JSONやその他ログ等の日本語マルチバイト文字デコード術

Emacsで扱うマルチバイト文字のデコードについて書きます
Emacs Advent Calendar 2016

エンコードされた日本語等を含むテキストをEmacsでデコードします。

バッファ内Unicodeの置換

Decode (\u3044)

;; 選択範囲(region) + M-x my-decode-backslash-u-string-in-region-and-replace
;; \U3044\U30fc\U307e\U3063\U304f\U3059 -> いーまっくす

(defun my-decode-backslash-u-string-in-region-and-replace ($beg $end)
  (interactive "r")
  (let (($bufstr (buffer-substring-no-properties $beg $end))
        $replaced)
    (with-temp-buffer
      (insert $bufstr)
      (goto-char (point-min))
      (while (re-search-forward "\\\\[uU]\\(.\\{4\\}\\)" nil t)
        (let (($hex (match-string 1)))
          (delete-region (match-beginning 0) (match-end 0))
          (insert (format "%c" (string-to-number $hex 16)))))
      (setq $replaced (buffer-string)))
    (delete-region $beg $end)
    (insert $replaced)))

Decode (&#x3044)

;; 選択範囲(region) + my-decode-ampersand-sharp-string-in-region-and-replace
;; いーまっくす -> いーまっくす

(defun my-decode-ampersand-sharp-string-in-region-and-replace ($beg $end)
  (interactive "r")
  (let (($bufstr (buffer-substring-no-properties $beg $end))
        $replaced)
    (with-temp-buffer
      (insert $bufstr)
      (goto-char (point-min))
      (while (re-search-forward "\\&\\#x\\(.\\{4\\}\\);" nil t)
        (let (($hex (match-string 1)))
          (delete-region (match-beginning 0) (match-end 0))
          (insert (format "%c" (string-to-number $hex 16)))))
      (setq $replaced (buffer-string)))
    (delete-region $beg $end)
    (insert $replaced)))

URLの変換

日本語(multibyte)ドメイン用には別途puny変換が必要なため, w3mの関数を利用しています。
Emacs25.2以降はデフォルトでpuny変換用の関数が搭載されるようです。

Encode

(my-encode-url-string "https://www.google.co.jp/search?q=いーまっくす")
;; => https://www.google.co.jp/search?q=%E3%81%84%E3%83%BC%E3%81%BE%E3%81%A3%E3%81%8F%E3%81%99

(defun my-encode-url-string ($url)
  (when (fboundp 'w3m-puny-encode-url)
    (setq $url (w3m-puny-encode-url $url)))
  (url-encode-url $url))

Decode

(my-decode-url-string "https://www.google.co.jp/search?q=%E3%81%84%E3%83%BC%E3%81%BE%E3%81%A3%E3%81%8F%E3%81%99")
;; => https://www.google.co.jp/search?q=いーまっくす

(defun my-decode-url-string ($url)
  (when (fboundp 'w3m-puny-decode-url)
    (setq $url (w3m-puny-decode-url $url)))
  (decode-coding-string (url-unhex-string $url) 'utf-8))

Multi-byte (\345\215\203)

バックスラッシュに3桁の8進数の文字

Encode

(my-encode-as-octet-string "emacsを始めよう！")
;; => emacs\343\202\222\345\247\213\343\202\201\343\202\210\343\201\206\357\274\201

(defun my-encode-as-octet-string ($str)
  (with-temp-buffer
    (insert $str)
    (encode-coding-region (point-min) (point-max) 'utf-8)
    (buffer-string)))

Decode

(my-decode-octet-string "emacs\343\202\222\345\247\213\343\202\201\343\202\210\343\201\206\357\274\201")
;; => emacsを始めよう！

(defun my-decode-octet-string ($str)
  (with-temp-buffer
    (insert $str)
    (decode-coding-region (point-min) (point-max) 'utf-8)
    (buffer-string)))

その他

(string-as-unibyte "千") ;; => \345\215\203
(string-as-multibyte "\345\215\203") ;; => 千
(toggle-enable-multibyte-characters) ;; バッファ内のマルチバイト文字置換

;; String -> Character
(string-to-char "あ") ;;=> 12354
?あ ;; => 12354

;; Character -> String
(char-to-string 12354) ;; => あ
(format "%c" 12354) ;; => あ

;; Unicode encode/decode
(my-char-to-entity ?あ) ;; => %3042
(my-entity-to-char "%3042" t)  ; => あ

(defun my-char-to-entity ($char)
  (interactive "cChar: ")
  (if (eq (type-of $char) 'string)
      (setq $char (string-to-char $char)))
  (let (($res (concat "%" (upcase (format "%x" $char)))))
    (when (called-interactively-p 'any)
      (message " %c -> %s " $char $res))
    $res))

(defun my-entity-to-char ($entity &optional $to-string)
  (interactive)
  (when (string-match "^%" $entity)
    (let* (($h (substring $entity 1))
           ($c (string-to-number $h 16))
           )
      (if $to-string
          (format "%c" $c)
        $c))))