| Age | Commit message (Collapse) | Author |
|---|
|
now build libraries with propolice enabled. Without this, existing
binaries (such as ports/packages) that link with any system library
other than libc will fail with an undefined symbol of "___guard"
(__guard on ELF).
Pointed out by markus@ and discussed with deraadt@
|
|
look for ncurses by checking for ncurses.h or libncurses instead of
checking for NCURSES_VERSION in curses.h. Closes PR 2942.
|
|
ok art@, miod@
|
|
|
|
|
|
|
|
|
|
hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.
|
|
|
|
|
|
|
|
|
|
|
|
is not set. Propagate "silent" flag in _nc_parse_entry() to _nc_get_token().
Set the silent flag when reading termcap files, including $TERMCAP.
|
|
o remove macro callPutChar() from tty_update.c, since this is no longer
needed (reported by Philippe Blain).
o add a null-pointer check in tic.c to handle the case when the input
file is really empty. Modify the next_char() function in comp_scan.c
to allow arbitrarily long lines, and incidentally supply a newline to
files that do not end in a newline. These changes improve tic's
recovery from attempts to read binary files, e.g., its output from
the terminfo database (reported by Bernhard Rosenkraenzer).
o revert change to c++/demo.cc from 20001209, which changed definition
of main() apparently to accommodate cygwin linker, but broke the demo
program.
|
|
as an alias" warning since it hits people with $TERMCAP set when
there is no terminfo file present. A better fix would be to make
_nc_get_token() take a quiet flag.
|
|
|
|
|
|
|
|
|
|
|
|
speed_t (u_int) to short to match historic termcap usage.
|
|
|
|
|
|
if root but not setugid.
Fix select usage to deal with an arbitrary number of fd's. This code
is not compiled since we use poll(2).
|
|
|
|
|
|
a bit to make it clear what is supposed to happen. Also fix some policy
wrt setugid in code we don't actually compile for consistency's sake.
|
|
behavior of checking ~/.terminfo.db and ~/.termcap.db
|
|
|
|
|
|
|
|
|
|
|
|
|
|
getcap(3) routines.
|
|
o change unctrl() to render C1 characters (128-159) as ~@, ~A, etc.
o trace() function is provided only if TRACE is defined, e.g., in the debug
library. Modify related calls to _tracechar() to use unctrl() instead.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
+ additional fixes for non-bce terminals (handling of delete_character)
to work when assume_default_colors() is not specified.
+ modify warning message from _nc_parse_entry() regarding extended
capability names to print only if tic/infocmp/toe have the -v flag
set, and not at all in ordinary user applications. Otherwise, this
warning would be shown for screen's extended capabilities in programs
that use the termcap interface.
+ modify use of _nc_tracing from programs such as tic so their debug
level is not in the same range as values set by trace() function.
+ small panel header cleanup (patch by Juergen Pfeifer).
|
|
+ add prototype for erase() to curses.h
+ add TRACE_DATABASE flag for trace().
+ formatting changes (style)
|
|
those explicitly listed in term.h. With extended names, non-standard
terminfo/termcap capabilities are passed through for those programs that
want to use them (screen for instance). This changes struct term and
therefor requires a library major increment. It also requires changing
how the terminfo.db file is parsed. Previously, it was parsed like:
foreach cap (all_possible_caps) if ( is_listed(cap) ) add_cap(cap)
However, now that we can't know the set of all possible capabilities
we need to convert the buffer we get back from cgetent into one that
looks more like a terminfo entry and pass it through to the ncurses
terminfo parsing routines. This also enables tic's -x flag.
|
|
+ make conversion to termcap ug (underline glitch) more consistently applied.
+ modify read_termtype(), fixing 'toe', which could dump core when it
found an incomplete entry such as "dumb" because it did not
initialize its buffer for _nc_read_file_entry().
+ use explicit VALID_NUMERIC() checks in a few places that had been
overlooked, and add a check to ensure that init_tabs is nonzero,
to avoid divide-by-zero (reported by Todd C Miller).
|
|
to be 'not present' since that has already been done for us in _nc_init_entry().
This fixes some core dumps caused by a divide by 0 bug deep within ncurses
that resulted from using the old way to specify that a capability was
'not present'.
|
|
|
|
|
