One of the tests was using `>/dev/null` to suppress the `type` output.
That needs to be `^/dev/null` now, but instead just go ahead and use the
new `-q` flag.
Use `functions -q` instead of searching the `functiosn -na` list for the
provided word. This may result in an automatically-loaded function being
sourced, but that happens anyway with the default output.
This change means the results of `test -q foo` can be relied upon to
indicate whether `foo` can actually be invoked. Previosly, if `foo` was
the name of an automatically-loaded function file but did not actually
define a function `foo`, and there was no execuable `foo`, then `type -q
foo` would lie and say `foo` can be invoked when it can't.
The --quiet flag is useful when only the exit status matters.
Fix the documentation for the -t flag to no longer claim that `type` can
print "keyword", as it never does that.
Stop printing a blank line for functions/builtins when the -p flag has
been passed. It's just not useful.
Track whether -a and -f have been supplied separately. That way both
`type -a -f command` and `type -f -a command` behaves correctly, as does
`type -a -f foo` where there are multiple executables named `foo` in the
$PATH.
Stop using getopt to parse flags. It's far more expensive than
necessary, and results in long flags not being parsed on OS X. This also
allows args starting with - after the options list to be properly
interpreted as a value to test.
Print the error message to stderr as is appropriate.
Use the new `command -p` functionality when the -a flag has not been
provided (`command` does not have any equivalent to the -a flag),
instead of using `which`. This is faster and also avoids any possible
disagreement between `which` and what fish thinks is valid.
Stop testing every path to see if it's executable, that test has already
been done by `which` or `command -p`.
The end result is `type -P ls` is roughly 250% faster, according to
profiling, on my OS X machine.
Instead of introducing a new local scope at the point of `set`, merely
push a new local scope at the end of env_init(). This means we have a
single toplevel local scope across the lifetime of the fish process,
which means that
set -l foo bar
echo $foo
behaves as expected, without modifying the global environment.
The mode restricts the scope in which the variable is searched for.
Use this new restricted scope functionality in the `set` builtin. This
fixes `set -g` to not show local shadowing variable values, and also
allows for scoped erasing of slices.
When attempting to set a readonly or electric variable in the local or
universal scopes, print an appropriate error. Similarly, print an error
when setting an electric variable as exported. In most cases this is
simply a nicer error instead of the 'read-only' one, but for the 'umask'
variable it prevents `set -l umask 0023` from silently changing the
global value.
They're dynamically calculated, so they qualify. This also removes them
from the list of exported global variables, because they're actually not
exported.
When using the `set` command with the -l flag, if we're at the top
level, create a temporary local scope. This makes query/assignment
behavior be consistent with the value-printing behavior.
This works by marking the current block as needing to pop the
environment if a local scope was pushed. I assume this is safe to do. I
also assume the current block is the right one to modify, rather than
trying to walk up the stack to the root.
env_exists() wasn't properly handling multiple scopes in some cases,
notably with readonly/electric variables. Rewrite it to operate in a
more straightforward fashion.
When initializing fish, ignore any inherited environment variables that
match any of the readonly or electric variable names.
This prevents really weird behavior when e.g. fish is launched with
COLUMNS already set to something. In that case, testing $COLUMNS within
fish behaves normally, but any subprocesses get the value that fish
itself had inherited.
