johnmastro

;; Turns out it's not necessary to define a new ibuffer column type. Both
;; `list-buffers' and `ibuffer' respect `list-buffers-directory', so just do
;; something like this:

(add-hook 'magit-mode-hook
          (lambda ()
            (let ((dir (abbreviate-file-name default-directory)))
              (setq list-buffers-directory dir))))

johnmastro

(defun maybe-cygwinize-drive-letter (file)
  "Return FILE in a Cygwin-friendly format.
For example, given \"c:/foo/bar\" return \"/foo/bar\", or given
\"e:/foo/bar\" return \"/e/foo/bar\"."
  (cond ((string-match "\\`[Cc]:/" file)
         (replace-match "/" t t file))
        ((string-match "\\`\\([A-Za-z]\\):/" file)
         (replace-match (concat "/" (match-string 1 file) "/")
                        t t file))
        (t file)))

johnmastro

;; Just doodling on the topic of somewhat automatically initializing
;; `mc/cmds-to-run-for-all'.
;;
;; I always find the prompts a bit distracting when working on a new computer
;; but the file `multiple-cursors' maintains, .mc-lists.el, doesn't seem very
;; well-suited to version control.

(require 'pcase)

(defvar mc-paredit&sp-cmds-to-run-for-all-regexp

johnmastro

(require 'pcase)

;; Based on Xah Lee's: http://ergoemacs.org/emacs/modernization_upcase-word.html

(defun cycle-case (beg end)
  "Cycle the word case of the active region or word at point."
  (interactive
   (pcase-let ((`(,beg . ,end) (if (use-region-p)
                                   (cons (region-beginning) (region-end))
                                 (bounds-of-thing-at-point 'word))))

johnmastro

;; An attempt at this Emacs SX question:
;; https://emacs.stackexchange.com/questions/10359/delete-portion-of-isearch-string-that-does-not-match-or-last-char-if-complete-m

(defun isearch-delete-something ()
  "Delete non-matching text or the last character."
  ;; Mostly copied from `isearch-del-char' and Drew's answer on the page above
  (interactive)
  (if (= 0 (length isearch-string))
      (ding)
    (setq isearch-string

johnmastro

" This is an updated, more powerful, version of the function discussed here:
" http://www.reddit.com/r/vim/comments/1rzvsm/do_any_of_you_redirect_results_of_i_to_the/
" that shows ]I, [I, ]D, [D, :ilist and :dlist results in the quickfix window, even spanning multiple files.

function! List(command, selection, start_at_cursor, ...)
    " derive the commands used below from the first argument
    let excmd   = a:command . "list"
    let normcmd = toupper(a:command)

    if a:selection

johnmastro

;;; iterm.el - Send text to a running iTerm instance

(require 'pcase)
(require 'thingatpt)

;; To match SublimeText's key binding:
;; (global-set-key (kbd "<C-return>") 'iterm-send-text)

(defvar iterm-default-thing 'line
  "The \"thing\" to send if no region is active.

johnmastro

(require 'slime)
(require 'dash)
(require 'trident-mode)

(defvar trident-target-process nil
  "The process that expansions will be sent to.")

(defun trident-list-process-buffers ()
  "Return a list of all buffers with an associated process."
  (-filter #'get-buffer-process (buffer-list)))

johnmastro

(defn filter-by-val-1
  [pred m]
  (into {} (filter (fn [[k v]] (pred v))
                   m)))

;; This one is a little faster
(defn filter-by-val-2
  [pred m]
  (persistent!
   (reduce-kv (fn [acc k v]

johnmastro

;; -*- lexical-binding: t -*-

;; Some code I've been playing around with to automate the process of of
;; creating a new Octopress page or post, with appropriate metadata, from a
;; buffer's content.

;; The `octopress-rake`, `octopress-extract-filename`, and `octopress-new`
;; functions are derived from the octopress-emacs code here:
;; github.com/gfreezy/octopress-emacs