tlikonen / wcheck-mode Goto Github PK

Is there any other way to show proposed corrections other than option selection. Can helm or company be used for that?

This is an attempt to use aspell instead of enchant, it doesn't work and no errors are reported, could docs show an example for aspell since this is a popular spell checker?

    (setq wcheck-language-data
      '
      (
        ("American English"
          (program . "/usr/bin/aspell")
          (args "-d" "en_US" "list")
          (action-program . "/usr/bin/aspell")
          (action-args "-a" "-d" "en_US" "list")
          (action-parser . wcheck-parser-ispell-suggestions)
          )
        )
      )

Buffer-local dictionaries are the killer feature of spell checkers such as flyspell-mode. It would be nice if someone could add to the wcheck-mode documentation how to set up a simple buffer-local dictionary.

Hi, I'm trying Wcheck and I really like it.

But there is a function that i miss from flyspell-correct -> flyspell-correct-wrapper

;; In most cases the last function is the most convenient, so don't forget to
;; bind it.
;;
;; (define-key flyspell-mode-map (kbd "C-;") 'flyspell-correct-wrapper)
;;
;; When invoked, it will show the list of corrections suggested by Flyspell.

With this function is very simple to correct misspelled word, just it C-; until the correct word is displayed.

So I'd like to ask if is it possible to have a similar function with wcheck, a sort of wrapper around wcheck-jump-backword and wcheck-actions .

EDIT:
After some test I ended with this function that does 90% of what I wanted, the only thing missing is the cycling through the candidates

;;;###autoload
(defun wcheck-actions-wrapper()
  (interactive)
  (save-excursion
    (wcheck-jump-backward)
    (wcheck-actions (point))))

Thank You for wcheck-mode! It works great, but in Ukrainian all words with apostrophe are marked as errors. enchant/hunspell from command line accept those words as correct.

wcheck marks all URLs in my text documents as mispellings. I would like to exclude URLs from being spell-checked.

Flyspell has flyspell-mode-predicate which directly serves this purpose.

I looked through the source and maybe regexp-start, regexp-body, regexp-end can serve this purpose, but it's not obvious how. If that's the mechanism you want to to use to support this use case, then documenting how to modify those variables to exclude URLs would be super helpful.

Hello,

I tried this package today. I tried the British English example from the README. It took me quite a while to find out that the example was wrong, at least, for my version of ispell. The "args" must include the "-a" option for ispell; else, nothing ever happens. Should we correct the doc?

While debugging, I got to know a lot of the code, btw - this seems to be a quite cool package!

I'm using the configuration snippets from README.org. When I activate wcheck-parser-ispell-suggestions on a marked word wcheck provides a list of possible corrections. If I abort the menu without choosing one of the suggestions, wcheck leaves a second window of one line height to the same buffer at the bottom of the frame, where the suggestions menu was.

This is only the case with emacs 24.0.50 as of 2011-09-07. This does not happen when using emacs 23.3.

Thanx, for providing wcheck, Gregor

Hello. With such config wcheck produces no errors at all, but if replace hunspell with enchant then works as expected:

    (setq-default wcheck-language      "English"
                  wcheck-language-data '(("English"
                                          (program . "hunspell")
                                          (args "-l" "-d" "en_US")
                                          (action-program . "hunspell")
                                          (action-args "-a" "-d" "en_US")
                                          (action-parser . wcheck-parser-ispell-suggestions)
                                          (read-or-skip-faces
                                           ((emacs-lisp-mode c-mode c++-mode python-mode)
                                            read font-lock-comment-face
                                            )
                                           (org-mode
                                            skip org-block-begin-line org-block-end-line org-meta-line org-link
                                            )
                                           (nil)
                                           )
                                          )

What i'm doing wrong?

The package guess-language guesses the language (using heuristics of trigrams) of the current paragraph (which provides for using different languages within one buffer). It could automatically change wchecks language configuration iff there was a hook in wcheck for the situation when an incorrect word is detected. This would be nice even if current buffers text is all in one language, because one wouldn't have to switch languages when e.g. writing several emails in different languages.

I therefore wish for such a hook (see it's issue #8 where this is discussed).

ie: if I say:

(text-mode read nil)

I also have to say:

(markdown-mode read nil)

Is this on purpose?

The enchant-add-to-dictionary function provided in the readme doesn't remove the mispelling marker from the word after it's been added to the user's personal dictionary, which I consider a bug and a common need when writing a function like this. The readme should document how to do this.

I skimmed the source but couldn't find a clear solution based on the function names and doc strings.

My workaround is to toggle wcheck-mode off and then on every time I add a word to my personal dictionary.

at least for the case of appending "word" variable to the private dictionary file.

This would be nice for users like me who don't know how to write lisp.

Thanx, gregor