combine boolean and compound options into a single allopt[] array for
processing in options.c.
move the definitions of the options into new include/optlist.h file which
uses a set of macros to define them appropriately.
during compile of options.c each option described in include/optlist.h:
1. automatically results in a function prototype for an optfn called
optfn_xxxx (xxxx is the option name).
2. automatically results in an opt_xxxx enum value for referencing
its index throughout options.c (xxxx is the option name).
3. is used to initialize an element of the allopt[] array at index
opt_xxxx (xxxx is the option name) based on the settings in the
NHOPTB, NHOPTC, NHOPTP macros. Those macros only live during the
compilation of include/optlist.h.
each optfn_xxxx() function can be called with a req id of: do_init, do_set,
get_val or do_handler.
req do_init is called from options_init, and if initialization or memory
allocation or other initialization for that particular option is needed,
it can be done in response to the init req.
req do_set is called from parseoptions() for each option it encounters
and the optfn_xxxx() function is expected to react and set the option
based on the string values that parseoptions() passes to it.
req get_val expects each optfn_xxxx() function to write the current
option value into the buffer it is passed.
req do_handler is called during doset() operations in response to player
selections most likely from the 'O' option-setting menu, but only if the
option is identified as having do_handler support in the allopts[]
'has_handler' boolean flag. Not every optfn_xxxx() does.
function special_handling() is eliminated. It's code has been redistributed
to individual handler functions for the option or purpose that they serve.
moved reglyph_darkroom() function from options.c to display.c
Preparation for moving enlightenment and conduct into new source
file insight.c. Right now it's a stub that shouldn't break anything
whether included or omitted. Once makefiles and project files have
been updated to compile and link it, the actual code will be moved.
unix/Makefile.src has been updated;
vms/Makefile.src and vmsbuild.com have been updated but not tested.
Add the new header file to HACKINCL which only matters to HSOURCES
which only matters for the 'tags' target.
'make depend' reordered some stuff that could have stayed as-is but
also found a missing dependency for nhlobj.{c,o}.
recover had deviated somewhat from NetHack in its
file expectations:
1) A couple of 3.7 fields needed to be accommodated.
2) hard-coded file size values had deviated.
The file sizes are now in an added header file named "filesize.h",
which is included at the bottom of config.h.
There will likely be another commit to write the filename size ahead
of the file name so that the precise number of characters can be read,
but since that will break existing saves, it can go in along with another
save-breaking commit.
This commit doesn't not alter savefiles written by nethack so does not
require an editlevel bump. It does alter the read-in expectation in
recover to match the game and this get recover working again.
In sys/unix/Makefile.{top,src}, move the 'all' target first so that
it's the default if no target is specified. Recent lua handling
unintentionally broke that for Makefile.top by putting 'lua_support'
target in front of 'all'.
In order for 'make depend' to be able to handle both Qt4/5 and Qt3,
they need to operate on different object file names.
renames qt*.o to qt3*.o for Qt3
renames qt*.cpp to qt3*.cpp for Qt3 (not essential but seems worthwhile)
moves Qt3's headers from include/qt*.h to win/Qt3/qt3*.h
copies include/qt_xpms.h (before rename) or win/Qt3/qt3_xpms.h (after)
to win/Qt/qt_xpms.h so that Qt4/5 no longer shares one header file
modifies win/Qt3/*.cpp and win/Qt3/qt3_win.h to reflect new header names
modifies Makefile.src to have Qt3 'moc' commands use new names
updates Makefile.src via re-running 'make depend'
'make depend' was only looking at include/*.h to find nested inclusion.
Now it will also look at win/*/*.h. That found a bunch of missing
dependencies for the old gnome sources and a few for Qt3.
Building without Qt still works. Building with it (any version) has
not been tested.
I deleted three lines
../include/../lib/lua-5.3.5/src/lua.h \
../include/../lib/lua-5.3.5/src/lualib.h \
../include/../lib/lua-5.3.5/src/lauxlib.h \
from $(CONFIG_H) but other than that, this is as-is with unmodified
'make depend'. depend.awk will need to be taught about suppressing
those lua headers.
Copy lib/lua-$(VERSION)/src/liblua.h to lib/lua/ rather than lib/.
Instead of copying any of the header files or telling the compiler
where to find the lua ones, generate include/nhlua.h on the fly and
restrict the knowledge of where they are to it (paths are relative
to include/).
|/* nhlua.h - generated by top Makefile */
|#include "../lib/lua-5.3.5/src/lua.h"
|LUA_API int (lua_error) (lua_State *L) NORETURN;
|#include "../lib/lua-5.3.5/src/lualib.h"
|#include "../lib/lua-5.3.5/src/lauxlib.h"
|/*nhlua.h*/
This might need to be redone (or augmented by having CFLAGS add back
'-I path-to-lua') if some compiler can't find '#include "luaconf.h"'
issued by lua.h.
Reduce the implied reliance of a specific version of lua.
Instead of copying liblua.a to src/, copy it to lib/. Instead of
telling the compiler to look for headers in lib/lua-5.3.5/src/ as
well as in include/, copy the relevant ones to lib/ and tell the
compiler to look for them there. 'make spotless' in src/ will
remove both the object library and the header files from lib/ but
there really should be a new Makefile.lib to take care of that
directory.
Update Makefile.src to be able to build lua in case someone starts
with 'make all' there instead of in the top Makefile. It doesn't
duplicate the option to fetch the lua source package though.
NHinternal/../genFiles.c has been updated to mention lib/liblua.a
and lib/lua*.h as 'generated at compile time' in Files and to skip
lib/lua-* entirely if it comes across that (so not operating on a
completely clean tree). But it won't be accurate unless/until
other ports stage their lua files there instead of in src/ and
lib/lua-$(VERSION)/src/.
I haven't tried 'make depend' to see what it makes of the numerous
changes....
Game is playable, and should compile on linux and Windows.
Assumes you have a lua 5.3 library available.
Removes level compiler and associated files.
Replaces special level des-files with lua scripts.
Exposes some NetHack internals to lua:
- des-table with commands to create special levels
- nh-table with NetHack core commands
- nhc-table with some constants
- u-table with some player-specific data (u-struct)
- selection userdata
Adds some rudimentary tests.
Adds new extended command #wizloadlua to run a specific script,
and #wizloaddes to run a specific level-creation script.
nhlib.lua is loaded for every lua script.
Download and untar lua:
mkdir lib
cd lib
curl -R -O http://www.lua.org/ftp/lua-5.3.5.tar.gz
tar zxf lua-5.3.5.tar.gz
Then make nethack normally.
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.
Separate the compiler flags used for compiling X11 code from the rest
of CFLAGS. Affects hints/macosx10.8 and later.
Add an explicit output argument to the generated compile rules.
Back out '#include "date.h"' so that cursinit.c won't be recompiled
every time any other file(s) need to be compiled. It doesn't need
patchlevel.h either. There is already a straightforward way to fetch
the copyright banner lines from version.c.
The splash screen (ascii art spelling "NetHack" preceding the normal
copyright lines) was invisible when showing white text on white-ish
background. Make it honor !guicolor.
"Shall I pick a character's role, race, gender and alignment for you?
[ynaq] (y) " was too wide to accept the answer on the same line on
an 80-column display so "(y) " was placed on the second line. That's
constructed in the core; change the construction to omit " a" when
using "character" rather than a role name. (tty shortens it by omitting
the default " (y)"; with " a" gone, it could revert to normal prompt.)
Also a bit of lint cleanup and some reformatting of cursinit.c....
People have been wondering how to change the tiles on the X11
version, and the old default of NetHack-specific binary tile data
isn't directly editable with image editing tools.
Also show in the #version info if xpm and graphic rip are enabled.
When make uses 'makedefs -v' to create date.h, force it to create
gitinfo.txt all the time instead of just when that doesn't already
exist. Use 'make GITINFO=0' to get the previous behavior.
To skip it entirely, you need to do that and also make sure that
some file by that name already exists. 'touch dat/gitinfo.txt' or
perhaps 'echo "#no git" > dat/gitinfo.txt' would suffice.
Hide the scary perl command during 'make all' feedback in src/.
I'm not a shell programmer but this works fine for me when skipping
NHgithook::NHversioning because dat/gitinfo.txt is already present,
when creating dat/gitinfo.txt successfully, and when creation fails
because perl can't find NHgithook.pm. It hasn't been tested when
perl itself is absent.
Originally by Ray Chason for 3.4.3, based on the Qt windowport by
Warwick Allison. The look and feel is mostly the same.
Some improvements over the Qt 3 interface are:
* Panes are resizable
* Full support for IBMgraphics, and walls and corridors are drawn with
graphical primitives for a continuous appearance no matter what the font
says
* Lots of irritating glitches fixed
* Menus support proportional fonts correctly
Adding this because the old Qt windowport cannot be compiled on Qt4,
even with Qt3 compatibility stuff.
TODO:
- background map glyphs
- status hilites
- menucolors
[Subject should mention Unix, but would exceed 50 characters.]
Explicit build rules ignore $(CPPFLAGS), but the implicit C rule
(at least in GNU make) specifies it. If user has a value for this
in the environment, that value would apply to building some source
files but not others. This patch gives it an explicit empty value,
so building via implicit rule should expand it to nothing and match
the fact that it's omitted from explicit rules.
There was one C++ file which relied on the implicit C++ rule. I've
added it to the files processed by 'make depend' and re-run that.
It now will get built via an explicit rule.
Also, a small amount of reformatting for HACKCSRC.
Update sys/unix/Makefile.src to force it to build monst.o and
objects.o before attempting to build makedefs, so that building
the latter doesn't use the rules for those two object files in
Makefile.utl. Prior to this, if 'make' was building multiple
targets in parallel monst.o and/or objects.o might be clobbered
when a process using util/Makefile tries to build them while its
parent is also building them via src/Makefile.
This includes a bit of reformatting which wasn't present in
yesterday's email attachment.
dungeon.o depending on lev.h is the only change found by 'make depend'.
(I'm a bit suspicious about that.)
I haven't attempted to reconcile the vms Makefiles with the Unix ones,
just put in this one new dependency. I know vms/Makefile.src lacks
handling for sys/share/*regex.c and vms/Makefile.top and install.com
both lack handling for 'sysconf'.
I started out cleaning up a bit of lint in the recent run-time options
handling and discovered that pmatchregex wasn't finished. Finish it and
also deal with the version lint. Argument declarations for function
definitions in pmatchregex.c have been switched to K&R style. (The ones
in posixregex.c have been left in ANSI style.)
There wasn't any build rule for pmatchregex.o; now there is (for Unix).
posixregex.o is still the default.
There isn't any build rule for cppregex.o (again, for Unix); the change
to cppregex.cpp is untested.
