lisp/thingatpt.el

   1 ;;; thingatpt.el --- get the `thing' at point
   2
   3 ;; Copyright (C) 1991,92,93,94,95,96,97,1998,2000
   4 ;;  Free Software Foundation, Inc.
   5
   6 ;; Author: Mike Williams <mikew@gopher.dosli.govt.nz>
   7 ;; Maintainer: FSF
   8 ;; Keywords: extensions, matching, mouse
   9 ;; Created: Thu Mar 28 13:48:23 1991
  10
  11 ;; This file is part of GNU Emacs.
  12
  13 ;; GNU Emacs is free software; you can redistribute it and/or modify
  14 ;; it under the terms of the GNU General Public License as published by
  15 ;; the Free Software Foundation; either version 2, or (at your option)
  16 ;; any later version.
  17
  18 ;; GNU Emacs is distributed in the hope that it will be useful,
  19 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  20 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  21 ;; GNU General Public License for more details.
  22
  23 ;;; Commentary:
  24
  25 ;; This file provides routines for getting the "thing" at the location of
  26 ;; point, whatever that "thing" happens to be.  The "thing" is defined by
  27 ;; its beginning and end positions in the buffer.
  28 ;;
  29 ;; The function bounds-of-thing-at-point finds the beginning and end
  30 ;; positions by moving first forward to the end of the "thing", and then
  31 ;; backwards to the beginning.  By default, it uses the corresponding
  32 ;; forward-"thing" operator (eg. forward-word, forward-line).
  33 ;;
  34 ;; Special cases are allowed for using properties associated with the named
  35 ;; "thing":
  36 ;;
  37 ;;   forward-op         Function to call to skip forward over a "thing" (or
  38 ;;                      with a negative argument, backward).
  39 ;;
  40 ;;   beginning-op       Function to call to skip to the beginning of a "thing".
  41 ;;   end-op             Function to call to skip to the end of a "thing".
  42 ;;
  43 ;; Reliance on existing operators means that many `things' can be accessed
  44 ;; without further code:  eg.
  45 ;;     (thing-at-point 'line)
  46 ;;     (thing-at-point 'page)
  47
  48 ;;; Code:
  49
  50 (provide 'thingatpt)
  51
  52 ;; Basic movement
  53
  54 ;;;###autoload
  55 (defun forward-thing (thing &optional n)
  56   "Move forward to the end of the next THING."
  57   (let ((forward-op (or (get thing 'forward-op)
  58                         (intern-soft (format "forward-%s" thing)))))
  59     (if (functionp forward-op)
  60         (funcall forward-op (or n 1))
  61       (error "Can't determine how to move over a %s" thing))))
  62
  63 ;; General routines
  64
  65 ;;;###autoload
  66 (defun bounds-of-thing-at-point (thing)
  67   "Determine the start and end buffer locations for the THING at point.
  68 THING is a symbol which specifies the kind of syntactic entity you want.
  69 Possibilities include `symbol', `list', `sexp', `defun', `filename', `url',
  70 `word', `sentence', `whitespace', `line', `page' and others.
  71
  72 See the file `thingatpt.el' for documentation on how to define
  73 a symbol as a valid THING.
  74
  75 The value is a cons cell (START . END) giving the start and end positions
  76 of the textual entity that was found."
  77   (if (get thing 'bounds-of-thing-at-point)
  78       (funcall (get thing 'bounds-of-thing-at-point))
  79     (let ((orig (point)))
  80       (condition-case nil
  81           (save-excursion
  82             ;; Try moving forward, then back.
  83             (let ((end (progn
  84                          (funcall
  85                           (or (get thing 'end-op)
  86                               (function (lambda () (forward-thing thing 1)))))
  87                          (point)))
  88                   (beg (progn
  89                          (funcall
  90                           (or (get thing 'beginning-op)
  91                               (function (lambda () (forward-thing thing -1)))))
  92                          (point))))
  93               (if (not (and beg (> beg orig)))
  94                   ;; If that brings us all the way back to ORIG,
  95                   ;; it worked.  But END may not be the real end.
  96                   ;; So find the real end that corresponds to BEG.
  97                   (let ((real-end
  98                          (progn
  99                            (funcall
 100                             (or (get thing 'end-op)
 101                                 (function (lambda () (forward-thing thing 1)))))
 102                            (point))))
 103                     (if (and beg real-end (<= beg orig) (<= orig real-end))
 104                         (cons beg real-end)))
 105                 (goto-char orig)
 106                 ;; Try a second time, moving backward first and then forward,
 107                 ;; so that we can find a thing that ends at ORIG.
 108                 (let ((beg (progn
 109                              (funcall
 110                               (or (get thing 'beginning-op)
 111                                   (function (lambda () (forward-thing thing -1)))))
 112                              (point)))
 113                       (end (progn
 114                              (funcall
 115                               (or (get thing 'end-op)
 116                                   (function (lambda () (forward-thing thing 1)))))
 117                              (point)))
 118                       (real-beg
 119                        (progn
 120                          (funcall
 121                           (or (get thing 'beginning-op)
 122                               (function (lambda () (forward-thing thing -1)))))
 123                          (point))))
 124                   (if (and real-beg end (<= real-beg orig) (<= orig end))
 125                       (cons real-beg end))))))
 126         (error nil)))))
 127
 128 ;;;###autoload
 129 (defun thing-at-point (thing)
 130   "Return the THING at point.
 131 THING is a symbol which specifies the kind of syntactic entity you want.
 132 Possibilities include `symbol', `list', `sexp', `defun', `filename', `url',
 133 `word', `sentence', `whitespace', `line', `page' and others.
 134
 135 See the file `thingatpt.el' for documentation on how to define
 136 a symbol as a valid THING."
 137   (if (get thing 'thing-at-point)
 138       (funcall (get thing 'thing-at-point))
 139     (let ((bounds (bounds-of-thing-at-point thing)))
 140       (if bounds
 141           (buffer-substring (car bounds) (cdr bounds))))))
 142
 143 ;; Go to beginning/end
 144
 145 (defun beginning-of-thing (thing)
 146   (let ((bounds (bounds-of-thing-at-point thing)))
 147     (or bounds (error "No %s here" thing))
 148     (goto-char (car bounds))))
 149
 150 (defun end-of-thing (thing)
 151   (let ((bounds (bounds-of-thing-at-point thing)))
 152     (or bounds (error "No %s here" thing))
 153     (goto-char (cdr bounds))))
 154
 155 ;;  Special cases
 156
 157 ;;  Lines
 158
 159 ;; bolp will be false when you click on the last line in the buffer
 160 ;; and it has no final newline.
 161
 162 (put 'line 'beginning-op
 163      (function (lambda () (if (bolp) (forward-line -1) (beginning-of-line)))))
 164
 165 ;;  Sexps
 166
 167 (defun in-string-p ()
 168   (let ((orig (point)))
 169     (save-excursion
 170       (beginning-of-defun)
 171       (nth 3 (parse-partial-sexp (point) orig)))))
 172
 173 (defun end-of-sexp ()
 174   (let ((char-syntax (char-syntax (char-after (point)))))
 175     (if (or (eq char-syntax ?\))
 176             (and (eq char-syntax ?\") (in-string-p)))
 177         (forward-char 1)
 178       (forward-sexp 1))))
 179
 180 (put 'sexp 'end-op 'end-of-sexp)
 181
 182 (defun beginning-of-sexp ()
 183   (let ((char-syntax (char-syntax (char-before (point)))))
 184     (if (or (eq char-syntax ?\()
 185             (and (eq char-syntax ?\") (in-string-p)))
 186         (forward-char -1)
 187       (forward-sexp -1))))
 188
 189 (put 'sexp 'beginning-op 'beginning-of-sexp)
 190
 191 ;;  Lists
 192
 193 (put 'list 'end-op (function (lambda () (up-list 1))))
 194 (put 'list 'beginning-op 'backward-sexp)
 195
 196 ;;  Filenames and URLs  www.com/foo%32bar
 197
 198 (defvar thing-at-point-file-name-chars "-~/[:alnum:]_.${}#%,:"
 199   "Characters allowable in filenames.")
 200
 201 (put 'filename 'end-op
 202      (lambda ()
 203        (re-search-forward (concat "\\=[" thing-at-point-file-name-chars "]*")
 204                           nil t)))
 205 (put 'filename 'beginning-op
 206      (lambda ()
 207        (if (re-search-backward (concat "[^" thing-at-point-file-name-chars "]")
 208                                nil t)
 209            (forward-char)
 210          (goto-char (point-min)))))
 211
 212 (defvar thing-at-point-url-path-regexp
 213   "[^]\t\n \"'()<>[^`{}]*[^]\t\n \"'()<>[^`{}.,;]+"
 214   "A regular expression probably matching the host and filename or e-mail part of a URL.")
 215
 216 (defvar thing-at-point-short-url-regexp
 217   (concat "[-A-Za-z0-9.]+" thing-at-point-url-path-regexp)
 218   "A regular expression probably matching a URL without an access scheme.
 219 Hostname matching is stricter in this case than for
 220 ``thing-at-point-url-regexp''.")
 221
 222 (defvar thing-at-point-uri-schemes
 223   ;; Officials from http://www.iana.org/assignments/uri-schemes
 224   '("ftp://" "http://" "gopher://" "mailto:" "news:" "nntp:"
 225     "telnet://" "wais://" "file:/" "prospero:" "z39.50s:" "z39.50r:"
 226     "cid:" "mid:" "vemmi:" "service:" "imap:" "nfs:" "acap:" "rtsp:"
 227     "tip:" "pop:" "data:" "dav:" "opaquelocktoken:" "sip:" "tel:" "fax:"
 228     "modem:" "ldap:" "https://" "soap.beep:" "soap.beeps:" "urn:" "go:"
 229     "afs:" "tn3270:" "mailserver:"
 230   ;; Compatibility
 231     "snews:")
 232   "Uniform Resource Identifier (URI) Schemes")
 233
 234 (defvar thing-at-point-url-regexp
 235   (concat "\\<\\(" (mapconcat 'identity thing-at-point-uri-schemes "\\|") "\\)"
 236           thing-at-point-url-path-regexp)
 237   "A regular expression probably matching a complete URL.")
 238
 239 (defvar thing-at-point-markedup-url-regexp
 240   "<URL:[^>]+>"
 241   "A regular expression matching a URL marked up per RFC1738.
 242 This may contain whitespace (including newlines) .")
 243
 244 (put 'url 'bounds-of-thing-at-point 'thing-at-point-bounds-of-url-at-point)
 245 (defun thing-at-point-bounds-of-url-at-point ()
 246   (let ((url "") short strip)
 247     (if (or (setq strip (thing-at-point-looking-at
 248                          thing-at-point-markedup-url-regexp))
 249             (thing-at-point-looking-at thing-at-point-url-regexp)
 250             ;; Access scheme omitted?
 251             (setq short (thing-at-point-looking-at
 252                          thing-at-point-short-url-regexp)))
 253         (let ((beginning (match-beginning 0))
 254               (end (match-end 0)))
 255           (cond (strip
 256                  (setq beginning (+ beginning 5))
 257                  (setq end (- end 1))))
 258           (cons beginning end)))))
 259
 260 (put 'url 'thing-at-point 'thing-at-point-url-at-point)
 261 (defun thing-at-point-url-at-point ()
 262   "Return the URL around or before point.
 263
 264 Search backwards for the start of a URL ending at or after point.  If
 265 no URL found, return nil.  The access scheme will be prepended if
 266 absent: \"mailto:\" if the string contains \"@\", \"ftp://\" if it
 267 starts with \"ftp\" and not \"ftp:/\", or \"http://\" by default."
 268
 269   (let ((url "") short strip)
 270     (if (or (setq strip (thing-at-point-looking-at
 271                          thing-at-point-markedup-url-regexp))
 272             (thing-at-point-looking-at thing-at-point-url-regexp)
 273             ;; Access scheme omitted?
 274             (setq short (thing-at-point-looking-at
 275                          thing-at-point-short-url-regexp)))
 276         (progn
 277           (setq url (buffer-substring-no-properties (match-beginning 0)
 278                                                     (match-end 0)))
 279           (and strip (setq url (substring url 5 -1))) ; Drop "<URL:" & ">"
 280           ;; strip whitespace
 281           (while (string-match "[ \t\n\r]+" url)
 282             (setq url (replace-match "" t t url)))
 283           (and short (setq url (concat (cond ((string-match "@" url)
 284                                               "mailto:")
 285                                              ;; e.g. ftp.swiss... or ftp-swiss...
 286                                              ((string-match "^ftp" url)
 287                                               "ftp://")
 288                                              (t "http://"))
 289                                        url)))
 290           (if (string-equal "" url)
 291               nil
 292             url)))))
 293
 294 ;; The normal thingatpt mechanism doesn't work for complex regexps.
 295 ;; This should work for almost any regexp wherever we are in the
 296 ;; match.  To do a perfect job for any arbitrary regexp would mean
 297 ;; testing every position before point.  Regexp searches won't find
 298 ;; matches that straddle the start position so we search forwards once
 299 ;; and then back repeatedly and then back up a char at a time.
 300
 301 (defun thing-at-point-looking-at (regexp)
 302   "Return non-nil if point is in or just after a match for REGEXP.
 303 Set the match data from the earliest such match ending at or after
 304 point."
 305   (save-excursion
 306     (let ((old-point (point)) match)
 307       (and (looking-at regexp)
 308            (>= (match-end 0) old-point)
 309            (setq match (point)))
 310       ;; Search back repeatedly from end of next match.
 311       ;; This may fail if next match ends before this match does.
 312       (re-search-forward regexp nil 'limit)
 313       (while (and (re-search-backward regexp nil t)
 314                   (or (> (match-beginning 0) old-point)
 315                       (and (looking-at regexp)  ; Extend match-end past search start
 316                            (>= (match-end 0) old-point)
 317                            (setq match (point))))))
 318       (if (not match) nil
 319         (goto-char match)
 320         ;; Back up a char at a time in case search skipped
 321         ;; intermediate match straddling search start pos.
 322         (while (and (not (bobp))
 323                     (progn (backward-char 1) (looking-at regexp))
 324                     (>= (match-end 0) old-point)
 325                     (setq match (point))))
 326         (goto-char match)
 327         (looking-at regexp)))))
 328
 329 (put 'url 'end-op
 330      (function (lambda ()
 331                  (let ((bounds (thing-at-point-bounds-of-url-at-point)))
 332                    (if bounds
 333                        (goto-char (cdr bounds))
 334                      (error "No URL here"))))))
 335 (put 'url 'beginning-op
 336      (function (lambda ()
 337                  (let ((bounds (thing-at-point-bounds-of-url-at-point)))
 338                    (if bounds
 339                        (goto-char (car bounds))
 340                      (error "No URL here"))))))
 341
 342 ;;  Whitespace
 343
 344 (defun forward-whitespace (arg)
 345   (interactive "p")
 346   (if (natnump arg)
 347       (re-search-forward "[ \t]+\\|\n" nil 'move arg)
 348     (while (< arg 0)
 349       (if (re-search-backward "[ \t]+\\|\n" nil 'move)
 350           (or (eq (char-after (match-beginning 0)) 10)
 351               (skip-chars-backward " \t")))
 352       (setq arg (1+ arg)))))
 353
 354 ;;  Buffer
 355
 356 (put 'buffer 'end-op (lambda () (goto-char (point-max))))
 357 (put 'buffer 'beginning-op (lambda () (goto-char (point-min))))
 358
 359 ;;  Symbols
 360
 361 (defun forward-symbol (arg)
 362   (interactive "p")
 363   (if (natnump arg)
 364       (re-search-forward "\\(\\sw\\|\\s_\\)+" nil 'move arg)
 365     (while (< arg 0)
 366       (if (re-search-backward "\\(\\sw\\|\\s_\\)+" nil 'move)
 367           (skip-syntax-backward "w_"))
 368       (setq arg (1+ arg)))))
 369
 370 ;;  Syntax blocks
 371
 372 (defun forward-same-syntax (&optional arg)
 373   (interactive "p")
 374   (while (< arg 0)
 375     (skip-syntax-backward
 376      (char-to-string (char-syntax (char-after (1- (point))))))
 377     (setq arg (1+ arg)))
 378   (while (> arg 0)
 379     (skip-syntax-forward (char-to-string (char-syntax (char-after (point)))))
 380     (setq arg (1- arg))))
 381
 382 ;;  Aliases
 383
 384 (defun word-at-point () (thing-at-point 'word))
 385 (defun sentence-at-point () (thing-at-point 'sentence))
 386
 387 (defun read-from-whole-string (str)
 388   "Read a lisp expression from STR.
 389 Signal an error if the entire string was not used."
 390   (let* ((read-data (read-from-string str))
 391          (more-left
 392           (condition-case nil
 393               ;; The call to `ignore' suppresses a compiler warning.
 394               (progn (ignore (read-from-string (substring str (cdr read-data))))
 395                      t)
 396             (end-of-file nil))))
 397     (if more-left
 398         (error "Can't read whole string")
 399       (car read-data))))
 400
 401 (defun form-at-point (&optional thing pred)
 402   (let ((sexp (condition-case nil
 403                   (read-from-whole-string (thing-at-point (or thing 'sexp)))
 404                 (error nil))))
 405     (if (or (not pred) (funcall pred sexp)) sexp)))
 406
 407 ;;;###autoload
 408 (defun sexp-at-point ()   (form-at-point 'sexp))
 409 ;;;###autoload
 410 (defun symbol-at-point () (form-at-point 'sexp 'symbolp))
 411 ;;;###autoload
 412 (defun number-at-point () (form-at-point 'sexp 'numberp))
 413 ;;;###autoload
 414 (defun list-at-point ()   (form-at-point 'list 'listp))
 415
 416 ;;; thingatpt.el ends here