On large files, performance of ansi-color-apply-on-region
is slow. Here's a solution that colors the current region and works with read-only buffers.
(require 'ansi-color)
(defun ansi-color-region ()
"Color the ANSI escape sequences in the acitve region.
Sequences start with an escape \033 (typically shown as \"^[\")
and end with \"m\", e.g. this is two sequences
^[[46;1mTEXT^[[0m
where the first sequence says to diplay TEXT as bold with
a cyan background and the second sequence turns it off.
This strips the ANSI escape sequences and if the buffer is saved,
the sequences will be lost."
(interactive)
(if (not (region-active-p))
(message "ansi-color-region: region is not active"))
(if buffer-read-only
;; read-only buffers may be pointing a read-only file system, so don't mark the buffer as
;; modified. If the buffer where to become modified, a warning will be generated when emacs
;; tries to autosave.
(let ((inhibit-read-only t)
(modified (buffer-modified-p)))
(ansi-color-apply-on-region (region-beginning) (region-end))
(set-buffer-modified-p modified))
(ansi-color-apply-on-region (region-beginning) (region-end))))
The one downside is that ansi-color-apply-on-region
removes the ANSI escape sequence characters from the buffer, so when you save, they are lost. I wonder if there's a way to hide the characters instead of stripping them?