Such errors are raised when SyntaxError nodes are executed, and are also
used for internal control flow.
The 'break' and 'continue' commands are currently only allowed inside
for loops, and outside function bodies.
This also adds a 'loop' keyword for infinite loops.
This patchset allows a match expression to have a list of names for its
glob parts, which are assigned to the matched values in the body of the
match.
For example,
```sh
stuff=foobarblahblah/target_{1..30}
for $stuff {
match $it {
*/* as (dir sub) {
echo "doing things with $sub in $dir"
make -C $dir $sub # or whatever...
}
}
}
```
With this, match expressions are now significantly more powerful!
This adds support for (basic) brace expansions with the following
syntaxes:
- `{expr?,expr?,expr?,...}` which is directly equivalent to `(expr expr
expr ...)`, with the missing expressions replaced with an empty string
literal.
- `{expr..expr}` which is a new range expansion, with two modes:
- if both expressions are one unicode code point long, the range is
equivalent to the two code points and all code points between the
two (numerically).
- if both expressions are numeric, the range is equivalent to both
numbers, and all numbers between the two.
- otherwise, it is equivalent to `(expr expr)`.
Closes#3832.
This commit adds an equivalent to the sh 'case' construct, except it's
much more pleasing to look at and write:
```sh
match "$something" {
p1 { echo "p1!" }
p2 { echo "p2!" }
* { echo "string catch-all!" }
}
```
is the equivalent of:
```sh
case $something in
p1)
echo "p1!"
;;
p2)
echo "p2!"
;;
*)
echo "catch-all!"
;;
esac
```
Since our shell does not treat lists as strings, matching lists is also
possible:
```sh
match (1foo 2foo foo3) {
(?foo 2* *) { echo wowzers! }
(* * *) { echo 3-element list catch-all }
}
```
This patchset makes the shell capable of lazily resolving and executing
sequences of commands, to allow for putting logical sequences in the
background.
In particular, it enables And/Or/Sequence nodes to be run in the background,
and consequently unmarks them as `would_execute`.
Doing so also fixes job control to an extent, as jobs are now capable of
having 'tails', so sequences can be put in the background while
preserving their following sequences.
A ListValue never stores null Values, so it makes sense to restrict it.
This also propagates use of NonnullRefPtr to the create() helpers.
There's a small bit of awkwardness around the use of initializer_list,
since we cannot directly construct a NonnullRefPtrVector from one.
Closes#2760.
This commit adds a 'for' loop, and tweaks the syntax slightly to make &&
bind more tightly than || (allowing for `expr && if_ok || if_bad`) :^)
This allows the parser to finally parse the entire source into a single
AST.
As a result of allowing comments inside sequences, Sequence is also
marked as would_execute if its left or right node would.
Take one small step towards #2357.
Handle completing barewords starting with '-' by piping the requests to
the Shell::complete_option(program_name, option) :^)
Also implements completion for a single builtin (setopt) until we figure out how
to handle #2357.
This builtin sets (and unsets) boolean flags that alter the behaviour of
the shell.
The only flags added are
- inline_exec_keep_empty_segments: Keep empty segments in the result of
splitting $(...) by $IFS
- verbose: Announce each command before executing it
It should be noted that the (rather extreme) verbosity of the names is
intentional, and will hopefully be alleviated by the next commit :^)
This commit also removes the ExecutionInputType and directly uses
RefPtr<Shell> instead, since nothing else is needed for execution
purposes, and also makes the shell refuse to evaluate commands with
any sort of syntax error.
