Mercurial > emacs
changeset 13814:f683f93b94e9
Initial revision
author | Karl Heuer <kwzh@gnu.org> |
---|---|
date | Thu, 21 Dec 1995 18:13:58 +0000 |
parents | 7d18fdd1acf4 |
children | 2a2eaca9d6bb |
files | lisp/uniquify.el |
diffstat | 1 files changed, 381 insertions(+), 0 deletions(-) [+] |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/lisp/uniquify.el Thu Dec 21 18:13:58 1995 +0000 @@ -0,0 +1,381 @@ +;;; uniquify.el --- unique buffer names dependent on pathname + +;; Copyright (c) 1989, 1995 Free Software Foundation, Inc. + +;; Author: Dick King <king@reasoning.com> +;; Maintainer: Michael Ernst <mernst@theory.lcs.mit.edu> +;; Created: 15 May 86 + +;; This file is part of GNU Emacs. + +;; GNU Emacs is free software; you can redistribute it and/or modify +;; it under the terms of the GNU General Public License as published by +;; the Free Software Foundation; either version 2, or (at your option) +;; any later version. + +;; GNU Emacs is distributed in the hope that it will be useful, +;; but WITHOUT ANY WARRANTY; without even the implied warranty of +;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +;; GNU General Public License for more details. + +;; You should have received a copy of the GNU General Public License +;; along with GNU Emacs; see the file COPYING. If not, write to +;; the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + +;;; Commentary: + +;; Emacs's standard method for making buffer names unique adds <2>, <3>, +;; etc. to the end of (all but one of) the buffers. This file replaces +;; that behavior, for buffers visiting files and dired buffers, with a +;; uniquification that adds parts of the pathname until the buffer names +;; are unique. For instance, buffers visiting /u/mernst/tmp/Makefile and +;; /usr/projects/zaphod/Makefile would be named Makefile|tmp and +;; Makefile|zaphod, respectively (instead of Makefile and Makefile<2>). +;; Other buffer name styles are also available. + +;; To use this file, just load it. +;; To disable it after loading, set variable uniquify-buffer-name-style to nil. +;; For other options, see "User-visible variables", below. + +;; uniquify.el works under Emacs, XEmacs, and InfoDock. + +;; Doesn't correctly handle buffer names created by M-x write-file in Emacs 18. + +;;; Change Log: + +;; Originally by Dick King <king@reasoning.com> 15 May 86 +;; Converted for Emacs 18 by Stephen Gildea <gildea@lcs.mit.edu> +;; Make uniquify-min-dir-content 0 truly non-invasive. gildea 23 May 89 +;; Some cleanup. uniquify-min-dir-content default 0. gildea 01 Jun 89 +;; Don't rename to "". Michael Ernst <mernst@theory.lcs.mit.edu> 15 Jun 94 +;; Add kill-buffer-hook. Kenneth Manheimer <ken.manheimer@nist.gov> 09 May 95 +;; Add advice for rename-buffer and create-file-buffer, handle dired buffers, +;; kill-buffer-rationalize-buffer-names-p, documentation. mernst 24 May 95 +;; Remove free variables, fix typos. mernst 5 Jun 95 +;; Efficiently support Emacs 19.27 & earlier. ken.manheimer, mernst 10 Jun 95 +;; Rename user options to "uniquify-...", add uniquify-reverse-dir-content-p, +;; add uniquify-ask-about-buffer-names-p. king, mernst 13 Jun 95 +;; Prefix functions by "uniquify-..."; rename mnemonic-buffer-names to +;; uniquify-buffer-name-style; add 'forward and 'post-forward-angle-brackets +;; styles; remove uniquify-reverse-dir-content-p; add +;; uniquify-trailing-separator-p. mernst 4 Aug 95 + +;; Valuable feedback was provided by +;; Paul Smith <psmith@baynetworks.com>, +;; Alastair Burt <burt@dfki.uni-kl.de>, +;; Bob Weiner <weiner@footloose.sps.mot.com>, +;; Albert L. Ting <alt@vlibs.com>, +;; gyro@reasoning.com. + + +;;; Code: + +(provide 'uniquify) + +;;; User-visible variables + +(defvar uniquify-buffer-name-style 'post-forward + "*If non-nil, buffer names are uniquified with parts of directory name. +The value determines the buffer name style and is one of 'forward, +'reverse, 'post-forward (the default), or 'post-forward-angle-brackets. +For example, files /foo/bar/mumble/name and /baz/quux/mumble/name +would be in the following buffers: + forward bar/mumble/name quux/mumble/name + reverse name\\mumble\\bar name\\mumble\\quux + post-forward name|bar/mumble name|quux/mumble + post-forward-angle-brackets name<bar/mumble> name<quux/mumble> + nil name name<2>") + +(defvar uniquify-after-kill-buffer-p nil + "*If non-nil, rerationalize buffer names after a buffer has been killed. +This can be dangerous if Emacs Lisp code is keeping track of buffers by their +names (rather than keeping pointers to the buffers themselves).") + +(defconst uniquify-ask-about-buffer-names-p nil + "*If non-nil, permit user to choose names for buffers with same base file. +If the user chooses to name a buffer, uniquification is preempted and no +other buffer names are changed.") + +(defvar uniquify-min-dir-content 0 + "*Minimum parts of directory pathname included in buffer name.") + +(defvar uniquify-separator nil + "*String separator for buffer name components. +When `uniquify-buffer-name-style' is 'post-forward, separates +base file name from directory part in buffer names (default \"|\"). +When `uniquify-buffer-name-style' is 'reverse, separates all +pathname components (default \"\\\").") + +(defvar uniquify-trailing-separator-p nil + "*If non-nil, add a pathname separator to dired buffer names. +If `uniquify-buffer-name-style' is 'forward, add the separator at the end; +if it's is 'reverse, add the separator at the beginning; otherwise, this +variable is ignored.") + + +;;; Utilities + +(defmacro uniquify-push (item list) + (` (setq (, list) (cons (, item) (, list))))) + +(defmacro uniquify-fix-list-base (a) + (` (car (, a)))) + +(defmacro uniquify-fix-list-filename (a) + (` (car (cdr (, a))))) + +(defmacro uniquify-fix-list-buffer (a) + (` (car (cdr (cdr (, a)))))) + +(defmacro uniquify-cadddr (a) + (` (car (cdr (cdr (cdr (, a))))))) + +;; Internal variables used free +(defvar uniquify-non-file-buffer-names nil) +(defvar uniquify-possibly-resolvable nil) + +;;; Main entry point. + +(defun uniquify-rationalize-file-buffer-names (&optional newbuffile newbuf) + "Makes file buffer names unique by adding segments from pathname. +If `uniquify-min-dir-content' > 0, always pulls that many +pathname elements. Arguments cause only a subset of buffers to be renamed." + (interactive) + (let (fix-list + uniquify-non-file-buffer-names + (depth uniquify-min-dir-content)) + (let ((buffers (buffer-list))) + (while buffers + (let* ((buffer (car buffers)) + (bfn (if (eq buffer newbuf) + (expand-file-name newbuffile) + (uniquify-buffer-file-name buffer))) + (rawname (and bfn (file-name-nondirectory bfn))) + (deserving (and rawname + (or (not newbuffile) + (equal rawname + (file-name-nondirectory newbuffile)))))) + (if deserving + (uniquify-push (list rawname bfn buffer nil) fix-list) + (uniquify-push (list (buffer-name buffer)) + uniquify-non-file-buffer-names))) + (setq buffers (cdr buffers)))) + ;; selects buffers whose names may need changing, and others that + ;; may conflict. + (setq fix-list + (sort fix-list 'uniquify-filename-sort)) + ;; bringing conflicting names together + (uniquify-rationalize-a-list fix-list depth) + (mapcar 'uniquify-unrationalized-buffer fix-list))) + +(defun uniquify-buffer-file-name (buffer) + "Return name of file BUFFER is visiting, or nil if none. +Works on dired buffers as well as ordinary file-visiting buffers." + (or (buffer-file-name buffer) + (save-excursion + (set-buffer buffer) + list-buffers-directory))) + +(defun uniquify-filename-sort (s1 s2) + (uniquify-filename-lessp + (uniquify-fix-list-filename s1) (uniquify-fix-list-filename s2))) + +(defun uniquify-filename-lessp (s1 s2) + (let ((s1f (file-name-nondirectory s1)) + (s2f (file-name-nondirectory s2))) + (and (not (equal s2f "")) + (or (string-lessp s1f s2f) + (and (equal s1f s2f) + (let ((s1d (file-name-directory s1)) + (s2d (file-name-directory s2))) + (and (not (<= (length s2d) 1)) + (or (<= (length s1d) 1) + (uniquify-filename-lessp + (substring s1d 0 -1) + (substring s2d 0 -1)))))))))) + +;; Was named do-the-buffers-you-couldnt-rationalize +(defun uniquify-unrationalized-buffer (item) + (or (uniquify-cadddr item) nil)) ;maybe better in the future + +(defun uniquify-rationalize-a-list (fix-list depth) + (let (conflicting-sublist + (old-name "") + proposed-name uniquify-possibly-resolvable) + (while fix-list + (let ((item (car fix-list))) + (setq proposed-name (uniquify-get-proposed-name item depth)) + (if (not (equal proposed-name old-name)) + (progn + (uniquify-rationalize-conflicting-sublist + conflicting-sublist old-name depth) + (setq conflicting-sublist nil))) + (uniquify-push item conflicting-sublist) + (setq old-name proposed-name)) + (setq fix-list (cdr fix-list))) + (uniquify-rationalize-conflicting-sublist + conflicting-sublist old-name depth))) + +(defun uniquify-get-proposed-name (item depth) + (let (index + (extra-string "") + (n depth) + (base (uniquify-fix-list-base item)) + (fn (uniquify-fix-list-filename item))) + (while (and (> n 0) + (setq index (string-match + (concat "\\(^\\|/[^/]*\\)/" + (regexp-quote extra-string) + (regexp-quote base) + "\\'") + fn))) + (setq extra-string (substring fn + (if (zerop index) 0 (1+ index)) + ;; (- (length base)) fails for base = "". + ;; Equivalently, we could have used + ;; (apply 'substring ... + ;; (and (not (string= "" base)) + ;; (list (- (length base))))) + (- (length fn) (length base))) + n (1- n))) + (if (zerop n) (setq uniquify-possibly-resolvable t)) + + + ;; Distinguish directories by adding extra separator. + (if (and uniquify-trailing-separator-p + (file-directory-p fn) + (not (string-equal base ""))) + (cond ((eq uniquify-buffer-name-style 'forward) + (setq base (concat base "/"))) + ((eq uniquify-buffer-name-style 'reverse) + (setq base (concat (or uniquify-separator "\\") base))))) + + ;; Trim trailing separator on directory part + (if (and (not (string-equal extra-string "")) + (or (eq uniquify-buffer-name-style 'post-forward) + (eq uniquify-buffer-name-style 'post-forward-angle-brackets))) + (setq extra-string (substring extra-string 0 + (- (length extra-string) 1)))) + + (cond ((string-equal extra-string "") + base) + ((string-equal base "") + extra-string) + ((eq uniquify-buffer-name-style 'forward) + (concat extra-string base)) + ((eq uniquify-buffer-name-style 'reverse) + (concat base (uniquify-reverse-components extra-string))) + ((eq uniquify-buffer-name-style 'post-forward) + (concat base (or uniquify-separator "|") extra-string)) + ((eq uniquify-buffer-name-style 'post-forward-angle-brackets) + (concat base "<" extra-string ">")) + (t (error "Bad value for uniquify-buffer-name-style: %s" + uniquify-buffer-name-style))))) + + +;; Deal with conflicting-sublist, which is set by uniquify-rationalize-a-list. +;; This is only called by uniquify-rationalize-a-list. +(defun uniquify-rationalize-conflicting-sublist (conflicting-sublist old-name depth) + (or (null conflicting-sublist) + (and (null (cdr conflicting-sublist)) + (not (assoc old-name uniquify-non-file-buffer-names)) + (or (and (not (string= old-name "")) + (uniquify-rename-buffer (car conflicting-sublist) old-name)) + t)) + (if uniquify-possibly-resolvable + (uniquify-rationalize-a-list conflicting-sublist (1+ depth))))) + +(defun uniquify-rename-buffer (item newname) + (let ((buffer (uniquify-fix-list-buffer item))) + (if (not (equal newname (buffer-name buffer))) + (let ((unset (current-buffer)) + ;; avoid hooks on rename-buffer + (uniquify-buffer-name-style nil)) + (set-buffer buffer) + (rename-buffer newname) + (set-buffer unset)))) + (rplaca (nthcdr 3 item) t)) + +(defun uniquify-reverse-components (instring) + (let ((sofar '()) + (cursor 0) + (len (length instring)) + (sep (or uniquify-separator "\\"))) + (while (< cursor len) + (if (= (aref instring cursor) ?/) + (setq sofar (cons sep sofar) + cursor (1+ cursor)) + (let ((first-slash (or (string-match "/" instring cursor) len))) + (setq sofar (cons (substring instring cursor first-slash) sofar) + cursor first-slash)))) + (apply (function concat) sofar))) + + +;;; Hooks from the rest of Emacs + +;; Emacs 19 (GNU Emacs or XEmacs) + +;; The logical place to put all this code is in generate-new-buffer-name. +;; It's written in C, so we would add a generate-new-buffer-name-function +;; which, if non-nil, would be called instead of the C. One problem with +;; that is that generate-new-buffer-name takes a potential buffer name as +;; its argument -- not other information, such as what file the buffer will +;; visit. + +;; The below solution works because generate-new-buffer-name is called +;; only by rename-buffer (which, as of 19.29, is never called from C) and +;; generate-new-buffer, which is called only by Lisp functions +;; create-file-buffer and rename-uniquely. Rename-uniquely generally +;; isn't used for buffers visiting files, so it's sufficient to hook +;; rename-buffer and create-file-buffer. (Setting find-file-hooks isn't +;; sufficient.) + +(defadvice rename-buffer (after rename-buffer-uniquify activate) + "Uniquify buffer names with parts of directory name." + (if (and uniquify-buffer-name-style + ;; UNIQUE argument + (ad-get-arg 1)) + (progn + (if uniquify-after-kill-buffer-p + ;; call with no argument; rationalize vs. old name as well as new + (uniquify-rationalize-file-buffer-names) + ;; call with argument: rationalize vs. new name only + (uniquify-rationalize-file-buffer-names + (uniquify-buffer-file-name (current-buffer)) (current-buffer))) + (setq ad-return-value (buffer-name (current-buffer)))))) + +(defadvice create-file-buffer (after create-file-buffer-uniquify activate) + "Uniquify buffer names with parts of directory name." + (if uniquify-buffer-name-style + (uniquify-rationalize-file-buffer-names (ad-get-arg 0) ad-return-value))) + +;; Buffer deletion +;; Rerationalize after a buffer is killed, to reduce coinciding buffer names. +;; This mechanism uses `kill-buffer-hook', which runs *before* deletion. +;; That means that the kill-buffer-hook function cannot just delete the +;; buffer -- it has to set something to do the rationalization *later*. +;; It actually puts another function on `post-command-hook'. This other +;; function runs the rationalization and then removes itself from the hook. +;; Is there a better way to accomplish this? +;; (This ought to set some global variables so the work is done only for +;; buffers with names similar to the deleted buffer. -MDE) + +(defun delay-uniquify-rationalize-file-buffer-names () + "Add `delayed-uniquify-rationalize-file-buffer-names' to `post-command-hook'. +For use on, eg, `kill-buffer-hook', to rationalize *after* buffer deletion." + (if (and uniquify-buffer-name-style + uniquify-after-kill-buffer-p) + (add-hook 'post-command-hook + 'delayed-uniquify-rationalize-file-buffer-names))) + +(defun delayed-uniquify-rationalize-file-buffer-names () + "Rerationalize buffer names and remove self from `post-command-hook'. +See also `delay-rationalize-file-buffer-names' for hook setter." + (uniquify-rationalize-file-buffer-names) + (remove-hook 'post-command-hook + 'delayed-uniquify-rationalize-file-buffer-names)) + +(add-hook 'kill-buffer-hook 'delay-uniquify-rationalize-file-buffer-names) + +;;; uniquify.el ends here +