1 . /////////////////////////////////////////////////////////////////////////////
2 . This is the primary source of the document that describes Exim's filtering
3 . facilities. It is an xfpt document that is converted into DocBook XML for
4 . subsequent conversion into printing and online formats. The markup used
5 . herein is "standard" xfpt markup, with some extras. The markup is summarized
6 . in a file called Markup.txt.
7 . /////////////////////////////////////////////////////////////////////////////
11 .include ./local_params
14 . /////////////////////////////////////////////////////////////////////////////
15 . These lines are processing instructions for the Simple DocBook Processor that
16 . Philip Hazel has developed as a less cumbersome way of making PostScript and
17 . PDFs than using xmlto and fop. They will be ignored by all other XML
19 . /////////////////////////////////////////////////////////////////////////////
23 foot_right_recto="&chaptertitle;"
24 foot_right_verso="&chaptertitle;"
25 table_warn_overflow="overprint"
26 toc_chapter_blanks="yes,yes"
27 toc_title="Exim's interfaces to mail filtering"
33 . ===========================================================================
34 . Additional xfpt markup used by this document, over and above the default
35 . provided in the xfpt library.
37 . Override the &$ flag to automatically insert a $ with the variable name
39 .flag &$ $& "<varname>$" "</varname>"
41 . A macro for the common 2-column tables
43 .macro table2 100pt 300pt
44 .itable none 0 0 2 $1 left $2 left
46 . ===========================================================================
49 . /////////////////////////////////////////////////////////////////////////////
50 . /////////////////////////////////////////////////////////////////////////////
52 . This preliminary stuff creates a <bookinfo> entry in the XML. This is removed
53 . when creating the PostScript/PDF output, because we do not want a full-blown
54 . title page created for those versions. When fop is being used to create
55 . PS/PDF, the stylesheet fudges up a title line to replace the text "Table of
56 . contents". When SDoP is being used, a processing instruction does this job.
57 . For the other forms of output, the <bookinfo> element is retained and used.
61 <title>Exim's interfaces to mail filtering</title>
62 <titleabbrev>Exim filtering</titleabbrev>
66 <author><firstname>Philip</firstname><surname>Hazel</surname></author>
67 <authorinitials>PH</authorinitials>
68 <revhistory><revision>
75 <authorinitials>PH</authorinitials>
76 </revision></revhistory>
79 </year><holder>University of Cambridge</holder></copyright>
83 . /////////////////////////////////////////////////////////////////////////////
84 . /////////////////////////////////////////////////////////////////////////////
87 .chapter "Forwarding and filtering in Exim" "CHAPforandfilt"
88 This document describes the user interfaces to Exim's in-built mail filtering
89 facilities, and is copyright © University of Cambridge &year(). It
90 corresponds to Exim version &version().
94 .section "Introduction" "SEC00"
95 Most Unix mail transfer agents (programs that deliver mail) permit individual
96 users to specify automatic forwarding of their mail, usually by placing a list
97 of forwarding addresses in a file called &_.forward_& in their home
98 directories. Exim extends this facility by allowing the forwarding instructions
99 to be a set of rules rather than just a list of addresses, in effect providing
100 &"&_.forward_& with conditions"&. Operating the set of rules is called
101 &'filtering'&, and the file that contains them is called a &'filter file'&.
103 Exim supports two different kinds of filter file. An &'Exim filter'& contains
104 instructions in a format that is unique to Exim. A &'Sieve filter'& contains
105 instructions in the Sieve format that is defined by RFC 3028. As this is a
106 standard format, Sieve filter files may already be familiar to some users.
107 Sieve files should also be portable between different environments. However,
108 the Exim filtering facility contains more features (such as variable
109 expansion), and better integration with the host environment (such as the use
110 of external processes and pipes).
112 The choice of which kind of filter to use can be left to the end-user, provided
113 that the system administrator has configured Exim appropriately for both kinds
114 of filter. However, if interoperability is important, Sieve is the only
117 The ability to use filtering or traditional forwarding has to be enabled by the
118 system administrator, and some of the individual facilities can be separately
119 enabled or disabled. A local document should be provided to describe exactly
120 what has been enabled. In the absence of this, consult your system
123 This document describes how to use a filter file and the format of its
124 contents. It is intended for use by end-users. Both Sieve filters and Exim
125 filters are covered. However, for Sieve filters, only issues that relate to the
126 Exim implementation are discussed, since Sieve itself is described elsewhere.
128 The contents of traditional &_.forward_& files are not described here. They
129 normally contain just a list of addresses, file names, or pipe commands,
130 separated by commas or newlines, but other types of item are also available.
131 The full details can be found in the chapter on the &(redirect)& router in the
132 Exim specification, which also describes how the system administrator can set
133 up and control the use of filtering.
137 .section "Filter operation" "SEC01"
138 It is important to realize that, in Exim, no deliveries are actually made while
139 a filter or traditional &_.forward_& file is being processed. Running a filter
140 or processing a traditional &_.forward_& file sets up future delivery
141 operations, but does not carry them out.
143 The result of filter or &_.forward_& file processing is a list of destinations
144 to which a message should be delivered. The deliveries themselves take place
145 later, along with all other deliveries for the message. This means that it is
146 not possible to test for successful deliveries while filtering. It also means
147 that any duplicate addresses that are generated are dropped, because Exim never
148 delivers the same message to the same address more than once.
153 .section "Testing a new filter file" "SECTtesting"
154 Filter files, especially the more complicated ones, should always be tested, as
155 it is easy to make mistakes. Exim provides a facility for preliminary testing
156 of a filter file before installing it. This tests the syntax of the file and
157 its basic operation, and can also be used with traditional &_.forward_& files.
159 Because a filter can do tests on the content of messages, a test message is
160 required. Suppose you have a new filter file called &_myfilter_& and a test
161 message in a file called &_test-message_&. Assuming that Exim is installed with
162 the conventional path name &_/usr/sbin/sendmail_& (some operating systems use
163 &_/usr/lib/sendmail_&), the following command can be used:
165 /usr/sbin/sendmail -bf myfilter <test-message
167 The &%-bf%& option tells Exim that the following item on the command line is
168 the name of a filter file that is to be tested. There is also a &%-bF%& option,
169 which is similar, but which is used for testing system filter files, as opposed
170 to user filter files, and which is therefore of use only to the system
173 The test message is supplied on the standard input. If there are no
174 message-dependent tests in the filter, an empty file (&_/dev/null_&) can be
175 used. A supplied message must start with header lines or the &"From&~"& message
176 separator line that is found in many multi-message folder files. Note that
177 blank lines at the start terminate the header lines. A warning is given if no
178 header lines are read.
180 The result of running this command, provided no errors are detected in the
181 filter file, is a list of the actions that Exim would try to take if presented
182 with the message for real. For example, for an Exim filter, the output
184 Deliver message to: gulliver@lilliput.fict.example
185 Save message to: /home/lemuel/mail/archive
187 means that one copy of the message would be sent to
188 &'gulliver@lilliput.fict.example'&, and another would be added to the file
189 &_/home/lemuel/mail/archive_&, if all went well.
191 The actions themselves are not attempted while testing a filter file in this
192 way; there is no check, for example, that any forwarding addresses are valid.
193 For an Exim filter, if you want to know why a particular action is being taken,
194 add the &%-v%& option to the command. This causes Exim to output the results of
195 any conditional tests and to indent its output according to the depth of
196 nesting of &(if)& commands. Further additional output from a filter test can be
197 generated by the &(testprint)& command, which is described below.
199 When Exim is outputting a list of the actions it would take, if any text
200 strings are included in the output, non-printing characters therein are
201 converted to escape sequences. In particular, if any text string contains a
202 newline character, this is shown as &"\n"& in the testing output.
204 When testing a filter in this way, Exim makes up an &"envelope"& for the
205 message. The recipient is by default the user running the command, and so is
206 the sender, but the command can be run with the &%-f%& option to supply a
207 different sender. For example,
209 /usr/sbin/sendmail -bf myfilter \
210 -f islington@never.where <test-message
212 Alternatively, if the &%-f%& option is not used, but the first line of the
213 supplied message is a &"From&~"& separator from a message folder file (not the
214 same thing as a &'From:'& header line), the sender is taken from there. If
215 &%-f%& is present, the contents of any &"From&~"& line are ignored.
217 The &"return path"& is the same as the envelope sender, unless the message
218 contains a &'Return-path:'& header, in which case it is taken from there. You
219 need not worry about any of this unless you want to test out features of a
220 filter file that rely on the sender address or the return path.
222 It is possible to change the envelope recipient by specifying further options.
223 The &%-bfd%& option changes the domain of the recipient address, while the
224 &%-bfl%& option changes the &"local part"&, that is, the part before the @
225 sign. An adviser could make use of these to test someone else's filter file.
227 The &%-bfp%& and &%-bfs%& options specify the prefix or suffix for the local
228 part. These are relevant only when support for multiple personal mailboxes is
229 implemented; see the description in section &<<SECTmbox>>& below.
232 .section "Installing a filter file" "SEC02"
233 A filter file is normally installed under the name &_.forward_& in your home
234 directory &-- it is distinguished from a conventional &_.forward_& file by its
235 first line (described below). However, the file name is configurable, and some
236 system administrators may choose to use some different name or location for
240 .section "Testing an installed filter file" "SEC03"
241 Testing a filter file before installation cannot find every potential problem;
242 for example, it does not actually run commands to which messages are piped.
243 Some &"live"& tests should therefore also be done once a filter is installed.
245 If at all possible, test your filter file by sending messages from some other
246 account. If you send a message to yourself from the filtered account, and
247 delivery fails, the error message will be sent back to the same account, which
248 may cause another delivery failure. It won't cause an infinite sequence of such
249 messages, because delivery failure messages do not themselves generate further
250 messages. However, it does mean that the failure won't be returned to you, and
251 also that the postmaster will have to investigate the stuck message.
253 If you have to test an Exim filter from the same account, a sensible precaution
254 is to include the line
256 if error_message then finish endif
258 as the first filter command, at least while testing. This causes filtering to
259 be abandoned for a delivery failure message, and since no destinations are
260 generated, the message goes on to be delivered to the original address. Unless
261 there is a good reason for not doing so, it is recommended that the above test
262 be left in all Exim filter files. (This does not apply to Sieve files.)
266 .section "Details of filtering commands" "SEC04"
267 The filtering commands for Sieve and Exim filters are completely different in
268 syntax and semantics. The Sieve mechanism is defined in RFC 3028; in the next
269 chapter we describe how it is integrated into Exim. The subsequent chapter
270 covers Exim filtering commands in detail.
274 .chapter "Sieve filter files" "CHAPsievefilter"
275 The code for Sieve filtering in Exim was contributed by Michael Haardt, and
276 most of the content of this chapter is taken from the notes he provided. Since
277 Sieve is an extensible language, it is important to understand &"Sieve"& in
278 this context as &"the specific implementation of Sieve for Exim"&.
280 This chapter does not contain a description of Sieve, since that can be found
281 in RFC 3028, which should be read in conjunction with these notes.
283 The Exim Sieve implementation offers the core as defined by RFC 3028,
284 comparison tests, the subaddress parameter, the &*copy*&, &*envelope*&,
285 &*fileinto*&, &*notify*&, and &*vacation*& extensions, but not the &*reject*&
286 extension. Exim does not support message delivery notifications (MDNs), so
287 adding it just to the Sieve filter (as required for &*reject*&) makes little
290 In order for Sieve to work properly in Exim, the system administrator needs to
291 make some adjustments to the Exim configuration. These are described in the
292 chapter on the &(redirect)& router in the full Exim specification.
295 .section "Recognition of Sieve filters" "SEC05"
296 A filter file is interpreted as a Sieve filter if its first line is
300 This is what distinguishes it from a conventional &_.forward_& file or an Exim
305 .section "Saving to specified folders" "SEC06"
306 If the system administrator has set things up as suggested in the Exim
307 specification, and you use &(keep)& or &(fileinto)& to save a mail into a
308 folder, absolute files are stored where specified, relative files are stored
309 relative to &$home$&, and &_inbox_& goes to the standard mailbox location.
313 .section "Strings containing header names" "SEC07"
314 RFC 3028 does not specify what happens if a string denoting a header field does
315 not contain a valid header name, for example, it contains a colon. This
316 implementation generates an error instead of ignoring the header field in order
317 to ease script debugging, which fits in with the common picture of Sieve.
321 .section "Exists test with empty list of headers" "SEC08"
322 The &*exists*& test succeeds only if all the specified headers exist. RFC 3028
323 does not explicitly specify what happens on an empty list of headers. This
324 implementation evaluates that condition as true, interpreting the RFC in a
329 .section "Header test with invalid MIME encoding in header" "SEC09"
330 Some MUAs process invalid base64 encoded data, generating junk. Others ignore
331 junk after seeing an equal sign in base64 encoded data. RFC 2047 does not
332 specify how to react in this case, other than stating that a client must not
333 forbid to process a message for that reason. RFC 2045 specifies that invalid
334 data should be ignored (apparently looking at end of line characters). It also
335 specifies that invalid data may lead to rejecting messages containing them (and
336 there it appears to talk about true encoding violations), which is a clear
337 contradiction to ignoring them.
339 RFC 3028 does not specify how to process incorrect MIME words. This
340 implementation treats them literally, as it does if the word is correct but its
341 character set cannot be converted to UTF-8.
345 .section "Address test for multiple addresses per header" "SEC10"
346 A header may contain multiple addresses. RFC 3028 does not explicitly specify
347 how to deal with them, but since the address test checks if anything matches
348 anything else, matching one address suffices to satisfy the condition. That
349 makes it impossible to test if a header contains a certain set of addresses and
350 no more, but it is more logical than letting the test fail if the header
351 contains an additional address besides the one the test checks for.
355 .section "Semantics of keep" "SEC11"
356 The &(keep)& command is equivalent to
360 It saves the message and resets the implicit keep flag. It does not set the
361 implicit keep flag; there is no command to set it once it has been reset.
365 .section "Semantics of fileinto" "SEC12"
366 RFC 3028 does not specify whether &(fileinto)& should try to create a mail
367 folder if it does not exist. This implementation allows the sysadmin to
368 configure that aspect using the &(appendfile)& transport options
369 &%create_directory%&, &%create_file%&, and &%file_must_exist%&. See the
370 &(appendfile)& transport in the Exim specification for details.
374 .section "Semantics of redirect" "SEC13"
375 Sieve scripts are supposed to be interoperable between servers, so this
376 implementation does not allow mail to be redirected to unqualified addresses,
377 because the domain would depend on the system being used. On systems with
378 virtual mail domains, the default domain is probably not what the user expects
383 .section "String arguments" "SEC14"
384 There has been confusion if the string arguments to &(require)& are to be
385 matched case-sensitively or not. This implementation matches them with the
386 match type &(:is)& (default, see section 2.7.1 of the RFC) and the comparator
387 &(i;ascii-casemap)& (default, see section 2.7.3 of the RFC). The RFC defines
388 the command defaults clearly, so any different implementations violate RFC
389 3028. The same is valid for comparator names, also specified as strings.
393 .section "Number units" "SEC15"
394 There is a mistake in RFC 3028: the suffix G denotes gibi-, not tebibyte.
395 The mistake is obvious, because RFC 3028 specifies G to denote 2^30
396 (which is gibi, not tebi), and that is what this implementation uses as
397 the scaling factor for the suffix G.
401 .section "RFC compliance" "SEC16"
402 Exim requires the first line of a Sieve filter to be
406 Of course the RFC does not specify that line. Do not expect examples to work
407 without adding it, though.
409 RFC 3028 requires the use of CRLF to terminate a line. The rationale was that
410 CRLF is universally used in network protocols to mark the end of the line. This
411 implementation does not embed Sieve in a network protocol, but uses Sieve
412 scripts as part of the Exim MTA. Since all parts of Exim use LF as the newline
413 character, this implementation does, too, by default, though the system
414 administrator may choose (at Exim compile time) to use CRLF instead.
416 Exim violates RFC 2822, section 3.6.8, by accepting 8-bit header names, so this
417 implementation repeats this violation to stay consistent with Exim. This is in
418 preparation for UTF-8 data.
420 Sieve scripts cannot contain NUL characters in strings, but mail headers could
421 contain MIME encoded NUL characters, which could never be matched by Sieve
422 scripts using exact comparisons. For that reason, this implementation extends
423 the Sieve quoted string syntax with \0 to describe a NUL character, violating
424 \0 being the same as 0 in RFC 3028. Even without using \0, the following tests
425 are all true in this implementation. Implementations that use C-style strings
426 will only evaluate the first test as true.
428 Subject: =?iso-8859-1?q?abc=00def
430 header :contains "Subject" ["abc"]
431 header :contains "Subject" ["def"]
432 header :matches "Subject" ["abc?def"]
434 Note that by considering Sieve to be an MUA, RFC 2047 can be interpreted in a
435 way that NUL characters truncating strings is allowed for Sieve
436 implementations, although not recommended. It is further allowed to use encoded
437 NUL characters in headers, but that's not recommended either. The above example
440 RFC 3028 states that if an implementation fails to convert a character set to
441 UTF-8, two strings cannot be equal if one contains octets greater than 127.
442 Assuming that all unknown character sets are one-byte character sets with the
443 lower 128 octets being US-ASCII is not sound, so this implementation violates
444 RFC 3028 and treats such MIME words literally. That way at least something
447 The folder specified by &(fileinto)& must not contain the character sequence
448 &".."& to avoid security problems. RFC 3028 does not specify the syntax of
449 folders apart from &(keep)& being equivalent to
453 This implementation uses &_inbox_& instead.
455 Sieve script errors currently cause messages to be silently filed into
456 &_inbox_&. RFC 3028 requires that the user is notified of that condition.
457 This may be implemented in the future by adding a header line to mails that
458 are filed into &_inbox_& due to an error in the filter.
462 .chapter "Exim filter files" "CHAPeximfilter"
463 This chapter contains a full description of the contents of Exim filter files.
466 .section "Format of Exim filter files" "SEC17"
467 Apart from leading white space, the first text in an Exim filter file must be
471 This is what distinguishes it from a conventional &_.forward_& file or a Sieve
472 filter file. If the file does not have this initial line (or the equivalent for
473 a Sieve filter), it is treated as a conventional &_.forward_& file, both when
474 delivering mail and when using the &%-bf%& testing mechanism. The white space
475 in the line is optional, and any capitalization may be used. Further text on
476 the same line is treated as a comment. For example, you could have
478 # Exim filter <<== do not edit or remove this line!
480 The remainder of the file is a sequence of filtering commands, which consist of
481 keywords and data values. For example, in the command
483 deliver gulliver@lilliput.fict.example
485 the keyword is &`deliver`& and the data value is
486 &`gulliver@lilliput.fict.example`&. White space or line breaks separate the
487 components of a command, except in the case of conditions for the &(if)&
488 command, where round brackets (parentheses) also act as separators. Complete
489 commands are separated from each other by white space or line breaks; there are
490 no special terminators. Thus, several commands may appear on one line, or one
491 command may be spread over a number of lines.
493 If the character # follows a separator anywhere in a command, everything from
494 # up to the next newline is ignored. This provides a way of including comments
498 .section "Data values in filter commands" "SEC18"
499 There are two ways in which a data value can be input:
502 If the text contains no white space, it can be typed verbatim. However, if it
503 is part of a condition, it must also be free of round brackets (parentheses),
504 as these are used for grouping in conditions.
506 Otherwise, text must be enclosed in double quotation marks. In this case, the
507 character \ (backslash) is treated as an &"escape character"& within the
508 string, causing the following character or characters to be treated specially:
510 &`\n`& is replaced by a newline
511 &`\r`& is replaced by a carriage return
512 &`\t`& is replaced by a tab
516 Backslash followed by up to three octal digits is replaced by the character
517 specified by those digits, and &`\x`& followed by up to two hexadecimal digits
518 is treated similarly. Backslash followed by any other character is replaced by
519 the second character, so that in particular, &`\"`& becomes &`"`& and &`\\`&
520 becomes &`\`&. A data item enclosed in double quotes can be continued onto the
521 next line by ending the first line with a backslash. Any leading white space at
522 the start of the continuation line is ignored.
524 In addition to the escape character processing that occurs when strings are
525 enclosed in quotes, most data values are also subject to &'string expansion'&
526 (as described in the next section), in which case the characters &`$`& and
527 &`\`& are also significant. This means that if a single backslash is actually
528 required in such a string, and the string is also quoted, &`\\\\`& has to be
531 The maximum permitted length of a data string, before expansion, is 1024
535 .section "String expansion" "SECTfilterstringexpansion"
536 Most data values are expanded before use. Expansion consists of replacing
537 substrings beginning with &`$`& with other text. The full expansion facilities
538 available in Exim are extensive. If you want to know everything that Exim can
539 do with strings, you should consult the chapter on string expansion in the Exim
542 In filter files, by far the most common use of string expansion is the
543 substitution of the contents of a variable. For example, the substring
547 is replaced by the address to which replies to the message should be sent. If
548 such a variable name is followed by a letter or digit or underscore, it must be
549 enclosed in curly brackets (braces), for example,
553 If a &`$`& character is actually required in an expanded string, it must be
554 escaped with a backslash, and because backslash is also an escape character in
555 quoted input strings, it must be doubled in that case. The following two
556 examples illustrate two different ways of testing for a &`$`& character in a
559 if $message_body contains \$ then ...
560 if $message_body contains "\\$" then ...
562 You can prevent part of a string from being expanded by enclosing it between
563 two occurrences of &`\N`&. For example,
565 if $message_body contains \N$$$$\N then ...
567 tests for a run of four dollar characters.
570 .section "Some useful general variables" "SEC19"
571 A complete list of the available variables is given in the Exim documentation.
572 This shortened list contains the ones that are most likely to be useful in
573 personal filter files:
575 &$body_linecount$&: The number of lines in the body of the message.
577 &$body_zerocount$&: The number of binary zero characters in the body of the
580 &$home$&: In conventional configurations, this variable normally contains the
581 user's home directory. The system administrator can, however, change this.
583 &$local_part$&: The part of the email address that precedes the @ sign &--
584 normally the user's login name. If support for multiple personal mailboxes is
585 enabled (see section &<<SECTmbox>>& below) and a prefix or suffix for the local
586 part was recognized, it is removed from the string in this variable.
588 &$local_part_prefix$&: If support for multiple personal mailboxes is enabled
589 (see section &<<SECTmbox>>& below), and a local part prefix was recognized,
590 this variable contains the prefix. Otherwise it contains an empty string.
592 &$local_part_suffix$&: If support for multiple personal mailboxes is enabled
593 (see section &<<SECTmbox>>& below), and a local part suffix was recognized,
594 this variable contains the suffix. Otherwise it contains an empty string.
596 &$message_body$&: The initial portion of the body of the message. By default,
597 up to 500 characters are read into this variable, but the system administrator
598 can configure this to some other value. Newlines in the body are converted into
601 &$message_body_end$&: The final portion of the body of the message, formatted
602 and limited in the same way as &$message_body$&.
604 &$message_body_size$&: The size of the body of the message, in bytes.
606 &$message_exim_id$&: The message's local identification string, which is unique
607 for each message handled by a single host.
609 &$message_headers$&: The header lines of the message, concatenated into a
610 single string, with newline characters between them.
612 &$message_size$&: The size of the entire message, in bytes.
614 &$original_local_part$&: When an address that arrived with the message is
615 being processed, this contains the same value as the variable &$local_part$&.
616 However, if an address generated by an alias, forward, or filter file is being
617 processed, this variable contains the local part of the original address.
619 &$reply_address$&: The contents of the &'Reply-to:'& header, if the message
620 has one; otherwise the contents of the &'From:'& header. It is the address to
621 which normal replies to the message should be sent.
623 &$return_path$&: The return path &-- that is, the sender field that will be
624 transmitted as part of the message's envelope if the message is sent to another
625 host. This is the address to which delivery errors are sent. In many cases,
626 this variable has the same value as &$sender_address$&, but if, for example,
627 an incoming message to a mailing list has been expanded, &$return_path$& may
628 have been changed to contain the address of the list maintainer.
630 &$sender_address$&: The sender address that was received in the envelope of
631 the message. This is not necessarily the same as the contents of the &'From:'&
632 or &'Sender:'& header lines. For delivery error messages (&"bounce messages"&)
633 there is no sender address, and this variable is empty.
635 &$tod_full$&: A full version of the time and date, for example: Wed, 18 Oct
636 1995 09:51:40 +0100. The timezone is always given as a numerical offset from
639 &$tod_log$&: The time and date in the format used for writing Exim's log files,
640 without the timezone, for example: 1995-10-12 15:32:29.
642 &$tod_zone$&: The local timezone offset, for example: +0100.
646 .section "Header variables" "SECTheadervariables"
647 There is a special set of expansion variables containing the header lines of
648 the message being processed. These variables have names beginning with
649 &$header_$& followed by the name of the header line, terminated by a colon.
655 The whole item, including the terminating colon, is replaced by the contents of
656 the message header line. If there is more than one header line with the same
657 name, their contents are concatenated. For header lines whose data consists of
658 a list of addresses (for example, &'From:'& and &'To:'&), a comma and newline
659 is inserted between each set of data. For all other header lines, just a
662 Leading and trailing white space is removed from header line data, and if there
663 are any MIME &"words"& that are encoded as defined by RFC 2047 (because they
664 contain non-ASCII characters), they are decoded and translated, if possible, to
665 a local character set. Translation is attempted only on operating systems that
666 have the &[iconv()]& function. This makes the header line look the same as it
667 would when displayed by an MUA. The default character set is ISO-8859-1, but
668 this can be changed by means of the &(headers)& command (see below).
670 If you want to see the actual characters that make up a header line, you can
671 specify &$rheader_$& instead of &$header_$&. This inserts the &"raw"&
672 header line, unmodified.
674 There is also an intermediate form, requested by &$bheader_$&, which removes
675 leading and trailing space and decodes MIME &"words"&, but does not do any
676 character translation. If an attempt to decode what looks superficially like a
677 MIME &"word"& fails, the raw string is returned. If decoding produces a binary
678 zero character, it is replaced by a question mark.
680 The capitalization of the name following &$header_$& is not significant.
681 Because any printing character except colon may appear in the name of a
682 message's header (this is a requirement of RFC 2822, the document that
683 describes the format of a mail message) curly brackets must &'not'& be used in
684 this case, as they will be taken as part of the header name. Two shortcuts are
685 allowed in naming header variables:
688 The initiating &$header_$&, &$rheader_$&, or &$bheader_$& can be
689 abbreviated to &$h_$&, &$rh_$&, or &$bh_$&, respectively.
691 The terminating colon can be omitted if the next character is white space. The
692 white space character is retained in the expanded string. However, this is not
693 recommended, because it makes it easy to forget the colon when it really is
697 If the message does not contain a header of the given name, an empty string is
698 substituted. Thus it is important to spell the names of headers correctly. Do
699 not use &$header_Reply_to$& when you really mean &$header_Reply-to$&.
702 .section "User variables" "SEC20"
703 There are ten user variables with names &$n0$& &-- &$n9$& that can be
704 incremented by the &(add)& command (see section &<<SECTadd>>&). These can be
705 used for &"scoring"& messages in various ways. If Exim is configured to run a
706 &"system filter"& on every message, the values left in these variables are
707 copied into the variables &$sn0$& &-- &$sn9$& at the end of the system filter,
708 thus making them available to users' filter files. How these values are used is
709 entirely up to the individual installation.
712 .section "Current directory" "SEC21"
713 The contents of your filter file should not make any assumptions about the
714 current directory. It is best to use absolute paths for file names; you can
715 normally make use of the &$home$& variable to refer to your home directory. The
716 &(save)& command automatically inserts &$home$& at the start of non-absolute
722 .section "Significant deliveries" "SECTsigdel"
723 When in the course of delivery a message is processed by a filter file, what
724 happens next, that is, after the filter file has been processed, depends on
725 whether or not the filter sets up any &'significant deliveries'&. If at least
726 one significant delivery is set up, the filter is considered to have handled
727 the entire delivery arrangements for the current address, and no further
728 processing of the address takes place. If, however, no significant deliveries
729 are set up, Exim continues processing the current address as if there were no
730 filter file, and typically sets up a delivery of a copy of the message into a
731 local mailbox. In particular, this happens in the special case of a filter file
732 containing only comments.
734 The delivery commands &(deliver)&, &(save)&, and &(pipe)& are by default
735 significant. However, if such a command is preceded by the word &"unseen"&, its
736 delivery is not considered to be significant. In contrast, other commands such
737 as &(mail)& and &(vacation)& do not set up significant deliveries unless
738 preceded by the word &"seen"&. The following example commands set up
739 significant deliveries:
741 deliver jack@beanstalk.example
742 pipe $home/bin/mymailscript
743 seen mail subject "message discarded"
746 The following example commands do not set up significant deliveries:
748 unseen deliver jack@beanstalk.example
749 unseen pipe $home/bin/mymailscript
750 mail subject "message discarded"
756 .section "Filter commands" "SEC222"
757 The filter commands that are described in subsequent sections are listed
758 below, with the section in which they are described in brackets:
761 .row &(add)& "&~&~increment a user variable (section &<<SECTadd>>&)"
762 .row &(deliver)& "&~&~deliver to an email address (section &<<SECTdeliver>>&)"
763 .row &(fail)& "&~&~force delivery failure (sysadmin use) (section &<<SECTfail>>&)"
764 .row &(finish)& "&~&~end processing (section &<<SECTfinish>>&)"
765 .row &(freeze)& "&~&~freeze message (sysadmin use) (section &<<SECTfreeze>>&)"
766 .row &(headers)& "&~&~set the header character set (section &<<SECTheaders>>&)"
767 .row &(if)& "&~&~test condition(s) (section &<<SECTif>>&)"
768 .row &(logfile)& "&~&~define log file (section &<<SECTlog>>&)"
769 .row &(logwrite)& "&~&~write to log file (section &<<SECTlog>>&)"
770 .row &(mail)& "&~&~send a reply message (section &<<SECTmail>>&)"
771 .row &(pipe)& "&~&~pipe to a command (section &<<SECTpipe>>&)"
772 .row &(save)& "&~&~save to a file (section &<<SECTsave>>&)"
773 .row &(testprint)& "&~&~print while testing (section &<<SECTtestprint>>&)"
774 .row &(vacation)& "&~&~tailored form of &(mail)& (section &<<SECTmail>>&)"
777 The &(headers)& command has additional parameters that can be used only in a
778 system filter. The &(fail)& and &(freeze)& commands are available only when
779 Exim's filtering facilities are being used as a system filter, and are
780 therefore usable only by the system administrator and not by ordinary users.
781 They are mentioned only briefly in this document; for more information, see the
782 main Exim specification.
786 .section "The add command" "SECTadd"
788 &` add `&<&'number'&>&` to `&<&'user variable'&>
792 There are 10 user variables of this type, with names &$n0$& &-- &$n9$&. Their
793 values can be obtained by the normal expansion syntax (for example &$n3$&) in
794 other commands. At the start of filtering, these variables all contain zero.
795 Both arguments of the &(add)& command are expanded before use, making it
796 possible to add variables to each other. Subtraction can be obtained by adding
801 .section "The deliver command" "SECTdeliver"
803 &` deliver`& <&'mail address'&>
804 &`e.g. deliver "Dr Livingstone <David@somewhere.africa.example>"`&
807 This command provides a forwarding operation. The delivery that it sets up is
808 significant unless the command is preceded by &"unseen"& (see section
809 &<<SECTsigdel>>&). The message is sent on to the given address, exactly as
810 happens if the address had appeared in a traditional &_.forward_& file. If you
811 want to deliver the message to a number of different addresses, you can use
812 more than one &(deliver)& command (each one may have only one address).
813 However, duplicate addresses are discarded.
815 To deliver a copy of the message to your normal mailbox, your login name can be
816 given as the address. Once an address has been processed by the filtering
817 mechanism, an identical generated address will not be so processed again, so
818 doing this does not cause a loop.
820 However, if you have a mail alias, you should &'not'& refer to it here. For
821 example, if the mail address &'L.Gulliver'& is aliased to &'lg303'& then all
822 references in Gulliver's &_.forward_& file should be to &'lg303'&. A reference
823 to the alias will not work for messages that are addressed to that alias,
824 since, like &_.forward_& file processing, aliasing is performed only once on an
825 address, in order to avoid looping.
827 Following the new address, an optional second address, preceded by
828 &"errors_to"& may appear. This changes the address to which delivery errors on
829 the forwarded message will be sent. Instead of going to the message's original
830 sender, they go to this new address. For ordinary users, the only value that is
831 permitted for this address is the user whose filter file is being processed.
832 For example, the user &'lg303'& whose mailbox is in the domain
833 &'lilliput.example'& could have a filter file that contains
835 deliver jon@elsewhere.example errors_to lg303@lilliput.example
837 Clearly, using this feature makes sense only in situations where not all
838 messages are being forwarded. In particular, bounce messages must not be
839 forwarded in this way, as this is likely to create a mail loop if something
844 .section "The save command" "SECTsave"
846 &` save `&<&'file name'&>
847 &`e.g. save $home/mail/bookfolder`&
850 This command specifies that a copy of the message is to be appended to the
851 given file (that is, the file is to be used as a mail folder). The delivery
852 that &(save)& sets up is significant unless the command is preceded by
853 &"unseen"& (see section &<<SECTsigdel>>&).
855 More than one &(save)& command may be obeyed; each one causes a copy of the
856 message to be written to its argument file, provided they are different
857 (duplicate &(save)& commands are ignored).
859 If the file name does not start with a / character, the contents of the
860 &$home$& variable are prepended, unless it is empty, or the system
861 administrator has disabled this feature. In conventional configurations, this
862 variable is normally set in a user filter to the user's home directory, but the
863 system administrator may set it to some other path. In some configurations,
864 &$home$& may be unset, or prepending may be disabled, in which case a
865 non-absolute path name may be generated. Such configurations convert this to an
866 absolute path when the delivery takes place. In a system filter, &$home$& is
869 The user must of course have permission to write to the file, and the writing
870 of the file takes place in a process that is running as the user, under the
871 user's primary group. Any secondary groups to which the user may belong are not
872 normally taken into account, though the system administrator can configure Exim
873 to set them up. In addition, the ability to use this command at all is
874 controlled by the system administrator &-- it may be forbidden on some systems.
876 An optional mode value may be given after the file name. The value for the mode
877 is interpreted as an octal number, even if it does not begin with a zero. For
880 save /some/folder 640
882 This makes it possible for users to override the system-wide mode setting for
883 file deliveries, which is normally 600. If an existing file does not have the
884 correct mode, it is changed.
886 An alternative form of delivery may be enabled on your system, in which each
887 message is delivered into a new file in a given directory. If this is the case,
888 this functionality can be requested by giving the directory name terminated by
889 a slash after the &(save)& command, for example
891 save separated/messages/
893 There are several different formats for such deliveries; check with your system
894 administrator or local documentation to find out which (if any) are available
895 on your system. If this functionality is not enabled, the use of a path name
896 ending in a slash causes an error.
900 .section "The pipe command" "SECTpipe"
902 &` pipe `&<&'command'&>
903 &`e.g. pipe "$home/bin/countmail $sender_address"`&
906 This command specifies that the message is to be delivered to the specified
907 command using a pipe. The delivery that it sets up is significant unless the
908 command is preceded by &"unseen"& (see section &<<SECTsigdel>>&). Remember,
909 however, that no deliveries are done while the filter is being processed. All
910 deliveries happen later on. Therefore, the result of running the pipe is not
911 available to the filter.
913 When the deliveries are done, a separate process is run, and a copy of the
914 message is passed on its standard input. The process runs as the user, under
915 the user's primary group. Any secondary groups to which the user may belong are
916 not normally taken into account, though the system administrator can configure
917 Exim to set them up. More than one &(pipe)& command may appear; each one causes
918 a copy of the message to be written to its argument pipe, provided they are
919 different (duplicate &(pipe)& commands are ignored).
921 When the time comes to transport the message, the command supplied to &(pipe)&
922 is split up by Exim into a command name and a number of arguments. These are
923 delimited by white space except for arguments enclosed in double quotes, in
924 which case backslash is interpreted as an escape, or in single quotes, in which
925 case no escaping is recognized. Note that as the whole command is normally
926 supplied in double quotes, a second level of quoting is required for internal
927 double quotes. For example:
929 pipe "$home/myscript \"size is $message_size\""
931 String expansion is performed on the separate components after the line has
932 been split up, and the command is then run directly by Exim; it is not run
933 under a shell. Therefore, substitution cannot change the number of arguments,
934 nor can quotes, backslashes or other shell metacharacters in variables cause
937 Documentation for some programs that are normally run via this kind of pipe
938 often suggest that the command should start with
942 This is a shell command, and should &'not'& be present in Exim filter files,
943 since it does not normally run the command under a shell.
945 However, there is an option that the administrator can set to cause a shell to
946 be used. In this case, the entire command is expanded as a single string and
947 passed to the shell for interpretation. It is recommended that this be avoided
948 if at all possible, since it can lead to problems when inserted variables
949 contain shell metacharacters.
951 The default PATH set up for the command is determined by the system
952 administrator, usually containing at least &_/bin_& and &_/usr/bin_& so that
953 common commands are available without having to specify an absolute file name.
954 However, it is possible for the system administrator to restrict the pipe
955 facility so that the command name must not contain any / characters, and must
956 be found in one of the directories in the configured PATH. It is also possible
957 for the system administrator to lock out the use of the &(pipe)& command
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:
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`&
979 LOCAL_PART, LOGNAME, and USER are all set to the same value, namely, your login
980 id. LOCAL_PART_PREFIX and LOCAL_PART_SUFFIX may be set if Exim is configured to
981 recognize prefixes or suffixes in the local parts of addresses. For example, a
982 message addressed to &'pat-suf2@domain.example'& may cause the filter for user
983 &'pat'& to be run. If this sets up a pipe delivery, LOCAL_PART_SUFFIX is
984 &`-suf2`& when the pipe command runs. The system administrator has to configure
985 Exim specially for this feature to be available.
987 If you run a command that is a shell script, be very careful in your use of
988 data from the incoming message in the commands in your script. RFC 2822 is very
989 generous in the characters that are permitted to appear in mail addresses, and
990 in particular, an address may begin with a vertical bar or a slash. For this
991 reason you should always use quotes round any arguments that involve data from
992 the message, like this:
994 /some/command '$SENDER'
996 so that inserted shell meta-characters do not cause unwanted effects.
998 Remember that, as was explained earlier, the pipe command is not run at the
999 time the filter file is interpreted. The filter just defines what deliveries
1000 are required for one particular addressee of a message. The deliveries
1001 themselves happen later, once Exim has decided everything that needs to be done
1004 A consequence of this is that you cannot inspect the return code from the pipe
1005 command from within the filter. Nevertheless, the code returned by the command
1006 is important, because Exim uses it to decide whether the delivery has succeeded
1009 The command should return a zero completion code if all has gone well. Most
1010 non-zero codes are treated by Exim as indicating a failure of the pipe. This is
1011 treated as a delivery failure, causing the message to be returned to its
1012 sender. However, there are some completion codes that are treated as temporary
1013 errors. The message remains on Exim's spool disk, and the delivery is tried
1014 again later, though it will ultimately time out if the delivery failures go on
1015 too long. The completion codes to which this applies can be specified by the
1016 system administrator; the default values are 73 and 75.
1018 The pipe command should not normally write anything to its standard output or
1019 standard error file descriptors. If it does, whatever is written is normally
1020 returned to the sender of the message as a delivery error, though this action
1021 can be varied by the system administrator.
1025 .section "Mail commands" "SECTmail"
1026 There are two commands that cause the creation of a new mail message, neither
1027 of which count as a significant delivery unless the command is preceded by the
1028 word &"seen"& (see section &<<SECTsigdel>>&). This is a powerful facility, but
1029 it should be used with care, because of the danger of creating infinite
1030 sequences of messages. The system administrator can forbid the use of these
1031 commands altogether.
1033 To help prevent runaway message sequences, these commands have no effect when
1034 the incoming message is a bounce (delivery error) message, and messages sent by
1035 this means are treated as if they were reporting delivery errors. Thus, they
1036 should never themselves cause a bounce message to be returned. The basic
1037 mail-sending command is
1039 &`mail [to `&<&'address-list'&>&`]`&
1040 &` [cc `&<&'address-list'&>&`]`&
1041 &` [bcc `&<&'address-list'&>&`]`&
1042 &` [from `&<&'address'&>&`]`&
1043 &` [reply_to `&<&'address'&>&`]`&
1044 &` [subject `&<&'text'&>&`]`&
1045 &` [extra_headers `&<&'text'&>&`]`&
1046 &` [text `&<&'text'&>&`]`&
1047 &` [[expand] file `&<&'filename'&>&`]`&
1048 &` [return message]`&
1049 &` [log `&<&'log file name'&>&`]`&
1050 &` [once `&<&'note file name'&>&`]`&
1051 &` [once_repeat `&<&'time interval'&>&`]`&
1052 &`e.g. mail text "Your message about $h_subject: has been received"`&
1054 Each <&'address-list'&> can contain a number of addresses, separated by commas,
1055 in the format of a &'To:'& or &'Cc:'& header line. In fact, the text you supply
1056 here is copied exactly into the appropriate header line. It may contain
1057 additional information as well as email addresses. For example:
1059 mail to "Julius Caesar <jc@rome.example>, \
1060 <ma@rome.example> (Mark A.)"
1062 Similarly, the texts supplied for &%from%& and &%reply_to%& are copied into
1063 their respective header lines.
1065 As a convenience for use in one common case, there is also a command called
1066 &(vacation)&. It behaves in the same way as &(mail)&, except that the defaults
1067 for the &%subject%&, &%file%&, &%log%&, &%once%&, and &%once_repeat%& options
1070 subject "On vacation"
1071 expand file .vacation.msg
1076 respectively. These are the same file names and repeat period used by the
1077 traditional Unix &(vacation)& command. The defaults can be overridden by
1078 explicit settings, but if a file name is given its contents are expanded only
1079 if explicitly requested.
1081 &*Warning*&: The &(vacation)& command should always be used conditionally,
1082 subject to at least the &(personal)& condition (see section &<<SECTpersonal>>&
1083 below) so as not to send automatic replies to non-personal messages from
1084 mailing lists or elsewhere. Sending an automatic response to a mailing list or
1085 a mailing list manager is an Internet Sin.
1087 For both commands, the key/value argument pairs can appear in any order. At
1088 least one of &%text%& or &%file%& must appear (except with &(vacation)&, where
1089 there is a default for &%file%&); if both are present, the text string appears
1090 first in the message. If &%expand%& precedes &%file%&, each line of the file is
1091 subject to string expansion before it is included in the message.
1093 Several lines of text can be supplied to &%text%& by including the escape
1094 sequence &"\n"& in the string wherever a newline is required. If the command is
1095 output during filter file testing, newlines in the text are shown as &"\n"&.
1097 Note that the keyword for creating a &'Reply-To:'& header is &%reply_to%&,
1098 because Exim keywords may contain underscores, but not hyphens. If the &%from%&
1099 keyword is present and the given address does not match the user who owns the
1100 forward file, Exim normally adds a &'Sender:'& header to the message, though it
1101 can be configured not to do this.
1103 The &%extra_headers%& keyword allows you to add custom header lines to the
1104 message. The text supplied must be one or more syntactically valid RFC 2822
1105 header lines. You can use &"\n"& within quoted text to specify newlines between
1106 headers, and also to define continued header lines. For example:
1108 extra_headers "h1: first\nh2: second\n continued\nh3: third"
1110 No newline should appear at the end of the final header line.
1112 If no &%to%& argument appears, the message is sent to the address in the
1113 &$reply_address$& variable (see section &<<SECTfilterstringexpansion>>& above).
1114 An &'In-Reply-To:'& header is automatically included in the created message,
1115 giving a reference to the message identification of the incoming message.
1117 If &%return message%& is specified, the incoming message that caused the filter
1118 file to be run is added to the end of the message, subject to a maximum size
1121 If a log file is specified, a line is added to it for each message sent.
1123 If a &%once%& file is specified, it is used to hold a database for remembering
1124 who has received a message, and no more than one message is ever sent to any
1125 particular address, unless &%once_repeat%& is set. This specifies a time
1126 interval after which another copy of the message is sent. The interval is
1127 specified as a sequence of numbers, each followed by the initial letter of one
1128 of &"seconds"&, &"minutes"&, &"hours"&, &"days"&, or &"weeks"&. For example,
1132 causes a new message to be sent if at least 5 days and 4 hours have elapsed
1133 since the last one was sent. There must be no white space in a time interval.
1135 Commonly, the file name specified for &%once%& is used as the base name for
1136 direct-access (DBM) file operations. There are a number of different DBM
1137 libraries in existence. Some operating systems provide one as a default, but
1138 even in this case a different one may have been used when building Exim. With
1139 some DBM libraries, specifying &%once%& results in two files being created,
1140 with the suffixes &_.dir_& and &_.pag_& being added to the given name. With
1141 some others a single file with the suffix &_.db_& is used, or the name is used
1144 Using a DBM file for implementing the &%once%& feature means that the file
1145 grows as large as necessary. This is not usually a problem, but some system
1146 administrators want to put a limit on it. The facility can be configured not to
1147 use a DBM file, but instead, to use a regular file with a maximum size. The
1148 data in such a file is searched sequentially, and if the file fills up, the
1149 oldest entry is deleted to make way for a new one. This means that some
1150 correspondents may receive a second copy of the message after an unpredictable
1151 interval. Consult your local information to see if your system is configured
1154 More than one &(mail)& or &(vacation)& command may be obeyed in a single filter
1155 run; they are all honoured, even when they are to the same recipient.
1159 .section "Logging commands" "SECTlog"
1160 A log can be kept of actions taken by a filter file. This facility is normally
1161 available in conventional configurations, but there are some situations where
1162 it might not be. Also, the system administrator may choose to disable it. Check
1163 your local information if in doubt.
1165 Logging takes place while the filter file is being interpreted. It does not
1166 queue up for later like the delivery commands. The reason for this is so that a
1167 log file need be opened only once for several write operations. There are two
1168 commands, neither of which constitutes a significant delivery. The first
1169 defines a file to which logging output is subsequently written:
1171 &` logfile `&<&'file name'&>
1172 &`e.g. logfile $home/filter.log`&
1174 The file name must be fully qualified. You can use &$home$&, as in this
1175 example, to refer to your home directory. The file name may optionally be
1176 followed by a mode for the file, which is used if the file has to be created.
1179 logfile $home/filter.log 0644
1181 The number is interpreted as octal, even if it does not begin with a zero.
1182 The default for the mode is 600. It is suggested that the &(logfile)& command
1183 normally appear as the first command in a filter file. Once a log file has
1184 been obeyed, the &(logwrite)& command can be used to write to it:
1186 &` logwrite "`&<&'some text string'&>&`"`&
1187 &`e.g. logwrite "$tod_log $message_id processed"`&
1189 It is possible to have more than one &(logfile)& command, to specify writing to
1190 different log files in different circumstances. Writing takes place at the end
1191 of the file, and a newline character is added to the end of each string if
1192 there isn't one already there. Newlines can be put in the middle of the string
1193 by using the &"\n"& escape sequence. Lines from simultaneous deliveries may get
1194 interleaved in the file, as there is no interlocking, so you should plan your
1195 logging with this in mind. However, data should not get lost.
1199 .section "The finish command" "SECTfinish"
1200 The command &(finish)&, which has no arguments, causes Exim to stop
1201 interpreting the filter file. This is not a significant action unless preceded
1202 by &"seen"&. A filter file containing only &"seen finish"& is a black hole.
1205 .section "The testprint command" "SECTtestprint"
1206 It is sometimes helpful to be able to print out the values of variables when
1207 testing filter files. The command
1209 &` testprint `&<&'text'&>
1210 &`e.g. testprint "home=$home reply_address=$reply_address"`&
1212 does nothing when mail is being delivered. However, when the filtering code is
1213 being tested by means of the &%-bf%& option (see section &<<SECTtesting>>&
1214 above), the value of the string is written to the standard output.
1217 .section "The fail command" "SECTfail"
1218 When Exim's filtering facilities are being used as a system filter, the
1219 &(fail)& command is available, to force delivery failure. Because this command
1220 is normally usable only by the system administrator, and not enabled for use by
1221 ordinary users, it is described in more detail in the main Exim specification
1222 rather than in this document.
1225 .section "The freeze command" "SECTfreeze"
1226 When Exim's filtering facilities are being used as a system filter, the
1227 &(freeze)& command is available, to freeze a message on the queue. Because this
1228 command is normally usable only by the system administrator, and not enabled
1229 for use by ordinary users, it is described in more detail in the main Exim
1230 specification rather than in this document.
1234 .section "The headers command" "SECTheaders"
1235 The &(headers)& command can be used to change the target character set that is
1236 used when translating the contents of encoded header lines for insertion by the
1237 &$header_$& mechanism (see section &<<SECTheadervariables>>& above). The
1238 default can be set in the Exim configuration; if not specified, ISO-8859-1 is
1239 used. The only currently supported format for the &(headers)& command in user
1240 filters is as in this example:
1242 headers charset "UTF-8"
1244 That is, &(headers)& is followed by the word &"charset"& and then the name of a
1245 character set. This particular example would be useful if you wanted to compare
1246 the contents of a header to a UTF-8 string.
1248 In system filter files, the &(headers)& command can be used to add or remove
1249 header lines from the message. These features are described in the main Exim
1254 .section "Obeying commands conditionally" "SECTif"
1255 Most of the power of filtering comes from the ability to test conditions and
1256 obey different commands depending on the outcome. The &(if)& command is used to
1257 specify conditional execution, and its general form is
1259 &`if `&<&'condition'&>
1260 &`then `&<&'commands'&>
1261 &`elif `&<&'condition'&>
1262 &`then `&<&'commands'&>
1263 &`else `&<&'commands'&>
1266 There may be any number of &(elif)& and &(then)& sections (including none) and
1267 the &(else)& section is also optional. Any number of commands, including nested
1268 &(if)& commands, may appear in any of the <&'commands'&> sections.
1270 Conditions can be combined by using the words &(and)& and &(or)&, and round
1271 brackets (parentheses) can be used to specify how several conditions are to
1272 combine. Without brackets, &(and)& is more binding than &(or)&. For example:
1275 $h_subject: contains "Make money" or
1276 $h_precedence: is "junk" or
1277 ($h_sender: matches ^\\d{8}@ and not personal) or
1278 $message_body contains "this is not spam"
1283 A condition can be preceded by &(not)& to negate it, and there are also some
1284 negative forms of condition that are more English-like.
1288 .section "String testing conditions" "SEC23"
1289 There are a number of conditions that operate on text strings, using the words
1290 &"begins"&, &"ends"&, &"is"&, &"contains"& and &"matches"&. If you want to
1291 apply the same test to more than one header line, you can easily concatenate
1292 them into a single string for testing, as in this example:
1294 if "$h_to:, $h_cc:" contains me@domain.example then ...
1296 If a string-testing condition name is written in lower case, the testing
1297 of letters is done without regard to case; if it is written in upper case
1298 (for example, &"CONTAINS"&), the case of letters is taken into account.
1301 &` `&<&'text1'&>&` begins `&<&'text2'&>
1302 &` `&<&'text1'&>&` does not begin `&<&'text2'&>
1303 &`e.g. $header_from: begins "Friend@"`&
1306 A &"begins"& test checks for the presence of the second string at the start of
1307 the first, both strings having been expanded.
1310 &` `&<&'text1'&>&` ends `&<&'text2'&>
1311 &` `&<&'text1'&>&` does not end `&<&'text2'&>
1312 &`e.g. $header_from: ends "public.com.example"`&
1315 An &"ends"& test checks for the presence of the second string at the end of
1316 the first, both strings having been expanded.
1319 &` `&<&'text1'&>&` is `&<&'text2'&>
1320 &` `&<&'text1'&>&` is not `&<&'text2'&>
1321 &`e.g. $local_part_suffix is "-foo"`&
1324 An &"is"& test does an exact match between the strings, having first expanded
1328 &` `&<&'text1'&>&` contains `&<&'text2'&>
1329 &` `&<&'text1'&>&` does not contain `&<&'text2'&>
1330 &`e.g. $header_subject: contains "evolution"`&
1333 A &"contains"& test does a partial string match, having expanded both strings.
1336 &` `&<&'text1'&>&` matches `&<&'text2'&>
1337 &` `&<&'text1'&>&` does not match `&<&'text2'&>
1338 &`e.g. $sender_address matches "(bill|john)@"`&
1341 For a &"matches"& test, after expansion of both strings, the second one is
1342 interpreted as a regular expression. Exim uses the PCRE regular expression
1343 library, which provides regular expressions that are compatible with Perl.
1345 The match succeeds if the regular expression matches any part of the first
1346 string. If you want a regular expression to match only at the start or end of
1347 the subject string, you must encode that requirement explicitly, using the
1348 &`^`& or &`$`& metacharacters. The above example, which is not so constrained,
1349 matches all these addresses:
1353 spoonbill@example.com
1354 littlejohn@example.com
1356 To match only the first two, you could use this:
1358 if $sender_address matches "^(bill|john)@" then ...
1360 Care must be taken if you need a backslash in a regular expression, because
1361 backslashes are interpreted as escape characters both by the string expansion
1362 code and by Exim's normal processing of strings in quotes. For example, if you
1363 want to test the sender address for a domain ending in &'.com'& the regular
1368 The backslash and dollar sign in that expression have to be escaped when used
1369 in a filter command, as otherwise they would be interpreted by the expansion
1370 code. Thus, what you actually write is
1372 if $sender_address matches \\.com\$
1374 An alternative way of handling this is to make use of the &`\N`& expansion
1375 flag for suppressing expansion:
1377 if $sender_address matches \N\.com$\N
1379 Everything between the two occurrences of &`\N`& is copied without change by
1380 the string expander (and in fact you do not need the final one, because it is
1381 at the end of the string). If the regular expression is given in quotes
1382 (mandatory only if it contains white space) you have to write either
1384 if $sender_address matches "\\\\.com\\$"
1388 if $sender_address matches "\\N\\.com$\\N"
1391 If the regular expression contains bracketed sub-expressions, numeric
1392 variable substitutions such as &$1$& can be used in the subsequent actions
1393 after a successful match. If the match fails, the values of the numeric
1394 variables remain unchanged. Previous values are not restored after &(endif)&.
1395 In other words, only one set of values is ever available. If the condition
1396 contains several sub-conditions connected by &(and)& or &(or)&, it is the
1397 strings extracted from the last successful match that are available in
1398 subsequent actions. Numeric variables from any one sub-condition are also
1399 available for use in subsequent sub-conditions, because string expansion of a
1400 condition occurs just before it is tested.
1403 .section "Numeric testing conditions" "SEC24"
1404 The following conditions are available for performing numerical tests:
1407 &` `&<&'number1'&>&` is above `&<&'number2'&>
1408 &` `&<&'number1'&>&` is not above `&<&'number2'&>
1409 &` `&<&'number1'&>&` is below `&<&'number2'&>
1410 &` `&<&'number1'&>&` is not below `&<&'number2'&>
1411 &`e.g. $message_size is not above 10k`&
1414 The <&'number'&> arguments must expand to strings of digits, optionally
1415 followed by one of the letters K or M (upper case or lower case) which cause
1416 multiplication by 1024 and 1024x1024 respectively.
1419 .section "Testing for significant deliveries" "SEC25"
1420 You can use the &(delivered)& condition to test whether or not any previously
1421 obeyed filter commands have set up a significant delivery. For example:
1423 if not delivered then save mail/anomalous endif
1425 &"Delivered"& is perhaps a poor choice of name for this condition, because the
1426 message has not actually been delivered; rather, a delivery has been set up for
1430 .section "Testing for error messages" "SEC26"
1431 The condition &(error_message)& is true if the incoming message is a bounce
1432 (mail delivery error) message. Putting the command
1434 if error_message then finish endif
1436 at the head of your filter file is a useful insurance against things going
1437 wrong in such a way that you cannot receive delivery error reports. &*Note*&:
1438 &(error_message)& is a condition, not an expansion variable, and therefore is
1439 not preceded by &`$`&.
1442 .section "Testing a list of addresses" "SEC27"
1443 There is a facility for looping through a list of addresses and applying a
1444 condition to each of them. It takes the form
1446 &`foranyaddress `&<&'string'&>&` (`&<&'condition'&>&`)`&
1448 where <&'string'&> is interpreted as a list of RFC 2822 addresses, as in a
1449 typical header line, and <&'condition'&> is any valid filter condition or
1450 combination of conditions. The &"group"& syntax that is defined for certain
1451 header lines that contain addresses is supported.
1453 The parentheses surrounding the condition are mandatory, to delimit it from
1454 possible further sub-conditions of the enclosing &(if)& command. Within the
1455 condition, the expansion variable &$thisaddress$& is set to the non-comment
1456 portion of each of the addresses in the string in turn. For example, if the
1459 B.Simpson <bart@sfld.example>, lisa@sfld.example (his sister)
1461 then &$thisaddress$& would take on the values &`bart@sfld.example`& and
1462 &`lisa@sfld.example`& in turn.
1464 If there are no valid addresses in the list, the whole condition is false. If
1465 the internal condition is true for any one address, the overall condition is
1466 true and the loop ends. If the internal condition is false for all addresses in
1467 the list, the overall condition is false. This example tests for the presence
1468 of an eight-digit local part in any address in a &'To:'& header:
1470 if foranyaddress $h_to: ( $thisaddress matches ^\\d{8}@ ) then ...
1472 When the overall condition is true, the value of &$thisaddress$& in the
1473 commands that follow &(then)& is the last value it took on inside the loop. At
1474 the end of the &(if)& command, the value of &$thisaddress$& is reset to what it
1475 was before. It is best to avoid the use of multiple occurrences of
1476 &(foranyaddress)&, nested or otherwise, in a single &(if)& command, if the
1477 value of &$thisaddress$& is to be used afterwards, because it isn't always
1478 clear what the value will be. Nested &(if)& commands should be used instead.
1480 Header lines can be joined together if a check is to be applied to more than
1481 one of them. For example:
1483 if foranyaddress $h_to:,$h_cc: ....
1485 This scans through the addresses in both the &'To:'& and the &'Cc:'& headers.
1488 .section "Testing for personal mail" "SECTpersonal"
1489 A common requirement is to distinguish between incoming personal mail and mail
1490 from a mailing list, or from a robot or other automatic process (for example, a
1491 bounce message). In particular, this test is normally required for &"vacation
1494 The &(personal)& condition checks that the message is not a bounce message and
1495 that the current user's email address appears in the &'To:'& header. It also
1496 checks that the sender is not the current user or one of a number of common
1497 daemons, and that there are no header lines starting &'List-'& in the message.
1498 Finally, it checks the content of the &'Precedence:'& header line, if there is
1501 You should always use the &(personal)& condition when generating automatic
1502 responses. This example shows the use of &(personal)& in a filter file that is
1503 sending out vacation messages:
1506 mail to $reply_address
1507 subject "I am on holiday"
1508 file $home/vacation/message
1509 once $home/vacation/once
1513 It is tempting, when writing commands like the above, to quote the original
1514 subject in the reply. For example:
1516 subject "Re: $h_subject:"
1518 There is a danger in doing this, however. It may allow a third party to
1519 subscribe you to an opt-in mailing list, provided that the list accepts bounce
1520 messages as subscription confirmations. (Messages sent from filters are always
1521 sent as bounce messages.) Well-managed lists require a non-bounce message to
1522 confirm a subscription, so the danger is relatively small.
1524 If prefixes or suffixes are in use for local parts &-- something which depends
1525 on the configuration of Exim (see section &<<SECTmbox>>& below) &-- the tests
1526 for the current user are done with the full address (including the prefix and
1527 suffix, if any) as well as with the prefix and suffix removed. If the system is
1528 configured to rewrite local parts of mail addresses, for example, to rewrite
1529 &`dag46`& as &`Dirk.Gently`&, the rewritten form of the address is also used in
1534 .section "Alias addresses for the personal condition" "SEC28"
1535 It is quite common for people who have mail accounts on a number of different
1536 systems to forward all their mail to one system, and in this case a check for
1537 personal mail should test all their various mail addresses. To allow for this,
1538 the &(personal)& condition keyword can be followed by
1540 &`alias `&<&'address'&>
1542 any number of times, for example:
1544 if personal alias smith@else.where.example
1545 alias jones@other.place.example
1548 The alias addresses are treated as alternatives to the current user's email
1549 address when testing the contents of header lines.
1552 .section "Details of the personal condition" "SEC29"
1553 The basic &(personal)& test is roughly equivalent to the following:
1555 not error_message and
1556 $message_headers does not contain "\nList-Id:" and
1557 $message_headers does not contain "\nList-Help:" and
1558 $message_headers does not contain "\nList-Subscribe:" and
1559 $message_headers does not contain "\nList-Unsubscribe:" and
1560 $message_headers does not contain "\nList-Post:" and
1561 $message_headers does not contain "\nList-Owner:" and
1562 $message_headers does not contain "\nList-Archive:" and
1564 "${if def:h_auto-submitted:{present}{absent}}" is "absent" or
1565 $header_auto-submitted: is "no"
1567 $header_precedence: does not contain "bulk" and
1568 $header_precedence: does not contain "list" and
1569 $header_precedence: does not contain "junk" and
1570 foranyaddress $header_to:
1571 ( $thisaddress contains "$local_part$domain" ) and
1572 not foranyaddress $header_from:
1574 $thisaddress contains "$local_part@$domain" or
1575 $thisaddress contains "server@" or
1576 $thisaddress contains "daemon@" or
1577 $thisaddress contains "root@" or
1578 $thisaddress contains "listserv@" or
1579 $thisaddress contains "majordomo@" or
1580 $thisaddress contains "-request@" or
1581 $thisaddress matches "^owner-[^@]+@"
1584 The variable &$local_part$& contains the local part of the mail address of
1585 the user whose filter file is being run &-- it is normally your login id. The
1586 &$domain$& variable contains the mail domain. As explained above, if aliases
1587 or rewriting are defined, or if prefixes or suffixes are in use, the tests for
1588 the current user are also done with alternative addresses.
1593 .section "Testing delivery status" "SEC30"
1594 There are two conditions that are intended mainly for use in system filter
1595 files, but which are available in users' filter files as well. The condition
1596 &(first_delivery)& is true if this is the first process that is attempting to
1597 deliver the message, and false otherwise. This indicator is not reset until the
1598 first delivery process successfully terminates; if there is a crash or a power
1599 failure (for example), the next delivery attempt is also a &"first delivery"&.
1601 In a user filter file &(first_delivery)& will be false if there was previously
1602 an error in the filter, or if a delivery for the user failed owing to, for
1603 example, a quota error, or if forwarding to a remote address was deferred for
1606 The condition &(manually_thawed)& is true if the message was &"frozen"& for
1607 some reason, and was subsequently released by the system administrator. It is
1608 unlikely to be of use in users' filter files.
1611 .section "Multiple personal mailboxes" "SECTmbox" "SEC31"
1612 The system administrator can configure Exim so that users can set up variants
1613 on their email addresses and handle them separately. Consult your system
1614 administrator or local documentation to see if this facility is enabled on your
1615 system, and if so, what the details are.
1617 The facility involves the use of a prefix or a suffix on an email address. For
1618 example, all mail addressed to &'lg303-'&<&'something'&> would be the property
1619 of user &'lg303'&, who could determine how it was to be handled, depending on
1620 the value of <&'something'&>.
1622 There are two possible ways in which this can be set up. The first possibility
1623 is the use of multiple &_.forward_& files. In this case, mail to &'lg303-foo'&,
1624 for example, is handled by looking for a file called &_.forward-foo_& in
1625 &'lg303'&'s home directory. If such a file does not exist, delivery fails
1626 and the message is returned to its sender.
1628 The alternative approach is to pass all messages through a single &_.forward_&
1629 file, which must be a filter file so that it can distinguish between the
1630 different cases by referencing the variables &$local_part_prefix$& or
1631 &$local_part_suffix$&, as in the final example in section &<<SECTex>>& below.
1633 It is possible to configure Exim to support both schemes at once. In this case,
1634 a specific &_.forward-foo_& file is first sought; if it is not found, the basic
1635 &_.forward_& file is used.
1637 The &(personal)& test (see section &<<SECTpersonal>>&) includes prefixes and
1638 suffixes in its checking.
1642 .section "Ignoring delivery errors" "SEC43"
1643 As was explained above, filtering just sets up addresses for delivery &-- no
1644 deliveries are actually done while a filter file is active. If any of the
1645 generated addresses subsequently suffers a delivery failure, an error message
1646 is generated in the normal way. However, if a filter command that sets up a
1647 delivery is preceded by the word &"noerror"&, errors for that delivery,
1648 and any deliveries consequent on it (that is, from alias, forwarding, or
1649 filter files it invokes) are ignored.
1653 .section "Examples of Exim filter commands" "SECTex"
1658 deliver baggins@rivendell.middle-earth.example
1661 Vacation handling using traditional means, assuming that the &_.vacation.msg_&
1662 and other files have been set up in your home directory:
1666 unseen pipe "/usr/ucb/vacation \"$local_part\""
1669 Vacation handling inside Exim, having first created a file called
1670 &_.vacation.msg_& in your home directory:
1674 if personal then vacation endif
1677 File some messages by subject:
1681 if $header_subject: contains "empire" or
1682 $header_subject: contains "foundation"
1688 Save all non-urgent messages by weekday:
1692 if $header_subject: does not contain "urgent" and
1693 $tod_full matches "^(...),"
1699 Throw away all mail from one site, except from postmaster:
1703 if $reply_address contains "@spam.site.example" and
1704 $reply_address does not contain "postmaster@"
1710 Handle multiple personal mailboxes:
1714 if $local_part_suffix is "-foo"
1717 elif $local_part_suffix is "-bar"