lisp/mh-e/mh-scan.el

   1 ;;; mh-scan.el --- MH-E scan line constants and utilities
   2
   3 ;; Copyright (C) 1993, 1995, 1997, 2000-2012  Free Software Foundation, Inc.
   4
   5 ;; Author: Bill Wohler <wohler@newt.com>
   6 ;; Maintainer: Bill Wohler <wohler@newt.com>
   7 ;; Keywords: mail
   8 ;; See: mh-e.el
   9
  10 ;; This file is part of GNU Emacs.
  11
  12 ;; GNU Emacs is free software: you can redistribute it and/or modify
  13 ;; it under the terms of the GNU General Public License as published by
  14 ;; the Free Software Foundation, either version 3 of the License, or
  15 ;; (at your option) any later version.
  16
  17 ;; GNU Emacs is distributed in the hope that it will be useful,
  18 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  19 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  20 ;; GNU General Public License for more details.
  21
  22 ;; You should have received a copy of the GNU General Public License
  23 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  24
  25 ;;; Commentary:
  26
  27 ;; This file contains constants and a few functions for interpreting
  28 ;; scan lines.
  29
  30 ;;; Change Log:
  31
  32 ;;; Code:
  33
  34 (require 'mh-e)
  35
  36 \f
  37
  38 ;;; Scan Formats
  39
  40 ;; The following scan formats are passed to the scan program if the setting of
  41 ;; `mh-scan-format-file' is t. They are identical except the later one makes
  42 ;; use of the nmh `decode' function to decode RFC 2047 encodings. If you just
  43 ;; want to change the column of the notations, use the `mh-set-cmd-note'
  44 ;; function.
  45
  46 (defvar mh-scan-format-mh
  47   (concat
  48    "%4(msg)"
  49    "%<(cur)+%| %>"
  50    "%<{replied}-"
  51    "%?(nonnull(comp{to}))%<(mymbox{to})t%>"
  52    "%?(nonnull(comp{cc}))%<(mymbox{cc})c%>"
  53    "%?(nonnull(comp{bcc}))%<(mymbox{bcc})b%>"
  54    "%?(nonnull(comp{newsgroups}))n%>"
  55    "%<(zero) %>"
  56    "%02(mon{date})/%02(mday{date})%<{date} %|*%>"
  57    "%<(mymbox{from})%<{to}To:%14(friendly{to})%>%>"
  58    "%<(zero)%17(friendly{from})%>  "
  59    "%{subject}%<{body}<<%{body}%>")
  60   "*Scan format string for MH.
  61 This string is passed to the scan program via the -format
  62 argument.  This format is identical to the default except that
  63 additional hints for fontification have been added to the fifth
  64 column (remember that in Emacs, the first column is 0).
  65
  66 The values of the fifth column, in priority order, are: \"-\" if
  67 the message has been replied to, t if an address on the To: line
  68 matches one of the mailboxes of the current user, \"c\" if the Cc:
  69 line matches, \"b\" if the Bcc: line matches, and \"n\" if a
  70 non-empty Newsgroups: header is present.")
  71
  72 (defvar mh-scan-format-nmh
  73   (concat
  74    "%4(msg)"
  75    "%<(cur)+%| %>"
  76    "%<{replied}-"
  77    "%?(nonnull(comp{to}))%<(mymbox{to})t%>"
  78    "%?(nonnull(comp{cc}))%<(mymbox{cc})c%>"
  79    "%?(nonnull(comp{bcc}))%<(mymbox{bcc})b%>"
  80    "%?(nonnull(comp{newsgroups}))n%>"
  81    "%<(zero) %>"
  82    "%02(mon{date})/%02(mday{date})%<{date} %|*%>"
  83    "%<(mymbox{from})%<{to}To:%14(decode(friendly{to}))%>%>"
  84    "%<(zero)%17(decode(friendly{from}))%>  "
  85    "%(decode{subject})%<{body}<<%{body}%>")
  86   "*Scan format string for nmh.
  87 This string is passed to the scan program via the -format arg.
  88 This format is identical to the default except that additional
  89 hints for fontification have been added to the fifth
  90 column (remember that in Emacs, the first column is 0).
  91
  92 The values of the fifth column, in priority order, are: \"-\" if
  93 the message has been replied to, t if an address on the To: field
  94 matches one of the mailboxes of the current user, \"c\" if the Cc:
  95 field matches, \"b\" if the Bcc: field matches, and \"n\" if a
  96 non-empty Newsgroups: field is present.")
  97
  98 \f
  99
 100 ;;; Regular Expressions
 101
 102 ;; Alphabetical.
 103
 104 (defvar mh-scan-body-regexp "\\(<<\\([^\n]+\\)?\\)"
 105   "This regular expression matches the message body fragment.
 106
 107 Note that the default setting of `mh-folder-font-lock-keywords'
 108 expects this expression to contain at least one parenthesized
 109 expression which matches the body text as in the default of
 110 \"\\\\(<<\\\\([^\\n]+\\\\)?\\\\)\".  If this regular expression is
 111 not correct, the body fragment will not be highlighted with the
 112 face `mh-folder-body'.")
 113
 114 (defvar mh-scan-blacklisted-msg-regexp "^\\( *[0-9]+\\)B"
 115   "This regular expression matches blacklisted (spam) messages.
 116
 117 It must match from the beginning of the line. Note that the
 118 default setting of `mh-folder-font-lock-keywords' expects this
 119 expression to contain at least one parenthesized expression which
 120 matches the message number as in the default of
 121
 122   \"^\\\\( *[0-9]+\\\\)B\".
 123
 124 This expression includes the leading space within parenthesis
 125 since it looks better to highlight it as well. The highlighting
 126 is done with the face `mh-folder-blacklisted'. This regular
 127 expression should be correct as it is needed by non-fontification
 128 functions. See also `mh-note-blacklisted'.")
 129
 130 (defvar mh-scan-cur-msg-number-regexp "^\\( *[0-9]+\\+\\).*"
 131   "This regular expression matches the current message.
 132
 133 It must match from the beginning of the line.  Note that the
 134 default setting of `mh-folder-font-lock-keywords' expects this
 135 expression to contain at least one parenthesized expression which
 136 matches the message number as in the default of
 137
 138   \"^\\\\( *[0-9]+\\\\+\\\\).*\".
 139
 140 This expression includes the leading space and current message
 141 marker \"+\" within the parenthesis since it looks better to
 142 highlight these items as well.  The highlighting is done with the
 143 face `mh-folder-cur-msg-number'.  This regular expression should
 144 be correct as it is needed by non-fontification functions.  See
 145 also `mh-note-cur'.")
 146
 147 (defvar mh-scan-date-regexp "\\([0-9][0-9]/[0-9][0-9]\\)"
 148   "This regular expression matches a valid date.
 149
 150 It must not be anchored to the beginning or the end of the line.
 151 Note that the default setting of `mh-folder-font-lock-keywords'
 152 expects this expression to contain only one parenthesized
 153 expression which matches the date field as in the default of
 154 \"\\\\([0-9][0-9]/[0-9][0-9]\\\\)\"}.  If this regular expression
 155 is not correct, the date will not be highlighted with the face
 156 `mh-folder-date'.")
 157
 158 (defvar mh-scan-deleted-msg-regexp "^\\( *[0-9]+\\)D"
 159   "This regular expression matches deleted messages.
 160
 161 It must match from the beginning of the line.  Note that the
 162 default setting of `mh-folder-font-lock-keywords' expects this
 163 expression to contain at least one parenthesized expression which
 164 matches the message number as in the default of
 165
 166   \"^\\\\( *[0-9]+\\\\)D\".
 167
 168 This expression includes the leading space within the parenthesis
 169 since it looks better to highlight it as well.  The highlighting
 170 is done with the face `mh-folder-deleted'.  This regular
 171 expression should be correct as it is needed by non-fontification
 172 functions.  See also `mh-note-deleted'.")
 173
 174 (defvar mh-scan-good-msg-regexp  "^\\( *[0-9]+\\)[^^DBW0-9]"
 175   "This regular expression matches \"good\" messages.
 176
 177 It must match from the beginning of the line.  Note that the
 178 default setting of `mh-folder-font-lock-keywords' expects this
 179 expression to contain at least one parenthesized expression which
 180 matches the message number as in the default of
 181
 182   \"^\\\\( *[0-9]+\\\\)[^^DBW0-9]\".
 183
 184 This expression includes the leading space within the parenthesis
 185 since it looks better to highlight it as well.  The highlighting
 186 is done with the face `mh-folder-msg-number'.  This regular
 187 expression should be correct as it is needed by non-fontification
 188 functions.")
 189
 190 (defvar mh-scan-msg-format-regexp "%\\([0-9]*\\)(msg)"
 191   "This regular expression finds the message number width in a scan format.
 192
 193 Note that the message number must be placed in a parenthesized
 194 expression as in the default of \"%\\\\([0-9]*\\\\)(msg)\".  This
 195 variable is only consulted if `mh-scan-format-file' is set to
 196 \"Use MH-E scan Format\".")
 197
 198 (defvar mh-scan-msg-format-string "%d"
 199   "This is a format string for width of the message number in a scan format.
 200
 201 Use \"0%d\" for zero-filled message numbers.  This variable is only
 202 consulted if `mh-scan-format-file' is set to \"Use MH-E scan
 203 Format\".")
 204
 205 (defvar mh-scan-msg-number-regexp "^ *\\([0-9]+\\)"
 206   "This regular expression extracts the message number.
 207
 208 It must match from the beginning of the line.  Note that the
 209 message number must be placed in a parenthesized expression as in
 210 the default of \"^ *\\\\([0-9]+\\\\)\".")
 211
 212 (defvar mh-scan-msg-overflow-regexp "^[?0-9][0-9]"
 213   "This regular expression matches overflowed message numbers.")
 214
 215 (defvar mh-scan-msg-search-regexp "^[^0-9]*%d[^0-9]"
 216   "This regular expression matches a particular message.
 217
 218 It is a format string; use \"%d\" to represent the location of the
 219 message number within the expression as in the default of
 220 \"^[^0-9]*%d[^0-9]\".")
 221
 222 (defvar mh-scan-rcpt-regexp  "\\(To:\\)\\(..............\\)"
 223   "This regular expression specifies the recipient in messages you sent.
 224
 225 Note that the default setting of `mh-folder-font-lock-keywords'
 226 expects this expression to contain two parenthesized expressions.
 227 The first is expected to match the \"To:\" that the default scan
 228 format file generates.  The second is expected to match the
 229 recipient's name as in the default of
 230 \"\\\\(To:\\\\)\\\\(..............\\\\)\".  If this regular
 231 expression is not correct, the \"To:\" string will not be
 232 highlighted with the face `mh-folder-to' and the recipient will
 233 not be highlighted with the face `mh-folder-address'")
 234
 235 (defvar mh-scan-refiled-msg-regexp  "^\\( *[0-9]+\\)\\^"
 236   "This regular expression matches refiled messages.
 237
 238 It must match from the beginning of the line.  Note that the
 239 default setting of `mh-folder-font-lock-keywords' expects this
 240 expression to contain at least one parenthesized expression which
 241 matches the message number as in the default of
 242
 243   \"^\\\\( *[0-9]+\\\\)\\\\^\".
 244
 245 This expression includes the leading space within the parenthesis
 246 since it looks better to highlight it as well.  The highlighting
 247 is done with the face `mh-folder-refiled'.  This regular
 248 expression should be correct as it is needed by non-fontification
 249 functions.  See also `mh-note-refiled'.")
 250
 251 (defvar mh-scan-sent-to-me-sender-regexp
 252   "^ *[0-9]+.\\([bct]\\).....[ ]*\\(..................\\)"
 253   "This regular expression matches messages sent to us.
 254
 255 Note that the default setting of `mh-folder-font-lock-keywords'
 256 expects this expression to contain at least two parenthesized
 257 expressions.  The first should match the fontification hint (see
 258 `mh-scan-format-nmh') and the second should match the user name
 259 as in the default of
 260
 261   ^ *[0-9]+.\\\\([bct]\\\\).....[ ]*\\\\(..................\\\\)
 262
 263 If this regular expression is not correct, the notation hints
 264 will not be highlighted with the face
 265 `mh-mh-folder-sent-to-me-hint' and the sender will not be
 266 highlighted with the face `mh-folder-sent-to-me-sender'.")
 267
 268 (defvar mh-scan-subject-regexp
 269   "^ *[0-9]+........[ ]*...................\\([Rr][Ee]\\(\\[[0-9]+\\]\\)?:\\s-*\\)*\\([^<\n]*\\)"
 270   "This regular expression matches the subject.
 271
 272 It must match from the beginning of the line.  Note that the
 273 default setting of `mh-folder-font-lock-keywords' expects this
 274 expression to contain at least three parenthesized expressions.
 275 The first is expected to match the \"Re:\" string, if any, and is
 276 highlighted with the face `mh-folder-followup'.  The second
 277 matches an optional bracketed number after \"Re:\", such as in
 278 \"Re[2]:\" (and is thus a sub-expression of the first expression)
 279 and the third is expected to match the subject line itself which
 280 is highlighted with the face `mh-folder-subject'.  For example,
 281 the default (broken on multiple lines for readability) is
 282
 283   ^ *[0-9]+........[ ]*...................
 284   \\\\([Rr][Ee]\\\\(\\\\\\=[[0-9]+\\\\]\\\\)?:\\\\s-*\\\\)*
 285   \\\\([^<\\n]*\\\\)
 286
 287 This regular expression should be correct as it is needed by
 288 non-fontification functions.")
 289
 290 (defvar mh-scan-valid-regexp "^ *[0-9]"
 291   "This regular expression describes a valid scan line.
 292
 293 This is used to eliminate error messages that are occasionally
 294 produced by \"inc\".")
 295
 296 (defvar mh-scan-whitelisted-msg-regexp "^\\( *[0-9]+\\)W"
 297   "This regular expression matches whitelisted (non-spam) messages.
 298
 299 It must match from the beginning of the line. Note that the
 300 default setting of `mh-folder-font-lock-keywords' expects this
 301 expression to contain at least one parenthesized expression which
 302 matches the message number as in the default of
 303
 304   \"^\\\\( *[0-9]+\\\\)W\".
 305
 306 This expression includes the leading space within parenthesis
 307 since it looks better to highlight it as well. The highlighting
 308 is done with the face `mh-folder-whitelisted'. This regular
 309 expression should be correct as it is needed by non-fontification
 310 functions. See also `mh-note-whitelisted'.")
 311
 312 \f
 313
 314 ;;; Widths, Offsets and Columns
 315
 316 (defvar mh-cmd-note 4
 317   "Column for notations.
 318
 319 This variable should be set with the function `mh-set-cmd-note'.
 320 This variable may be updated dynamically if
 321 `mh-adaptive-cmd-note-flag' is on.
 322
 323 Note that columns in Emacs start with 0.")
 324 (make-variable-buffer-local 'mh-cmd-note)
 325
 326 (defvar mh-scan-cmd-note-width 1
 327   "Number of columns consumed by the cmd-note field in `mh-scan-format'.
 328
 329 This column will have one of the values: \" \", \"^\", \"D\", \"B\", \"W\", \"+\", where
 330
 331   \" \" is the default value,
 332   \"^\" is the `mh-note-refiled' character,
 333   \"D\" is the `mh-note-deleted' character,
 334   \"B\" is the `mh-note-blacklisted' character,
 335   \"W\" is the `mh-note-whitelisted' character, and
 336   \"+\" is the `mh-note-cur' character.")
 337
 338 (defvar mh-scan-destination-width 1
 339   "Number of columns consumed by the destination field in `mh-scan-format'.
 340
 341 This column will have one of \" \", \"%\", \"-\", \"t\", \"c\", \"b\", or \"n\"
 342 in it.
 343
 344   \" \" blank space is the default character.
 345   \"%\" indicates that the message in a named MH sequence.
 346   \"-\" indicates that the message has been annotated with a replied field.
 347   \"t\" indicates that the message contains mymbox in the To: field.
 348   \"c\" indicates that the message contains mymbox in the Cc: field.
 349   \"b\" indicates that the message contains mymbox in the Bcc: field.
 350   \"n\" indicates that the message contains a Newsgroups: field.")
 351
 352 (defvar mh-scan-date-width 5
 353   "Number of columns consumed by the date field in `mh-scan-format'.
 354 This column will typically be of the form mm/dd.")
 355
 356 (defvar mh-scan-date-flag-width 1
 357   "Number of columns consumed to flag (in)valid dates in `mh-scan-format'.
 358 This column will have \" \" for valid and \"*\" for invalid or
 359 missing dates.")
 360
 361 (defvar mh-scan-from-mbox-width 17
 362   "Number of columns consumed with the \"From:\" line in `mh-scan-format'.
 363 This column will have a friendly name or e-mail address of the
 364 originator, or a \"To: address\" for outgoing e-mail messages.")
 365
 366 (defvar mh-scan-from-mbox-sep-width 2
 367   "Number of columns consumed by whitespace after from-mbox in `mh-scan-format'.
 368 This column will only ever have spaces in it.")
 369
 370 (defvar mh-scan-field-destination-offset
 371   (+ mh-scan-cmd-note-width)
 372   "The offset from the `mh-cmd-note' for the destination column.")
 373
 374 (defvar mh-scan-field-from-start-offset
 375   (+ mh-scan-cmd-note-width
 376      mh-scan-destination-width
 377      mh-scan-date-width
 378      mh-scan-date-flag-width)
 379   "The offset from the `mh-cmd-note' to find the start of \"From:\" address.")
 380
 381 (defvar mh-scan-field-from-end-offset
 382   (+ mh-scan-field-from-start-offset mh-scan-from-mbox-width)
 383   "The offset from the `mh-cmd-note' to find the end of \"From:\" address.")
 384
 385 (defvar mh-scan-field-subject-start-offset
 386   (+ mh-scan-cmd-note-width
 387      mh-scan-destination-width
 388      mh-scan-date-width
 389      mh-scan-date-flag-width
 390      mh-scan-from-mbox-width
 391      mh-scan-from-mbox-sep-width)
 392   "The offset from the `mh-cmd-note' to find the start of the subject.")
 393
 394 \f
 395
 396 ;;; Notation
 397
 398 ;; Alphabetical.
 399
 400 (defvar mh-note-blacklisted ?B
 401   "Messages that have been blacklisted are marked by this character.
 402 See also `mh-scan-blacklisted-msg-regexp'.")
 403
 404 (defvar mh-note-cur ?+
 405   "The current message (in MH, not in MH-E) is marked by this character.
 406 See also `mh-scan-cur-msg-number-regexp'.")
 407
 408 (defvar mh-note-copied ?C
 409   "Messages that have been copied are marked by this character.")
 410
 411 (defvar mh-note-deleted ?D
 412   "Messages that have been deleted are marked by this character.
 413 See also `mh-scan-deleted-msg-regexp'.")
 414
 415 (defvar mh-note-dist ?R
 416   "Messages that have been redistributed are marked by this character.")
 417
 418 (defvar mh-note-forw ?F
 419   "Messages that have been forwarded are marked by this character.")
 420
 421 (defvar mh-note-printed ?P
 422   "Messages that have been printed are marked by this character.")
 423
 424 (defvar mh-note-refiled ?^
 425   "Messages that have been refiled are marked by this character.
 426 See also `mh-scan-refiled-msg-regexp'.")
 427
 428 (defvar mh-note-repl ?-
 429   "Messages that have been replied to are marked by this character.")
 430
 431 (defvar mh-note-seq ?%
 432   "Messages in a user-defined sequence are marked by this character.
 433
 434 Messages in the \"search\" sequence are marked by this character as
 435 well.")
 436
 437 (defvar mh-note-whitelisted ?W
 438   "Messages that have been whitelisted are marked by this character.
 439 See also `mh-scan-whitelisted-msg-regexp'.")
 440
 441 \f
 442
 443 ;;; Utilities
 444
 445 ;;;###mh-autoload
 446 (defun mh-scan-msg-number-regexp ()
 447   "Return value of variable `mh-scan-msg-number-regexp'."
 448   mh-scan-msg-number-regexp)
 449
 450 ;;;###mh-autoload
 451 (defun mh-scan-msg-search-regexp ()
 452   "Return value of variable `mh-scan-msg-search-regexp'."
 453   mh-scan-msg-search-regexp)
 454
 455 ;;;###mh-autoload
 456 (defun mh-set-cmd-note (column)
 457   "Set `mh-cmd-note' to COLUMN.
 458 Note that columns in Emacs start with 0."
 459   (setq mh-cmd-note column))
 460
 461 ;;;###mh-autoload
 462 (defun mh-scan-format ()
 463   "Return the output format argument for the scan program."
 464   (if (equal mh-scan-format-file t)
 465       (list "-format" (if (mh-variant-p 'nmh 'gnu-mh)
 466                           (list (mh-update-scan-format
 467                                  mh-scan-format-nmh mh-cmd-note))
 468                         (list (mh-update-scan-format
 469                                mh-scan-format-mh mh-cmd-note))))
 470     (if (not (equal mh-scan-format-file nil))
 471         (list "-form" mh-scan-format-file))))
 472
 473 (defun mh-update-scan-format (fmt width)
 474   "Return a scan format with the (msg) width in the FMT replaced with WIDTH.
 475
 476 The message number width portion of the format is discovered
 477 using `mh-scan-msg-format-regexp'.  Its replacement is controlled
 478 with `mh-scan-msg-format-string'."
 479   (or (and
 480        (string-match mh-scan-msg-format-regexp fmt)
 481        (let ((begin (match-beginning 1))
 482              (end (match-end 1)))
 483          (concat (substring fmt 0 begin)
 484                  (format mh-scan-msg-format-string width)
 485                  (substring fmt end))))
 486       fmt))
 487
 488 ;;;###mh-autoload
 489 (defun mh-msg-num-width (folder)
 490   "Return the width of the largest message number in this FOLDER."
 491   (or mh-progs (mh-find-path))
 492   (let ((tmp-buffer (get-buffer-create mh-temp-buffer))
 493         (width 0))
 494     (with-current-buffer tmp-buffer
 495       (erase-buffer)
 496       (apply 'call-process
 497              (expand-file-name mh-scan-prog mh-progs) nil '(t nil) nil
 498              (list folder "last" "-format" "%(msg)"))
 499       (goto-char (point-min))
 500       (if (re-search-forward mh-scan-msg-number-regexp nil 0 1)
 501           (setq width (length (buffer-substring
 502                                (match-beginning 1) (match-end 1))))))
 503     width))
 504
 505 ;;;###mh-autoload
 506 (defun mh-msg-num-width-to-column (width)
 507   "Return the column for notations given message number WIDTH.
 508 Note that columns in Emacs start with 0.
 509
 510 If `mh-scan-format-file' is set to \"Use MH-E scan Format\" this
 511 means that either `mh-scan-format-mh' or `mh-scan-format-nmh' are
 512 in use.  This function therefore assumes that the first column is
 513 empty (to provide room for the cursor), the following WIDTH
 514 columns contain the message number, and the column for notations
 515 comes after that."
 516   (if (eq mh-scan-format-file t)
 517       (max (1+ width) 2)
 518     (error "%s %s" "Can't call `mh-msg-num-width-to-column' when"
 519            "`mh-scan-format-file' is not set to \"Use MH-E scan Format\"")))
 520
 521 (provide 'mh-scan)
 522
 523 ;; Local Variables:
 524 ;; indent-tabs-mode: nil
 525 ;; sentence-end-double-space: nil
 526 ;; End:
 527
 528 ;;; mh-scan.el ends here