| 1 | /// |
| 2 | $Cambridge: exim/doc/doc-docbook/filter.ascd,v 1.2 2005/11/10 12:30:13 ph10 Exp $ |
| 3 | |
| 4 | This file contains the Asciidoc source for the document that describes Exim's |
| 5 | filtering facilities from a user's point of view. See the file AdMarkup.txt for |
| 6 | an explanation of the markup that is used. It is more or less standard |
| 7 | Asciidoc, but with a few changes and additions. |
| 8 | /// |
| 9 | |
| 10 | |
| 11 | /// |
| 12 | This preliminary stuff creates a <bookinfo> entry in the XML. This is removed |
| 13 | when creating the PostScript/PDF output, because we do not want a full-blown |
| 14 | title page created for those versions. The stylesheet fudges up a title line to |
| 15 | replace the text "Table of contents". However, for the other forms of output, |
| 16 | the <bookinfo> element is retained and used. |
| 17 | /// |
| 18 | |
| 19 | Exim's interfaces to mail filtering |
| 20 | =================================== |
| 21 | :author: Philip Hazel |
| 22 | :copyright: University of Cambridge |
| 23 | :cpyear: 2005 |
| 24 | :date: 06 October 2005 |
| 25 | :doctitleabbrev: Exim filtering |
| 26 | :revision: 4.60 |
| 27 | |
| 28 | |
| 29 | ////////////////////////////////////////////////////////////////////////////// |
| 30 | ***WARNING*** Do not put anything, not even a titleabbrev setting, before |
| 31 | the first chapter (luckily it does not need one) because if you do, AsciiDoc |
| 32 | creates an empty <preface> element, which we do not want. |
| 33 | ////////////////////////////////////////////////////////////////////////////// |
| 34 | |
| 35 | |
| 36 | Forwarding and filtering in Exim |
| 37 | -------------------------------- |
| 38 | |
| 39 | This document describes the user interfaces to Exim's in-built mail filtering |
| 40 | facilities, and is copyright (C) University of Cambridge 2005. It corresponds |
| 41 | to Exim version 4.60. |
| 42 | |
| 43 | |
| 44 | |
| 45 | Introduction |
| 46 | ~~~~~~~~~~~~ |
| 47 | Most Unix mail transfer agents (programs that deliver mail) permit individual |
| 48 | users to specify automatic forwarding of their mail, usually by placing a list |
| 49 | of forwarding addresses in a file called '.forward' in their home directories. |
| 50 | Exim extends this facility by allowing the forwarding instructions to be a set |
| 51 | of rules rather than just a list of addresses, in effect providing ``'.forward' |
| 52 | with conditions''. Operating the set of rules is called 'filtering', and the |
| 53 | file that contains them is called a 'filter file'. |
| 54 | |
| 55 | Exim supports two different kinds of filter file. An 'Exim filter' contains |
| 56 | instructions in a format that is unique to Exim. A 'Sieve filter' contains |
| 57 | instructions in the Sieve format that is defined by RFC 3028. As this is a |
| 58 | standard format, Sieve filter files may already be familiar to some users. |
| 59 | Sieve files should also be portable between different environments. However, |
| 60 | the Exim filtering facility contains more features (such as variable |
| 61 | expansion), and better integration with the host environment (such as the use |
| 62 | of external processes and pipes). |
| 63 | |
| 64 | The choice of which kind of filter to use can be left to the end-user, provided |
| 65 | that the system administrator has configured Exim appropriately for both kinds |
| 66 | of filter. However, if interoperability is important, Sieve is the only |
| 67 | choice. |
| 68 | |
| 69 | The ability to use filtering or traditional forwarding has to be enabled by the |
| 70 | system administrator, and some of the individual facilities can be separately |
| 71 | enabled or disabled. A local document should be provided to describe exactly |
| 72 | what has been enabled. In the absence of this, consult your system |
| 73 | administrator. |
| 74 | |
| 75 | This document describes how to use a filter file and the format of its |
| 76 | contents. It is intended for use by end-users. Both Sieve filters and Exim |
| 77 | filters are covered. However, for Sieve filters, only issues that relate to the |
| 78 | Exim implementation are discussed, since Sieve itself is described elsewhere. |
| 79 | |
| 80 | The contents of traditional '.forward' files are not described here. They |
| 81 | normally contain just a list of addresses, file names, or pipe commands, |
| 82 | separated by commas or newlines, but other types of item are also available. |
| 83 | The full details can be found in the chapter on the ^redirect^ router in the |
| 84 | Exim specification, which also describes how the system administrator can set |
| 85 | up and control the use of filtering. |
| 86 | |
| 87 | |
| 88 | |
| 89 | Filter operation |
| 90 | ~~~~~~~~~~~~~~~~ |
| 91 | It is important to realize that, in Exim, no deliveries are actually made while |
| 92 | a filter or traditional '.forward' file is being processed. Running a filter |
| 93 | or processing a traditional '.forward' file sets up future delivery |
| 94 | operations, but does not carry them out. |
| 95 | |
| 96 | The result of filter or '.forward' file processing is a list of destinations |
| 97 | to which a message should be delivered. The deliveries themselves take place |
| 98 | later, along with all other deliveries for the message. This means that it is |
| 99 | not possible to test for successful deliveries while filtering. It also means |
| 100 | that any duplicate addresses that are generated are dropped, because Exim never |
| 101 | delivers the same message to the same address more than once. |
| 102 | |
| 103 | |
| 104 | |
| 105 | |
| 106 | [[SECTtesting]] |
| 107 | Testing a new filter file |
| 108 | ~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 109 | Filter files, especially the more complicated ones, should always be tested, as |
| 110 | it is easy to make mistakes. Exim provides a facility for preliminary testing |
| 111 | of a filter file before installing it. This tests the syntax of the file and |
| 112 | its basic operation, and can also be used with traditional '.forward' files. |
| 113 | |
| 114 | Because a filter can do tests on the content of messages, a test message is |
| 115 | required. Suppose you have a new filter file called 'myfilter' and a test |
| 116 | message called 'test-message'. Assuming that Exim is installed with the |
| 117 | conventional path name '/usr/sbin/sendmail' (some operating systems use |
| 118 | '/usr/lib/sendmail'), the following command can be used: |
| 119 | |
| 120 | /usr/sbin/sendmail -bf myfilter <test-message |
| 121 | |
| 122 | The %-bf% option tells Exim that the following item on the command line is the |
| 123 | name of a filter file that is to be tested. There is also a %-bF% option, |
| 124 | which is similar, but which is used for testing system filter files, as opposed |
| 125 | to user filter files, and which is therefore of use only to the system |
| 126 | administrator. |
| 127 | |
| 128 | The test message is supplied on the standard input. If there are no |
| 129 | message-dependent tests in the filter, an empty file ('/dev/null') can be |
| 130 | used. A supplied message must start with header lines or the ``From'' message |
| 131 | separator line which is found in many multi-message folder files. Note that |
| 132 | blank lines at the start terminate the header lines. A warning is given if no |
| 133 | header lines are read. |
| 134 | |
| 135 | The result of running this command, provided no errors are detected in the |
| 136 | filter file, is a list of the actions that Exim would try to take if presented |
| 137 | with the message for real. |
| 138 | For example, for an Exim filter, the output |
| 139 | |
| 140 | Deliver message to: gulliver@lilliput.fict.example |
| 141 | Save message to: /home/lemuel/mail/archive |
| 142 | |
| 143 | means that one copy of the message would be sent to |
| 144 | 'gulliver@lilliput.fict.example', and another would be added to the file |
| 145 | _/home/lemuel/mail/archive_, if all went well. |
| 146 | |
| 147 | The actions themselves are not attempted while testing a filter file in this |
| 148 | way; there is no check, for example, that any forwarding addresses are valid. |
| 149 | For an Exim filter, |
| 150 | if you want to know why a particular action is being taken, add the %-v% |
| 151 | option to the command. This causes Exim to output the results of any |
| 152 | conditional tests and to indent its output according to the depth of nesting of |
| 153 | ^if^ commands. Further additional output from a filter test can be generated |
| 154 | by the ^testprint^ command, which is described below. |
| 155 | |
| 156 | When Exim is outputting a list of the actions it would take, if any text |
| 157 | strings are included in the output, non-printing characters therein are |
| 158 | converted to escape sequences. In particular, if any text string contains a |
| 159 | newline character, this is shown as ``\n'' in the testing output. |
| 160 | |
| 161 | When testing a filter in this way, Exim makes up an ``envelope'' for the message. |
| 162 | The recipient is by default the user running the command, and so is the sender, |
| 163 | but the command can be run with the %-f% option to supply a different sender. |
| 164 | For example, |
| 165 | |
| 166 | .... |
| 167 | /usr/sbin/sendmail -bf myfilter \ |
| 168 | -f islington@never.where <test-message |
| 169 | .... |
| 170 | |
| 171 | Alternatively, if the %-f% option is not used, but the first line of the |
| 172 | supplied message is a ``From'' separator from a message folder file (not the same |
| 173 | thing as a 'From:' header line), the sender is taken from there. If %-f% is |
| 174 | present, the contents of any ``From'' line are ignored. |
| 175 | |
| 176 | The ``return path'' is the same as the envelope sender, unless the message |
| 177 | contains a 'Return-path:' header, in which case it is taken from there. You |
| 178 | need not worry about any of this unless you want to test out features of a |
| 179 | filter file that rely on the sender address or the return path. |
| 180 | |
| 181 | It is possible to change the envelope recipient by specifying further options. |
| 182 | The %-bfd% option changes the domain of the recipient address, while the |
| 183 | %-bfl% option changes the ``local part'', that is, the part before the @ sign. |
| 184 | An adviser could make use of these to test someone else's filter file. |
| 185 | |
| 186 | The %-bfp% and %-bfs% options specify the prefix or suffix for the local part. |
| 187 | These are relevant only when support for multiple personal mailboxes is |
| 188 | implemented; see the description in section <<SECTmbox>> below. |
| 189 | |
| 190 | |
| 191 | Installing a filter file |
| 192 | ~~~~~~~~~~~~~~~~~~~~~~~~ |
| 193 | A filter file is normally installed under the name '.forward' in your home |
| 194 | directory -- it is distinguished from a conventional '.forward' file by its |
| 195 | first line (described below). However, the file name is configurable, and some |
| 196 | system administrators may choose to use some different name or location for |
| 197 | filter files. |
| 198 | |
| 199 | |
| 200 | Testing an installed filter file |
| 201 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 202 | Testing a filter file before installation cannot find every potential problem; |
| 203 | for example, it does not actually run commands to which messages are piped. |
| 204 | Some ``live'' tests should therefore also be done once a filter is installed. |
| 205 | |
| 206 | If at all possible, test your filter file by sending messages from some other |
| 207 | account. If you send a message to yourself from the filtered account, and |
| 208 | delivery fails, the error message will be sent back to the same account, which |
| 209 | may cause another delivery failure. It won't cause an infinite sequence of such |
| 210 | messages, because delivery failure messages do not themselves generate further |
| 211 | messages. However, it does mean that the failure won't be returned to you, and |
| 212 | also that the postmaster will have to investigate the stuck message. |
| 213 | |
| 214 | If you have to test an Exim filter from the same account, a sensible precaution |
| 215 | is to include the line |
| 216 | |
| 217 | if error_message then finish endif |
| 218 | |
| 219 | as the first filter command, at least while testing. This causes filtering to |
| 220 | be abandoned for a delivery failure message, and since no destinations are |
| 221 | generated, the message goes on to be delivered to the original address. Unless |
| 222 | there is a good reason for not doing so, it is recommended that the above test |
| 223 | be left in all Exim filter files. |
| 224 | (This does not apply to Sieve files.) |
| 225 | |
| 226 | |
| 227 | |
| 228 | Details of filtering commands |
| 229 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 230 | The filtering commands for Sieve and Exim filters are completely different in |
| 231 | syntax and semantics. The Sieve mechanism is defined in RFC 3028; in the next |
| 232 | chapter we describe how it is integrated into Exim. The subsequent chapter |
| 233 | covers Exim filtering commands in detail. |
| 234 | |
| 235 | |
| 236 | |
| 237 | [[CHAPsievefilter]] |
| 238 | Sieve filter files |
| 239 | ------------------ |
| 240 | The code for Sieve filtering in Exim was contributed by Michael Haardt, and |
| 241 | most of the content of this chapter is taken from the notes he provided. Since |
| 242 | Sieve is an extensible language, it is important to understand ``Sieve'' in |
| 243 | this context as ``the specific implementation of Sieve for Exim''. |
| 244 | |
| 245 | This chapter does not contain a description of Sieve, since that can be found |
| 246 | in RFC 3028, which should be read in conjunction with these notes. |
| 247 | |
| 248 | The Exim Sieve implementation offers the core as defined by RFC 3028, |
| 249 | comparison tests, the *copy*, *envelope*, *fileinto*, and *vacation* |
| 250 | extensions, but not the *reject* extension. Exim does not support message |
| 251 | delivery notifications (MDNs), so adding it just to the Sieve filter (as |
| 252 | required for *reject*) makes little sense. |
| 253 | |
| 254 | In order for Sieve to work properly in Exim, the system administrator needs to |
| 255 | make some adjustments to the Exim configuration. These are described in the |
| 256 | chapter on the ^redirect^ router in the full Exim specification. |
| 257 | |
| 258 | |
| 259 | Recognition of Sieve filters |
| 260 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 261 | A filter file is interpreted as a Sieve filter if its first line is |
| 262 | |
| 263 | # Sieve filter |
| 264 | |
| 265 | This is what distinguishes it from a conventional '.forward' file or an Exim |
| 266 | filter file. |
| 267 | |
| 268 | |
| 269 | |
| 270 | Saving to specified folders |
| 271 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 272 | If the system administrator has set things up as suggested in the Exim |
| 273 | specification, and you use *keep* or *fileinto* to save a mail into a |
| 274 | folder, absolute files are stored where specified, relative files are stored |
| 275 | relative to $home$, and *inbox* goes to the standard mailbox location. |
| 276 | |
| 277 | |
| 278 | |
| 279 | Strings containing header names |
| 280 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 281 | RFC 3028 does not specify what happens if a string denoting a header field does |
| 282 | not contain a valid header name, for example, it contains a colon. This |
| 283 | implementation generates an error instead of ignoring the header field in order |
| 284 | to ease script debugging, which fits in with the common picture of Sieve. |
| 285 | |
| 286 | |
| 287 | |
| 288 | Exists test with empty list of headers |
| 289 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 290 | The *exists* test succeeds only if all the specified headers exist. RFC 3028 |
| 291 | does not explicitly specify what happens on an empty list of headers. This |
| 292 | implementation evaluates that condition as true, interpreting the RFC in a |
| 293 | strict sense. |
| 294 | |
| 295 | |
| 296 | |
| 297 | Header test with invalid MIME encoding in header |
| 298 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 299 | Some MUAs process invalid base64 encoded data, generating junk. |
| 300 | Others ignore junk after seeing an equal sign in base64 encoded data. |
| 301 | RFC 2047 does not specify how to react in this case, other than stating |
| 302 | that a client must not forbid to process a message for that reason. |
| 303 | RFC 2045 specifies that invalid data should be ignored (apparently |
| 304 | looking at end of line characters). It also specifies that invalid data |
| 305 | may lead to rejecting messages containing them (and there it appears to |
| 306 | talk about true encoding violations), which is a clear contradiction to |
| 307 | ignoring them. |
| 308 | |
| 309 | RFC 3028 does not specify how to process incorrect MIME words. |
| 310 | This implementation treats them literally, as it does if the word is |
| 311 | correct but its character set cannot be converted to UTF-8. |
| 312 | |
| 313 | |
| 314 | |
| 315 | Address test for multiple addresses per header |
| 316 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 317 | A header may contain multiple addresses. RFC 3028 does not explicitly |
| 318 | specify how to deal with them, but since the address test checks if |
| 319 | anything matches anything else, matching one address suffices to |
| 320 | satisfy the condition. That makes it impossible to test if a header |
| 321 | contains a certain set of addresses and no more, but it is more logical |
| 322 | than letting the test fail if the header contains an additional address |
| 323 | besides the one the test checks for. |
| 324 | |
| 325 | |
| 326 | |
| 327 | Semantics of keep |
| 328 | ~~~~~~~~~~~~~~~~~ |
| 329 | The *keep* command is equivalent to |
| 330 | |
| 331 | fileinto "inbox"; |
| 332 | |
| 333 | It saves the message and resets the implicit keep flag. It does not set the |
| 334 | implicit keep flag; there is no command to set it once it has been reset. |
| 335 | |
| 336 | |
| 337 | |
| 338 | Semantics of fileinto |
| 339 | ~~~~~~~~~~~~~~~~~~~~~ |
| 340 | RFC 3028 does not specify whether %fileinto% should try to create a mail folder |
| 341 | if it does not exist. This implementation allows the sysadmin to configure that |
| 342 | aspect using the ^appendfile^ transport options %create_directory%, |
| 343 | %create_file%, and %file_must_exist%. See the ^appendfile^ transport in |
| 344 | the Exim specification for details. |
| 345 | |
| 346 | |
| 347 | |
| 348 | Semantics of redirect |
| 349 | ~~~~~~~~~~~~~~~~~~~~~ |
| 350 | Sieve scripts are supposed to be interoperable between servers, so this |
| 351 | implementation does not allow mail to be redirected to unqualified addresses, |
| 352 | because the domain would depend on the system being used. On systems with |
| 353 | virtual mail domains, the default domain is probably not what the user expects |
| 354 | it to be. |
| 355 | |
| 356 | |
| 357 | |
| 358 | String arguments |
| 359 | ~~~~~~~~~~~~~~~~ |
| 360 | There has been confusion if the string arguments to *require* are to be matched |
| 361 | case-sensitively or not. This implementation matches them with the match type |
| 362 | ^:is^ (default, see section 2.7.1 of the RFC) and the comparator |
| 363 | ^i;ascii-casemap^ (default, see section 2.7.3 of the RFC). The RFC defines the |
| 364 | command defaults clearly, so any different implementations violate RFC 3028. |
| 365 | The same is valid for comparator names, also specified as strings. |
| 366 | |
| 367 | |
| 368 | |
| 369 | Number units |
| 370 | ~~~~~~~~~~~~ |
| 371 | There is a mistake in RFC 3028: the suffix G denotes gibi-, not tebibyte. |
| 372 | The mistake is obvious, because RFC 3028 specifies G to denote 2^30 |
| 373 | (which is gibi, not tebi), and that is what this implementation uses as |
| 374 | the scaling factor for the suffix G. |
| 375 | |
| 376 | |
| 377 | |
| 378 | RFC compliance |
| 379 | ~~~~~~~~~~~~~~ |
| 380 | Exim requires the first line of a Sieve filter to be |
| 381 | |
| 382 | # Sieve filter |
| 383 | |
| 384 | Of course the RFC does not specify that line. Do not expect examples to work |
| 385 | without adding it, though. |
| 386 | |
| 387 | RFC 3028 requires the use of CRLF to terminate a line. |
| 388 | The rationale was that CRLF is universally used in network protocols |
| 389 | to mark the end of the line. This implementation does not embed Sieve |
| 390 | in a network protocol, but uses Sieve scripts as part of the Exim MTA. |
| 391 | Since all parts of Exim use LF as the newline character, this implementation |
| 392 | does, too, by default, though the system administrator may choose (at Exim |
| 393 | compile time) to use CRLF instead. |
| 394 | |
| 395 | Exim violates RFC 2822, section 3.6.8, by accepting 8-bit header names, so |
| 396 | this implementation repeats this violation to stay consistent with Exim. |
| 397 | This is in preparation for UTF-8 data. |
| 398 | |
| 399 | Sieve scripts cannot contain NUL characters in strings, but mail |
| 400 | headers could contain MIME encoded NUL characters, which could never |
| 401 | be matched by Sieve scripts using exact comparisons. For that reason, |
| 402 | this implementation extends the Sieve quoted string syntax with \0 |
| 403 | to describe a NUL character, violating \0 being the same as 0 in |
| 404 | RFC 3028. Even without using \0, the following tests are all true in |
| 405 | this implementation. Implementations that use C-style strings will only |
| 406 | evaluate the first test as true. |
| 407 | |
| 408 | Subject: =?iso-8859-1?q?abc=00def |
| 409 | |
| 410 | header :contains "Subject" ["abc"] |
| 411 | header :contains "Subject" ["def"] |
| 412 | header :matches "Subject" ["abc?def"] |
| 413 | |
| 414 | Note that by considering Sieve to be an MUA, RFC 2047 can be interpreted |
| 415 | in a way that NUL characters truncating strings is allowed for Sieve |
| 416 | implementations, although not recommended. It is further allowed to use |
| 417 | encoded NUL characters in headers, but that's not recommended either. |
| 418 | The above example shows why. |
| 419 | |
| 420 | RFC 3028 states that if an implementation fails to convert a character |
| 421 | set to UTF-8, two strings cannot be equal if one contains octets greater |
| 422 | than 127. Assuming that all unknown character sets are one-byte character |
| 423 | sets with the lower 128 octets being US-ASCII is not sound, so this |
| 424 | implementation violates RFC 3028 and treats such MIME words literally. |
| 425 | That way at least something could be matched. |
| 426 | |
| 427 | The folder specified by *fileinto* must not contain the character |
| 428 | sequence ``##`..`##'' to avoid security problems. RFC 3028 does not specify the |
| 429 | syntax of folders apart from *keep* being equivalent to |
| 430 | |
| 431 | fileinto "INBOX"; |
| 432 | |
| 433 | This implementation uses _inbox_ instead. |
| 434 | |
| 435 | Sieve script errors currently cause messages to be silently filed into |
| 436 | _inbox_. RFC 3028 requires that the user is notified of that condition. |
| 437 | This may be implemented in the future by adding a header line to mails that |
| 438 | are filed into _inbox_ due to an error in the filter. |
| 439 | |
| 440 | |
| 441 | |
| 442 | [[CHAPeximfilter]] |
| 443 | Exim filter files |
| 444 | ----------------- |
| 445 | This chapter contains a full description of the contents of Exim filter files. |
| 446 | |
| 447 | |
| 448 | Format of Exim filter files |
| 449 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 450 | Apart from leading white space, the first text in an Exim filter file must be |
| 451 | |
| 452 | # Exim filter |
| 453 | |
| 454 | This is what distinguishes it from a conventional '.forward' file or a Sieve |
| 455 | filter file. If the file does not have this initial line (or the equivalent for |
| 456 | a Sieve filter), it is treated as a conventional '.forward' file, both when |
| 457 | delivering mail and when using the %-bf% testing mechanism. The white space in |
| 458 | the line is optional, and any capitalization may be used. Further text on the |
| 459 | same line is treated as a comment. For example, you could have |
| 460 | |
| 461 | # Exim filter <<== do not edit or remove this line! |
| 462 | |
| 463 | The remainder of the file is a sequence of filtering commands, which consist of |
| 464 | keywords and data values. For example, in the command |
| 465 | |
| 466 | deliver gulliver@lilliput.fict.example |
| 467 | |
| 468 | the keyword is `deliver` and the data value is |
| 469 | `gulliver@lilliput.fict.example`. White space or line breaks separate the |
| 470 | components of a command, except in the case of conditions for the ^if^ command, |
| 471 | where round brackets (parentheses) also act as separators. Complete commands |
| 472 | are separated from each other by white space or line breaks; there are no |
| 473 | special terminators. Thus, several commands may appear on one line, or one |
| 474 | command may be spread over a number of lines. |
| 475 | |
| 476 | If the character # follows a separator anywhere in a command, everything from |
| 477 | # up to the next newline is ignored. This provides a way of including comments |
| 478 | in a filter file. |
| 479 | |
| 480 | |
| 481 | Data values in filter commands |
| 482 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 483 | There are two ways in which a data value can be input: |
| 484 | |
| 485 | - If the text contains no white space then it can be typed verbatim. However, if |
| 486 | it is part of a condition, it must also be free of round brackets |
| 487 | (parentheses), as these are used for grouping in conditions. |
| 488 | |
| 489 | - Otherwise, it must be enclosed in double quotation marks. In this case, the |
| 490 | character \ (backslash) is treated as an ``escape character'' within the string, |
| 491 | causing the following character or characters to be treated specially: |
| 492 | + |
| 493 | &&&& |
| 494 | `\n` is replaced by a newline |
| 495 | `\r` is replaced by a carriage return |
| 496 | `\t` is replaced by a tab |
| 497 | &&&& |
| 498 | |
| 499 | Backslash followed by up to three octal digits is replaced by the character |
| 500 | specified by those digits, and \x followed by up to two hexadecimal digits is |
| 501 | treated similarly. Backslash followed by any other character is replaced |
| 502 | by the second character, so that in particular, \\" becomes " and \\ becomes |
| 503 | \. A data item enclosed in double quotes can be continued onto the next line |
| 504 | by ending the first line with a backslash. Any leading white space at the start |
| 505 | of the continuation line is ignored. |
| 506 | |
| 507 | In addition to the escape character processing that occurs when strings are |
| 508 | enclosed in quotes, most data values are also subject to 'string expansion' |
| 509 | (as described in the next section), in which case the characters `\$` and `\` |
| 510 | are also significant. This means that if a single backslash is actually |
| 511 | required in such a string, and the string is also quoted, \\\\ has to be |
| 512 | entered. |
| 513 | |
| 514 | The maximum permitted length of a data string, before expansion, is 1024 |
| 515 | characters. |
| 516 | |
| 517 | |
| 518 | [[SECTfilterstringexpansion]] |
| 519 | String expansion |
| 520 | ~~~~~~~~~~~~~~~~ |
| 521 | Most data values are expanded before use. Expansion consists of replacing |
| 522 | substrings beginning with `\$` with other text. The full expansion facilities |
| 523 | available in Exim are extensive. If you want to know everything that Exim can |
| 524 | do with strings, you should consult the chapter on string expansion in the Exim |
| 525 | documentation. |
| 526 | |
| 527 | In filter files, by far the most common use of string expansion is the |
| 528 | substitution of the contents of a variable. For example, the substring |
| 529 | |
| 530 | $reply_address |
| 531 | |
| 532 | is replaced by the address to which replies to the message should be sent. If |
| 533 | such a variable name is followed by a letter or digit or underscore, it must be |
| 534 | enclosed in curly brackets (braces), for example, |
| 535 | |
| 536 | ${reply_address} |
| 537 | |
| 538 | If a `\$` character is actually required in an expanded string, it must be |
| 539 | escaped with a backslash, and because backslash is also an escape character in |
| 540 | quoted input strings, it must be doubled in that case. The following two |
| 541 | examples illustrate two different ways of testing for a `\$` character in a |
| 542 | message: |
| 543 | |
| 544 | if $message_body contains \$ then ... |
| 545 | if $message_body contains "\\$" then ... |
| 546 | |
| 547 | You can prevent part of a string from being expanded by enclosing it between |
| 548 | two occurrences of `\N`. For example, |
| 549 | |
| 550 | if $message_body contains \N$$$$\N then ... |
| 551 | |
| 552 | tests for a run of four dollar characters. |
| 553 | |
| 554 | |
| 555 | Some useful general variables |
| 556 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 557 | A complete list of the available variables is given in the Exim documentation. |
| 558 | This shortened list contains the ones that are most likely to be useful in |
| 559 | personal filter files: |
| 560 | |
| 561 | $body_linecount$: The number of lines in the body of the message. |
| 562 | |
| 563 | $body_zerocount$: The number of binary zero characters in the body of the |
| 564 | message. |
| 565 | |
| 566 | |
| 567 | $home$: In conventional configurations, this variable normally contains the |
| 568 | user's home directory. The system administrator can, however, change this. |
| 569 | |
| 570 | $local_part$: The part of the email address that precedes the @ sign -- |
| 571 | normally the user's login name. If support for multiple personal mailboxes is |
| 572 | enabled (see section <<SECTmbox>> below) and a prefix or suffix for the local |
| 573 | part was recognized, it is removed from the string in this variable. |
| 574 | |
| 575 | $local_part_prefix$: If support for multiple personal mailboxes is enabled |
| 576 | (see section <<SECTmbox>> below), and a local part prefix was recognized, |
| 577 | this variable contains the prefix. Otherwise it contains an empty string. |
| 578 | |
| 579 | $local_part_suffix$: If support for multiple personal mailboxes is enabled |
| 580 | (see section <<SECTmbox>> below), and a local part suffix was recognized, |
| 581 | this variable contains the suffix. Otherwise it contains an empty string. |
| 582 | |
| 583 | $message_body$: The initial portion of the body of the message. By default, |
| 584 | up to 500 characters are read into this variable, but the system administrator |
| 585 | can configure this to some other value. Newlines in the body are converted into |
| 586 | single spaces. |
| 587 | |
| 588 | $message_body_end$: The final portion of the body of the message, formatted |
| 589 | and limited in the same way as $message_body$. |
| 590 | |
| 591 | $message_body_size$: The size of the body of the message, in bytes. |
| 592 | |
| 593 | $message_headers$: The header lines of the message, concatenated into a |
| 594 | single string, with newline characters between them. |
| 595 | |
| 596 | $message_id$: The message's local identification string, which is unique for |
| 597 | each message handled by a single host. |
| 598 | |
| 599 | $message_size$: The size of the entire message, in bytes. |
| 600 | |
| 601 | $original_local_part$: When an address that arrived with the message is |
| 602 | being processed, this contains the same value as the variable $local_part$. |
| 603 | However, if an address generated by an alias, forward, or filter file is being |
| 604 | processed, this variable contains the local part of the original address. |
| 605 | |
| 606 | $reply_address$: The contents of the 'Reply-to:' header, if the message |
| 607 | has one; otherwise the contents of the 'From:' header. It is the address to |
| 608 | which normal replies to the message should be sent. |
| 609 | |
| 610 | $return_path$: The return path -- that is, the sender field that will be |
| 611 | transmitted as part of the message's envelope if the message is sent to another |
| 612 | host. This is the address to which delivery errors are sent. In many cases, |
| 613 | this variable has the same value as $sender_address$, but if, for example, |
| 614 | an incoming message to a mailing list has been expanded, $return_path$ may |
| 615 | have been changed to contain the address of the list maintainer. |
| 616 | |
| 617 | $sender_address$: The sender address that was received in the envelope of |
| 618 | the message. This is not necessarily the same as the contents of the 'From:' |
| 619 | or 'Sender:' header lines. For delivery error messages (``bounce messages'') |
| 620 | there is no sender address, and this variable is empty. |
| 621 | |
| 622 | $tod_full$: A full version of the time and date, for example: Wed, 18 Oct |
| 623 | 1995 09:51:40 +0100. The timezone is always given as a numerical offset from |
| 624 | GMT. |
| 625 | |
| 626 | $tod_log$: The time and date in the format used for writing Exim's log files, |
| 627 | without the timezone, for example: 1995-10-12 15:32:29. |
| 628 | |
| 629 | $tod_zone$: The local timezone offset, for example: +0100. |
| 630 | |
| 631 | |
| 632 | |
| 633 | [[SECTheadervariables]] |
| 634 | Header variables |
| 635 | ~~~~~~~~~~~~~~~~ |
| 636 | There is a special set of expansion variables containing the header lines of |
| 637 | the message being processed. These variables have names beginning with |
| 638 | $header_$ followed by the name of the header line, terminated by a colon. |
| 639 | For example, |
| 640 | |
| 641 | $header_from: |
| 642 | $header_subject: |
| 643 | |
| 644 | The whole item, including the terminating colon, is replaced by the contents of |
| 645 | the message header line. If there is more than one header line with the same |
| 646 | name, their contents are concatenated. For header lines whose data consists of |
| 647 | a list of addresses (for example, 'From:' and 'To:'), a comma and newline is |
| 648 | inserted between each set of data. For all other header lines, just a newline |
| 649 | is used. |
| 650 | |
| 651 | Leading and trailing white space is removed from header line data, and if there |
| 652 | are any MIME ``words'' that are encoded as defined by RFC 2047 (because they |
| 653 | contain non-ASCII characters), they are decoded and translated, if possible, to |
| 654 | a local character set. Translation is attempted only on operating systems that |
| 655 | have the ^^iconv()^^ function. This makes the header line look the same as it |
| 656 | would when displayed by an MUA. The default character set is ISO-8859-1, but |
| 657 | this can be changed by means of the ^headers^ command (see below). |
| 658 | |
| 659 | If you want to see the actual characters that make up a header line, you can |
| 660 | specify $rheader_$ instead of $header_$. This inserts the ``raw'' |
| 661 | header line, unmodified. |
| 662 | |
| 663 | There is also an intermediate form, requested by $bheader_$, which removes |
| 664 | leading and trailing space and decodes MIME ``words'', but does not do any |
| 665 | character translation. If an attempt to decode what looks superficially like a |
| 666 | MIME ``word'' fails, the raw string is returned. If decoding produces a binary |
| 667 | zero character, it is replaced by a question mark. |
| 668 | |
| 669 | The capitalization of the name following $header_$ is not significant. |
| 670 | Because any printing character except colon may appear in the name of a |
| 671 | message's header (this is a requirement of RFC 2822, the document that |
| 672 | describes the format of a mail message) curly brackets must 'not' be used in |
| 673 | this case, as they will be taken as part of the header name. Two shortcuts are |
| 674 | allowed in naming header variables: |
| 675 | |
| 676 | - The initiating $header_$, $rheader_$, or $bheader_$ can be |
| 677 | abbreviated to $h_$, $rh_$, or $bh_$, respectively. |
| 678 | |
| 679 | - The terminating colon can be omitted if the next character is white space. The |
| 680 | white space character is retained in the expanded string. However, this is not |
| 681 | recommended, because it makes it easy to forget the colon when it really is |
| 682 | needed. |
| 683 | |
| 684 | If the message does not contain a header of the given name, an empty string is |
| 685 | substituted. Thus it is important to spell the names of headers correctly. Do |
| 686 | not use $header_Reply_to$ when you really mean $header_Reply-to$. |
| 687 | |
| 688 | |
| 689 | User variables |
| 690 | ~~~~~~~~~~~~~~ |
| 691 | There are ten user variables with names $n0$ -- $n9$ that can be |
| 692 | incremented by the ^add^ command (see section <<SECTadd>>). These can be used |
| 693 | for ``scoring'' messages in various ways. If Exim is configured to run a |
| 694 | ``system filter'' on every message, the values left in these variables are |
| 695 | copied into the variables $sn0$ -- $sn9$ at the end of the system filter, thus |
| 696 | making them available to users' filter files. How these values are used is |
| 697 | entirely up to the individual installation. |
| 698 | |
| 699 | |
| 700 | Current directory |
| 701 | ~~~~~~~~~~~~~~~~~ |
| 702 | The contents of your filter file should not make any assumptions about the |
| 703 | current directory. It is best to use absolute paths for file names; you |
| 704 | can normally make use of the $home$ variable to refer to your home directory. |
| 705 | The ^save^ command automatically inserts $home$ at the start of non-absolute |
| 706 | paths. |
| 707 | |
| 708 | |
| 709 | |
| 710 | |
| 711 | [[SECTsigdel]] |
| 712 | Significant deliveries |
| 713 | ~~~~~~~~~~~~~~~~~~~~~~ |
| 714 | When in the course of delivery a message is processed by a filter file, what |
| 715 | happens next, that is, after the filter file has been processed, depends on |
| 716 | whether or not the filter sets up any 'significant deliveries'. If at least |
| 717 | one significant delivery is set up, the filter is considered to have handled |
| 718 | the entire delivery arrangements for the current address, and no further |
| 719 | processing of the address takes place. If, however, no significant deliveries |
| 720 | are set up, Exim continues processing the current address as if there were no |
| 721 | filter file, and typically sets up a delivery of a copy of the message into a |
| 722 | local mailbox. In particular, this happens in the special case of a filter file |
| 723 | containing only comments. |
| 724 | |
| 725 | The delivery commands ^deliver^, ^save^, and ^pipe^ are by default |
| 726 | significant. However, if such a command is preceded by the word ^unseen^, its |
| 727 | delivery is not considered to be significant. In contrast, other commands such |
| 728 | as ^mail^ and ^vacation^ do not set up significant deliveries unless |
| 729 | preceded by the word ^seen^. |
| 730 | |
| 731 | The following example commands set up significant deliveries: |
| 732 | |
| 733 | deliver jack@beanstalk.example |
| 734 | pipe $home/bin/mymailscript |
| 735 | seen mail subject "message discarded" |
| 736 | seen finish |
| 737 | |
| 738 | The following example commands do not set up significant deliveries: |
| 739 | |
| 740 | unseen deliver jack@beanstalk.example |
| 741 | unseen pipe $home/bin/mymailscript |
| 742 | mail subject "message discarded" |
| 743 | finish |
| 744 | |
| 745 | |
| 746 | |
| 747 | |
| 748 | Filter commands |
| 749 | ~~~~~~~~~~~~~~~ |
| 750 | The filter commands that are described in subsequent sections are listed |
| 751 | below, with the section in which they are described in brackets: |
| 752 | |
| 753 | [frame="none"] |
| 754 | `-------------`----------------------------------------------- |
| 755 | ^add^ ~~increment a user variable (section <<SECTadd>>) |
| 756 | ^deliver^ ~~deliver to an email address (section <<SECTdeliver>>) |
| 757 | ^fail^ ~~force delivery failure (sysadmin use) (section <<SECTfail>>) |
| 758 | ^finish^ ~~end processing (section <<SECTfinish>>) |
| 759 | ^freeze^ ~~freeze message (sysadmin use) (section <<SECTfreeze>>) |
| 760 | ^headers^ ~~set the header character set (section <<SECTheaders>>) |
| 761 | ^if^ ~~test condition(s) (section <<SECTif>>) |
| 762 | ^logfile^ ~~define log file (section <<SECTlog>>) |
| 763 | ^logwrite^ ~~write to log file (section <<SECTlog>>) |
| 764 | ^mail^ ~~send a reply message (section <<SECTmail>>) |
| 765 | ^pipe^ ~~pipe to a command (section <<SECTpipe>>) |
| 766 | ^save^ ~~save to a file (section <<SECTsave>>) |
| 767 | ^testprint^ ~~print while testing (section <<SECTtestprint>>) |
| 768 | ^vacation^ ~~tailored form of ^mail^ (section <<SECTmail>>) |
| 769 | -------------------------------------------------------------- |
| 770 | |
| 771 | The ^headers^ command has additional parameters that can be used only in a |
| 772 | system filter. The ^fail^ and ^freeze^ commands are available only when |
| 773 | Exim's filtering facilities are being used as a system filter, and are |
| 774 | therefore usable only by the system administrator and not by ordinary users. |
| 775 | They are mentioned only briefly in this document; for more information, see the |
| 776 | main Exim specification. |
| 777 | |
| 778 | |
| 779 | |
| 780 | [[SECTadd]] |
| 781 | The add command |
| 782 | ~~~~~~~~~~~~~~~ |
| 783 | &&& |
| 784 | ` add `<'number'>` to `<'user variable'> |
| 785 | `e.g. add 2 to n3` |
| 786 | &&& |
| 787 | |
| 788 | There are 10 user variables of this type, with names $n0$ -- $n9$. Their |
| 789 | values can be obtained by the normal expansion syntax (for example $n3$) in |
| 790 | other commands. At the start of filtering, these variables all contain zero. |
| 791 | Both arguments of the ^add^ command are expanded before use, making it |
| 792 | possible to add variables to each other. Subtraction can be obtained by adding |
| 793 | negative numbers. |
| 794 | |
| 795 | |
| 796 | |
| 797 | [[SECTdeliver]] |
| 798 | The deliver command |
| 799 | ~~~~~~~~~~~~~~~~~~~ |
| 800 | |
| 801 | &&& |
| 802 | ` deliver` <'mail address'> |
| 803 | `e.g. deliver "Dr Livingstone <David@somewhere.africa.example>"` |
| 804 | &&& |
| 805 | |
| 806 | This command provides a forwarding operation. The delivery that it sets up is |
| 807 | significant unless the command is preceded by ^unseen^ (see section |
| 808 | <<SECTsigdel>>). The message is sent on to the given address, exactly as |
| 809 | happens if the address had appeared in a traditional '.forward' file. If you |
| 810 | want to deliver the message to a number of different addresses, you can use |
| 811 | more than one ^deliver^ command (each one may have only one address). However, |
| 812 | duplicate addresses are discarded. |
| 813 | |
| 814 | To deliver a copy of the message to your normal mailbox, your login name can be |
| 815 | given as the address. Once an address has been processed by the filtering |
| 816 | mechanism, an identical generated address will not be so processed again, so |
| 817 | doing this does not cause a loop. |
| 818 | |
| 819 | However, if you have a mail alias, you should 'not' refer to it here. For |
| 820 | example, if the mail address 'L.Gulliver' is aliased to 'lg303' then all |
| 821 | references in Gulliver's '.forward' file should be to 'lg303'. A reference |
| 822 | to the alias will not work for messages that are addressed to that alias, |
| 823 | since, like '.forward' file processing, aliasing is performed only once on an |
| 824 | address, in order to avoid looping. |
| 825 | |
| 826 | Following the new address, an optional second address, preceded by |
| 827 | ^errors_to^ may appear. This changes the address to which delivery errors on |
| 828 | the forwarded message will be sent. Instead of going to the message's original |
| 829 | sender, they go to this new address. For ordinary users, the only value that is |
| 830 | permitted for this address is the user whose filter file is being processed. |
| 831 | For example, the user 'lg303' whose mailbox is in the domain |
| 832 | 'lilliput.example' could have a filter file that contains |
| 833 | |
| 834 | deliver jon@elsewhere.example errors_to lg303@lilliput.example |
| 835 | |
| 836 | Clearly, using this feature makes sense only in situations where not all |
| 837 | messages are being forwarded. In particular, bounce messages must not be |
| 838 | forwarded in this way, as this is likely to create a mail loop if something |
| 839 | goes wrong. |
| 840 | |
| 841 | |
| 842 | |
| 843 | [[SECTsave]] |
| 844 | The save command |
| 845 | ~~~~~~~~~~~~~~~~ |
| 846 | &&& |
| 847 | ` save `<'file name'> |
| 848 | `e.g. save $home/mail/bookfolder` |
| 849 | &&& |
| 850 | |
| 851 | This command specifies that a copy of the message is to be appended to the |
| 852 | given file (that is, the file is to be used as a mail folder). The delivery |
| 853 | that ^save^ sets up is significant unless the command is preceded by |
| 854 | ^unseen^ (see section <<SECTsigdel>>). |
| 855 | |
| 856 | More than one ^save^ command may be obeyed; each one causes a copy of the |
| 857 | message to be written to its argument file, provided they are different |
| 858 | (duplicate ^save^ commands are ignored). |
| 859 | |
| 860 | If the file name does not start with a / character, the contents of the |
| 861 | $home$ variable are prepended, unless it is empty. In conventional |
| 862 | configurations, this variable is normally set in a user filter to the user's |
| 863 | home directory, but the system administrator may set it to some other path. In |
| 864 | some configurations, $home$ may be unset, in which case a non-absolute path |
| 865 | name may be generated. Such configurations convert this to an absolute path |
| 866 | when the delivery takes place. In a system filter, $home$ is never set. |
| 867 | |
| 868 | The user must of course have permission to write to the file, and the writing |
| 869 | of the file takes place in a process that is running as the user, under the |
| 870 | user's primary group. Any secondary groups to which the user may belong are not |
| 871 | normally taken into account, though the system administrator can configure Exim |
| 872 | to set them up. In addition, the ability to use this command at all is |
| 873 | controlled by the system administrator -- it may be forbidden on some systems. |
| 874 | |
| 875 | An optional mode value may be given after the file name. The value for the mode |
| 876 | is interpreted as an octal number, even if it does not begin with a zero. For |
| 877 | example: |
| 878 | |
| 879 | save /some/folder 640 |
| 880 | |
| 881 | This makes it possible for users to override the system-wide mode setting for |
| 882 | file deliveries, which is normally 600. If an existing file does not have the |
| 883 | correct mode, it is changed. |
| 884 | |
| 885 | An alternative form of delivery may be enabled on your system, in which each |
| 886 | message is delivered into a new file in a given directory. If this is the case, |
| 887 | this functionality can be requested by giving the directory name terminated by |
| 888 | a slash after the ^save^ command, for example |
| 889 | |
| 890 | save separated/messages/ |
| 891 | |
| 892 | There are several different formats for such deliveries; check with your system |
| 893 | administrator or local documentation to find out which (if any) are available |
| 894 | on your system. If this functionality is not enabled, the use of a path name |
| 895 | ending in a slash causes an error. |
| 896 | |
| 897 | |
| 898 | |
| 899 | [[SECTpipe]] |
| 900 | The pipe command |
| 901 | ~~~~~~~~~~~~~~~~ |
| 902 | &&& |
| 903 | ` pipe `<'command'> |
| 904 | `e.g. pipe "$home/bin/countmail $sender_address"` |
| 905 | &&& |
| 906 | |
| 907 | This command specifies that the message is to be delivered to the specified |
| 908 | command using a pipe. The delivery that it sets up is significant unless the |
| 909 | command is preceded by ^unseen^ (see section <<SECTsigdel>>). Remember, |
| 910 | however, that no deliveries are done while the filter is being processed. All |
| 911 | deliveries happen later on. Therefore, the result of running the pipe is not |
| 912 | available to the filter. |
| 913 | |
| 914 | When the deliveries are done, a separate process is run, and a copy of the |
| 915 | message is passed on its standard input. The process runs as the user, under |
| 916 | the user's primary group. Any secondary groups to which the user may belong are |
| 917 | not normally taken into account, though the system administrator can configure |
| 918 | Exim to set them up. More than one ^pipe^ command may appear; each one causes |
| 919 | a copy of the message to be written to its argument pipe, provided they are |
| 920 | different (duplicate ^pipe^ commands are ignored). |
| 921 | |
| 922 | When the time comes to transport the message, |
| 923 | the command supplied to ^pipe^ is split up by Exim into a command name and a |
| 924 | number of arguments. These are delimited by white space except for arguments |
| 925 | enclosed in double quotes, in which case backslash is interpreted as an escape, |
| 926 | or in single quotes, in which case no escaping is recognized. Note that as the |
| 927 | whole command is normally supplied in double quotes, a second level of quoting |
| 928 | is required for internal double quotes. For example: |
| 929 | |
| 930 | pipe "$home/myscript \"size is $message_size\"" |
| 931 | |
| 932 | String expansion is performed on the separate components after the line has |
| 933 | been split up, and the command is then run directly by Exim; it is not run |
| 934 | under a shell. Therefore, substitution cannot change the number of arguments, |
| 935 | nor can quotes, backslashes or other shell metacharacters in variables cause |
| 936 | confusion. |
| 937 | |
| 938 | Documentation for some programs that are normally run via this kind of pipe |
| 939 | often suggest that the command should start with |
| 940 | |
| 941 | IFS=" " |
| 942 | |
| 943 | This is a shell command, and should 'not' be present in Exim filter files, |
| 944 | since it does not normally run the command under a shell. |
| 945 | |
| 946 | However, there is an option that the administrator can set to cause a shell to |
| 947 | be used. In this case, the entire command is expanded as a single string and |
| 948 | passed to the shell for interpretation. It is recommended that this be avoided |
| 949 | if at all possible, since it can lead to problems when inserted variables |
| 950 | contain shell metacharacters. |
| 951 | |
| 952 | The default PATH set up for the command is determined by the system |
| 953 | administrator, usually containing at least _/usr/bin_ so that common commands |
| 954 | are available without having to specify an absolute file name. However, it is |
| 955 | possible for the system administrator to restrict the pipe facility so that the |
| 956 | command name must not contain any / characters, and must be found in one of the |
| 957 | directories in the configured PATH. It is also possible for the system |
| 958 | administrator to lock out the use of the ^pipe^ command altogether. |
| 959 | |
| 960 | When the command is run, a number of environment variables are set up. The |
| 961 | complete list for pipe deliveries may be found in the Exim reference manual. |
| 962 | Those that may be useful for pipe deliveries from user filter files are: |
| 963 | |
| 964 | &&& |
| 965 | `DOMAIN ` the domain of the address |
| 966 | `HOME ` your home directory |
| 967 | `LOCAL_PART ` see below |
| 968 | `LOCAL_PART_PREFIX ` see below |
| 969 | `LOCAL_PART_SUFFIX ` see below |
| 970 | `LOGNAME ` your login name |
| 971 | `MESSAGE_ID ` the unique id of the message |
| 972 | `PATH ` the command search path |
| 973 | `RECIPIENT ` the complete recipient address |
| 974 | `SENDER ` the sender of the message |
| 975 | `SHELL ` `/bin/sh` |
| 976 | `USER ` see below |
| 977 | &&& |
| 978 | |
| 979 | LOCAL_PART, LOGNAME, and USER are all set to the same value, |
| 980 | namely, your login id. LOCAL_PART_PREFIX and LOCAL_PART_SUFFIX may |
| 981 | be set if Exim is configured to recognize prefixes or suffixes in the local |
| 982 | parts of addresses. For example, a message addressed to |
| 983 | 'pat-suf2@domain.example' may cause the filter for user 'pat' to be run. If |
| 984 | this sets up a pipe delivery, LOCAL_PART_SUFFIX is `-suf2` when the |
| 985 | pipe command runs. The system administrator has to configure Exim specially for |
| 986 | this feature to be available. |
| 987 | |
| 988 | If you run a command that is a shell script, be very careful in your use of |
| 989 | data from the incoming message in the commands in your script. RFC 2822 is very |
| 990 | generous in the characters that are permitted to appear in mail addresses, and |
| 991 | in particular, an address may begin with a vertical bar or a slash. For this |
| 992 | reason you should always use quotes round any arguments that involve data from |
| 993 | the message, like this: |
| 994 | |
| 995 | /some/command '$SENDER' |
| 996 | |
| 997 | so that inserted shell meta-characters do not cause unwanted effects. |
| 998 | |
| 999 | Remember that, as was explained earlier, the pipe command is not run at the |
| 1000 | time the filter file is interpreted. The filter just defines what deliveries |
| 1001 | are required for one particular addressee of a message. The deliveries |
| 1002 | themselves happen later, once Exim has decided everything that needs to be done |
| 1003 | for the message. |
| 1004 | |
| 1005 | A consequence of this is that you cannot inspect the return code from the pipe |
| 1006 | command from within the filter. Nevertheless, the code returned by the command |
| 1007 | is important, because Exim uses it to decide whether the delivery has succeeded |
| 1008 | or failed. |
| 1009 | |
| 1010 | The command should return a zero completion code if all has gone well. Most |
| 1011 | non-zero codes are treated by Exim as indicating a failure of the pipe. This is |
| 1012 | treated as a delivery failure, causing the message to be returned to its |
| 1013 | sender. However, there are some completion codes that are treated as temporary |
| 1014 | errors. The message remains on Exim's spool disk, and the delivery is tried |
| 1015 | again later, though it will ultimately time out if the delivery failures go on |
| 1016 | too long. The completion codes to which this applies can be specified by the |
| 1017 | system administrator; the default values are 73 and 75. |
| 1018 | |
| 1019 | The pipe command should not normally write anything to its standard output or |
| 1020 | standard error file descriptors. If it does, whatever is written is normally |
| 1021 | returned to the sender of the message as a delivery error, though this action |
| 1022 | can be varied by the system administrator. |
| 1023 | |
| 1024 | |
| 1025 | |
| 1026 | [[SECTmail]] |
| 1027 | Mail commands |
| 1028 | ~~~~~~~~~~~~~ |
| 1029 | There are two commands that cause the creation of a new mail message, neither |
| 1030 | of which count as a significant delivery unless the command is preceded by the |
| 1031 | word ^seen^ (see section <<SECTsigdel>>). This is a powerful facility, but it |
| 1032 | should be used with care, because of the danger of creating infinite sequences |
| 1033 | of messages. The system administrator can forbid the use of these commands |
| 1034 | altogether. |
| 1035 | |
| 1036 | To help prevent runaway message sequences, these commands have no effect when |
| 1037 | the incoming message is a bounce (delivery error) message, and messages sent by |
| 1038 | this means are treated as if they were reporting delivery errors. Thus, they |
| 1039 | should never themselves cause a bounce message to be returned. The basic |
| 1040 | mail-sending command is |
| 1041 | |
| 1042 | &&& |
| 1043 | `mail [to `<'address-list'>`]` |
| 1044 | ` [cc `<'address-list'>`]` |
| 1045 | ` [bcc `<'address-list'>`]` |
| 1046 | ` [from `<'address'>`]` |
| 1047 | ` [reply_to `<'address'>`]` |
| 1048 | ` [subject `<'text'>`]` |
| 1049 | ` [extra_headers `<'text'>`]` |
| 1050 | ` [text `<'text'>`]` |
| 1051 | ` [[expand] file `<'filename'>`]` |
| 1052 | ` [return message]` |
| 1053 | ` [log `<'log file name'>`]` |
| 1054 | ` [once `<'note file name'>`]` |
| 1055 | ` [once_repeat `<'time interval'>`]` |
| 1056 | |
| 1057 | `e.g. mail text "Your message about $h_subject: has been received"` |
| 1058 | &&& |
| 1059 | |
| 1060 | Each <'address-list'> can contain a number of addresses, separated by commas, |
| 1061 | in the format of a 'To:' or 'Cc:' header line. In fact, the text you supply |
| 1062 | here is copied exactly into the appropriate header line. It may contain |
| 1063 | additional information as well as email addresses. For example: |
| 1064 | |
| 1065 | .... |
| 1066 | mail to "Julius Caesar <jc@rome.example>, \ |
| 1067 | <ma@rome.example> (Mark A.)" |
| 1068 | .... |
| 1069 | |
| 1070 | Similarly, the texts supplied for ^from^ and ^reply_to^ are copied into |
| 1071 | their respective header lines. |
| 1072 | |
| 1073 | As a convenience for use in one common case, there is also a command called |
| 1074 | ^vacation^. It behaves in the same way as ^mail^, except that the defaults for |
| 1075 | the %subject%, %file%, %log%, %once%, and %once_repeat% options are |
| 1076 | |
| 1077 | subject "On vacation" |
| 1078 | expand file .vacation.msg |
| 1079 | log .vacation.log |
| 1080 | once .vacation |
| 1081 | once_repeat 7d |
| 1082 | |
| 1083 | respectively. These are the same file names and repeat period used by the |
| 1084 | traditional Unix ^vacation^ command. The defaults can be overridden by |
| 1085 | explicit settings, but if a file name is given its contents are expanded only |
| 1086 | if explicitly requested. |
| 1087 | |
| 1088 | *Warning*: The ^vacation^ command should always be used conditionally, |
| 1089 | subject to at least the ^personal^ condition (see section <<SECTpersonal>> |
| 1090 | below) so as not to send automatic replies to non-personal messages from |
| 1091 | mailing lists or elsewhere. Sending an automatic response to a mailing list or |
| 1092 | a mailing list manager is an Internet Sin. |
| 1093 | |
| 1094 | For both commands, the key/value argument pairs can appear in any order. At |
| 1095 | least one of ^text^ or ^file^ must appear (except with ^vacation^, where |
| 1096 | there is a default for ^file^); if both are present, the text string appears |
| 1097 | first in the message. If ^expand^ precedes ^file^, each line of the file is |
| 1098 | subject to string expansion before it is included in the message. |
| 1099 | |
| 1100 | Several lines of text can be supplied to ^text^ by including the escape |
| 1101 | sequence ``\n'' in the string wherever a newline is required. If the command is |
| 1102 | output during filter file testing, newlines in the text are shown as ``\n''. |
| 1103 | |
| 1104 | Note that the keyword for creating a 'Reply-To:' header is ^reply_to^, |
| 1105 | because Exim keywords may contain underscores, but not hyphens. If the ^from^ |
| 1106 | keyword is present and the given address does not match the user who owns the |
| 1107 | forward file, Exim normally adds a 'Sender:' header to the message, |
| 1108 | though it can be configured not to do this. |
| 1109 | |
| 1110 | The %extra_headers% keyword allows you to add custom header lines to the |
| 1111 | message. The text supplied must be one or more syntactically valid RFC 2822 |
| 1112 | header lines. You can use ``\n'' within quoted text to specify newlines between |
| 1113 | headers, and also to define continued header lines. For example: |
| 1114 | |
| 1115 | extra_headers "h1: first\nh2: second\n continued\nh3: third" |
| 1116 | |
| 1117 | No newline should appear at the end of the final header line. |
| 1118 | |
| 1119 | If no ^to^ argument appears, the message is sent to the address in the |
| 1120 | $reply_address$ variable (see section <<SECTfilterstringexpansion>> above). |
| 1121 | An 'In-Reply-To:' header is automatically included in the created message, |
| 1122 | giving a reference to the message identification of the incoming message. |
| 1123 | |
| 1124 | If ^return message^ is specified, the incoming message that caused the filter |
| 1125 | file to be run is added to the end of the message, subject to a maximum size |
| 1126 | limitation. |
| 1127 | |
| 1128 | If a log file is specified, a line is added to it for each message sent. |
| 1129 | |
| 1130 | If a ^once^ file is specified, it is used to hold a database for remembering |
| 1131 | who has received a message, and no more than one message is ever sent to any |
| 1132 | particular address, unless ^once_repeat^ is set. This specifies a time |
| 1133 | interval after which another copy of the message is sent. The interval is |
| 1134 | specified as a sequence of numbers, each followed by the initial letter of one |
| 1135 | of ``seconds'', ``minutes'', ``hours'', ``days'', or ``weeks''. For example, |
| 1136 | |
| 1137 | once_repeat 5d4h |
| 1138 | |
| 1139 | causes a new message to be sent if 5 days and 4 hours have elapsed since the |
| 1140 | last one was sent. There must be no white space in a time interval. |
| 1141 | |
| 1142 | Commonly, the file name specified for ^once^ is used as the base name for |
| 1143 | direct-access (DBM) file operations. There are a number of different DBM |
| 1144 | libraries in existence. Some operating systems provide one as a default, but |
| 1145 | even in this case a different one may have been used when building Exim. With |
| 1146 | some DBM libraries, specifying ^once^ results in two files being created, |
| 1147 | with the suffixes _.dir_ and _.pag_ being added to the given name. With |
| 1148 | some others a single file with the suffix _.db_ is used, or the name is used |
| 1149 | unchanged. |
| 1150 | |
| 1151 | Using a DBM file for implementing the ^once^ feature means that the file |
| 1152 | grows as large as necessary. This is not usually a problem, but some system |
| 1153 | administrators want to put a limit on it. The facility can be configured not to |
| 1154 | use a DBM file, but instead, to use a regular file with a maximum size. The |
| 1155 | data in such a file is searched sequentially, and if the file fills up, the |
| 1156 | oldest entry is deleted to make way for a new one. This means that some |
| 1157 | correspondents may receive a second copy of the message after an unpredictable |
| 1158 | interval. Consult your local information to see if your system is configured |
| 1159 | this way. |
| 1160 | |
| 1161 | More than one ^mail^ or ^vacation^ command may be obeyed in a single filter |
| 1162 | run; they are all honoured, even when they are to the same recipient. |
| 1163 | |
| 1164 | |
| 1165 | |
| 1166 | [[SECTlog]] |
| 1167 | Logging commands |
| 1168 | ~~~~~~~~~~~~~~~~ |
| 1169 | A log can be kept of actions taken by a filter file. This facility is normally |
| 1170 | available in conventional configurations, but there are some situations where |
| 1171 | it might not be. Also, the system administrator may choose to disable it. Check |
| 1172 | your local information if in doubt. |
| 1173 | |
| 1174 | Logging takes place while the filter file is being interpreted. It does not |
| 1175 | queue up for later like the delivery commands. The reason for this is so that a |
| 1176 | log file need be opened only once for several write operations. There are two |
| 1177 | commands, neither of which constitutes a significant delivery. The first |
| 1178 | defines a file to which logging output is subsequently written: |
| 1179 | |
| 1180 | &&& |
| 1181 | ` logfile `<'file name'> |
| 1182 | `e.g. logfile $home/filter.log` |
| 1183 | &&& |
| 1184 | |
| 1185 | The file name must be fully qualified. You can use $home$, as in this |
| 1186 | example, to refer to your home directory. The file name may optionally be |
| 1187 | followed by a mode for the file, which is used if the file has to be created. |
| 1188 | For example, |
| 1189 | |
| 1190 | logfile $home/filter.log 0644 |
| 1191 | |
| 1192 | The number is interpreted as octal, even if it does not begin with a zero. |
| 1193 | The default for the mode is 600. It is suggested that the ^logfile^ command |
| 1194 | normally appear as the first command in a filter file. Once ^logfile^ has |
| 1195 | been obeyed, the ^logwrite^ command can be used to write to the log file: |
| 1196 | |
| 1197 | &&& |
| 1198 | ` logwrite "`<'some text string'>`"` |
| 1199 | `e.g. logwrite "$tod_log $message_id processed"` |
| 1200 | &&& |
| 1201 | |
| 1202 | It is possible to have more than one ^logfile^ command, to specify writing to |
| 1203 | different log files in different circumstances. Writing takes place at the end |
| 1204 | of the file, and a newline character is added to the end of each string if |
| 1205 | there isn't one already there. Newlines can be put in the middle of the string |
| 1206 | by using the ``\n'' escape sequence. Lines from simultaneous deliveries may get |
| 1207 | interleaved in the file, as there is no interlocking, so you should plan your |
| 1208 | logging with this in mind. However, data should not get lost. |
| 1209 | |
| 1210 | |
| 1211 | |
| 1212 | [[SECTfinish]] |
| 1213 | The finish command |
| 1214 | ~~~~~~~~~~~~~~~~~~ |
| 1215 | The command ^finish^, which has no arguments, causes Exim to stop |
| 1216 | interpreting the filter file. This is not a significant action unless preceded |
| 1217 | by ^seen^. A filter file containing only ^seen finish^ is a black hole. |
| 1218 | |
| 1219 | |
| 1220 | [[SECTtestprint]] |
| 1221 | The testprint command |
| 1222 | ~~~~~~~~~~~~~~~~~~~~~ |
| 1223 | It is sometimes helpful to be able to print out the values of variables when |
| 1224 | testing filter files. The command |
| 1225 | |
| 1226 | &&& |
| 1227 | ` testprint `<'text'> |
| 1228 | `e.g. testprint "home=$home reply_address=$reply_address"` |
| 1229 | &&& |
| 1230 | |
| 1231 | does nothing when mail is being delivered. However, when the filtering code is |
| 1232 | being tested by means of the %-bf% option (see section <<SECTtesting>> above), |
| 1233 | the value of the string is written to the standard output. |
| 1234 | |
| 1235 | |
| 1236 | [[SECTfail]] |
| 1237 | The fail command |
| 1238 | ~~~~~~~~~~~~~~~~ |
| 1239 | When Exim's filtering facilities are being used as a system filter, the |
| 1240 | ^fail^ command is available, to force delivery failure. Because this command |
| 1241 | is normally usable only by the system administrator, and not enabled for use by |
| 1242 | ordinary users, it is described in more detail in the main Exim specification |
| 1243 | rather than in this document. |
| 1244 | |
| 1245 | |
| 1246 | [[SECTfreeze]] |
| 1247 | The freeze command |
| 1248 | ~~~~~~~~~~~~~~~~~~ |
| 1249 | When Exim's filtering facilities are being used as a system filter, the |
| 1250 | ^freeze^ command is available, to freeze a message on the queue. Because this |
| 1251 | command is normally usable only by the system administrator, and not enabled |
| 1252 | for use by ordinary users, it is described in more detail in the main Exim |
| 1253 | specification rather than in this document. |
| 1254 | |
| 1255 | |
| 1256 | |
| 1257 | [[SECTheaders]] |
| 1258 | The headers command |
| 1259 | ~~~~~~~~~~~~~~~~~~~ |
| 1260 | The ^headers^ command can be used to change the target character set that is |
| 1261 | used when translating the contents of encoded header lines for insertion by the |
| 1262 | $header_$ mechanism (see section <<SECTheadervariables>> above). The default |
| 1263 | can be set in the Exim configuration; if not specified, ISO-8859-1 is used. The |
| 1264 | only currently supported format for the ^headers^ command in user filters is as |
| 1265 | in this example: |
| 1266 | |
| 1267 | headers charset "UTF-8" |
| 1268 | |
| 1269 | That is, ^headers^ is followed by the word ^charset^ and then the name of a |
| 1270 | character set. This particular example would be useful if you wanted to compare |
| 1271 | the contents of a header to a UTF-8 string. |
| 1272 | |
| 1273 | In system filter files, the ^headers^ command can be used to add or remove |
| 1274 | header lines from the message. These features are described in the main Exim |
| 1275 | specification. |
| 1276 | |
| 1277 | |
| 1278 | |
| 1279 | [[SECTif]] |
| 1280 | Obeying commands conditionally |
| 1281 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1282 | Most of the power of filtering comes from the ability to test conditions and |
| 1283 | obey different commands depending on the outcome. The ^if^ command is used to |
| 1284 | specify conditional execution, and its general form is |
| 1285 | |
| 1286 | &&& |
| 1287 | `if `<'condition'> |
| 1288 | `then `<'commands'> |
| 1289 | `elif `<'condition'> |
| 1290 | `then `<'commands'> |
| 1291 | `else `<'commands'> |
| 1292 | `endif` |
| 1293 | &&& |
| 1294 | |
| 1295 | There may be any number of ^elif^ and ^then^ sections (including none) and |
| 1296 | the ^else^ section is also optional. Any number of commands, including nested |
| 1297 | ^if^ commands, may appear in any of the <'commands'> sections. |
| 1298 | |
| 1299 | Conditions can be combined by using the words ^and^ and ^or^, and round |
| 1300 | brackets (parentheses) can be used to specify how several conditions are to |
| 1301 | combine. Without brackets, ^and^ is more binding than ^or^. |
| 1302 | For example, |
| 1303 | |
| 1304 | if |
| 1305 | $h_subject: contains "Make money" or |
| 1306 | $h_precedence: is "junk" or |
| 1307 | ($h_sender: matches ^\\d{8}@ and not personal) or |
| 1308 | $message_body contains "this is not spam" |
| 1309 | then |
| 1310 | seen finish |
| 1311 | endif |
| 1312 | |
| 1313 | A condition can be preceded by ^not^ to negate it, and there are also some |
| 1314 | negative forms of condition that are more English-like. |
| 1315 | |
| 1316 | |
| 1317 | |
| 1318 | String testing conditions |
| 1319 | ~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1320 | There are a number of conditions that operate on text strings, using the words |
| 1321 | ``begins'', ``ends'', ``is'', ``contains'' and ``matches''. If you want to apply the same |
| 1322 | test to more than one header line, you can easily concatenate them into a |
| 1323 | single string for testing, as in this example: |
| 1324 | |
| 1325 | if "$h_to:, $h_cc:" contains me@domain.example then ... |
| 1326 | |
| 1327 | If a string-testing condition name is written in lower case, the testing |
| 1328 | of letters is done without regard to case; if it is written in upper case |
| 1329 | (for example, ``CONTAINS''), the case of letters is taken into account. |
| 1330 | |
| 1331 | &&& |
| 1332 | ` `<'text1'>` begins `<'text2'> |
| 1333 | ` `<'text1'>` does not begin `<'text2'> |
| 1334 | `e.g. $header_from: begins "Friend@"` |
| 1335 | &&& |
| 1336 | |
| 1337 | A ``begins'' test checks for the presence of the second string at the start of |
| 1338 | the first, both strings having been expanded. |
| 1339 | |
| 1340 | &&& |
| 1341 | ` `<'text1'>` ends `<'text2'> |
| 1342 | ` `<'text1'>` does not end `<'text2'> |
| 1343 | `e.g. $header_from: ends "public.com.example"` |
| 1344 | &&& |
| 1345 | |
| 1346 | An ``ends'' test checks for the presence of the second string at the end of |
| 1347 | the first, both strings having been expanded. |
| 1348 | |
| 1349 | &&& |
| 1350 | ` `<'text1'>` is `<'text2'> |
| 1351 | ` `<'text1'>` is not `<'text2'> |
| 1352 | `e.g. $local_part_suffix is "-foo"` |
| 1353 | &&& |
| 1354 | |
| 1355 | An ``is'' test does an exact match between the strings, having first expanded |
| 1356 | both strings. |
| 1357 | |
| 1358 | &&& |
| 1359 | ` `<'text1'>` contains `<'text2'> |
| 1360 | ` `<'text1'>` does not contain `<'text2'> |
| 1361 | `e.g. $header_subject: contains "evolution"` |
| 1362 | &&& |
| 1363 | |
| 1364 | A ``contains'' test does a partial string match, having expanded both strings. |
| 1365 | |
| 1366 | &&& |
| 1367 | ` `<'text1'>` matches `<'text2'> |
| 1368 | ` `<'text1'>` does not match `<'text2'> |
| 1369 | `e.g. $sender_address matches "(bill|john)@"` |
| 1370 | &&& |
| 1371 | |
| 1372 | For a ``matches'' test, after expansion of both strings, the second one is |
| 1373 | interpreted as a regular expression. Exim uses the PCRE regular expression |
| 1374 | library, which provides regular expressions that are compatible with Perl. |
| 1375 | |
| 1376 | The match succeeds if the regular expression matches any part of the first |
| 1377 | string. If you want a regular expression to match only at the start or end of |
| 1378 | the subject string, you must encode that requirement explicitly, using the `^` |
| 1379 | or `$` metacharacters. The above example, which is not so constrained, matches |
| 1380 | all these addresses: |
| 1381 | |
| 1382 | bill@test.example |
| 1383 | john@some.example |
| 1384 | spoonbill@example.com |
| 1385 | littlejohn@example.com |
| 1386 | |
| 1387 | To match only the first two, you could use this: |
| 1388 | |
| 1389 | if $sender_address matches "^(bill|john)@" then ... |
| 1390 | |
| 1391 | Care must be taken if you need a backslash in a regular expression, because |
| 1392 | backslashes are interpreted as escape characters both by the string expansion |
| 1393 | code and by Exim's normal processing of strings in quotes. For example, if you |
| 1394 | want to test the sender address for a domain ending in '.com' the regular |
| 1395 | expression is |
| 1396 | |
| 1397 | \.com$ |
| 1398 | |
| 1399 | The backslash and dollar sign in that expression have to be escaped when used |
| 1400 | in a filter command, as otherwise they would be interpreted by the expansion |
| 1401 | code. Thus, what you actually write is |
| 1402 | |
| 1403 | if $sender_address matches \\.com\$ |
| 1404 | |
| 1405 | An alternative way of handling this is to make use of the `\N` expansion |
| 1406 | flag for suppressing expansion: |
| 1407 | |
| 1408 | if $sender_address matches \N\.com$\N |
| 1409 | |
| 1410 | Everything between the two occurrences of `\N` is copied without change by |
| 1411 | the string expander (and in fact you do not need the final one, because it is |
| 1412 | at the end of the string). If the regular expression is given in quotes |
| 1413 | (mandatory only if it contains white space) you have to write either |
| 1414 | |
| 1415 | if $sender_address matches "\\\\.com\\$" |
| 1416 | |
| 1417 | or |
| 1418 | |
| 1419 | if $sender_address matches "\\N\\.com$\\N" |
| 1420 | |
| 1421 | |
| 1422 | If the regular expression contains bracketed sub-expressions, numeric |
| 1423 | variable substitutions such as $1$ can be used in the subsequent actions |
| 1424 | after a successful match. If the match fails, the values of the numeric |
| 1425 | variables remain unchanged. Previous values are not restored after ^endif^. |
| 1426 | In other words, only one set of values is ever available. If the condition |
| 1427 | contains several sub-conditions connected by ^and^ or ^or^, it is the |
| 1428 | strings extracted from the last successful match that are available in |
| 1429 | subsequent actions. Numeric variables from any one sub-condition are also |
| 1430 | available for use in subsequent sub-conditions, because string expansion of a |
| 1431 | condition occurs just before it is tested. |
| 1432 | |
| 1433 | |
| 1434 | Numeric testing conditions |
| 1435 | ~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1436 | The following conditions are available for performing numerical tests: |
| 1437 | |
| 1438 | &&& |
| 1439 | ` `<'number1'>` is above `<'number2'> |
| 1440 | ` `<'number1'>` is not above `<'number2'> |
| 1441 | ` `<'number1'>` is below `<'number2'> |
| 1442 | ` `<'number1'>` is not below `<'number2'> |
| 1443 | `e.g. $message_size is not above 10k` |
| 1444 | &&& |
| 1445 | |
| 1446 | The <'number'> arguments must expand to strings of digits, optionally followed |
| 1447 | by one of the letters K or M (upper case or lower case) which cause |
| 1448 | multiplication by 1024 and 1024x1024 respectively. |
| 1449 | |
| 1450 | |
| 1451 | Testing for significant deliveries |
| 1452 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1453 | You can use the ^delivered^ condition to test whether or not any previously |
| 1454 | obeyed filter commands have set up a significant delivery. For example: |
| 1455 | |
| 1456 | if not delivered then save mail/anomalous endif |
| 1457 | |
| 1458 | ``Delivered'' is perhaps a poor choice of name for this condition, because the |
| 1459 | message has not actually been delivered; rather, a delivery has been set up for |
| 1460 | later processing. |
| 1461 | |
| 1462 | |
| 1463 | Testing for error messages |
| 1464 | ~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1465 | The condition ^error_message^ is true if the incoming message is a bounce |
| 1466 | (mail delivery error) message. Putting the command |
| 1467 | |
| 1468 | if error_message then finish endif |
| 1469 | |
| 1470 | at the head of your filter file is a useful insurance against things going |
| 1471 | wrong in such a way that you cannot receive delivery error reports. *Note*: |
| 1472 | ^error_message^ is a condition, not an expansion variable, and therefore is |
| 1473 | not preceded by `$`. |
| 1474 | |
| 1475 | |
| 1476 | Testing a list of addresses |
| 1477 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1478 | There is a facility for looping through a list of addresses and applying a |
| 1479 | condition to each of them. It takes the form |
| 1480 | |
| 1481 | &&& |
| 1482 | `foranyaddress `<'string'>` (`<'condition'>`)` |
| 1483 | &&& |
| 1484 | |
| 1485 | where <'string'> is interpreted as a list of RFC 2822 addresses, as in a |
| 1486 | typical header line, and <'condition'> is any valid filter condition or |
| 1487 | combination of conditions. The ``group'' syntax that is defined for certain |
| 1488 | header lines that contain addresses is supported. |
| 1489 | |
| 1490 | The parentheses surrounding the condition are mandatory, to delimit it from |
| 1491 | possible further sub-conditions of the enclosing ^if^ command. Within the |
| 1492 | condition, the expansion variable $thisaddress$ is set to the non-comment |
| 1493 | portion of each of the addresses in the string in turn. For example, if the |
| 1494 | string is |
| 1495 | |
| 1496 | B.Simpson <bart@sfld.example>, lisa@sfld.example (his sister) |
| 1497 | |
| 1498 | then $thisaddress$ would take on the values `bart@sfld.example` and |
| 1499 | `lisa@sfld.example` in turn. |
| 1500 | |
| 1501 | If there are no valid addresses in the list, the whole condition is false. If |
| 1502 | the internal condition is true for any one address, the overall condition is |
| 1503 | true and the loop ends. If the internal condition is false for all addresses in |
| 1504 | the list, the overall condition is false. This example tests for the presence |
| 1505 | of an eight-digit local part in any address in a 'To:' header: |
| 1506 | |
| 1507 | if foranyaddress $h_to: ( $thisaddress matches ^\\d{8}@ ) then ... |
| 1508 | |
| 1509 | When the overall condition is true, the value of $thisaddress$ in the |
| 1510 | commands that follow ^then^ is the last value it took on inside the loop. At |
| 1511 | the end of the ^if^ command, the value of $thisaddress$ is reset to what it |
| 1512 | was before. It is best to avoid the use of multiple occurrences of |
| 1513 | ^foranyaddress^, nested or otherwise, in a single ^if^ command, if the |
| 1514 | value of $thisaddress$ is to be used afterwards, because it isn't always |
| 1515 | clear what the value will be. Nested ^if^ commands should be used instead. |
| 1516 | |
| 1517 | Header lines can be joined together if a check is to be applied to more than |
| 1518 | one of them. For example: |
| 1519 | |
| 1520 | if foranyaddress $h_to:,$h_cc: .... |
| 1521 | |
| 1522 | scans through the addresses in both the 'To:' and the 'Cc:' headers. |
| 1523 | |
| 1524 | |
| 1525 | [[SECTpersonal]] |
| 1526 | Testing for personal mail |
| 1527 | ~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1528 | A common requirement is to distinguish between incoming personal mail and mail |
| 1529 | from a mailing list, or from a robot or other automatic process (for example, a |
| 1530 | bounce message). In particular, this test is normally required for ``vacation |
| 1531 | messages''. |
| 1532 | |
| 1533 | The ^personal^ condition checks that the message is not a bounce message and |
| 1534 | that the current user's email address appears in the 'To:' header. It also |
| 1535 | checks that the sender is not the current user or one of a number of common |
| 1536 | daemons, and that there are no header lines starting 'List-' in the message. |
| 1537 | Finally, it checks the content of the 'Precedence:' header line, if there is |
| 1538 | one. |
| 1539 | |
| 1540 | You should always use the ^personal^ condition when generating automatic |
| 1541 | responses. This example shows the use of ^personal^ in a filter file that is |
| 1542 | sending out vacation messages: |
| 1543 | |
| 1544 | if personal then |
| 1545 | mail to $reply_address |
| 1546 | subject "I am on holiday" |
| 1547 | file $home/vacation/message |
| 1548 | once $home/vacation/once |
| 1549 | once_repeat 10d |
| 1550 | endif |
| 1551 | |
| 1552 | It is tempting, when writing commands like the above, to quote the original |
| 1553 | subject in the reply. For example: |
| 1554 | |
| 1555 | subject "Re: $h_subject:" |
| 1556 | |
| 1557 | There is a danger in doing this, however. It may allow a third party to |
| 1558 | subscribe you to an opt-in mailing list, provided that the list accepts bounce |
| 1559 | messages as subscription confirmations. (Messages sent from filters are always |
| 1560 | sent as bounce messages.) Well-managed lists require a non-bounce message to |
| 1561 | confirm a subscription, so the danger is relatively small. |
| 1562 | |
| 1563 | If prefixes or suffixes are in use for local parts -- something which depends |
| 1564 | on the configuration of Exim (see section <<SECTmbox>> below) -- the tests for |
| 1565 | the current user are done with the full address (including the prefix and |
| 1566 | suffix, if any) as well as with the prefix and suffix removed. If the system is |
| 1567 | configured to rewrite local parts of mail addresses, for example, to rewrite |
| 1568 | `dag46` as `Dirk.Gently`, the rewritten form of the address is also used in |
| 1569 | the tests. |
| 1570 | |
| 1571 | |
| 1572 | |
| 1573 | Alias addresses for the personal condition |
| 1574 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1575 | It is quite common for people who have mail accounts on a number of different |
| 1576 | systems to forward all their mail to one system, and in this case a check for |
| 1577 | personal mail should test all their various mail addresses. To allow for this, |
| 1578 | the ^personal^ condition keyword can be followed by |
| 1579 | |
| 1580 | &&& |
| 1581 | `alias `<'address'> |
| 1582 | &&& |
| 1583 | |
| 1584 | any number of times, for example |
| 1585 | |
| 1586 | if personal alias smith@else.where.example |
| 1587 | alias jones@other.place.example |
| 1588 | then ... |
| 1589 | |
| 1590 | The alias addresses are treated as alternatives to the current user's email |
| 1591 | address when testing the contents of header lines. |
| 1592 | |
| 1593 | |
| 1594 | Details of the personal condition |
| 1595 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1596 | The basic ^personal^ test is roughly equivalent to the following: |
| 1597 | |
| 1598 | not error_message and |
| 1599 | $message_headers does not contain "\nList-Id:" and |
| 1600 | $message_headers does not contain "\nList-Help:" and |
| 1601 | $message_headers does not contain "\nList-Subscribe:" and |
| 1602 | $message_headers does not contain "\nList-Unsubscribe:" and |
| 1603 | $message_headers does not contain "\nList-Post:" and |
| 1604 | $message_headers does not contain "\nList-Owner:" and |
| 1605 | $message_headers does not contain "\nList-Archive:" and |
| 1606 | ( |
| 1607 | "${if def h_auto-submitted:{present}{absent}}" is "absent" or |
| 1608 | $header_auto-submitted: is "no" |
| 1609 | ) and |
| 1610 | $header_precedence: does not contain "bulk" and |
| 1611 | $header_precedence: does not contain "list" and |
| 1612 | $header_precedence: does not contain "junk" and |
| 1613 | foranyaddress $header_to: |
| 1614 | ( $thisaddress contains "$local_part$domain" ) and |
| 1615 | not foranyaddress $header_from: |
| 1616 | ( |
| 1617 | $thisaddress contains "$local_partdomain" or |
| 1618 | $thisaddress contains "server" or |
| 1619 | $thisaddress contains "daemon" or |
| 1620 | $thisaddress contains "root" or |
| 1621 | $thisaddress contains "listserv" or |
| 1622 | $thisaddress contains "majordomo" or |
| 1623 | $thisaddress contains "-request" or |
| 1624 | $thisaddress matches "^owner-[^]+" |
| 1625 | ) |
| 1626 | |
| 1627 | The variable $local_part$ contains the local part of the mail address of |
| 1628 | the user whose filter file is being run -- it is normally your login id. The |
| 1629 | $domain$ variable contains the mail domain. As explained above, if aliases |
| 1630 | or rewriting are defined, or if prefixes or suffixes are in use, the tests for |
| 1631 | the current user are also done with alternative addresses. |
| 1632 | |
| 1633 | |
| 1634 | |
| 1635 | |
| 1636 | Testing delivery status |
| 1637 | ~~~~~~~~~~~~~~~~~~~~~~~ |
| 1638 | There are two conditions that are intended mainly for use in system filter |
| 1639 | files, but which are available in users' filter files as well. The condition |
| 1640 | ^first_delivery^ is true if this is the first process that is attempting to |
| 1641 | deliver the message, and false otherwise. This indicator is not reset until the |
| 1642 | first delivery process successfully terminates; if there is a crash or a power |
| 1643 | failure (for example), the next delivery attempt is also a ``first delivery''. |
| 1644 | |
| 1645 | In a user filter file ^first_delivery^ will be false if there was previously an |
| 1646 | error in the filter, or if a delivery for the user failed owing to, for |
| 1647 | example, a quota error, or if forwarding to a remote address was deferred for |
| 1648 | some reason. |
| 1649 | |
| 1650 | The condition ^manually_thawed^ is true if the message was ``frozen'' for some |
| 1651 | reason, and was subsequently released by the system administrator. It is |
| 1652 | unlikely to be of use in users' filter files. |
| 1653 | |
| 1654 | |
| 1655 | [[SECTmbox]] |
| 1656 | Multiple personal mailboxes |
| 1657 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1658 | The system administrator can configure Exim so that users can set up variants |
| 1659 | on their email addresses and handle them separately. Consult your system |
| 1660 | administrator or local documentation to see if this facility is enabled on your |
| 1661 | system, and if so, what the details are. |
| 1662 | |
| 1663 | The facility involves the use of a prefix or a suffix on an email address. For |
| 1664 | example, all mail addressed to 'lg303-'<'something'> would be the property of |
| 1665 | user 'lg303', who could determine how it was to be handled, depending on the |
| 1666 | value of <'something'>. |
| 1667 | |
| 1668 | There are two possible ways in which this can be set up. The first possibility |
| 1669 | is the use of multiple '.forward' files. In this case, mail to 'lg303-foo', |
| 1670 | for example, is handled by looking for a file called _.forward-foo_ in |
| 1671 | 'lg303'{ap}s home directory. If such a file does not exist, delivery fails and the |
| 1672 | message is returned to its sender. |
| 1673 | |
| 1674 | The alternative approach is to pass all messages through a single _.forward_ |
| 1675 | file, which must be a filter file so that it can distinguish between the |
| 1676 | different cases by referencing the variables $local_part_prefix$ or |
| 1677 | $local_part_suffix$, as in the final example in section <<SECTex>> below. |
| 1678 | |
| 1679 | It is possible to configure Exim to support both schemes at once. In this case, |
| 1680 | a specific _.forward-foo_ file is first sought; if it is not found, the basic |
| 1681 | _.forward_ file is used. |
| 1682 | |
| 1683 | The ^personal^ test (see section <<SECTpersonal>>) includes prefixes and |
| 1684 | suffixes in its checking. |
| 1685 | |
| 1686 | |
| 1687 | |
| 1688 | Ignoring delivery errors |
| 1689 | ~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1690 | As was explained above, filtering just sets up addresses for delivery -- no |
| 1691 | deliveries are actually done while a filter file is active. If any of the |
| 1692 | generated addresses subsequently suffers a delivery failure, an error message |
| 1693 | is generated in the normal way. However, if a filter command that sets up a |
| 1694 | delivery is preceded by the word ^noerror^, errors for that delivery, |
| 1695 | 'and any deliveries consequent on it' (that is, from alias, forwarding, or |
| 1696 | filter files it invokes) are ignored. |
| 1697 | |
| 1698 | |
| 1699 | |
| 1700 | [[SECTex]] |
| 1701 | Examples of Exim filter commands |
| 1702 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 1703 | Simple forwarding: |
| 1704 | |
| 1705 | # Exim filter |
| 1706 | deliver baggins@rivendell.middle-earth.example |
| 1707 | |
| 1708 | Vacation handling using traditional means, assuming that the _.vacation.msg_ |
| 1709 | and other files have been set up in your home directory: |
| 1710 | |
| 1711 | # Exim filter |
| 1712 | unseen pipe "/usr/ucb/vacation \"$local_part\"" |
| 1713 | |
| 1714 | Vacation handling inside Exim, having first created a file called |
| 1715 | _.vacation.msg_ in your home directory: |
| 1716 | |
| 1717 | # Exim filter |
| 1718 | if personal then vacation endif |
| 1719 | |
| 1720 | File some messages by subject: |
| 1721 | |
| 1722 | # Exim filter |
| 1723 | if $header_subject: contains "empire" or |
| 1724 | $header_subject: contains "foundation" |
| 1725 | then |
| 1726 | save $home/mail/f+e |
| 1727 | endif |
| 1728 | |
| 1729 | Save all non-urgent messages by weekday: |
| 1730 | |
| 1731 | # Exim filter |
| 1732 | if $header_subject: does not contain "urgent" and |
| 1733 | $tod_full matches "^(...)," |
| 1734 | then |
| 1735 | save $home/mail/$1 |
| 1736 | endif |
| 1737 | |
| 1738 | Throw away all mail from one site, except from postmaster: |
| 1739 | |
| 1740 | # Exim filter |
| 1741 | if $reply_address contains "@spam.site.example" and |
| 1742 | $reply_address does not contain "postmaster@" |
| 1743 | then |
| 1744 | seen finish |
| 1745 | endif |
| 1746 | |
| 1747 | Handle multiple personal mailboxes: |
| 1748 | |
| 1749 | # Exim filter |
| 1750 | if $local_part_suffix is "-foo" |
| 1751 | then |
| 1752 | save $home/mail/foo |
| 1753 | elif $local_part_suffix is "-bar" |
| 1754 | then |
| 1755 | save $home/mail/bar |
| 1756 | endif |
| 1757 | |
| 1758 | |