lisp/regi.el

   1 ;;; regi.el --- REGular expression Interpreting engine
   2
   3 ;; Copyright (C) 1993 Free Software Foundation, Inc.
   4
   5 ;; Author: 1993 Barry A. Warsaw, Century Computing, Inc. <bwarsaw@cen.com>
   6 ;; Maintainer:    bwarsaw@cen.com
   7 ;; Created:       24-Feb-1993
   8 ;; Version:       1.8
   9 ;; Last Modified: 1993/06/01 21:33:00
  10 ;; Keywords:      extensions, matching
  11
  12 ;; This file is part of GNU Emacs.
  13
  14 ;; GNU Emacs is free software; you can redistribute it and/or modify
  15 ;; it under the terms of the GNU General Public License as published by
  16 ;; the Free Software Foundation; either version 2, or (at your option)
  17 ;; any later version.
  18
  19 ;; GNU Emacs is distributed in the hope that it will be useful,
  20 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  21 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  22 ;; GNU General Public License for more details.
  23
  24 ;; You should have received a copy of the GNU General Public License
  25 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  26 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  27 ;; Boston, MA 02111-1307, USA.
  28
  29 ;;; Code:
  30
  31 \f
  32 (defun regi-pos (&optional position col-p)
  33   "Return the character position at various buffer positions.
  34 Optional POSITION can be one of the following symbols:
  35
  36 `bol'  == beginning of line
  37 `boi'  == beginning of indentation
  38 `eol'  == end of line [default]
  39 `bonl' == beginning of next line
  40 `bopl' == beginning of previous line
  41
  42 Optional COL-P non-nil returns `current-column' instead of character position."
  43   (save-excursion
  44     (cond
  45      ((eq position 'bol)  (beginning-of-line))
  46      ((eq position 'boi)  (back-to-indentation))
  47      ((eq position 'bonl) (forward-line 1))
  48      ((eq position 'bopl) (forward-line -1))
  49      (t (end-of-line)))
  50     (if col-p (current-column) (point))))
  51
  52 (defun regi-mapcar (predlist func &optional negate-p case-fold-search-p)
  53   "Build a regi frame where each element of PREDLIST appears exactly once.
  54 The frame contains elements where each member of PREDLIST is
  55 associated with FUNC, and optionally NEGATE-P and CASE-FOLD-SEARCH-P."
  56   (let (frame tail)
  57     (if (or negate-p case-fold-search-p)
  58         (setq tail (list negate-p)))
  59     (if case-fold-search-p
  60         (setq tail (append tail (list case-fold-search-p))))
  61     (while predlist
  62       (let ((element (list (car predlist) func)))
  63         (if tail
  64             (setq element (append element tail)))
  65         (setq frame (append frame (list element))
  66               predlist (cdr predlist))
  67         ))
  68     frame))
  69
  70 \f
  71 (defun regi-interpret (frame &optional start end)
  72   "Interpret the regi frame FRAME.
  73 If optional START and END are supplied, they indicate the region of
  74 interest, and the buffer is narrowed to the beginning of the line
  75 containing START, and beginning of the line after the line containing
  76 END.  Otherwise, point and mark are not set and processing continues
  77 until your FUNC returns the `abort' symbol (see below).  Beware!  Not
  78 supplying a START or END could put you in an infinite loop.
  79
  80 A regi frame is a list of entries of the form:
  81
  82      (PRED FUNC [NEGATE-P [CASE-FOLD-SEARCH]])
  83
  84 PRED is a predicate against which each line in the region is tested,
  85 and if a match occurs, FUNC is `eval'd.  Point is then moved to the
  86 beginning of the next line, the frame is reset and checking continues.
  87 If a match doesn't occur, the next entry is checked against the
  88 current line until all entries in the frame are checked.  At this
  89 point, if no match occurred, the frame is reset and point is moved to
  90 the next line.  Checking continues until every line in the region is
  91 checked.  Optional NEGATE-P inverts the result of PRED before FUNC is
  92 called and `case-fold-search' is bound to the optional value of
  93 CASE-FOLD-SEARCH for the PRED check.
  94
  95 PRED can be a string, variable, function or one of the following
  96 symbols: t, nil, `begin', `end', and `every'.  If PRED is a string, or
  97 a variable or list that evaluates to a string, it is interpreted as a
  98 regular expression and is matched against the current line (from the
  99 beginning) using `looking-at'.  If PRED does not evaluate to a string,
 100 it is interpreted as a binary value (nil or non-nil).
 101
 102 PRED can also be one of the following symbols:
 103
 104 t       -- always produces a true outcome
 105 `begin' -- always executes before anything else
 106 `end'   -- always executes after everything else
 107 `every' -- execute after frame is matched on a line
 108
 109 Note that NEGATE-P and CASE-FOLD-SEARCH are meaningless if PRED is one
 110 of these special symbols.  Only the first occurrence of each symbol in
 111 a frame entry is used, the rest are ignored.
 112
 113 Your FUNC can return values which control regi processing.  If a list
 114 is returned from your function, it can contain any combination of the
 115 following elements:
 116
 117 the symbol `continue'
 118      Tells regi to continue processing frame-entries after a match,
 119      instead of resetting to the first entry and advancing to the next
 120      line, as is the default behavior.  When returning this symbol,
 121      you must take care not to enter an infinite loop.
 122
 123 the symbol `abort'
 124      Tells regi to terminate processing this frame.  any end
 125      frame-entry is still processed.
 126
 127 the list `(frame . NEWFRAME)'
 128      Tells regi to use NEWFRAME as its current frame.  In other words,
 129      your FUNC can modify the executing regi frame on the fly.
 130
 131 the list `(step . STEP)'
 132      Tells regi to move STEP number of lines forward during normal
 133      processing.  By default, regi moves forward 1 line.  STEP can be
 134      negative, but be careful of infinite loops.
 135
 136 You should usually take care to explicitly return nil from your
 137 function if no action is to take place.  Your FUNC will always be
 138 `eval'ed.  The following variables will be temporarily bound to some
 139 useful information:
 140
 141 `curline'
 142      the current line in the buffer, as a string
 143
 144 `curframe'
 145      the full, current frame being executed
 146
 147 `curentry'
 148      the current frame entry being executed."
 149
 150   (save-excursion
 151     (save-restriction
 152       (let (begin-tag end-tag every-tag current-frame working-frame donep)
 153
 154         ;; set up the narrowed region
 155         (and start
 156              end
 157              (let* ((tstart start)
 158                     (start (min start end))
 159                     (end   (max start end)))
 160                (narrow-to-region
 161                 (progn (goto-char end) (regi-pos 'bonl))
 162                 (progn (goto-char start) (regi-pos 'bol)))))
 163
 164         ;; lets find the special tags and remove them from the working
 165         ;; frame. note that only the last special tag is used.
 166         (mapcar
 167          (function
 168           (lambda (entry)
 169             (let ((pred (car entry))
 170                   (func (car (cdr entry))))
 171               (cond
 172                ((eq pred 'begin) (setq begin-tag func))
 173                ((eq pred 'end)   (setq end-tag func))
 174                ((eq pred 'every) (setq every-tag func))
 175                (t
 176                 (setq working-frame (append working-frame (list entry))))
 177                ) ; end-cond
 178               )))
 179          frame) ; end-mapcar
 180
 181         ;; execute the begin entry
 182         (eval begin-tag)
 183
 184         ;; now process the frame
 185         (setq current-frame working-frame)
 186         (while (not (or donep (eobp)))
 187           (let* ((entry            (car current-frame))
 188                  (pred             (nth 0 entry))
 189                  (func             (nth 1 entry))
 190                  (negate-p         (nth 2 entry))
 191                  (case-fold-search (nth 3 entry))
 192                  match-p)
 193             (catch 'regi-throw-top
 194               (cond
 195                ;; we are finished processing the frame for this line
 196                ((not current-frame)
 197                 (setq current-frame working-frame) ;reset frame
 198                 (forward-line 1)
 199                 (throw 'regi-throw-top t))
 200                ;; see if predicate evaluates to a string
 201                ((stringp (setq match-p (eval pred)))
 202                 (setq match-p (looking-at match-p)))
 203                ) ; end-cond
 204
 205               ;; now that we've done the initial matching, check for
 206               ;; negation of match
 207               (and negate-p
 208                    (setq match-p (not match-p)))
 209
 210               ;; if the line matched, package up the argument list and
 211               ;; funcall the FUNC
 212               (if match-p
 213                    (let* ((curline (buffer-substring
 214                                     (regi-pos 'bol)
 215                                     (regi-pos 'eol)))
 216                           (curframe current-frame)
 217                           (curentry entry)
 218                           (result (eval func))
 219                           (step (or (cdr (assq 'step result)) 1))
 220                           )
 221                      ;; changing frame on the fly?
 222                      (if (assq 'frame result)
 223                          (setq working-frame (cdr (assq 'frame result))))
 224
 225                      ;; continue processing current frame?
 226                      (if (memq 'continue result)
 227                          (setq current-frame (cdr current-frame))
 228                        (forward-line step)
 229                        (setq current-frame working-frame))
 230
 231                      ;; abort current frame?
 232                      (if (memq 'abort result)
 233                          (progn
 234                            (setq donep t)
 235                            (throw 'regi-throw-top t)))
 236                      ) ; end-let
 237
 238                 ;; else if no match occurred, then process the next
 239                 ;; frame-entry on the current line
 240                 (setq current-frame (cdr current-frame))
 241
 242                 ) ; end-if match-p
 243               ) ; end catch
 244             ) ; end let
 245
 246           ;; after every cycle, evaluate every-tag
 247           (eval every-tag)
 248           ) ; end-while
 249
 250         ;; now process the end entry
 251         (eval end-tag)))))
 252
 253 \f
 254 (provide 'regi)
 255 ;;; regi.el ends here