A function defined into an other function gives an error

Here's a real-life example from meta-openembedded: https://github.com/openembedded/meta-openembedded/blob/4dbbef7a39ad18206ca6cebf7a1e08aebe5b5a65/meta-oe/recipes-support/opencv/opencv_4.8.0.bb#L46

Here's a minimal example:

func() {
    nestedfunc() {
        echo "test"
    }
}

output:

(recipe [0, 0] - [5, 0]
  (ERROR [0, 0] - [4, 1]
    (function_definition [0, 0] - [3, 5]
      (identifier [0, 0] - [0, 4])
      (ERROR [0, 7] - [1, 16]
        (ERROR [1, 4] - [1, 14]))
      (shell_content [2, 8] - [2, 19]))))

Thank you

I noticed the line continuation with backslashes generates errors in some cases. Unfortunately, I failed to understand exactly when it happens. Most of the time, it seems not to be an issue.

Here two real life examples taken from Poky:
https://git.yoctoproject.org/poky/tree/meta/classes/recipe_sanity.bbclass#n45
https://git.yoctoproject.org/poky/tree/meta/classes/copyleft_filter.bbclass#n76

Here an other example I made myself:

def example():
    print('test', \
        True
    )

Curiously, in this example, the parsing works properly if True is replaced by a string.

Thank you

An unused identifier is invalid code and is expected to give errors. However, when some valid code come after it, the errors get weird and it make it difficult to provide something useful to the user.

This one is fine:

BAR='bar'
FOO

(recipe [0, 0] - [1, 3]
  (variable_assignment [0, 0] - [0, 9]
    (identifier [0, 0] - [0, 3])
    (literal [0, 4] - [0, 9]
      (string [0, 4] - [0, 9]
        (string_content [0, 5] - [0, 8]))))
  (ERROR [1, 0] - [1, 3]
    (identifier [1, 0] - [1, 3])))

These are not:
1.

FOO
BAR='bar'

(recipe [0, 0] - [1, 9]
  (variable_assignment [0, 0] - [1, 9]
    (identifier [0, 0] - [0, 3])
    (ERROR [1, 0] - [1, 3]
      (python_identifier [1, 0] - [1, 3]))
    (literal [1, 4] - [1, 9]
      (string [1, 4] - [1, 9]
        (string_content [1, 5] - [1, 8])))))

FOO
BAR() {
    
}

(recipe [0, 0] - [3, 1]
  (function_definition [0, 0] - [3, 1]
    (identifier [0, 0] - [0, 3])
    (ERROR [1, 0] - [1, 3]
      (python_identifier [1, 0] - [1, 3]))))

FOO
python bar() {

}

(recipe [0, 0] - [3, 1]
  (function_definition [0, 0] - [3, 1]
    (identifier [0, 0] - [0, 3])
    (ERROR [1, 0] - [1, 10]
      (python_identifier [1, 0] - [1, 6])
      (ERROR [1, 7] - [1, 10]))))

In each cases the FOO identifier is wrongly put inside a function_definition. A python_identifier also comes out of nowhere in 1. and 2.

I was wondering if it'd be possible to support the 3.x syntax. For example, _append instead of :append.

A bracket inside a string inside a function gives an error

d() {
  t = '{'
}

A bracket inside a comment gives an error

do() {
  # {
}

It does work inside python functions, though

Thank you

Hello!

I noticed inline_python won't be detected when inside a shell function

Here's a somewhat minimal example:

foo() {
  echo ${@"bar"}
}

Here a real-life example from poky: https://git.yoctoproject.org/poky/tree/meta/recipes-support/icu/icu_73-2.bb#n126

Thank you, we love this library

Version: 1.0.1
Currently, the ${} with an identifier will be treated as variable expansion:
variable_expansion: $ => seq('${', $.identifier, '}'),
when there is an empty ${}, the generated tree node is unrelated to variable expansion.

Can we make the identifier optional such that ${} alone can be treated as variable expansion?
The language extension will be able to provide the desired completion even with empty ${}

In the line inherit update-rc.d, everything after (including) period is parsed as error. It should probably handle all characters valid in a unix filename?
It stopped working after I updated treesitter, but I think, it previously used only nvim's regex highlighting.

Version:
1.0.1

Current behaviour:
The same overrides syntax in tree-sitter is treated differently as shown in these screenshots. When it is not in the last line in the file, it is treated as the override (even though it is still picking up the content following). When it is in the last line, it is treated as an identifier and nothing related to override anymore.

Expectation:
When it encounters MYVAR:a: in the last line of the file, it still treats it as overrides or partially. This is an override statement yet to be completed and it will benefit the completion in language extension.

Good to have:
When it encounters MYVAR:a: not in the last line of the file, the current one is enough to identify an override syntax to perform further actions. But it will be nice if it won't pick up the content following as part of the override syntax.

It seems fakeroot is not supported on python functions and gives an error:

fakeroot python() {
  pass
}

Here's a real life example: https://git.yoctoproject.org/poky/tree/meta/recipes-core/meta/uninative-tarball.bb#n45

However, it does work on bash functions

Thank you