lisp/mh-e/mh-scan.el

   1 ;;; mh-scan.el --- MH-E scan line constants and utilities
   2
   3 ;; Copyright (C) 1993, 1995, 1997,
   4 ;;   2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
   5 ;;   Free Software Foundation, Inc.
   6
   7 ;; Author: Bill Wohler <wohler@newt.com>
   8 ;; Maintainer: Bill Wohler <wohler@newt.com>
   9 ;; Keywords: mail
  10 ;; See: mh-e.el
  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 3 of the License, or
  17 ;; (at your option) 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.  If not, see <http://www.gnu.org/licenses/>.
  26
  27 ;;; Commentary:
  28
  29 ;; This file contains constants and a few functions for interpreting
  30 ;; scan lines.
  31
  32 ;;; Change Log:
  33
  34 ;;; Code:
  35
  36 (require 'mh-e)
  37
  38 \f
  39
  40 ;;; Scan Formats
  41
  42 ;; The following scan formats are passed to the scan program if the setting of
  43 ;; `mh-scan-format-file' is t. They are identical except the later one makes
  44 ;; use of the nmh `decode' function to decode RFC 2047 encodings. If you just
  45 ;; want to change the column of the notations, use the `mh-set-cmd-note'
  46 ;; function.
  47
  48 (defvar mh-scan-format-mh
  49   (concat
  50    "%4(msg)"
  51    "%<(cur)+%| %>"
  52    "%<{replied}-"
  53    "%?(nonnull(comp{to}))%<(mymbox{to})t%>"
  54    "%?(nonnull(comp{cc}))%<(mymbox{cc})c%>"
  55    "%?(nonnull(comp{bcc}))%<(mymbox{bcc})b%>"
  56    "%?(nonnull(comp{newsgroups}))n%>"
  57    "%<(zero) %>"
  58    "%02(mon{date})/%02(mday{date})%<{date} %|*%>"
  59    "%<(mymbox{from})%<{to}To:%14(friendly{to})%>%>"
  60    "%<(zero)%17(friendly{from})%>  "
  61    "%{subject}%<{body}<<%{body}%>")
  62   "*Scan format string for MH.
  63 This string is passed to the scan program via the -format
  64 argument.  This format is identical to the default except that
  65 additional hints for fontification have been added to the fifth
  66 column (remember that in Emacs, the first column is 0).
  67
  68 The values of the fifth column, in priority order, are: \"-\" if
  69 the message has been replied to, t if an address on the To: line
  70 matches one of the mailboxes of the current user, \"c\" if the Cc:
  71 line matches, \"b\" if the Bcc: line matches, and \"n\" if a
  72 non-empty Newsgroups: header is present.")
  73
  74 (defvar mh-scan-format-nmh
  75   (concat
  76    "%4(msg)"
  77    "%<(cur)+%| %>"
  78    "%<{replied}-"
  79    "%?(nonnull(comp{to}))%<(mymbox{to})t%>"
  80    "%?(nonnull(comp{cc}))%<(mymbox{cc})c%>"
  81    "%?(nonnull(comp{bcc}))%<(mymbox{bcc})b%>"
  82    "%?(nonnull(comp{newsgroups}))n%>"
  83    "%<(zero) %>"
  84    "%02(mon{date})/%02(mday{date})%<{date} %|*%>"
  85    "%<(mymbox{from})%<{to}To:%14(decode(friendly{to}))%>%>"
  86    "%<(zero)%17(decode(friendly{from}))%>  "
  87    "%(decode{subject})%<{body}<<%{body}%>")
  88   "*Scan format string for nmh.
  89 This string is passed to the scan program via the -format arg.
  90 This format is identical to the default except that additional
  91 hints for fontification have been added to the fifth
  92 column (remember that in Emacs, the first column is 0).
  93
  94 The values of the fifth column, in priority order, are: \"-\" if
  95 the message has been replied to, t if an address on the To: field
  96 matches one of the mailboxes of the current user, \"c\" if the Cc:
  97 field matches, \"b\" if the Bcc: field matches, and \"n\" if a
  98 non-empty Newsgroups: field is present.")
  99
 100 \f
 101
 102 ;;; Regular Expressions
 103
 104 ;; Alphabetical.
 105
 106 (defvar mh-scan-body-regexp "\\(<<\\([^\n]+\\)?\\)"
 107   "This regular expression matches the message body fragment.
 108
 109 Note that the default setting of `mh-folder-font-lock-keywords'
 110 expects this expression to contain at least one parenthesized
 111 expression which matches the body text as in the default of
 112 \"\\\\(<<\\\\([^\\n]+\\\\)?\\\\)\".  If this regular expression is
 113 not correct, the body fragment will not be highlighted with the
 114 face `mh-folder-body'.")
 115
 116 (defvar mh-scan-cur-msg-number-regexp "^\\( *[0-9]+\\+\\).*"
 117   "This regular expression matches the current message.
 118
 119 It must match from the beginning of the line.  Note that the
 120 default setting of `mh-folder-font-lock-keywords' expects this
 121 expression to contain at least one parenthesized expression which
 122 matches the message number as in the default of
 123
 124   \"^\\\\( *[0-9]+\\\\+\\\\).*\".
 125
 126 This expression includes the leading space and current message
 127 marker \"+\" within the parenthesis since it looks better to
 128 highlight these items as well.  The highlighting is done with the
 129 face `mh-folder-cur-msg-number'.  This regular expression should
 130 be correct as it is needed by non-fontification functions.  See
 131 also `mh-note-cur'.")
 132
 133 (defvar mh-scan-date-regexp "\\([0-9][0-9]/[0-9][0-9]\\)"
 134   "This regular expression matches a valid date.
 135
 136 It must not be anchored to the beginning or the end of the line.
 137 Note that the default setting of `mh-folder-font-lock-keywords'
 138 expects this expression to contain only one parenthesized
 139 expression which matches the date field as in the default of
 140 \"\\\\([0-9][0-9]/[0-9][0-9]\\\\)\"}.  If this regular expression
 141 is not correct, the date will not be highlighted with the face
 142 `mh-folder-date'.")
 143
 144 (defvar mh-scan-deleted-msg-regexp "^\\( *[0-9]+\\)D"
 145   "This regular expression matches deleted messages.
 146
 147 It must match from the beginning of the line.  Note that the
 148 default setting of `mh-folder-font-lock-keywords' expects this
 149 expression to contain at least one parenthesized expression which
 150 matches the message number as in the default of
 151
 152   \"^\\\\( *[0-9]+\\\\)D\".
 153
 154 This expression includes the leading space within the parenthesis
 155 since it looks better to highlight it as well.  The highlighting
 156 is done with the face `mh-folder-deleted'.  This regular
 157 expression should be correct as it is needed by non-fontification
 158 functions.  See also `mh-note-deleted'.")
 159
 160 (defvar mh-scan-good-msg-regexp  "^\\( *[0-9]+\\)[^D^0-9]"
 161   "This regular expression matches \"good\" messages.
 162
 163 It must match from the beginning of the line.  Note that the
 164 default setting of `mh-folder-font-lock-keywords' expects this
 165 expression to contain at least one parenthesized expression which
 166 matches the message number as in the default of
 167
 168   \"^\\\\( *[0-9]+\\\\)[^D^0-9]\".
 169
 170 This expression includes the leading space within the parenthesis
 171 since it looks better to highlight it as well.  The highlighting
 172 is done with the face `mh-folder-msg-number'.  This regular
 173 expression should be correct as it is needed by non-fontification
 174 functions.")
 175
 176 (defvar mh-scan-msg-format-regexp "%\\([0-9]*\\)(msg)"
 177   "This regular expression finds the message number width in a scan format.
 178
 179 Note that the message number must be placed in a parenthesized
 180 expression as in the default of \"%\\\\([0-9]*\\\\)(msg)\".  This
 181 variable is only consulted if `mh-scan-format-file' is set to
 182 \"Use MH-E scan Format\".")
 183
 184 (defvar mh-scan-msg-format-string "%d"
 185   "This is a format string for width of the message number in a scan format.
 186
 187 Use \"0%d\" for zero-filled message numbers.  This variable is only
 188 consulted if `mh-scan-format-file' is set to \"Use MH-E scan
 189 Format\".")
 190
 191 (defvar mh-scan-msg-number-regexp "^ *\\([0-9]+\\)"
 192   "This regular expression extracts the message number.
 193
 194 It must match from the beginning of the line.  Note that the
 195 message number must be placed in a parenthesized expression as in
 196 the default of \"^ *\\\\([0-9]+\\\\)\".")
 197
 198 (defvar mh-scan-msg-overflow-regexp "^[?0-9][0-9]"
 199   "This regular expression matches overflowed message numbers.")
 200
 201 (defvar mh-scan-msg-search-regexp "^[^0-9]*%d[^0-9]"
 202   "This regular expression matches a particular message.
 203
 204 It is a format string; use \"%d\" to represent the location of the
 205 message number within the expression as in the default of
 206 \"^[^0-9]*%d[^0-9]\".")
 207
 208 (defvar mh-scan-rcpt-regexp  "\\(To:\\)\\(..............\\)"
 209   "This regular expression specifies the recipient in messages you sent.
 210
 211 Note that the default setting of `mh-folder-font-lock-keywords'
 212 expects this expression to contain two parenthesized expressions.
 213 The first is expected to match the \"To:\" that the default scan
 214 format file generates.  The second is expected to match the
 215 recipient's name as in the default of
 216 \"\\\\(To:\\\\)\\\\(..............\\\\)\".  If this regular
 217 expression is not correct, the \"To:\" string will not be
 218 highlighted with the face `mh-folder-to' and the recipient will
 219 not be highlighted with the face `mh-folder-address'")
 220
 221 (defvar mh-scan-refiled-msg-regexp  "^\\( *[0-9]+\\)\\^"
 222   "This regular expression matches refiled messages.
 223
 224 It must match from the beginning of the line.  Note that the
 225 default setting of `mh-folder-font-lock-keywords' expects this
 226 expression to contain at least one parenthesized expression which
 227 matches the message number as in the default of
 228
 229   \"^\\\\( *[0-9]+\\\\)\\\\^\".
 230
 231 This expression includes the leading space within the parenthesis
 232 since it looks better to highlight it as well.  The highlighting
 233 is done with the face `mh-folder-refiled'.  This regular
 234 expression should be correct as it is needed by non-fontification
 235 functions.  See also `mh-note-refiled'.")
 236
 237 (defvar mh-scan-sent-to-me-sender-regexp
 238   "^ *[0-9]+.\\([bct]\\).....[ ]*\\(..................\\)"
 239   "This regular expression matches messages sent to us.
 240
 241 Note that the default setting of `mh-folder-font-lock-keywords'
 242 expects this expression to contain at least two parenthesized
 243 expressions.  The first should match the fontification hint (see
 244 `mh-scan-format-nmh') and the second should match the user name
 245 as in the default of
 246
 247   ^ *[0-9]+.\\\\([bct]\\\\).....[ ]*\\\\(..................\\\\)
 248
 249 If this regular expression is not correct, the notation hints
 250 will not be highlighted with the face
 251 `mh-mh-folder-sent-to-me-hint' and the sender will not be
 252 highlighted with the face `mh-folder-sent-to-me-sender'.")
 253
 254 (defvar mh-scan-subject-regexp
 255   "^ *[0-9]+........[ ]*...................\\([Rr][Ee]\\(\\[[0-9]+\\]\\)?:\\s-*\\)*\\([^<\n]*\\)"
 256   "This regular expression matches the subject.
 257
 258 It must match from the beginning of the line.  Note that the
 259 default setting of `mh-folder-font-lock-keywords' expects this
 260 expression to contain at least three parenthesized expressions.
 261 The first is expected to match the \"Re:\" string, if any, and is
 262 highlighted with the face `mh-folder-followup'.  The second
 263 matches an optional bracketed number after \"Re:\", such as in
 264 \"Re[2]:\" (and is thus a sub-expression of the first expression)
 265 and the third is expected to match the subject line itself which
 266 is highlighted with the face `mh-folder-subject'.  For example,
 267 the default (broken on multiple lines for readability) is
 268
 269   ^ *[0-9]+........[ ]*...................
 270   \\\\([Rr][Ee]\\\\(\\\\\\=[[0-9]+\\\\]\\\\)?:\\\\s-*\\\\)*
 271   \\\\([^<\\n]*\\\\)
 272
 273 This regular expression should be correct as it is needed by
 274 non-fontification functions.")
 275
 276 (defvar mh-scan-valid-regexp "^ *[0-9]"
 277   "This regular expression describes a valid scan line.
 278
 279 This is used to eliminate error messages that are occasionally
 280 produced by \"inc\".")
 281
 282 \f
 283
 284 ;;; Widths, Offsets and Columns
 285
 286 (defvar mh-cmd-note 4
 287   "Column for notations.
 288
 289 This variable should be set with the function `mh-set-cmd-note'.
 290 This variable may be updated dynamically if
 291 `mh-adaptive-cmd-note-flag' is on.
 292
 293 Note that columns in Emacs start with 0.")
 294 (make-variable-buffer-local 'mh-cmd-note)
 295
 296 (defvar mh-scan-cmd-note-width 1
 297   "Number of columns consumed by the cmd-note field in `mh-scan-format'.
 298
 299 This column will have one of the values: \" \", \"D\", \"^\", \"+\", where
 300
 301   \" \" is the default value,
 302   \"D\" is the `mh-note-deleted' character,
 303   \"^\" is the `mh-note-refiled' character, and
 304   \"+\" is the `mh-note-cur' character.")
 305
 306 (defvar mh-scan-destination-width 1
 307   "Number of columns consumed by the destination field in `mh-scan-format'.
 308
 309 This column will have one of \" \", \"%\", \"-\", \"t\", \"c\", \"b\", or \"n\"
 310 in it.
 311
 312   \" \" blank space is the default character.
 313   \"%\" indicates that the message in in a named MH sequence.
 314   \"-\" indicates that the message has been annotated with a replied field.
 315   \"t\" indicates that the message contains mymbox in the To: field.
 316   \"c\" indicates that the message contains mymbox in the Cc: field.
 317   \"b\" indicates that the message contains mymbox in the Bcc: field.
 318   \"n\" indicates that the message contains a Newsgroups: field.")
 319
 320 (defvar mh-scan-date-width 5
 321   "Number of columns consumed by the date field in `mh-scan-format'.
 322 This column will typically be of the form mm/dd.")
 323
 324 (defvar mh-scan-date-flag-width 1
 325   "Number of columns consumed to flag (in)valid dates in `mh-scan-format'.
 326 This column will have \" \" for valid and \"*\" for invalid or
 327 missing dates.")
 328
 329 (defvar mh-scan-from-mbox-width 17
 330   "Number of columns consumed with the \"From:\" line in `mh-scan-format'.
 331 This column will have a friendly name or e-mail address of the
 332 originator, or a \"To: address\" for outgoing e-mail messages.")
 333
 334 (defvar mh-scan-from-mbox-sep-width 2
 335   "Number of columns consumed by whitespace after from-mbox in `mh-scan-format'.
 336 This column will only ever have spaces in it.")
 337
 338 (defvar mh-scan-field-destination-offset
 339   (+ mh-scan-cmd-note-width)
 340   "The offset from the `mh-cmd-note' for the destination column.")
 341
 342 (defvar mh-scan-field-from-start-offset
 343   (+ mh-scan-cmd-note-width
 344      mh-scan-destination-width
 345      mh-scan-date-width
 346      mh-scan-date-flag-width)
 347   "The offset from the `mh-cmd-note' to find the start of \"From:\" address.")
 348
 349 (defvar mh-scan-field-from-end-offset
 350   (+ mh-scan-field-from-start-offset mh-scan-from-mbox-width)
 351   "The offset from the `mh-cmd-note' to find the end of \"From:\" address.")
 352
 353 (defvar mh-scan-field-subject-start-offset
 354   (+ mh-scan-cmd-note-width
 355      mh-scan-destination-width
 356      mh-scan-date-width
 357      mh-scan-date-flag-width
 358      mh-scan-from-mbox-width
 359      mh-scan-from-mbox-sep-width)
 360   "The offset from the `mh-cmd-note' to find the start of the subject.")
 361
 362 \f
 363
 364 ;;; Notation
 365
 366 ;; Alphabetical.
 367
 368 (defvar mh-note-cur ?+
 369   "The current message (in MH, not in MH-E) is marked by this character.
 370 See also `mh-scan-cur-msg-number-regexp'.")
 371
 372 (defvar mh-note-copied ?C
 373   "Messages that have been copied are marked by this character.")
 374
 375 (defvar mh-note-deleted ?D
 376   "Messages that have been deleted are marked by this character.
 377 See also `mh-scan-deleted-msg-regexp'.")
 378
 379 (defvar mh-note-dist ?R
 380   "Messages that have been redistributed are marked by this character.")
 381
 382 (defvar mh-note-forw ?F
 383   "Messages that have been forwarded are marked by this character.")
 384
 385 (defvar mh-note-printed ?P
 386   "Messages that have been printed are marked by this character.")
 387
 388 (defvar mh-note-refiled ?^
 389   "Messages that have been refiled are marked by this character.
 390 See also `mh-scan-refiled-msg-regexp'.")
 391
 392 (defvar mh-note-repl ?-
 393   "Messages that have been replied to are marked by this character.")
 394
 395 (defvar mh-note-seq ?%
 396   "Messages in a user-defined sequence are marked by this character.
 397
 398 Messages in the \"search\" sequence are marked by this character as
 399 well.")
 400
 401 \f
 402
 403 ;;; Utilities
 404
 405 ;;;###mh-autoload
 406 (defun mh-scan-msg-number-regexp ()
 407   "Return value of variable `mh-scan-msg-number-regexp'."
 408   mh-scan-msg-number-regexp)
 409
 410 ;;;###mh-autoload
 411 (defun mh-scan-msg-search-regexp ()
 412   "Return value of variable `mh-scan-msg-search-regexp'."
 413   mh-scan-msg-search-regexp)
 414
 415 ;;;###mh-autoload
 416 (defun mh-set-cmd-note (column)
 417   "Set `mh-cmd-note' to COLUMN.
 418 Note that columns in Emacs start with 0."
 419   (setq mh-cmd-note column))
 420
 421 ;;;###mh-autoload
 422 (defun mh-scan-format ()
 423   "Return the output format argument for the scan program."
 424   (if (equal mh-scan-format-file t)
 425       (list "-format" (if (mh-variant-p 'nmh 'gnu-mh)
 426                           (list (mh-update-scan-format
 427                                  mh-scan-format-nmh mh-cmd-note))
 428                         (list (mh-update-scan-format
 429                                mh-scan-format-mh mh-cmd-note))))
 430     (if (not (equal mh-scan-format-file nil))
 431         (list "-form" mh-scan-format-file))))
 432
 433 (defun mh-update-scan-format (fmt width)
 434   "Return a scan format with the (msg) width in the FMT replaced with WIDTH.
 435
 436 The message number width portion of the format is discovered
 437 using `mh-scan-msg-format-regexp'.  Its replacement is controlled
 438 with `mh-scan-msg-format-string'."
 439   (or (and
 440        (string-match mh-scan-msg-format-regexp fmt)
 441        (let ((begin (match-beginning 1))
 442              (end (match-end 1)))
 443          (concat (substring fmt 0 begin)
 444                  (format mh-scan-msg-format-string width)
 445                  (substring fmt end))))
 446       fmt))
 447
 448 ;;;###mh-autoload
 449 (defun mh-msg-num-width (folder)
 450   "Return the width of the largest message number in this FOLDER."
 451   (or mh-progs (mh-find-path))
 452   (let ((tmp-buffer (get-buffer-create mh-temp-buffer))
 453         (width 0))
 454     (with-current-buffer tmp-buffer
 455       (erase-buffer)
 456       (apply 'call-process
 457              (expand-file-name mh-scan-prog mh-progs) nil '(t nil) nil
 458              (list folder "last" "-format" "%(msg)"))
 459       (goto-char (point-min))
 460       (if (re-search-forward mh-scan-msg-number-regexp nil 0 1)
 461           (setq width (length (buffer-substring
 462                                (match-beginning 1) (match-end 1))))))
 463     width))
 464
 465 ;;;###mh-autoload
 466 (defun mh-msg-num-width-to-column (width)
 467   "Return the column for notations given message number WIDTH.
 468 Note that columns in Emacs start with 0.
 469
 470 If `mh-scan-format-file' is set to \"Use MH-E scan Format\" this
 471 means that either `mh-scan-format-mh' or `mh-scan-format-nmh' are
 472 in use.  This function therefore assumes that the first column is
 473 empty (to provide room for the cursor), the following WIDTH
 474 columns contain the message number, and the column for notations
 475 comes after that."
 476   (if (eq mh-scan-format-file t)
 477       (max (1+ width) 2)
 478     (error "%s %s" "Can't call `mh-msg-num-width-to-column' when"
 479            "`mh-scan-format-file' is not set to \"Use MH-E scan Format\"")))
 480
 481 (provide 'mh-scan)
 482
 483 ;; Local Variables:
 484 ;; indent-tabs-mode: nil
 485 ;; sentence-end-double-space: nil
 486 ;; End:
 487
 488 ;; arch-tag: 5ab35d46-101e-443b-a2b6-5a908cf97528
 489 ;;; mh-scan.el ends here