Changes typeset such that ${newparam-notset} yields "notset" and
"typeset -p newparam" does not show an assignment to the parameter. This
is similar to the default behavior of bash and ksh, with minor differences
in typeset output.
Also add tests for some POSIX incompatibilities plus minor changes for test
harness robustness.
zsh typically runs the final command in a pipeline in the main shell
instead of a subshell. However, POSIX specifies that all commands in a
pipeline run in a subshell, but permits zsh's behavior as an extension.
The default /bin/sh implementations on various Linux distros and the
BSDs always use a subshell for all components of a pipeline.
Since zsh may be used as /bin/sh in some cases (such as macOS Catalina),
it makes sense to have the common sh behavior when emulating sh, so do
that by checking for being the final item of a multi-item pipeline and
creating a subshell in that case.
From the comment above execpline(), we know the following:
last1 is a flag that this command is the last command in a shell that
is about to exit, so we can exec instead of forking. It gets passed
all the way down to execcmd() which actually makes the decision. A 0
is always passed if the command is not the last in the pipeline. […]
If last1 is zero but the command is at the end of a pipeline, we pass
2 down to execcmd().
So there are three cases to consider in this code:
• last1 is 0, which means we are not at the end of a pipeline, in which
case we should not change behavior.
• last1 is 1, which means we are effectively running in a subshell,
because nothing that happens due to the exec is going to affect the
actual shell, since it will have been replaced. So there is nothing
to do here.
• last1 is 2, which means our command is at the end of the pipeline, so
in sh mode we should create a subshell by forking.
input is nonzero if the input to this process is a pipe that we've
opened. At the end of a multi-stage pipeline, it will necessarily be
nonzero.
Note that several of the tests may appear bizarre, since most developers
do not place useless variable assignments directly at the end of a
pipeline. However, as the function tests demonstrate, there are cases
where assignments may occur when a shell function is used at the end of
a command. The remaining assignment tests simply test additional cases,
such as the use of local, that would otherwise be untested.
The inconsistency caused test failures where TERM is e.g. rxvt-unicode.
This also makes a couple of bits available in zattr by removing flags
indicating whether to use termcap which is not an attribute as such.
This change modifies the zsh binary safety check surrounding execve() so
it can run shell scripts having concatenated binary content. We're using
the same safety check as FreeBSD /bin/sh [1]. POSIX was recently revised
to require this behavior:
"The input file may be of any type, but the initial portion of the
file intended to be parsed according to the shell grammar (XREF to
XSH 2.10.2 Shell Grammar Rules) shall consist of characters and
shall not contain the NUL character. The shell shall not enforce
any line length limits."
"Earlier versions of this standard required that input files to the
shell be text files except that line lengths were unlimited.
However, that was overly restrictive in relation to the fact that
shells can parse a script without a trailing newline, and in
relation to a common practice of concatenating a shell script
ending with an 'exit' or 'exec $command' with a binary data payload
to form a single-file self-extracting archive." [2] [3]
One example use case of such scripts, is the Cosmopolitan C Library [4]
which configuse the GNU Linker to output a polyglot shell+binary format
that runs on Linux / Mac / Windows / FreeBSD / OpenBSD.
[1] 9a1cd36331
[2] http://austingroupbugs.net/view.php?id=1250
[3] http://austingroupbugs.net/view.php?id=1226#c4394
[4] https://justine.lol/cosmopolitan/index.html
If ~/foo does not exist and `zf_mkdir -p zf_mkdir -p` is executed
concurrently in multiple shells, it was possible prior to this patch
for the command to fail with EEXIST.
Fixed a regression introduced by workers/30496 (5.0.3) whereby %2K would
no longer be the equivalent of %K{2} (%K{green}) in prompt expansion.
That was one missing case where the is_fg flag was not passed along to
match_colour() after code factorisation.
Add tests for the different syntax variants, using echoti
as a reference.
The --enable-wgetch-events codepath is experimental (according to
ncurses-6.2/INSTALL) and off by default (according to
ncurses-6.2/configure.in). With that codepath disabled, the macro
KEY_EVENT is not provided, which (before this commit) manifested as a
build-time error:
[ 245s] gcc -c -I. -I../../Src -I../../Src -I../../Src/Zle -I. -DHAVE_CONFIG_H -DMODULE -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector-strong -funwind-tables -fasynchronous-unwind-tables -fstack-clash-protection -Werror=return-type -flto=auto -g -fPIE -D_DEFAULT_SOURCE -D_XOPEN_SOURCE=600 -I/usr/include/ncursesw -fPIC -o curses..o curses.c
[ 246s] In file included from curses.c:210:
[ 246s] curses_keys.h:93:15: error: 'KEY_EVENT' undeclared here (not in a function); did you mean 'KEY_RESET'?
[ 246s] 93 | {"EVENT", KEY_EVENT},
[ 246s] | ^~~~~~~~~
[ 246s] | KEY_RESET
curses_keys.h is only used for setting the "kevent" output parameter of
'zcurses input' (and the associated $zcurses_keycodes special variable),
so there's no harm in just leaving KEY_EVENT out of it. (That codepath
deals gracefully with numeric values that don't correspond to any of the
known compile-time values, as that can happen whenever the build- and
run-time versions of ncurses don't provide the same set of KEY_* macros,
with or without relation to that configure flag.)
Reported by Martin Liska.
