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 Free Software Foundation, Inc.
   5
   6 ;; Author: Bill Wohler <wohler@newt.com>
   7 ;; Maintainer: Bill Wohler <wohler@newt.com>
   8 ;; Keywords: mail
   9 ;; See: mh-e.el
  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 3 of the License, or
  16 ;; (at your option) 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 ;; You should have received a copy of the GNU General Public License
  24 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  25
  26 ;;; Commentary:
  27
  28 ;; This file contains constants and a few functions for interpreting
  29 ;; scan lines.
  30
  31 ;;; Change Log:
  32
  33 ;;; Code:
  34
  35 (require 'mh-e)
  36
  37 \f
  38
  39 ;;; Scan Formats
  40
  41 ;; The following scan formats are passed to the scan program if the setting of
  42 ;; `mh-scan-format-file' is t. They are identical except the later one makes
  43 ;; use of the nmh `decode' function to decode RFC 2047 encodings. If you just
  44 ;; want to change the column of the notations, use the `mh-set-cmd-note'
  45 ;; function.
  46
  47 (defvar mh-scan-format-mh
  48   (concat
  49    "%4(msg)"
  50    "%<(cur)+%| %>"
  51    "%<{replied}-"
  52    "%?(nonnull(comp{to}))%<(mymbox{to})t%>"
  53    "%?(nonnull(comp{cc}))%<(mymbox{cc})c%>"
  54    "%?(nonnull(comp{bcc}))%<(mymbox{bcc})b%>"
  55    "%?(nonnull(comp{newsgroups}))n%>"
  56    "%<(zero) %>"
  57    "%02(mon{date})/%02(mday{date})%<{date} %|*%>"
  58    "%<(mymbox{from})%<{to}To:%14(friendly{to})%>%>"
  59    "%<(zero)%17(friendly{from})%>  "
  60    "%{subject}%<{body}<<%{body}%>")
  61   "*Scan format string for MH.
  62 This string is passed to the scan program via the -format
  63 argument. This format is identical to the default except that
  64 additional hints for fontification have been added to the fifth
  65 column (remember that in Emacs, the first column is 0).
  66
  67 The values of the fifth column, in priority order, are: \"-\" if
  68 the message has been replied to, t if an address on the To: line
  69 matches one of the mailboxes of the current user, \"c\" if the Cc:
  70 line matches, \"b\" if the Bcc: line matches, and \"n\" if a
  71 non-empty Newsgroups: header is present.")
  72
  73 (defvar mh-scan-format-nmh
  74   (concat
  75    "%4(msg)"
  76    "%<(cur)+%| %>"
  77    "%<{replied}-"
  78    "%?(nonnull(comp{to}))%<(mymbox{to})t%>"
  79    "%?(nonnull(comp{cc}))%<(mymbox{cc})c%>"
  80    "%?(nonnull(comp{bcc}))%<(mymbox{bcc})b%>"
  81    "%?(nonnull(comp{newsgroups}))n%>"
  82    "%<(zero) %>"
  83    "%02(mon{date})/%02(mday{date})%<{date} %|*%>"
  84    "%<(mymbox{from})%<{to}To:%14(decode(friendly{to}))%>%>"
  85    "%<(zero)%17(decode(friendly{from}))%>  "
  86    "%(decode{subject})%<{body}<<%{body}%>")
  87   "*Scan format string for nmh.
  88 This string is passed to the scan program via the -format arg.
  89 This format is identical to the default except that additional
  90 hints for fontification have been added to the fifth
  91 column (remember that in Emacs, the first column is 0).
  92
  93 The values of the fifth column, in priority order, are: \"-\" if
  94 the message has been replied to, t if an address on the To: field
  95 matches one of the mailboxes of the current user, \"c\" if the Cc:
  96 field matches, \"b\" if the Bcc: field matches, and \"n\" if a
  97 non-empty Newsgroups: field is present.")
  98
  99 \f
 100
 101 ;;; Regular Expressions
 102
 103 ;; Alphabetical.
 104
 105 (defvar mh-scan-body-regexp "\\(<<\\([^\n]+\\)?\\)"
 106   "This regular expression matches the message body fragment.
 107
 108 Note that the default setting of `mh-folder-font-lock-keywords'
 109 expects this expression to contain at least one parenthesized
 110 expression which matches the body text as in the default of
 111 \"\\\\(<<\\\\([^\\n]+\\\\)?\\\\)\". If this regular expression is
 112 not correct, the body fragment will not be highlighted with the
 113 face `mh-folder-body'.")
 114
 115 (defvar mh-scan-cur-msg-number-regexp "^\\( *[0-9]+\\+\\).*"
 116   "This regular expression matches the current message.
 117
 118 It must match from the beginning of the line. Note that the
 119 default setting of `mh-folder-font-lock-keywords' expects this
 120 expression to contain at least one parenthesized expression which
 121 matches the message number as in the default of
 122
 123   \"^\\\\( *[0-9]+\\\\+\\\\).*\".
 124
 125 This expression includes the leading space and current message
 126 marker \"+\" within the parenthesis since it looks better to
 127 highlight these items as well. The highlighting is done with the
 128 face `mh-folder-cur-msg-number'. This regular expression should
 129 be correct as it is needed by non-fontification functions. See
 130 also `mh-note-cur'.")
 131
 132 (defvar mh-scan-date-regexp "\\([0-9][0-9]/[0-9][0-9]\\)"
 133   "This regular expression matches a valid date.
 134
 135 It must not be anchored to the beginning or the end of the line.
 136 Note that the default setting of `mh-folder-font-lock-keywords'
 137 expects this expression to contain only one parenthesized
 138 expression which matches the date field as in the default of
 139 \"\\\\([0-9][0-9]/[0-9][0-9]\\\\)\"}. If this regular expression
 140 is not correct, the date will not be highlighted with the face
 141 `mh-folder-date'.")
 142
 143 (defvar mh-scan-deleted-msg-regexp "^\\( *[0-9]+\\)D"
 144   "This regular expression matches deleted messages.
 145
 146 It must match from the beginning of the line. Note that the
 147 default setting of `mh-folder-font-lock-keywords' expects this
 148 expression to contain at least one parenthesized expression which
 149 matches the message number as in the default of
 150
 151   \"^\\\\( *[0-9]+\\\\)D\".
 152
 153 This expression includes the leading space within the parenthesis
 154 since it looks better to highlight it as well. The highlighting
 155 is done with the face `mh-folder-deleted'. This regular
 156 expression should be correct as it is needed by non-fontification
 157 functions. See also `mh-note-deleted'.")
 158
 159 (defvar mh-scan-good-msg-regexp  "^\\( *[0-9]+\\)[^D^0-9]"
 160   "This regular expression matches \"good\" messages.
 161
 162 It must match from the beginning of the line. Note that the
 163 default setting of `mh-folder-font-lock-keywords' expects this
 164 expression to contain at least one parenthesized expression which
 165 matches the message number as in the default of
 166
 167   \"^\\\\( *[0-9]+\\\\)[^D^0-9]\".
 168
 169 This expression includes the leading space within the parenthesis
 170 since it looks better to highlight it as well. The highlighting
 171 is done with the face `mh-folder-msg-number'. This regular
 172 expression should be correct as it is needed by non-fontification
 173 functions.")
 174
 175 (defvar mh-scan-msg-format-regexp "%\\([0-9]*\\)(msg)"
 176   "This regular expression finds the message number width in a scan format.
 177
 178 Note that the message number must be placed in a parenthesized
 179 expression as in the default of \"%\\\\([0-9]*\\\\)(msg)\". This
 180 variable is only consulted if `mh-scan-format-file' is set to
 181 \"Use MH-E scan Format\".")
 182
 183 (defvar mh-scan-msg-format-string "%d"
 184   "This is a format string for width of the message number in a scan format.
 185
 186 Use \"0%d\" for zero-filled message numbers. This variable is only
 187 consulted if `mh-scan-format-file' is set to \"Use MH-E scan
 188 Format\".")
 189
 190 (defvar mh-scan-msg-number-regexp "^ *\\([0-9]+\\)"
 191   "This regular expression extracts the message number.
 192
 193 It must match from the beginning of the line. Note that the
 194 message number must be placed in a parenthesized expression as in
 195 the default of \"^ *\\\\([0-9]+\\\\)\".")
 196
 197 (defvar mh-scan-msg-overflow-regexp "^[?0-9][0-9]"
 198   "This regular expression matches overflowed message numbers.")
 199
 200 (defvar mh-scan-msg-search-regexp "^[^0-9]*%d[^0-9]"
 201   "This regular expression matches a particular message.
 202
 203 It is a format string; use \"%d\" to represent the location of the
 204 message number within the expression as in the default of
 205 \"^[^0-9]*%d[^0-9]\".")
 206
 207 (defvar mh-scan-rcpt-regexp  "\\(To:\\)\\(..............\\)"
 208   "This regular expression specifies the recipient in messages you sent.
 209
 210 Note that the default setting of `mh-folder-font-lock-keywords'
 211 expects this expression to contain two parenthesized expressions.
 212 The first is expected to match the \"To:\" that the default scan
 213 format file generates. The second is expected to match the
 214 recipient's name as in the default of
 215 \"\\\\(To:\\\\)\\\\(..............\\\\)\". If this regular
 216 expression is not correct, the \"To:\" string will not be
 217 highlighted with the face `mh-folder-to' and the recipient will
 218 not be highlighted with the face `mh-folder-address'")
 219
 220 (defvar mh-scan-refiled-msg-regexp  "^\\( *[0-9]+\\)\\^"
 221   "This regular expression matches refiled messages.
 222
 223 It must match from the beginning of the line. Note that the
 224 default setting of `mh-folder-font-lock-keywords' expects this
 225 expression to contain at least one parenthesized expression which
 226 matches the message number as in the default of
 227
 228   \"^\\\\( *[0-9]+\\\\)\\\\^\".
 229
 230 This expression includes the leading space within the parenthesis
 231 since it looks better to highlight it as well. The highlighting
 232 is done with the face `mh-folder-refiled'. This regular
 233 expression should be correct as it is needed by non-fontification
 234 functions. See also `mh-note-refiled'.")
 235
 236 (defvar mh-scan-sent-to-me-sender-regexp
 237   "^ *[0-9]+.\\([bct]\\).....[ ]*\\(..................\\)"
 238   "This regular expression matches messages sent to us.
 239
 240 Note that the default setting of `mh-folder-font-lock-keywords'
 241 expects this expression to contain at least two parenthesized
 242 expressions. The first should match the fontification hint (see
 243 `mh-scan-format-nmh') and the second should match the user name
 244 as in the default of
 245
 246   ^ *[0-9]+.\\\\([bct]\\\\).....[ ]*\\\\(..................\\\\)
 247
 248 If this regular expression is not correct, the notation hints
 249 will not be highlighted with the face
 250 `mh-mh-folder-sent-to-me-hint' and the sender will not be
 251 highlighted with the face `mh-folder-sent-to-me-sender'.")
 252
 253 (defvar mh-scan-subject-regexp
 254   "^ *[0-9]+........[ ]*...................\\([Rr][Ee]\\(\\[[0-9]+\\]\\)?:\\s-*\\)*\\([^<\n]*\\)"
 255   "This regular expression matches the subject.
 256
 257 It must match from the beginning of the line. Note that the
 258 default setting of `mh-folder-font-lock-keywords' expects this
 259 expression to contain at least three parenthesized expressions.
 260 The first is expected to match the \"Re:\" string, if any, and is
 261 highlighted with the face `mh-folder-followup'. The second
 262 matches an optional bracketed number after \"Re:\", such as in
 263 \"Re[2]:\" (and is thus a sub-expression of the first expression)
 264 and the third is expected to match the subject line itself which
 265 is highlighted with the face `mh-folder-subject'. For example,
 266 the default (broken on multiple lines for readability) is
 267
 268   ^ *[0-9]+........[ ]*...................
 269   \\\\([Rr][Ee]\\\\(\\\\\\=[[0-9]+\\\\]\\\\)?:\\\\s-*\\\\)*
 270   \\\\([^<\\n]*\\\\)
 271
 272 This regular expression should be correct as it is needed by
 273 non-fontification functions.")
 274
 275 (defvar mh-scan-valid-regexp "^ *[0-9]"
 276   "This regular expression describes a valid scan line.
 277
 278 This is used to eliminate error messages that are occasionally
 279 produced by \"inc\".")
 280
 281 \f
 282
 283 ;;; Widths, Offsets and Columns
 284
 285 (defvar mh-cmd-note 4
 286   "Column for notations.
 287
 288 This variable should be set with the function `mh-set-cmd-note'.
 289 This variable may be updated dynamically if
 290 `mh-adaptive-cmd-note-flag' is on.
 291
 292 Note that columns in Emacs start with 0.")
 293 (make-variable-buffer-local 'mh-cmd-note)
 294
 295 (defvar mh-scan-cmd-note-width 1
 296   "Number of columns consumed by the cmd-note field in `mh-scan-format'.
 297
 298 This column will have one of the values: \" \", \"D\", \"^\", \"+\" and
 299 where \" \" is the default value,
 300
 301   \"D\" is the `mh-note-deleted' character,
 302   \"^\" is the `mh-note-refiled' character, and
 303   \"+\" is the `mh-note-cur' character.")
 304
 305 (defvar mh-scan-destination-width 1
 306   "Number of columns consumed by the destination field in `mh-scan-format'.
 307
 308 This column will have one of \" \", \"%\", \"-\", \"t\", \"c\", \"b\", or \"n\"
 309 in it.
 310
 311   \" \" blank space is the default character.
 312   \"%\" indicates that the message in in a named MH sequence.
 313   \"-\" indicates that the message has been annotated with a replied field.
 314   \"t\" indicates that the message contains mymbox in the To: field.
 315   \"c\" indicates that the message contains mymbox in the Cc: field.
 316   \"b\" indicates that the message contains mymbox in the Bcc: field.
 317   \"n\" indicates that the message contains a Newsgroups: field.")
 318
 319 (defvar mh-scan-date-width 5
 320   "Number of columns consumed by the date field in `mh-scan-format'.
 321 This column will typically be of the form mm/dd.")
 322
 323 (defvar mh-scan-date-flag-width 1
 324   "Number of columns consumed to flag (in)valid dates in `mh-scan-format'.
 325 This column will have \" \" for valid and \"*\" for invalid or
 326 missing dates.")
 327
 328 (defvar mh-scan-from-mbox-width 17
 329   "Number of columns consumed with the \"From:\" line in `mh-scan-format'.
 330 This column will have a friendly name or e-mail address of the
 331 originator, or a \"To: address\" for outgoing e-mail messages.")
 332
 333 (defvar mh-scan-from-mbox-sep-width 2
 334   "Number of columns consumed by whitespace after from-mbox in `mh-scan-format'.
 335 This column will only ever have spaces in it.")
 336
 337 (defvar mh-scan-field-destination-offset
 338   (+ mh-scan-cmd-note-width)
 339   "The offset from the `mh-cmd-note' for the destination column.")
 340
 341 (defvar mh-scan-field-from-start-offset
 342   (+ mh-scan-cmd-note-width
 343      mh-scan-destination-width
 344      mh-scan-date-width
 345      mh-scan-date-flag-width)
 346   "The offset from the `mh-cmd-note' to find the start of \"From:\" address.")
 347
 348 (defvar mh-scan-field-from-end-offset
 349   (+ mh-scan-field-from-start-offset mh-scan-from-mbox-width)
 350   "The offset from the `mh-cmd-note' to find the end of \"From:\" address.")
 351
 352 (defvar mh-scan-field-subject-start-offset
 353   (+ mh-scan-cmd-note-width
 354      mh-scan-destination-width
 355      mh-scan-date-width
 356      mh-scan-date-flag-width
 357      mh-scan-from-mbox-width
 358      mh-scan-from-mbox-sep-width)
 359   "The offset from the `mh-cmd-note' to find the start of the subject.")
 360
 361 \f
 362
 363 ;;; Notation
 364
 365 ;; Alphabetical.
 366
 367 (defvar mh-note-cur ?+
 368   "The current message (in MH, not in MH-E) is marked by this character.
 369 See also `mh-scan-cur-msg-number-regexp'.")
 370
 371 (defvar mh-note-copied ?C
 372   "Messages that have been copied are marked by this character.")
 373
 374 (defvar mh-note-deleted ?D
 375   "Messages that have been deleted are marked by this character.
 376 See also `mh-scan-deleted-msg-regexp'.")
 377
 378 (defvar mh-note-dist ?R
 379   "Messages that have been redistributed are marked by this character.")
 380
 381 (defvar mh-note-forw ?F
 382   "Messages that have been forwarded are marked by this character.")
 383
 384 (defvar mh-note-printed ?P
 385   "Messages that have been printed are marked by this character.")
 386
 387 (defvar mh-note-refiled ?^
 388   "Messages that have been refiled are marked by this character.
 389 See also `mh-scan-refiled-msg-regexp'.")
 390
 391 (defvar mh-note-repl ?-
 392   "Messages that have been replied to are marked by this character.")
 393
 394 (defvar mh-note-seq ?%
 395   "Messages in a user-defined sequence are marked by this character.
 396
 397 Messages in the \"search\" sequence are marked by this character as
 398 well.")
 399
 400 \f
 401
 402 ;;; Utilities
 403
 404 ;;;###mh-autoload
 405 (defun mh-scan-msg-number-regexp ()
 406   "Return value of variable `mh-scan-msg-number-regexp'."
 407   mh-scan-msg-number-regexp)
 408
 409 ;;;###mh-autoload
 410 (defun mh-scan-msg-search-regexp ()
 411   "Return value of variable `mh-scan-msg-search-regexp'."
 412   mh-scan-msg-search-regexp)
 413
 414 ;;;###mh-autoload
 415 (defun mh-set-cmd-note (column)
 416   "Set `mh-cmd-note' to COLUMN.
 417 Note that columns in Emacs start with 0."
 418   (setq mh-cmd-note column))
 419
 420 ;;;###mh-autoload
 421 (defun mh-scan-format ()
 422   "Return the output format argument for the scan program."
 423   (if (equal mh-scan-format-file t)
 424       (list "-format" (if (mh-variant-p 'nmh 'mu-mh)
 425                           (list (mh-update-scan-format
 426                                  mh-scan-format-nmh mh-cmd-note))
 427                         (list (mh-update-scan-format
 428                                mh-scan-format-mh mh-cmd-note))))
 429     (if (not (equal mh-scan-format-file nil))
 430         (list "-form" mh-scan-format-file))))
 431
 432 (defun mh-update-scan-format (fmt width)
 433   "Return a scan format with the (msg) width in the FMT replaced with WIDTH.
 434
 435 The message number width portion of the format is discovered
 436 using `mh-scan-msg-format-regexp'. Its replacement is controlled
 437 with `mh-scan-msg-format-string'."
 438   (or (and
 439        (string-match mh-scan-msg-format-regexp fmt)
 440        (let ((begin (match-beginning 1))
 441              (end (match-end 1)))
 442          (concat (substring fmt 0 begin)
 443                  (format mh-scan-msg-format-string width)
 444                  (substring fmt end))))
 445       fmt))
 446
 447 ;;;###mh-autoload
 448 (defun mh-msg-num-width (folder)
 449   "Return the width of the largest message number in this FOLDER."
 450   (or mh-progs (mh-find-path))
 451   (let ((tmp-buffer (get-buffer-create mh-temp-buffer))
 452         (width 0))
 453     (save-excursion
 454       (set-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