Call ripgrep-all in emacs with consult

consult-ripgrep-all.el

;; New versions of consult make this dramatically simpler than it used to be. Just one function that you can bind to a key:

;;;###autoload
(defun consult-ripgrep-all (&optional dir initial)
  (interactive "P")
  (let ((consult-ripgrep-args "rga --null --line-buffered --color=never --max-columns=1000  --smart-case --no-heading --with-filename --line-number"))
    (consult-ripgrep dir initial)))

That would definitely be a usability improvement! I'm open to suggestions.

I don't have much knowledge of elisp, consult or embark. So I don't have good suggestions. :)

same, same.

ty for posting! probably saved me an hour or more 👍

Thanks for sharing this! This still works.

@jthaman @fountainer

For the PDF page upon selection, you can try this (requires pdf-tools):

(defun pdf-consult-ripgrep-all (&optional dir initial)
  (interactive "P")
  (let* ((consult-ripgrep-args "rga --null --line-buffered --color=never --max-columns=1000 --smart-case --no-heading --with-filename --line-number")
	 (result (consult-ripgrep dir initial)))
    (when (and result (string-match "\\([^:]+\\.pdf\\).*page \\([0-9]+\\)" result))
      (pdf-view-goto-page (string-to-number (match-string 2 result))))))

There's still room for improvement. It does not take you to the page during the preview. I'll leave that for later.

@jthaman Thank you for this.

I noticed that ripgrep by default considers .gitignore. As binary files like pdf are usually part of .gitignore it was not searched by ripgrep-all.

If someone is facing the same issue, then include --no-ignore-vcs flag as one of the args in consult-ripgrep-args to resolve this. But note that there will be performance penalty in big repos.