Converting backslash sequences to (non-ascii) characters

[Eduardo Ochs]

Hi Erich,

I won't be able to approve posts with big attachments until I
recover my moderator password for the eev mailing list, so this
is a public answer to your e-mail that was intended to be public
but that I only received in private... you said that you were
trying to generate links to a PDF using this,

  http://angg.twu.net/eev-intros/find-pdf-like-intro.html#10
  (find-pdf-like-intro "10. Generating a pair with the page number")

but you were getting sexps like these

  # (find-leo9page 1 "Th\351orie de la Science")
  # (find-leo9text 1 "Th\351orie de la Science")
  # (find-leo9page (+ 0 1) "Th\351orie de la Science")
  # (find-leo9text (+ 0 1) "Th\351orie de la Science")

instead of:

  # (find-leo9page 1 "Théorie de la Science")
  # (find-leo9text 1 "Théorie de la Science")
  # (find-leo9page (+ 0 1) "Théorie de la Science")
  # (find-leo9text (+ 0 1) "Théorie de la Science")



I thought a bit about making `M-h M-p' convert some of its
backslash sequences to the corresponding characters and realized
that it wouldn't be easy to find a robust way to do that...

...so here is a workaround. If you mark a region and then run
`M-x bsl', the `bsl' will convert the right (???) backslash
sequences in that region to the corresponding characters. This
code is something that I can test and maintain easily, and I will
be able to reuse its structure in other things that to do.

Hope it helps! =)
  Cheers,
    Eduardo



--snip--snip-- Here is the code: --snip--snip--

;; (find-elnode "Basic Char Syntax" "\\a")
;; (find-elnode "General Escape Syntax" "\\uXXXX")
;; (find-elnode "General Escape Syntax" "\\x41")
;; (find-elnode "General Escape Syntax" "octal")
;; (find-elnode "Rx Notation")
;; (find-elnode "Rx Constructs" "Capture groups")
;; (find-elnode "Search and Replace")

(setq ee-elisp-bsl-re
      (rx "\\" (or (group-n 1 (any "abtnvfres\\d"))
   (group-n 2 (regexp "[0-7][0-7][0-7]"))
   (and (any "x") (group-n 3 hex hex))
   (and (any "uU") (group-n 4 hex hex hex hex
    (opt hex hex) (opt hex hex)))
   )))

(defun ee-elisp-bsl-convert1 (str)
  (save-match-data
    (string-match ee-elisp-bsl-re str)
    (let ((char  (match-string 1 str))
          (octal (match-string 2 str))
          (hexa  (match-string 3 str))
          (hexb  (match-string 4 str)))
      (cond (char  (match-string 0 str))
    (octal (format "%c" (string-to-number octal 8)))
    (hexa  (format "%c" (string-to-number hexa 16)))
    (hexb  (format "%c" (string-to-number hexb 16)))
    ))))

(defun ee-elisp-bsl-replace (s e)
  "Replace some backslash sequences in the region.
The sequences \\ooo, \\xhh and \\uhhhh, where the `o's are octal
digits and the `h's are hex digits, are replaced by their
corresponding characters."
  (interactive "r")
  (save-excursion
    (save-restriction
      (narrow-to-region s e)
      (goto-char (point-min))
      (while (re-search-forward ee-elisp-bsl-re nil 'noerror)
(replace-match (ee-elisp-bsl-convert1 (match-string 0)) 'fixedcase 'literal)
))))

(defalias 'bsl 'ee-elisp-bsl-replace)

;; Tests:
;; (ee-elisp-bsl-convert1 "\\n")
;; (ee-elisp-bsl-convert1 "\\234")
;; (ee-elisp-bsl-convert1 "\\351")
;; (ee-elisp-bsl-convert1 "_\\351_")
;; (ee-elisp-bsl-convert1 "\\x41")
;; (ee-elisp-bsl-convert1 "\\u2022")
;; (find-einsert '((0 255)))
;;
;; ...and mark the next line and run `M-x bsl' on it:
;; "Foo \n\\\234\u2022\xa2"