In curses, selecting a partial stack, then unselecting the entry,
and then selecting it normally, the entry still kept the quantity
from the partial selection. Make it behave like all the other
windowports by resetting the quantity when the entry is unselected.
This is a better workaround for the missing message window border when
when restoring with align_status:left. Unlike the previous hack, this
may even be the correct way to handle it in the first place.
I haven't tracked down when the problem started.
This fixes the missing message window border when restoring with
align_status:left (which narrows and forces that window to the right).
It can't possibly be the correct way to fix things but does work.
This was not a problem with 3.6.4 (the most recent pre-3.7 playground
I have on hand) but I don't have the energy to use 'git bisect' to
track done when it started to whether the cause is discernable. The
message history is put into place without going through putstr() so
differs from normal message handling.
Dynamically resizing the terminal window during play by dragging its
edges was beeping (a lot if you dragged slowly). Recogize the key
code sent for that instead of complaining about unrecognized input.
Reported by jeremyhetzler: with recently revised curses interface,
when dismissing a menu or prompt with ESC the screen flashed.
This was caused by calling beep() when the terminal is set for
'visible bell' (of flagged as incapable of 'audible bell'); the
curses library flashed the screen deliberately. We don't want that.
Change the function key handling to not call beep() when an ESC
is typed by itself rather than being the leading part of a
multi-character escape sequence. beep() will still be called if
you type an arrow key (or other function key) when nethack is
expecting text input. That's what the recent change intended.
This also removes an early return from parse_escape_sequence()
when a number pad key generates an escape sequence. I don't have
a number pad to verify that this bit works as intended.
Closes#1002
While watching a ttyrec, I noticed strange behaviour in
the cursor updates; it was moved to approximately middle
of the map window every so often, usually when doing ranged
attacks. This wasn't really noticeable in normal gameplay,
as it was moving the cursor there, and then almost immediately
to where it was supposed to be.
I managed to trace it down to the refresh() in curses_delay_output().
That call updates the terminal to match the ncurses stdscr window,
but the stdscr window cursor position wasn't updated by NetHack.
Some bits made when attempting and failing to figure out the curses
problem (which turned out to be an early return skipping reset of
input timeout and fixed by paxed).
parse_escape_sequence changed getch timeout, but when receiving
a certain escape sequence, it returned without changing the
timeout back. This caused all subsequent getch calls be
non-blocking, and getch returns ERR when there's no input
waiting in that mode.
NetHack interprets ERR as the terminal going away, so performs
an automatic save.
In some scenarios where a message prompted a 'more' (in curses shown as
'>>'), the '>>' could overwrite the last one or two characters of the
message.
This could happen if a single message on its own was the length of the
message space minus 2 (e.g. for an 80 character terminal, "h - a concave
amulet named this is an amulet name, it is very long, and so on."), or
if an addition to an existing line brought it to the length of the
message space minus 1 or 2 (e.g. "You materialize on a different level!
You remember this level as testlevlname."). The two scenarios had
slightly different causes.
I think this fixes both of those scenarios. I believe this is the
cause of the problem described in GitHub issue #990 but there isn't
enough detail to know that for a fact.
Meta-key fix for curses interface running on top of ncurses library.
Previously only digits and lower case letters would produce a meta
character when combined with Alt (or Option on Apple keyboards), now
it should work for any basic character (not arrows or other function
keys). It only works on terminals that send two characters ESC k
for Alt+k but that is not a change in behavior.
curses interface running on top of PDcurses library uses different
code which isn't fixed by this. The alt key fixup it does have was
already present in curses_read_char() and recently got duplicated in
curses_convert_keys(). At least one other routine calls the latter
so it was necessary, but curses_read_char() calls that routine so
doesn't need to keep its own copy of the fixup.
They weren't working at all because the values of ALT_A through ALT_Z
were out of the normal char range with PDCurses, and caught by the default
case in the switch, where reject got set.
/* use key as-is unless it's out of normal char range */
reject = ((uchar) ret < 1 || ret > 255);
Issue reported by jeremyhetzler: left and right arrows produced
unexpected characters when trying to use them to edit text that is
being entered.
The curses interface converts arrow keys and function keys related
to the keypad into movement keys (hjkl or 4286 depending on the
number_pad setting). But it was doing that all the time, not just
when nethack wanted movement keys. This extends the existing
program_state.getting_a_command flag to getdir() so that it can be
used for controlling that in addition to 'altmeta' support.
Typing an arrow when interacting with the map (actual command or
getpos, now getdir too) will still work. Typing one when making a
wish or naming a pet will issue a beep and be treated as if '\0' had
been typed, and that normally gets treated as if ESC had been typed.
[Possible room for improvement there. Losing the whole text when
trying to back up a character feels a bit harsh.]
Treating left arrow as escape rather than as h or 4 will probably be
enough to train players not to try to edit text with it after they
get burned by that a time or two.
Bonus fix: curses' keystroke conversion only supported traditional
number_pad behavior, not the inverted phone number pad layout.
Closes#987
From the newsgroup: identifying by menu pops up multiple menus in
succession if the player picks fewer invent entries than are being
granted, but the second and subsequent ones could cover up the
message window and hide the feedback from prior ones.
If multiple popup menus are needed when identifying, issue --More--
before each menu after the first. The code seemed to be trying to
do this already, but it should have used wait_synch() rather than
mark_synch(), or perhaps used display_nhwindow(WIN_MESSAGE, TRUE)
instead of either one of those. For curses, both mark_synch() and
wait_synch() were no-ops. Now they do something. X11's behavior
wasn't right either; it seemed to be lagging one message behind
(something I had noticed recently and then forgotten about; I still
don't remember the context then so don't know whether this fixes
that earlier situation).
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.
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
^ ^
../win/curses/cursinit.c:102:9: warning: variable 'min_message_height' set but not used [-Wunused-but-set-variable]
int min_message_height = 1;
^
1 warning generated.
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() */
../win/curses/cursmain.c: In function 'curses_init_nhwindows':
../win/curses/cursmain.c:157:17: warning: unused variable 'pdc_font' [-Wunused-variable]
157 | static char pdc_font[BUFSZ] = "";
| ^~~~~~~~
../win/curses/cursmain.c: At top level:
../win/curses/cursmain.c:157:17: warning: 'pdc_font' defined but not used [-Wunused-variable]
A change to the curses interface from three years ago to make sure
that round-off didn't make the horizontal and vertical clipped map
indicators appear to not be clipped was using ROWNO for both instead
of COLNO for the horizontal one. For modest clipping the mistake
was unnoticeable; I don't know whether that remained true for more
extreme clipping.
[Not fixed: the curses scrollbar stuff ignores the fact that map
column 0 is unused.]
* When saving: curses_exit_nhwindows calls curses_uncurse_terminal,
which calls endwin. curses_exit_nhwindows then calls raw_print,
which calls more Curses functions after endwin has been called.
Fix this by having curses_raw_print use puts if window_inited
is false.
* When dying, quitting, etc.: really_done opens the "Goodbye" window,
which refreshes the other windows when it closes. But the status
window (and possibly the map and message windows) are gone by that
point. The window pointers are properly NULLed, but the NULL is then
passed to touchwin. Fix this by checking window pointers for NULL.
The definition of enc_stat[] got changed by a pull request nearly a
year ago ('const char *enc_stat[]' -> 'const char *const enc_stat[]')
but the separate declarations for it weren't changed to match.
Make the same change for hu_stat[]. Not sure why the pull request
didn't include it since the old declaration and the usage are same.
The curses one is in code that isn't used.
My /usr/include/curses.h has various A_attribute macros but A_ITALIC
isn't one of them. Compiling cursmisc.c failed because one of the
uses of that wasn't guarded by #ifdef A_ITALIC. Instead of adding the
ommitted #if, substitute A_UNDERLINE for A_ITALIC when that's missing.
The select attribute menu when adding a menu color or a status hilite
now shows an entry for italic that's underlined (as expected) but the
underline entry itself does not display any sort of attribute. I
didn't pursue that.
Redo the fix for using doset(#optionsfull) to toggle perm_invent
under curses. Move it to curses code and let the core be unaware
of it. It does the perm_invent update twice when creating the
window for that.
Enabling perm_invent with 'O' ('m O' these days) with curses used to
work but stopped at some point. Analysis by entrez has attributed
the change to the g.program_state.in_docrt flag in docrt(). When
curses creates the perm_invent window for update_inventory(), it
calls docrt() to have nethack redraw the screen.
docrt() -> update_inventory() -> curses_update_inventory() -> ...
-> curs_reset_windows() -> doredraw() -> docrt() [early return]
resulted in room for the persistent inventory window but it was
blank.
This also replaces a couple of doredraw() calls with direct calls to
docrt() (one in code that isn't used). doredraw() implements a user
command; docrt() does the actual redrawing.
This replaces the old pushq/saveq arrays (which were used to save
the keys pressed by the user for repeating a previous command)
with a new command queue. This means there's no hard-coded limit
to the saved keys, and it can repeat extended commands which are
not bound to any key.
The 'wizmgender' option is flagged as 'wizonly' in optlist.h but that
doesn't prevent it from being set in NETHACKOPTIONS or .nethackrc.
Apply the fix from entrez to only honor it when running in wizard
mode.
The wizard-mode option to highlight female monsters stopped having any
in-game effect after cb0c21e. Formerly it caused female monsters to be
highlighted with a red background (red color + inverse); this commit
uses inverse video only without overriding their color. Ensuring the
color override works consistently with the ENHANCED_SYMBOLS 24-bit color
doesn't seem worth it for what is a very niche debugging option, and I
think inverse video should probably suffice.
It also used to be a TTY-only option, but this enables it in curses as
well.
Change the inner workings of the experimental TTY_PERM_INVENT.
Switch to delivering the content to tty for the experimental perm_invent
via the existing window port interface (start_menu(), add_menu(), end_menu).
This also adds a new window port interface call ctrl_nhwindow() for
delivering information to the window port, and/or obtaining specific
information from the window port. The information and requests can
be extended as required. To be documented later once the changes settle
down.
Due to the intrusive nature of these changes and the possibility of
some bugs in the new code, I'm going to leave TTY_PERM_INVENT commented
out in the repository for a day or two. Anyone wishing to test it out
can do so by uncommenting TTY_PERM_INVENT in config.h.
One of the drivers of this change was that screen coordinates require a
type that can hold values greater than 127. Parameters to the window
port routines require a large type in order to be able to have values
a fair bit larger than COLNO and ROWNO passed to them, particularly for
their use to the right of the map window.
This splits the uses of xchar into 3 different situations, and adjusts
their type and size:
xchar
|
-----------------------
| | |
coordxy xint16 xint8
coordxy: Actual x or y coordinates for various things (moved to 16-bits).
xint16: Same data size as coordxy, but for non-coordinate use (16-bits).
xint8: There are only a few use cases initially, where it was very
plain to see that the variable could remain as 8-bits, rather
than be bumped to 16-bits. There are probably more such cases
that could be changed after additional review.
Note: This first changed all xchar variables to coordxy. Some were
reviewed and got changed to xint16 or xint8 when it became apparent that
their usage was not for coordinates.
This increments EDITLEVEL in patchlevel.h
Add a non-string identifier to window_procs for use in runtime
identification of the current window port being used.
Use a macro WPID to add the identification at the top of the
various existing window_procs declarations. It expands to the
existing text string, as well as the newly added field wp_id
with a wp_ identifier.
For example, WPID(tty) expands to: "tty", wp_tty
The generated wp_tty must be present in the wp_ids enum at
the top of include/winprocs.h.
The WINDOWPORT(x) macro has been updated to expand to a simple
value comparison (port.wp_id == wp_x), instead of a
string comparison.
