
(if that is even possible; the types of creatures which can hide
under things can't zap wands or cast spells; maybe there are some
exceptions?). Zapping up used to hit only the top item, but zapping
down hit the whole stack. Now up still hits only the top, but down
skips the top and hits the rest.
Caveat: not adquately tested.
Redo the details about giving or suppressing "with great effort you
push the boulder". It works the same except that if push a different
boulder than previously, you'll get a new message. If you do it
while riding, you have the same lack-of-message for successive pushes
instead of getting a message every turn.
Don't exercise strength when pushing a boulder if poly'd into a giant.
I don't know whether we'll ever run the code through clang-format
again, but in case we do, end-of-line comments which span lines
should begin continuation lines with '*', otherwise they get turned
into block comments beginning on the next line.
|code; /* start
| end */
gets changed to
|code;
|/* start end */
but
|code; /* start
| * end */
stays as-is.
The consolidation of global variables from scattered source
files into decl.c and declared in decl.h was begun in 3.7.0.
Their placement in common files was done for centralized
initialization and potential re-initialization during a
"play again" scenario.
It wasn't really necessary for all of them to be housed in a
single huge structure to meet the "play again" requirement,
and the single huge structure has been a little unwieldy when
it comes to maintenance.
Following this commit, instead of one single extremely large structure
named 'g' to house all of the relocated global variables, they
are distributed into several ga through gz.
To make things easy for the developer, each variable is placed
into the struct corresponding to the starting letter of the variable.
That way, no lookup is required in order to know which struct houses
a particular variable, it is a simple match to the starting letter
for all the centralized global variables.
A global variable named 'amulets', would be found in ga.
ga.amulets
^ ^
A global varable named 'move', would be found in gm.
gm.moves
^ ^
A global variable named 'val_for_n_or_more' would be found in gv.
gv.val_for_n_or_more
^ ^
A global variable named 'youmonst' would be found in gy.
gy.youmonst
^ ^
It turns out that there are some objects marked unpaid that aren't
carried by the hero, so the recent sanity check for unpaid/no_charge
could complain. Unpaid items dropped on the shop boundary (gap in
shop wall, doorway, shk's free spot) stayed unpaid when dropped onto
the floor, similar to recent change for pushed shop-owned boulders.
Don't give sanity complaints for those. They could be all the way
inside a shop too, where unpaid items in a gap in the shop wall got
pushed into the shop when the wall was repaired. (Possibly those
should come off the bill instead of remaining unpaid.)
Teleporting items out of a shop was marking them unpaid instead of
treating that as robbery. That's a bug caught by the sanity check.
rloco() was also marking shop items which got teleported from one
spot inside the shop to another spot inside the same shop as unpaid.
Fix both of those things. Also, if an unpaid item on the boundary
gets teleported all the way inside, take it off the bill.
Change 'I u' to mention whether there are additional unpaid items on
the floor somewhere since they won't be part of unpaid inventory and
they're not on the used-up bill either. It might occasionally help
the player figure out why the shopkeeper won't let the hero out of
the shop.
If paranoid_confirm settings include praying, don't put the answer
to "are you sure you want to pray?" into the do-again buffer where ^A
would use it to ignore confirmation if prayer is repeated. And for
wizard mode, when confirmation is 'y' then the answer to "force the
gods to be pleased?" has to be suppressed from the do-again buffer too
or it would be used by subsequent ^A to answer "are you sure?".
This is basically a band-aid just for #pray. There are probably other
confirmations that should be suppressed from do-again instead of being
reusable. The rest of the paranoid_confirm ones should be ok because
they require "yes" and that doesn't end up in the do-again buffer, but
there are bound to be other confirmations that shouldn't automatically
be re-used during repetition.
Although gcc specifies support for declaring a function as
noreturn after the function name and parameters, other compilers
do so via an attribute at the start of the declaration. Add some
macro support for the attribute-at-the-beginning method:
o MS Visual Studio compiler
o Upcoming C23 standard (untested at this point)
Handle items in gaps of a wall shared between adjacent shops.
Make handling of shop boundaries more explicit: walls, the door,
and the "free spot" by the door aren't classified as 'costly' but
obj->unpaid and obj->no_charge are valid there.
Move unpaid/no_charge checking into its own routine to unclutter
objlist_sanity().
Pushing a shop-owned boulder to the free spot or doorway or gap in
wall triggers the sanity check for the time being.
When dist2() got changed to use coordxy parameters, a macro that uses
it in its definition was overlooked and it had (int) casts in it.
That caused a warning about possible data loss when the int
then got converted to coordxy for the dist2() call.
Give online2() coordxy parameters instead of int, like its bretheren.
Avoid a couple of implicit conversion warnings where ints were being assigned
to smaller uchar or ints being assigned to smaller short.
A couple of signed vs unsigned warnings on some rumor processing.
Avoid some signed vs unsigned warnings in mdlib/makedefs where a signed int
param eventually got used in an external call that took size_t.
Eliminate all of it by just having the outer NetHack routine also take
a size_t.
Lastly, insert some default C99 alternative time-related code
in mdlib/makedefs since asctime() and ctime() are being flagged as
deprecated in the upcoming C23 standard and will now start to trigger
warnings for anyone using a C23-compliant compiler.
Pushing a shop-owned boulder out of the shop wouldn't charge the hero
anything. Remedy this (and remove the boulder from the bill if the hero
then pushes it back in). Also tried to handle a couple other uncharged
boulder "theft" scenarios: pushing a boulder into lava or water, into a
trapdoor or hole, or into a level teleporter (various other traps
already charged for the boulder -- it was pretty inconsistent).
I externified onbill() for this, since relying on otmp->unpaid by itself
impossibles if you push a boulder through a gap in a wall between two
adjoining shops.
Short for distu(mtmp->mx, mtmp->my) (i.e. the distance between the hero
and the specified monster), which is a very common use of distu(). The
idea is that this would be a convenient shorthand for it; I actually
thought it (or something very similar) existed already, but couldn't
find it when I tried to use it earlier. Based on the number of uses of
fully-spelled-out 'distu(mtmp->mx, mtmp->my)' replaced in this commit
I'm guessing I just imagined it.
Instead of using a compile-time macro to suppress inclusion of the
menu entry to show UNIX command-line usage in the help menu, use a
sysconf setting instead.
Default is HIDEUSAGE=0, to include the entry for command-line usage.
Set HIDEUSAGE=1 to exclude that. Does not affect 'nethack --usage'
if player actually has access to the command-line.
Write up a description of how the command line works on UNIX and put
that in new file dat/usagehlp. Add support for
|nethack --usage | --help | -? | ?
to display it and exit.
Also add a menu entry for nethack's help command to show it during
play. That can be suppressed by uncommenting new '#define HIDE_USAGE'
in config.h since it won't be useful on servers that don't give
players access to command lines.
New genl_display_file() just writes to stdout. opt_usage(), which
calls it, might need some suid/sgid handling to make sure the output
is done as the player rather than as nethack.
doc/nethack.6 is already out of date again.
1. remove all window interface bits from compiler.370, and have
the preceding include files set some variables to control
the behavior of compiler.370 when it comes to c++.
2. some more common Makefile lines into sys/unix/hints/include/multiw-3.370.
3. make it so you can pass cppregex=1 on the Make command line to build with
sys/share/cppregex.cpp instead of posixregex.c
4. fix sys/share/cppregex.cpp so that it will build with clang compiler
(required an additional header include). I don't know if it would have
worked with g++ without that change. The include can be placed into an #ifdef
block if there's an issue with the change on other compilers.
5. Anything that needs to compile using c++ (Qt, sys/share/cppregex.cpp) can
just ensure that CPLUSPLUS_NEEDED Makefile variable is set above the lines
in compiler.370 to ensure that things get set up for c++. It no longer
checks specifically for Qt. That is what sys/unix/hints/include/multiw-2.370
does now.
Condense the setup of PCHAR/PCHAR2 and OBJCLASS/OBJCLASS2 (last one
renamed from OBJCLASS7) so that it's easier to see the variations
at once on an ordinary size terminal/window. Revise some of the
indentation and other spacing, also to try to enhance readability
a little.
Unrelated: remove a trailing space that crept in with a recent pull
request.
Now, the only usage of GCC_WARN is for the guard of PRINTF_F in wincurs.h.
This guard can be removed safely, as PRINTF_F is already used unconditionally in extern.h.
We don't know at compile time whether the X server even supports
Unicode. So take the configured map font, change the registry to
ISO10646 and try the resulting font string. Keep the configured
font if that doesn't work.
GCCs older than 3.1 understand __attribute__(printf(...)), but only
with functions; it doesn't work with function pointers. This change
uses PRINTF_F_PTR to remove the attribute from two function pointers.
This change establishes GCC 3.0 as the minimum version to build
NetHack. Older versions have trouble with the variadic macros and
variable declarations in mid-block.
I made more things in dump_enums() static and/or const. In the
process I discovered both compile problems for NODUMPENUMS and when
fixed, link problems for NODUMPENUMS+ENHANCED_SYMBOLS.
The uft8map.c portion has no changes, just reformatting.
Option parsing rejected
|OPTIONS=!cond_X
for all valid X.
Using the menu to unselect all condition options treated that as not
having made any choice and didn't make any changes. That would be
reasonable if nothing was preselected, but things are so unselecting
all of them is a choice. (A bizarre one, but still should be viable.)
Mostly this deals with including cond_X options when #saveoptions is
used to write a new RC file. It now produces something like
|OPTIONS=!cond_barehanded,cond_blind,!cond_busy,cond_conf,!cond_deaf,\
| cond_iron,cond_fly,cond_foodPois,!cond_glowhands,cond_grab,\
| cond_hallucinat,!cond_held,!cond_ice,cond_lava,cond_levitate,\
| !cond_paralyzed,cond_ride,!cond_sleep,cond_slime,!cond_slip,\
| cond_stone,cond_strngl,cond_stun,!cond_submerged,cond_termIll,\
| !cond_tethered,!cond_trap,!cond_unconscious,!cond_woundedlegs,\
| !cond_holding
after the last alphabetical option and before the bound keys, menu
colors, and others which aren't simple OPTIONS=X settings. This only
happens if there is already one or more OPTIONS=cond_X entries in the
old file when it was read or if 'mO' gets used to make any changes.
Not fixed: after my RC had something similar to the above and before
I changed status conditions to accept negation, I was getting several
"the cond_ option may not both have a value and be negated" messages
written to stdout instead of the config file error handler. So they
vanished when the screen was initialized without providing a --More--
prompt to acknowledge that they have been seen.
Instead of using index() macro defined to strchr, use C99 strchr.
Instead of using rindex() macro defined to strrchr, use C99 strrchr.
If you want to try building on a platform that doesn't offer those
two functions, these are available:
define NOT_C99 /* to make some non-C99 code available */
define NEED_INDEX /* to define a macro for index() */
define NEED_RINDX /* to define a macro for rindex() */
