lsp-grammarly
lsp-grammarly copied to clipboard
Newlines affect suggestions
Hello and thank you for this package!
I have experienced the following issue. When splitting a sentence across multiple lines (I have been using auto-fill in my buffers) the suggestions are wrong.
MWE:
The successive construction of the model from finer to coarser blocks not only improves the efficiency but also increases the quality.
The successive construction of the model from finer to coarser
blocks not only improves the efficiency but also increases the
quality.
The first example raises no issues.
In the second case, I get errors on improves
and increases
because it identifies the blocks
as the subject.
Is there any way to remove single-line breaks from the parser?
Thanks!
Thanks for the bug report! ❤️
Hmm... I am not able to reproduce this error. I have tested on both Emacs (lsp-grammarly) and VSCode, but neither of them can reproduce this error. Can you place the recipe to reproduce this? Thanks!
I am testing on Emacs (lsp-grammarly) on macOS
; I am using doom-emacs
as my emacs setup.
I am adding the output of doom info
generated Aug 07, 2022 18:02:22
system MacOS 11.6.7 Darwin 20.6.0 x86_64
emacs 28.1 ~/.emacs.d/ -> ~/.emacs.d/
doom 3.0.0-dev HEAD -> master, origin/master, origin/HEAD 33c5f3721
2022-07-09 21:10:40 +0200 ~/.doom.d/ -> ~/.doom.d/
shell /bin/zsh
features ACL GMP GNUTLS JSON LIBXML2 MODULES NOTIFY KQUEUE PDUMPER THREADS
TOOLKIT_SCROLL_BARS XIM ZLIB
traits batch server-running envvar-file
modules :completion company (vertico +icons) :ui doom doom-dashboard hl-todo
minimap modeline ophints (popup +defaults) treemacs vc-gutter
vi-tilde-fringe workspaces :editor file-templates snippets :emacs
(dired +icons) electric vc :term vterm :checkers syntax (spell
+flyspell) :tools (eval +overlay) lookup lsp magit :os macos :lang
(cc +lsp) emacs-lisp julia latex markdown org sh web yaml :config
(default +bindings +smartparens)
packages (lsp-grammarly) (julia-snail) (julia-repl :disable t) (matlab-mode)
I am adding a screenshot of my Emacs buffer.
Thanks for the screenshot and information! These really help a lot! ❤️
Can you also try it in VSCode? If it produces the same result, I think the culprit should be the upstream (language server).
I tried VSCode, it produces the same result.
Any ideas where to forward this issue? Thanks!
It seems like an upstream issue, you can post an issue in https://github.com/znck/grammarly (this is their link).
Before you do, make sure you are using the latest lsp-grammarly
and language server. Thanks! ❤️
This seems like an upstream issue, so I am closing this now! Thanks for reporting this issue! :)