Missing context.scm and folds.scm

[pidgeon777]

I think the following files:

• context.scm
• folds.scm

are missing from here:

https://github.com/jpt13653903/tree-sitter-vhdl/tree/master/queries/Neovim

I have some obsolete files which I paste here:

context.scm:

(entity_declaration) @context
(generic_clause) @context
(port_clause) @context
(architecture_body) @context
(function_body) @context
(full_type_declaration) @context
(if_statement) @context
(if) @context
(else) @context
(elsif) @context
(process_statement) @context
(case_statement) @context
(case_statement_alternative) @context
(entity_instantiation) @context
(port_map_aspect) @context
(generic_map_aspect) @context
(if_generate_statement) @context
(for_generate_statement) @context

folds.scm:

[
    (entity_declaration)
    (entity_header)
    (architecture_body)
    (generic_clause)
    (port_clause)
    (if_statement)
    (case_statement)
    (concurrent_statement_part)
    (declarative_part)
    (sequence_of_statements)
    (process_statement)
    (record_type_definition)
    (component_declaration)
    (component_instantiation_statement)
    (port_map_aspect)
    (generic_map_aspect)
] @fold

[pidgeon777]

Also I see there is a reference to a file called injections.scm here:

e5fecf9#diff-3bd6f6646d0ce51e37b9c71eb325ce2036b821b5061c8983257e0fdc5413646e

[jpt13653903]

Your reference is the develop branch. I'll merge to master eventually -- give it 2 or 3 weeks, we're still in alpha-testing.

My suggestion would be to run the develop branch in your config, use it as your daily driver and report bugs as you find them.

[jpt13653903]

@superzanti is on it: ref: #2

[jpt13653903]

By the way -- the "obsolete" files you posted are indeed obsolete. They're from the old fork. This repo has a new parser, with new syntax tree node names and structure, so we're building new fold, context and indent queries.

[pidgeon777]

Thank you for the prompt reply.

I just created this new version of context.scm, which seems to work well (tested with nvim-treesitter-context).

context.scm:

(entity_declaration) @context
(generic_clause) @context
(port_clause) @context
; (architecture_body) @context
; Replaces the above:
(architecture_definition) @context
; Added:
(BEGIN) @context
; Added:
(concurrent_block) @context
; (function_body) @context
; Replaces the above:
(subprogram_definition) @context
; Added:
(function_specification) @context
; Added:
(subprogram_head) @context
; Added:
(subprogram_declaration) @context
; Added:
(sequential_block) @context
; Added:
(loop_statement) @context
; (full_type_declaration) @context
; Replaces the above:
(type_declaration) @context
(if_statement) @context
; (if) @context
; Replaces the above:
(IF) @context
; Added:
(else_statements) @context
; (else) @context
; Replaces the above:
(ELSE) @context
; Added:
(elsif_statements) @context
; (elsif) @context
; Replaces the above:
(ELSIF) @context
(process_statement) @context
(case_statement) @context
(case_statement_alternative) @context
; Added:
(when_element) @context
; (entity_instantiation) @context
; Replaces the above:
(component_instantiation_statement) @context
(port_map_aspect) @context
(generic_map_aspect) @context
(if_generate_statement) @context
(for_generate_statement) @context

Replaces the above means that I replace the previous rule (no longer valid) with the new one, based on your new tree-sitter grammar.

Please note that the schema could be further improved, but for now, it meets my needs.

[superzanti]

Update - this just seems to be an error with changes in a single buffer. Closing out of the buffer then re-opening it when no changes have occurred seems to fix this issue.

@jpt13653903 This could be worth testing on your end to see if it's a bug with how trees are parsed when the buffer is modified.

[jpt13653903]

Interesting...
Please could you commit both options and push to your fork? I'll test and see if I can figure it out, and hopefully fix it.

[jpt13653903]

And if you could open a bug report and tell me how to reproduce the problem, even better...

[pidgeon777]

You can make it a "little" bit better by adding context.end captures, for example:

[
  (entity_declaration)
  (generic_clause)
  (port_clause)
  (architecture_definition)
  (concurrent_block)
  (subprogram_definition)
  (function_specification)
  (subprogram_head)
  (subprogram_declaration)
  (sequential_block)
  (type_declaration)
  (if_statement
    (then_statements) @context.end)
  (elsif_statements
    (then_statements) @context.end)
  (else_statements)
  (loop_statement
    (loop_body) @context.end)
  (process_statement)
  (case_statement)
  (case_statement_alternative)
  (when_element)
  (component_instantiation_statement)
  (port_map_aspect)
  (generic_map_aspect)
  (if_generate_statement)
  (for_generate_statement)
] @context

Is this based on the context.scm I posted above? In any cases, which should be the differences compared to my version, if any?

[jpt13653903]

It's the same as yours -- I just added the context.end captures

[superzanti]

If anyone is interested in the old SCM files I have them here:

context.scm
folds.scm
highlights.scm
textobjects.scm

I should be finished with the updates for the new parser tomorrow.

[jpt13653903]

That textobjects one also sounds useful. Do you use nvim-treesitter-textobjects?

[pidgeon777]

I'm also curious about it.

[superzanti]

I do use the nvim-treesitter-text objects. I'll be updating it too.
I'll try to get it done today.

[jpt13653903]

Great - thank you.
Then we can throw that into the alpha-test pool as well.
If I can update to v1.0 and issue the various PRs for nvim-treesitter in 2 weeks or so that'll be awesome.

[jpt13653903]

@superzanti

I see that you haven't forked the repo yet... Is it possible to share a link to your work-in-progress? Maybe we can help out - you don't have to go at it alone...

Alternatively, you can open up a draft PR so long, then we can have a central place where we can iterate towards a solution.

[superzanti]

Yeah, with the latest updates these all need to be updated:

textobjects.scm.txt
context.scm.txt
folds.scm.txt

But if people are itching to get started - here you go.

[superzanti]

Okay, I just updated it to the latest commit. It's got a bit to go, but expect a pull request on Monday or Tuesday.

context.scm.txt
folds.scm.txt
textobjects.scm.txt

[jpt13653903]

Great -- thank you. I'll wait for the PR, because I want to give you the credit and don't like committing impersonations (even though I do have all your details from this commit.

By the way -- where are you committing backups of the work, seeing as you don't have a fork of the repo yet...?

[superzanti]

Thanks I appreciate it.

You've done a lot of hard work on this.

Since this has been a constantly updating project I've just been updating those 3 files in my query directory - as that's the quickest to edit and debug. I haven't forked the repo yet, but I will on monday/tuesday.

[jpt13653903]

See PR #18