Provide a little more information when dumplog is unavailable.
While testing various permutations, I encountered a couple of
problems with conditionally declared variables.
The Windows data file path has to be constructed because
Windows defines VERSION_IN_DLB_FILENAME.
Keep the personal configuration file details as the last
information displayed.
Have the --showpaths feedback mention whether dlb is in use or not,
and show the container file name(s) when it is. Users of prebuilt
binaries or who build with a hints file instead of picking and
choosing things in config.h might not know, and vms (if it ever
catches up with --showpaths) uses a different container name from
everybody else ("nh-data.dlb" instead of "nhdat").
While not a path exactly, the dumplog file isn't placed somewhere
fixed so being able to see where it is placed could be useful.
This cascaded a bit during testing. Fix one of the warnings from
hardfought (fqn_prefix_names[]). And a few more that came up with
SYSCF disabled (panictrace_gdb, two unused variables if files.c).
Fix a couple of warnings about unused variables but mainly handle
OSX's multiple-choice for configuration file name. I tried moving
my ~/.nethackrc to '~/Library/Prefences/NetHack Defaults.txt' and
both regular play and --showpaths are finding it successfully.
Also an untested fix for VMS. "nethackini" is not supposed to have
any path information attached.
Add
--showpaths
early option to show where NetHack is expecting to find certain files
without starting up a game. It exits afterwards.
Windows sample (for illustration only, locations may differ for you):
Variable playground locations:
[hackdir ]="C:\Users\JohnDoe\NetHack\3.6\"
[leveldir ]="C:\Users\JohnDoe\AppData\Local\NetHack\3.6\"
[savedir ]="C:\Users\JohnDoe\AppData\Local\NetHack\3.6\"
[bonesdir ]="C:\ProgramData\NetHack\3.6\"
[datadir ]="C:\personal\nhdev\363\test\binary\"
[scoredir ]="C:\ProgramData\NetHack\3.6\"
[lockdir ]="C:\ProgramData\NetHack\3.6\"
[sysconfdir]="C:\ProgramData\NetHack\3.6\"
[configdir ]="C:\Users\JohnDoe\NetHack\3.6\"
[troubledir]="C:\Users\JohnDoe\NetHack\3.6\"
Your system configuration file (in sysconfdir):
"C:\Users\JohnDoe\NetHack\3.6\sysconf"
Your system symbols file (in sysconfdir):
"C:\Users\JohnDoe\NetHack\3.6\symbols"
Your personal configuration file (in configdir):
"C:\Users\JohnDoe\NetHack\3.6\.nethackrc"
Linux (for illustration only, locations may differ for you):
Your system configuration file:
"/home/johndoe/nh/install/games/lib/nethackdir/sysconf"
Your system symbols file:
"/home/johndoe/nh/install/games/lib/nethackdir/symbols"
Your personal configuration file:
"/home/johndoe/.nethackrc"
Reported directly to devteam, the POSIX_TYPES subset (most? all
these days?) of Unix that defines USE_FCNTL was unlocking lock file
'perm' when done with it but wasn't explicitly closing it unless
the unlocking failed. Triggered a valgrind complaint and could have
posed a problem if restart gets implemented for this configuraiton.
A few symbol-related modifications:
- fulfill a request from a blind player to allow them to
specify a unique/recognizable character for all pets and/or
the player in the config file for use when using a screen
reader (S_player_override, S_pet_override). Requires sysconf
setting ACCESSIBILITY to be set to have an effect, although
they can still be specified in the config file.
- Config file SYMBOLS entries were not working properly on
the rogue level. Allow ROGUESYMBOLS as well as SYMBOLS to be
specified in the config file independently.
- When values are moved into showsyms[], the overriding SYMBOLS
or ROGUESYMBOLS entry from the config file is used if there is
one; if there is no overriding value for a particular symbol,
the loaded symset value is used; if there is no symset entry
loaded for the symbol then a default symbol is used.
Fixes#235
For initial options under curses, specifying 'DECgraphics' as a
boolean rather than as 'symset:DECgraphics' wasn't overriding the
new default 'symset:curses'. Since previously DECgraphics was
rejected for curses, it's possible that no one noticed.
Change the way symbol sets are loaded to make them have the same order
as they appear in the symbols file rather than being reversed.
Revise dat/symbols so that the new ordering yields a result similar
to the old ordering, more or less. I've added a few set descriptions.
The only substantive change is marking DECgraphics as primary-only
(not available on rogue level) and adding new set DECgraphics_2 which
is commented out near the end.
Define symbol handling H_MAC since one of the sets specifies
'handling: MAC'. All H_MAC is used for now is to avoid showing
MACgraphics as a symset when compiled without MAC_GRAPHICS_ENV (which
was used for pre-OSX Mac by the old code in sys/mac/), so it will be
hidden for everyone.
I left handling H_CURS even though curses doesn't implement anything
for it. It could do something when rendering the map or assign a
function to 'cursesgraphics_mode_callback' for special init or both
but hasn't needed to. Since curses is now supporting DECgraphics,
define 'decgraphics_mode_callback' for it. No value is being
assigned so that doesn't do anything; curses seems to be setting up
the primary character set as text and secondary one as line-drawing
without the need for that hook.
With the added set descriptions, 'O's symset menu looked horrible for
curses due to the way curses decides to set the width of menus and
the resulting line wrapping which took place because of a too-narrow
menu. I've added a chunk of code to the options handling code which
shouldn't really be there but makes the menu much easier to read.
Lastly, do some formatting cleanup in files.c.
Fixes#200
The Guidebook claims that there's no symbol for 'S_strange_object'
which is literally true, but there is one for S_strange_obj. It has
been in place longer than the paragraph claiming that there's no way
to customize that symbol. I'm not sure why variant spelling was used.
Also, files.c doesn't use loadsyms[], it calls a routine which returns
a pointer to a specific element in that array.
Observed on a text file with crlf endings on Linux, where the
so-called blank lines weren't being ignored as they should
have been, despite there being a line to remove \r from the
end.
A pointer was being pre-decremented before the check for a
matching whitespace character.
Changing an inventory item's bknown flag wasn't followed by a call to
update_inventory() in many circumstances, so information which should
have appeared wasn't showing up until some other event triggered an
update.
I ran the fuzzer with MONITOR_HEAP enabled and heaputil found a dozen
or so un-free'd allocations, all made by the same dupstr() call in
special_handling() for "symset" and "roguesymset". (Reproducible with
a few tens of thousands of fuzzer moves, although you have to take
over from the fuzzer and make a clean exit rather than just interrupt
it or there'll be lots of other un-free'd memory.) I haven't actually
figured out how/why it was leaking, but reorganizing the code has made
the leak go away (according to a couple of even longer fuzzer runs) so
I'm settling for that.
There was a spurious seli-colon after an if's test, making a boundary
check be ineffective. When looking at that, I noticed that the 'O'
command's display of the current value for mouse_support ("0=off" and
so forth) was relying on implicit concatenation of adjacent string
literals, which would break K&R compilation. Do that concatenation
the old fashioned way....
While testing (after temporarily adding WC_MOUSE_SUPPORT to tty's
window_procs), I also noticed that wording used by config_error_add
looked strange when it was in response to giving a bad value via 'O'
command. Suppress its "config_error_add: " prefix is that situation.
New: call to panic() in impossible() used arbitrary string as a
format so was vulnerable to percent signs in that string. (This
potentially serious problem is not limited to USE_OLDARGS.)
Old: revised message string for impossible ("save/restore might fix
this" instead of "perhaps you'd better quit") passed wrong number of
arguments to pline() when using the clumsy VA_PASSx() mechanism (was
missing arg 0 for the fixed-arg format argument).
Old: varargs config_error_add() in files.c wouldn't compile for
USE_OLDARGS. Evidently no one has been impacted by that but this
fixes it anyway. (Two problems: prototype used FDECL() when it
should have been using VDECL(), and calls to config_error_add() in
the same file would need the VA_PASSx() stuff to force presence of
all optional args. I moved it instead of adding the latter.)
Remove trailing spaces, and remove tabs from the files that had
trailing spaces.
Also, rndorcname() was using a random value to terminate a loop
and was recalculating a new one each iteration.
When entering a new menucolor via options, show regex error
immediately afterwards, instead of asking for color and attribute
before showing the error.
Also actually show config errors even if config error handler
hasn't been initialized.
The function handles comments, empty lines, config sections,
CHOOSE, and line continuation, while calling another function
for other lines.
Currently used for config file, sysconf, and WIZKIT.
