Include note about exicyclog and keeping more than 99 back copies.
[exim.git] / doc / doc-src / filter.src
1. $Cambridge: exim/doc/doc-src/filter.src,v 1.1 2004/10/07 15:04:35 ph10 Exp $
3.if !set style
4.library "a4ps"
5.linelength ~~sys.linelength + 0.2in
6.emphasis ~~sys.linelength + 0.1in
7.pagedepth ~~sys.pagedepth - 0.2in
8.bindfont 53 "atl/Times-Roman" 7
9.set ssspaceb 1.50
12.include ""
14.set sgcal true
15.set html false
16.set texinfo false
19.if ~~sys.fancy
20.flag $sm{ "$push$g0$f53"
23.pagedepth ~~sys.pagedepth - 1ld
24.linelength 75em
25.emphasis 77em
26.footdepth 0
27.disable formfeed
28.backspace none
29.set chapspaceb 24
30.set sspacea 24
31.flag $sm{ "$push"
34.macro tabs 6
35.if ~~sys.fancy
36.tabset ~~1em
38.set temp (~~1 * 5)/4
39.tabset ~~temp em
43.macro startitems
46.indent 3em
49.macro enditems
54.macro item "item"
56.if ~~sys.leftonpage < 5ld
59.tempindent 0
64.macro index
67.set contents false
68.set displayindent 2em
71. ======================================================
74.if ~~sys.fancy
75.footdepth 2ld
82.set chapter -1
83.chapter Exim's interfaces to mail filtering -2ld
85Exim is a mail transfer agent for Unix-like systems. This document describes
86the user interfaces to its in-built mail filtering facilities, and is copyright
87(c) University of Cambridge 2004. It corresponds to Exim version 4.40.
90. ---------------------------------------------------------------------------
91. Some clever jiggery-pokery here. The contents list is known to be less than
92. one page long, so we arrange for it to get onto the rest of the first page.
93. Because we aren't doing any indexing, the z-rawindex file will contain only
94. the TOC entries. The Makefile arranges for it to be empty at the start, then
95. runs SGCAL twice so on the second pass it gets inserted automatically.
97.if ~~sgcal 1ld
99$chead{Contents} 1ld
101.tabset 2em 2em
103.linedepth ~~sys.linedepth - 1
104.include "z-rawindex"
108.set contents true
110. ---------------------------------------------------------------------------
113.set chapter 0
114.chapter Forwarding and filtering in Exim
117.section Introduction
118Most Unix mail transfer agents (programs that deliver mail) permit individual
119users to specify automatic forwarding of their mail, usually by placing a list
120of forwarding addresses in a file called \(.forward)\ in their home directories.
121Exim extends this facility by allowing the forwarding instructions to be a set
122of rules rather than just a list of addresses, in effect providing `\(.forward)\
123with conditions'. Operating the set of rules is called $it{filtering}, and the
124file that contains them is called a $it{filter file}.
126Exim supports two different kinds of filter file. An \*Exim filter*\ contains
127instructions in a format that is unique to Exim. A \*Sieve filter*\ contains
128instructions in the Sieve format that is defined by RFC 3028. As this is a
129standard format, Sieve filter files may already be familiar to some users.
130Sieve files should also be portable between different environments. However,
131the Exim filtering facility contains more features (such as variable
132expansion), and better integration with the host environment (such as the use
133of external processes and pipes).
135The choice of which kind of filter to use can be left to the end-user, provided
136that the system administrator has configured Exim appropriately for both kinds
137of filter. However, if interoperability is important, Sieve is the only
140The ability to use filtering or traditional forwarding has to be enabled by the
141system administrator, and some of the individual facilities can be separately
142enabled or disabled. A local document should be provided to describe exactly
143what has been enabled. In the absence of this, consult your system
146It is important to realize that, in Exim, no deliveries are actually made while
147a filter or traditional \(.forward)\ file is being processed. The result of
148such processing is a list of destinations to which a message should be
149delivered -- the deliveries themselves take place later, along with all other
150deliveries for the message. This means that it is not possible to test for
151successful deliveries while filtering. It also means that any duplicate
152addresses that are generated are dropped, since Exim never delivers the same
153message to the same address more than once.
155This document describes how to use a filter file and the format of its
156contents. It is intended for use by end-users. Both Sieve filters and Exim
157filters are covered. However, for Sieve filters, only issues that relate to the
158Exim implementation are discussed, since Sieve itself is described elsewhere.
160The contents of traditional \(.forward)\ files are not described here. They
161normally contain just a list of addresses, file names, or pipe commands,
162separated by commas or newlines, but other types of item are also available.
163The full details can be found in the chapter on the \%redirect%\ router in the
164Exim specification, which also describes how the system administrator can set
165up and control the use of filtering.
167.section Testing a new filter file
168.rset SECTtesting "~~chapter.~~section"
169Filter files, especially the more complicated ones, should always be tested, as
170it is easy to make mistakes. Exim provides a facility for preliminary testing
171of a filter file before installing it. This tests the syntax of the file and
172its basic operation, and can also be used with traditional \(.forward)\ files.
174Because a filter can do tests on the content of messages, a test message is
175required. Suppose you have a new filter file called \(myfilter)\ and a test
176message called \(test-message)\. Assuming that Exim is installed with the
177conventional path name \(/usr/sbin/sendmail)\ (some operating systems use
178\(/usr/lib/sendmail)\), the following command can be used:
179.display asis
180/usr/sbin/sendmail -bf myfilter <test-message
182The \-bf-\ option tells Exim that the following item on the command line is the
183name of a filter file that is to be tested. There is also a \-bF-\ option,
184which is similar, but which is used for testing system filter files, as opposed
185to user filter files, and which is therefore of use only to the system
188The test message is supplied on the standard input. If there are no
189message-dependent tests in the filter, an empty file (\(/dev/null)\) can be
190used. A supplied message must start with header lines or the `From' message
191separator line which is found in many multi-message folder files. Note that
192blank lines at the start terminate the header lines. A warning is given if no
193header lines are read.
195The result of running this command, provided no errors are detected in the
196filter file, is a list of the actions that Exim would try to take if presented
197with the message for real.
198For example, for an Exim filter, the output
199.display asis
200Deliver message to: gulliver@lilliput.fict.example
201Save message to: /home/lemuel/mail/archive
203means that one copy of the message would be sent to
204\gulliver@@lilliput.fict.example\, and another would be added to the file
205\(/home/lemuel/mail/archive)\, if all went well.
207The actions themselves are not attempted while testing a filter file in this
208way; there is no check, for example, that any forwarding addresses are valid.
209For an Exim filter,
210if you want to know why a particular action is being taken, add the \-v-\
211option to the command. This causes Exim to output the results of any
212conditional tests and to indent its output according to the depth of nesting of
213\"if"\ commands. Further additional output from a filter test can be generated
214by the \"testprint"\ command, which is described below.
216When Exim is outputting a list of the actions it would take, if any text
217strings are included in the output, non-printing characters therein are
218converted to escape sequences. In particular, if any text string contains a
219newline character, this is shown as `@\n' in the testing output.
221When testing a filter in this way, Exim makes up an `envelope' for the message.
222The recipient is by default the user running the command, and so is the sender,
223but the command can be run with the \-f-\ option to supply a different sender.
224For example,
226.indent 0
227/usr/sbin/sendmail -bf myfilter -f islington@@never.where <test-message
229Alternatively, if the \-f-\ option is not used, but the first line of the
230supplied message is a `From' separator from a message folder file (not the same
231thing as a \"From:"\ header line), the sender is taken from there. If \-f-\ is
232present, the contents of any `From' line are ignored.
234The `return path' is the same as the envelope sender, unless the message
235contains a \"Return-path:"\ header, in which case it is taken from there. You
236need not worry about any of this unless you want to test out features of a
237filter file that rely on the sender address or the return path.
239It is possible to change the envelope recipient by specifying further options.
240The \-bfd-\ option changes the domain of the recipient address, while the
241\-bfl-\ option changes the `local part', that is, the part before the @@ sign.
242An adviser could make use of these to test someone else's filter file.
244The \-bfp-\ and \-bfs-\ options specify the prefix or suffix for the local part.
245These are relevant only when support for multiple personal mailboxes is
246implemented; see the description in section ~~SECTmbox below.
248.section Installing a filter file
249A filter file is normally installed under the name \(.forward)\ in your home
250directory -- it is distinguished from a conventional \(.forward)\ file by its
251first line (described below). However, the file name is configurable, and some
252system administrators may choose to use some different name or location for
253filter files.
255.section Testing an installed filter file
256Testing a filter file before installation cannot find every potential problem;
257for example, it does not actually run commands to which messages are piped.
258Some `live' tests should therefore also be done once a filter is installed.
260If at all possible, test your filter file by sending messages from some other
261account. If you send a message to yourself from the filtered account, and
262delivery fails, the error message will be sent back to the same account, which
263may cause another delivery failure. It won't cause an infinite sequence of such
264messages, because delivery failure messages do not themselves generate further
265messages. However, it does mean that the failure won't be returned to you, and
266also that the postmaster will have to investigate the stuck message.
268If you have to test an Exim filter from the same account, a sensible precaution
269is to include the line
270.display asis
271if error_message then finish endif
273as the first filter command, at least while testing. This causes filtering to
274be abandoned for a delivery failure message, and since no destinations are
275generated, the message goes on to be delivered to the original address. Unless
276there is a good reason for not doing so, it is recommended that the above test
277be left in all Exim filter files.
278(This does not apply to Sieve files.)
281.section Details of filtering commands
282The filtering commands for Sieve and Exim filters are completely different in
283syntax and semantics. The Sieve mechanism is defined in RFC 3028; in the next
284chapter we describe how it is integrated into Exim. The subsequent chapter
285covers Exim filtering commands in detail.
293.chapter Sieve filter files
294.rset CHAPsievefilter "~~chapter"
295The code for Sieve filtering in Exim was contributed by Michael Haardt, and
296most of the content of this chapter is taken from the notes he provided. Since
297Sieve is a extensible language, it is important to understand `Sieve' in this
298context as `the specific implementation of Sieve for Exim'.
300This chapter does not contain a description of Sieve, since that can be found
301in RFC 3028, which should be read in conjunction with these notes.
303The Exim Sieve implementation offers the core as defined by RFC 3028, the
304\%envelope%\ and the \%fileinto%\ extensions, but not the \%reject%\ extension.
305Exim does not support message delivery notifications (MDNs), so adding it just
306to the Sieve filter (as required for \%reject%\) makes little sense.
308In order for Sieve to work properly in Exim, the system administrator needs to
309make some adjustments to the Exim configuration. These are described in the
310chapter on the \%redirect%\ router in the full Exim specification.
312.section Recognition of Sieve filters
313A filter file is interpreted as a Sieve filter if its first line is
314.display asis
315# Sieve filter
317This is what distinguishes it from a conventional \(.forward)\ file or an Exim
318filter file.
321.section Saving to specified folders
322If the system administrator has set things up as suggested in the Exim
323specification, and you use \%keep%\ or \%fileinto%\ to save a mail into a
324folder, absolute files are stored where specified, relative files are stored
325relative to \$home$\, and \%inbox%\ goes to the standard mailbox location.
328.section Strings containing header names
329RFC 3028 does not specify what happens if a string denoting a header field does
330not contain a valid header name, for example, it contains a colon. This
331implementation generates an error instead of ignoring the header field in order
332to ease script debugging, which fits in the common picture of Sieve.
335.section Exists test with empty list of headers
336The \%exists%\ test succeeds only if all specified headers exist. RFC 3028
337does not explicitly specify what happens on an empty list of headers. This
338implementation evaluates that condition as true, interpreting the RFC in a
339strict sense.
342.section Header test with invalid MIME encoding in header
343Some MUAs process invalid base64 encoded data, generating junk.
344Others ignore junk after seeing an equal sign in base64 encoded data.
345RFC 2047 does not specify how to react in this case, other than stating
346that a client must not forbid to process a message for that reason.
347RFC 2045 specifies that invalid data should be ignored (apparently
348looking at end of line characters). It also specifies that invalid data
349may lead to rejecting messages containing them (and there it appears to
350talk about true encoding violations), which is a clear contradiction to
351ignoring them.
353RFC 3028 does not specify how to process incorrect MIME words.
354This implementation treats them literally, as it does if the word is
355correct but its character set cannot be converted to UTF-8.
358.section Address test for multiple addresses per header
359A header may contain multiple addresses. RFC 3028 does not explicitly
360specify how to deal with them, but since the address test checks if
361anything matches anything else, matching one address suffices to
362satisfy the condition. That makes it impossible to test if a header
363contains a certain set of addresses and no more, but it is more logical
364than letting the test fail if the header contains an additional address
365besides the one the test checks for.
368.section Semantics of keep
369The \%keep%\ command is equivalent to
371fileinto "inbox";
373It saves the message and resets the implicit keep flag. It does not set the
374implicit keep flag; there is no command to set it once it has been reset.
377.section Semantics of fileinto
378RFC 3028 does not specify whether \fileinto\ should try to create a mail folder
379if it does not exist. This implementation allows the sysadmin to configure that
380aspect using the \%appendfile%\ transport options \create@_directory\,
381\create@_file\, and \file@_must@_exist\. See the \%appendfile%\ transport in
382the Exim specification for details.
385.section Semantics of redirect
386Sieve scripts are supposed to be interoperable between servers, so this
387implementation does not allow mail to be redirected to unqualified addresses,
388because the domain would depend on the system being used. On systems with
389virtual mail domains, the default domain is probably not what the user expects
390it to be.
393.section String arguments
394There has been confusion if the string arguments to \%require%\ are to be
395matched case-sensitively or not. This implementation matches them with
396the match type \":is"\ (default, see section 2.7.1) and the comparator
397\"i;ascii-casemap"\ (default, see section 2.7.3). The RFC defines the
398command defaults clearly, so any different implementations violate RFC
3993028. The same is valid for comparator names, also specified as strings.
402.section Number units
403There is a mistake in RFC 3028: the suffix G denotes gibi-, not tebibyte.
404The mistake is obvious, because RFC 3028 specifies G to denote 2@^30
405(which is gibi, not tebi), and that is what this implementation uses as
406scaling factor for the suffix G.
409.section RFC compliance
410Exim requires the first line of a Sieve filter to be
411.display asis
412# Sieve filter
414Of course the RFC does not specify that line. Do not expect examples to work
415without adding it, though.
417RFC 3028 requires the use of CRLF to terminate a line.
418The rationale was that CRLF is universally used in network protocols
419to mark the end of the line. This implementation does not embed Sieve
420in a network protocol, but uses Sieve scripts as part of the Exim MTA.
421Since all parts of Exim use LF as newline character, this implementation
422does, too, by default, though the system administrator may choose (at Exim
423compile time) to use CRLF instead.
425Exim violates RFC 2822, section 3.6.8, by accepting 8-bit header names, so
426this implementation repeats this violation to stay consistent with Exim.
427This is in preparation to UTF-8 data.
429Sieve scripts cannot contain NUL characters in strings, but mail
430headers could contain MIME encoded NUL characters, which could never
431be matched by Sieve scripts using exact comparisons. For that reason,
432this implementation extends the Sieve quoted string syntax with @\0
433to describe a NUL character, violating @\0 being the same as 0 in
434RFC 3028. Even without using @\0, the following tests are all true in
435this implementation. Implementations that use C-style strings will only
436evaluate the first test as true.
437.display asis
438Subject: =?iso-8859-1?q?abc=00def
440header :contains "Subject" ["abc"]
441header :contains "Subject" ["def"]
442header :matches "Subject" ["abc?def"]
445Note that by considering Sieve to be a MUA, RFC 2047 can be interpreted
446in a way that NUL characters truncating strings is allowed for Sieve
447implementations, although not recommended. It is further allowed to use
448encoded NUL characters in headers, but that's not recommended either.
449The above example shows why.
451RFC 3028 states that if an implementation fails to convert a character
452set to UTF-8, two strings cannot be equal if one contains octets greater
453than 127. Assuming that all unknown character sets are one-byte character
454sets with the lower 128 octets being US-ASCII is not sound, so this
455implementation violates RFC 3028 and treats such MIME words literally.
456That way at least something could be matched.
458The folder specified by \%fileinto%\ must not contain the character
459sequence \".."\ to avoid security problems. RFC 3028 does not specify the
460syntax of folders apart from \%keep%\ being equivalent to
461.display asis
462fileinto "INBOX";
464This implementation uses \"inbox"\ instead.
466Sieve script errors currently cause messages to be silently filed into
467\"inbox"\. RFC 3028 requires that the user is notified of that condition.
468This may be implemented in future by adding a header line to mails that
469are filed into \"inbox"\ due to an error in the filter.
477.chapter Exim filter files
478.rset CHAPeximfilter "~~chapter"
479This chapter contains a full description of the contents of Exim filter files.
481.section Format of Exim filter files
482Apart from leading white space, the first text in a filter file must be
483.display asis
484# Exim filter
486This is what distinguishes it from a conventional \(.forward)\ file or a Sieve
487filter file. If the file does not have this initial line (or the equivalent for
488a Sieve filter), it is treated as a
489conventional \(.forward)\ file, both when delivering mail and when using the
490\-bf-\ testing mechanism. The white space in the line is optional, and any
491capitalization may be used. Further text on the same line is treated as a
492comment. For example, you could have
493.display asis
494# Exim filter <<== do not edit or remove this line!
496The remainder of the file is a sequence of filtering commands, which consist of
497keywords and data values. For example, in the command
498.display asis
499deliver gulliver@lilliput.fict.example
501the keyword is \"deliver"\ and the data value is
503White space or line breaks separate the components of a command, except in the
504case of conditions for the \"if"\ command, where round brackets (parentheses)
505also act as separators. Complete commands are separated from each other by
506white space or line breaks; there are no special terminators. Thus, several
507commands may appear on one line, or one command may be spread over a number of
510If the character @# follows a separator anywhere in a command, everything from
511@# up to the next newline is ignored. This provides a way of including comments
512in a filter file.
514.section Data values in filter commands
515There are two ways in which a data value can be input:
516.numberpars $.
517If the text contains no white space then it can be typed verbatim. However, if
518it is part of a condition, it must also be free of round brackets
519(parentheses), as these are used for grouping in conditions.
521Otherwise, it must be enclosed in double quotation marks. In this case, the
522character @\ (backslash) is treated as an `escape character' within the string,
523causing the following character or characters to be treated specially:
524.display rm
525.tabs 8
526@\n $t is replaced by a newline
527@\r $t is replaced by a carriage return
528@\t $t is replaced by a tab
530Backslash followed by up to three octal digits is replaced by the character
531specified by those digits, and @\x followed by up to two hexadecimal digits is
532treated similarly. Backslash followed by any other character is replaced
533by the second character, so that in particular, @\" becomes " and @\@\ becomes
534@\$<. A data item enclosed in double quotes can be continued onto the next line
535by ending the first line with a backslash. Any leading white space at the start
536of the continuation line is ignored.
538In addition to the escape character processing that occurs when strings are
539enclosed in quotes, most data values are also subject to $it{string expansion}
540(as described in the next section), in which case the characters \@$\ and \@\\
541are also significant. This means that if a single backslash is actually
542required in such a string, and the string is also quoted, @\@\@\@\ has to be
545The maximum permitted length of a data string, before expansion, is 1024
549.section String expansion
550.rset SECTfilterstringexpansion "~~chapter.~~section"
551Most data values are expanded before use. Expansion consists of replacing
552substrings beginning with \"@$"\ with other text. The full expansion facilities
553available in Exim are extensive. If you want to know everything that Exim can
554do with strings, you should consult the chapter on string expansion in the Exim
557In filter files, by far the most common use of string expansion is the
558substitution of the contents of a variable. For example, the substring
559.display asis
562is replaced by the address to which replies to the message should be sent. If
563such a variable name is followed by a letter or digit or underscore, it must be
564enclosed in curly brackets (braces), for example,
565.display asis
568If a \"@$"\ character is actually required in an expanded string, it must be
569escaped with a backslash, and because backslash is also an escape character in
570quoted input strings, it must be doubled in that case. The following two
571examples illustrate two different ways of testing for a \"@$"\ character in a
573.display asis
574if $message_body contains \$ then ...
575if $message_body contains "\\$" then ...
577You can prevent part of a string from being expanded by enclosing it between
578two occurrences of \"@\N"\. For example,
579.display asis
580if $message_body contains \N$$$$\N then ...
582tests for a run of four dollar characters.
584.section Some useful general variables
585A complete list of the available variables is given in the Exim documentation.
586This shortened list contains the ones that are most likely to be useful in
587personal filter files:
589\$body@_linecount$\: The number of lines in the body of the message.
591\$home$\: In conventional configurations, this variable normally contains the
592user's home directory. The system administrator can, however, change this.
594\$local@_part$\: The part of the email address that precedes the @@ sign --
595normally the user's login name. If support for multiple personal mailboxes is
596enabled (see section ~~SECTmbox below) and a prefix or suffix for the local
597part was recognized, it is removed from the string in this variable.
599\$local@_part@_prefix$\: If support for multiple personal mailboxes is enabled
600(see section ~~SECTmbox below), and a local part prefix was recognized,
601this variable contains the prefix. Otherwise it contains an empty string.
603\$local@_part@_suffix$\: If support for multiple personal mailboxes is enabled
604(see section ~~SECTmbox below), and a local part suffix was recognized,
605this variable contains the suffix. Otherwise it contains an empty string.
607\$message@_body$\: The initial portion of the body of the message. By default,
608up to 500 characters are read into this variable, but the system administrator
609can configure this to some other value. Newlines in the body are converted into
610single spaces.
612\$message@_body@_end$\: The final portion of the body of the message, formatted
613and limited in the same way as \$message@_body$\.
615\$message@_body@_size$\: The size of the body of the message, in bytes.
617\$message@_headers$\: The header lines of the message, concatenated into a
618single string, with newline characters between them.
620\$message@_id$\: The message's local identification string, which is unique for
621each message handled by a single host.
623\$message@_size$\: The size of the entire message, in bytes.
625\$original@_local@_part$\: When an address that arrived with the message is
626being processed, this contains the same value as the variable \$local@_part$\.
627However, if an address generated by an alias, forward, or filter file is being
628processed, this variable contains the local part of the original address.
630\$reply@_address$\: The contents of the \"Reply-to:"\ header, if the message
631has one; otherwise the contents of the \"From:"\ header. It is the address to
632which normal replies to the message should be sent.
634\$return@_path$\: The return path -- that is, the sender field that will be
635transmitted as part of the message's envelope if the message is sent to another
636host. This is the address to which delivery errors are sent. In many cases,
637this variable has the same value as \$sender@_address$\, but if, for example,
638an incoming message to a mailing list has been expanded, \$return@_path$\ may
639have been changed to contain the address of the list maintainer.
641\$sender@_address$\: The sender address that was received in the envelope of
642the message. This is not necessarily the same as the contents of the \"From:"\
643or \"Sender:"\ header lines. For delivery error messages (`bounce messages')
644there is no sender address, and this variable is empty.
646\$tod@_full$\: A full version of the time and date, for example: Wed, 18 Oct
6471995 09:51:40 +0100. The timezone is always given as a numerical offset from
650\$tod@_log$\: The time and date in the format used for writing Exim's log files,
651without the timezone, for example: 1995-10-12 15:32:29.
653\$tod@_zone$\: The local timezone offset, for example: +0100.
656.section Header variables
657.rset SECTheadervariables "~~chapter.~~section"
658There is a special set of expansion variables containing the header lines of
659the message being processed. These variables have names beginning with
660\"@$header@_"\ followed by the name of the header line, terminated by a colon.
661For example,
662.display asis
666The whole item, including the terminating colon, is replaced by the contents of
667the message header line. If there is more than one header line with the same
668name, their contents are concatenated. For header lines whose data consists of
669a list of addresses (for example, ::From:: and ::To::), a comma and newline is
670inserted between each set of data. For all other header lines, just a newline
671is used.
673Leading and trailing white space is removed from header line data, and if there
674are any MIME `words' that are encoded as defined by RFC 2047 (because they
675contain non-ASCII characters), they are decoded and translated, if possible, to
676a local character set. Translation is attempted only on operating systems that
677have the \iconv(@)\ function. This makes the header line look the same as it
678would when displayed by an MUA. The default character set is ISO-8859-1, but
679this can be changed by means of the \"headers"\ command (see below).
681If you want to see the actual characters that make up a header line, you can
682specify \"@$rheader@_"\ instead of \"@$header@_"\. This inserts the `raw'
683header line, unmodified.
685There is also an intermediate form, requested by \"@$bheader@_"\, which removes
686leading and trailing space and decodes MIME `words', but does not do any
687character translation. If an attempt to decode what looks superficially like a
688MIME `word' fails, the raw string is returned. If decoding produces a binary
689zero character, it is replaced by a question mark.
691The capitalization of the name following \"@$header@_"\ is not significant.
692Because any printing character except colon may appear in the name of a
693message's header (this is a requirement of RFC 2822, the document that
694describes the format of a mail message) curly brackets must $it{not} be used in
695this case, as they will be taken as part of the header name. Two shortcuts are
696allowed in naming header variables:
697.numberpars $.
698The initiating \"@$header@_"\, \"@$rheader@_"\, or \"@$bheader@_"\ can be
699abbreviated to \"@$h@_"\, \"@$rh@_"\, or \"@$bh@_"\, respectively.
701The terminating colon can be omitted if the next character is white space. The
702white space character is retained in the expanded string. However, this is not
703recommended, because it makes it easy to forget the colon when it really is
706If the message does not contain a header of the given name, an empty string is
707substituted. Thus it is important to spell the names of headers correctly. Do
708not use \"@$header@_Reply@_to"\ when you really mean \"@$header@_Reply-to"\.
710.section User variables
711There are ten user variables with names \$n0$\ -- \$n9$\ that can be
712incremented by the \"add"\ command (see section ~~SECTadd). These can be used
713for `scoring' messages in various ways. If Exim is configured to run a `system
714filter' on every message, the values left in these variables are copied into
715the variables \$sn0$\ -- \$sn9$\ at the end of the system filter, thus making
716them available to users' filter files. How these values are used is entirely up
717to the individual installation.
719.section Current directory
720The contents of your filter file should not make any assumptions about the
721current directory. It is best to use absolute paths for file names; you
722can normally make use of the \$home$\ variable to refer to your home directory.
723The \save\ command automatically inserts \$home$\ at the start of non-absolute
728.section Significant deliveries
729When in the course of delivery a message is processed by a filter file, what
730happens next, that is, after the whole filter file has been processed, depends
731on whether the filter has set up any $it{significant deliveries} or not. If
732there is at least one significant delivery, the filter is considered to
733have handled the entire delivery arrangements for the current address, and no
734further processing of the address takes place. If, however, no significant
735deliveries have been set up, Exim continues processing the current address as
736if there were no filter file, and typically sets up a delivery of a copy of the
737message into a local mailbox. In particular, this happens in the special case
738of a filter file containing only comments.
740The delivery commands \"deliver"\, \"save"\, and \"pipe"\ are by default
741significant. However, if such a command is preceded by the word \"unseen"\, its
742delivery is not considered to be significant. In contrast, other commands such
743as \"mail"\ and \"vacation"\ do not count as significant deliveries unless
744preceded by the word \"seen"\.
747.section Filter commands
748The filter commands which are described in subsequent sections are listed
749below, with the section in which they are described in brackets:
750.display rm
751.tabs 15
752\add\ $t increment a user variable (section ~~SECTadd)
753\deliver\ $t deliver to an email address (section ~~SECTdeliver)
754\fail\ $t force delivery failure (sysadmin use) (section ~~SECTfail)
755\finish\ $t end processing (section ~~SECTfinish)
756\freeze\ $t freeze message (sysadmin use) (section ~~SECTfreeze)
757\headers\ $t set the header character set (section ~~SECTheaders)
758\if\ $t test condition(s) (section ~~SECTif)
759\logfile\ $t define log file (section ~~SECTlog)
760\logwrite\ $t write to log file (section ~~SECTlog)
761\mail\ $t send a reply message (section ~~SECTmail)
762\pipe\ $t pipe to a command (section ~~SECTpipe)
763\save\ $t save to a file (section ~~SECTsave)
764\testprint\ $t print while testing (section ~~SECTtestprint)
765\vacation\ $t tailored form of \mail\ (section ~~SECTmail)
767In addition, when Exim's filtering facilities are being used as a system
768filter, the \"fail"\, \"freeze"\, and \"headers"\ commands are available.
769However, since they are usable only by the system administrator and not by
770ordinary users, they are described in the main Exim specification rather than
771in this document.
773.section The add command
774.rset SECTadd "~~chapter.~~section"
776 add <<number>> to <<user variable>>
777e.g. add 2 to n3
779There are 10 user variables of this type, and their values can be obtained by
780the normal expansion syntax (for example \$n3$\) in other commands. At the
781start of filtering, these variables all contain zero. Both arguments of the
782\"add"\ command are expanded before use, making it possible to add variables to
783each other. Subtraction can be obtained by adding negative numbers.
786.section The deliver command
787.rset SECTdeliver "~~chapter.~~section"
789 deliver <<mail address>>
790e.g. deliver "Dr Livingstone <>"
792This provides a forwarding operation. The message is sent on to the given
793address, exactly as happens if the address had appeared in a traditional
794\(.forward)\ file. If you want to deliver the message to a number of different
795addresses, you can use more than one \"deliver"\ command (each one may have
796only one address). However, duplicate addresses are discarded.
798To deliver a copy of the message to your normal mailbox, your login name can be
799given as the address. Once an address has been processed by the filtering
800mechanism, an identical generated address will not be so processed again, so
801doing this does not cause a loop.
803However, if you have a mail alias, you should $it{not} refer to it here. For
804example, if the mail address \"L.Gulliver"\ is aliased to \"lg103"\ then all
805references in Gulliver's \(.forward)\ file should be to \"lg103"\. A reference
806to the alias will not work for messages that are addressed to that alias,
807since, like \(.forward)\ file processing, aliasing is performed only once on an
808address, in order to avoid looping.
810Following the new address, an optional second address, preceded by
811\"errors@_to"\ may appear. This changes the address to which delivery errors on
812the forwarded message will be sent. Instead of going to the message's original
813sender, they go to this new address. For ordinary users, the only value that is
814permitted for this address is the user whose filter file is being processed.
815For example, the user \"lg103"\ whose mailbox is in the domain
816\lilliput.example\ could have a filter file that contains
817.display asis
818 deliver jon@elsewhere.example errors_to lg103@lilliput.example
820Clearly, using this feature makes sense only in situations where not all
821messages are being forwarded. In particular, bounce messages must not be
822forwarded in this way, as this is likely to create a mail loop if something
823goes wrong.
826.section The save command
827.rset SECTsave "~~chapter.~~section"
829 save <<file name>>
830e.g. save @$home/mail/bookfolder
832This causes a copy of the message to be appended to the given file (that is,
833the file is used as a mail folder). More than one \"save"\ command may appear;
834each one causes a copy of the message to be written to its argument file,
835provided they are different (duplicate \"save"\ commands are ignored).
837If the file name does not start with a / character, the contents of the
838\$home$\ variable are prepended, unless it is empty. In conventional
839configurations, this variable is normally set in a user filter to the user's
840home directory, but the system administrator may set it to some other path. In
841some configurations, \$home$\ may be unset, in which case a non-absolute path
842name may be generated. Such configurations convert this to an absolute path
843when the delivery takes place. In a system filter, \$home$\ is never set.
845The user must of course have permission to write to the file, and the writing
846of the file takes place in a process that is running as the user, under the
847user's primary group. Any secondary groups to which the user may belong are not
848normally taken into account, though the system administrator can configure Exim
849to set them up. In addition, the ability to use this command at all is
850controlled by the system administrator -- it may be forbidden on some systems.
852An optional mode value may be given after the file name. The value for the mode
853is interpreted as an octal number, even if it does not begin with a zero. For
856 save /some/folder 640
858This makes it possible for users to override the system-wide mode setting for
859file deliveries, which is normally 600. If an existing file does not have the
860correct mode, it is changed.
862An alternative form of delivery may be enabled on your system, in which each
863message is delivered into a new file in a given directory. If this is the case,
864this functionality can be requested by giving the directory name terminated by
865a slash after the \"save"\ command, for example
867 save separated/messages/
869There are several different formats for such deliveries; check with your system
870administrator or local documentation to find out which (if any) are available
871on your system. If this functionality is not enabled, the use of a path name
872ending in a slash causes an error.
875.section The pipe command
876.rset SECTpipe "~~chapter.~~section"
878 pipe <<command>>
879e.g. pipe "@$home/bin/countmail @$sender@_address"
881This command sets up delivery to a specified command using a pipe. Remember,
882however, that no deliveries are done while the filter is being processed. All
883deliveries happen later on. Therefore, the result of running the pipe is not
884available to the filter.
886When the deliveries are done, a separate process is run, and a copy of the
887message is passed on its standard input. The process runs as the user, under
888the user's primary group. Any secondary groups to which the user may belong are
889not normally taken into account, though the system administrator can configure
890Exim to set them up. More than one \"pipe"\ command may appear; each one causes
891a copy of the message to be written to its argument pipe, provided they are
892different (duplicate \"pipe"\ commands are ignored).
894When the time comes to transport the message,
895the command supplied to \"pipe"\ is split up by Exim into a command name and a
896number of arguments. These are delimited by white space except for arguments
897enclosed in double quotes, in which case backslash is interpreted as an escape,
898or in single quotes, in which case no escaping is recognized. Note that as the
899whole command is normally supplied in double quotes, a second level of quoting
900is required for internal double quotes. For example:
901.display asis
902 pipe "$home/myscript \"size is $message_size\""
904String expansion is performed on the separate components after the line has
905been split up, and the command is then run directly by Exim; it is not run
906under a shell. Therefore, substitution cannot change the number of arguments,
907nor can quotes, backslashes or other shell metacharacters in variables cause
910Documentation for some programs that are normally run via this kind of pipe
911often suggest that the command should start with
912.display asis
913IFS=" "
915This is a shell command, and should $it{not} be present in Exim filter files,
916since it does not normally run the command under a shell.
918However, there is an option that the administrator can set to cause a shell to
919be used. In this case, the entire command is expanded as a single string and
920passed to the shell for interpretation. It is recommended that this be avoided
921if at all possible, since it can lead to problems when inserted variables
922contain shell metacharacters.
924The default \\PATH\\ set up for the command is determined by the system
925administrator, usually containing at least \/usr/bin\ so that common commands
926are available without having to specify an absolute file name. However, it is
927possible for the system administrator to restrict the pipe facility so that the
928command name must not contain any / characters, and must be found in one of the
929directories in the configured \\PATH\\. It is also possible for the system
930administrator to lock out the use of the \"pipe"\ command altogether.
932When the command is run, a number of environment variables are set up. The
933complete list for pipe deliveries may be found in the Exim reference manual.
934Those that may be useful for pipe deliveries from user filter files are:
936.tabs 20
937DOMAIN $t $rm{the domain of the address}
938HOME $t $rm{your home directory}
939LOCAL@_PART $t $rm{see below}
940LOCAL@_PART@_PREFIX $t $rm{see below}
941LOCAL@_PART@_SUFFIX $t $rm{see below}
942LOGNAME $t $rm{your login name}
943MESSAGE@_ID $t $rm{the message's unique id}
944PATH $t $rm{the command search path}
945RECIPIENT $t $rm{the complete recipient address}
946SENDER $t $rm{the sender of the message}
947SHELL $t $bf{/bin/sh}
948USER $t $rm{see below}
950\\LOCAL@_PART\\, \\LOGNAME\\, and \\USER\\ are all set to the same value,
951namely, your login id. \\LOCAL@_PART@_PREFIX\\ and \\LOCAL@_PART@_SUFFIX\\ may
952be set if Exim is configured to recognize prefixes or suffixes in the local
953parts of addresses. For example, a message addressed to
954\*pat-suf2@@domain.example*\ may cause user \*pat*\'s filter file to be run. If
955this sets up a pipe delivery, \\LOCAL@_PART@_SUFFIX\\ is \"-suf2"\ when the
956pipe command runs. The system administrator has to configure Exim specially for
957this feature to be available.
959If you run a command that is a shell script, be very careful in your use of
960data from the incoming message in the commands in your script. RFC 2822 is very
961generous in the characters that are legally permitted to appear in mail
962addresses, and in particular, an address may begin with a vertical bar or a
963slash. For this reason you should always use quotes round any arguments that
964involve data from the message, like this:
965.display asis
966/some/command '$SENDER'
968so that inserted shell meta-characters do not cause unwanted effects.
970Remember that, as was explained earlier, the pipe command is not run at the
971time the filter file is interpreted. The filter just defines what deliveries
972are required for one particular addressee of a message. The deliveries
973themselves happen later, once Exim has decided everything that needs to be done
974for the message.
976A consequence of this is that you cannot inspect the return code from the pipe
977command from within the filter. Nevertheless, the code returned by the command
978is important, because Exim uses it to decide whether the delivery has succeeded
979or failed.
981The command should return a zero completion code if all has gone well. Most
982non-zero codes are treated by Exim as indicating a failure of the pipe. This is
983treated as a delivery failure, causing the message to be returned to its
984sender. However, there are some completion codes which are treated as temporary
985errors. The message remains on Exim's spool disk, and the delivery is tried
986again later, though it will ultimately time out if the delivery failures go on
987too long. The completion codes to which this applies can be specified by the
988system administrator; the default values are 73 and 75.
990The pipe command should not normally write anything to its standard output or
991standard error file descriptors. If it does, whatever is written is normally
992returned to the sender of the message as a delivery error, though this action
993can be varied by the system administrator.
996.section Mail commands
997.rset SECTmail "~~chapter.~~section"
998There are two commands which cause the creation of a new mail message, neither
999of which count as a significant delivery unless the command is preceded by the
1000word \"seen"\. This is a powerful facility, but it should be used with care,
1001because of the danger of creating infinite sequences of messages. The system
1002administrator can forbid the use of these commands altogether.
1004To help prevent runaway message sequences, these commands have no effect when
1005the incoming message is a delivery error message, and messages sent by this
1006means are treated as if they were reporting delivery errors. Thus they should
1007never themselves cause a delivery error message to be returned. The basic
1008mail-sending command is
1010 mail [to <<address-list>>]
1011 [cc <<address-list>>]
1012 [bcc <<address-list>>]
1013 [from <<address>>]
1014 [reply@_to <<address>>]
1015 [subject <<text>>]
1016 [extra@_headers <<text>>]
1017 [text <<text>>]
1018 [[expand] file <<filename>>]
1019 [return message]
1020 [log <<log file name>>]
1021 [once <<note file name>>]
1022 [once@_repeat <<time interval>>]
1024e.g. mail text "Your message about @$h@_subject: has been received"
1027Each <<address-list>> can contain a number of addresses, separated by commas,
1028in the format of a ::To:: or ::Cc:: header line. In fact, the text you supply
1029here is copied exactly into the appropriate header line. Thus, it may contain
1030additional information as well as email addresses. For example:
1031.display asis
1032mail to "Julius Caesar <jc@rome.example>, \
1033 <ma@rome.example> (Mark A.)"
1035Similarly, the texts supplied for ::From:: and ::Reply-to:: are copied into
1036their respective header lines.
1038As a convenience for use in one common case, there is also a command called
1039\vacation\. It behaves in the same way as \mail\, except that the defaults for
1042\"file"\, \"log"\, \"once"\, and \"once@_repeat"\ options are
1044subject "On vacation"
1045expand file .vacation.msg
1046log .vacation.log
1047once .vacation
1048once@_repeat 7d
1050respectively. These are the same file names and repeat period used by the
1051traditional Unix \"vacation"\ command. The defaults can be overridden by
1052explicit settings, but if a file name is given its contents are expanded only
1053if explicitly requested.
1055\**Warning**\: The \"vacation"\ command should always be used conditionally,
1056subject to at least the \"personal"\ condition (see section ~~SECTpersonal
1057below) so as not to send automatic replies to non-personal messages from
1058mailing lists or elsewhere. Sending an automatic response to a mailing list or
1059a mailing list manager is an Internet Sin.
1061For both commands, the key/value argument pairs can appear in any order. At
1062least one of \"text"\ or \"file"\ must appear (except with \"vacation"\); if
1063both are present, the text string appears first in the message. If \"expand"\
1064precedes \"file"\, each line of the file is subject to string expansion as
1065it is included in the message.
1067Several lines of text can be supplied to \"text"\ by including the escape
1068sequence `@\n' in the string where newlines are required. If the command is
1069output during filter file testing, newlines in the text are shown as `@\n'.
1071Note that the keyword for creating a \"Reply-To:"\ header is \reply@_to\,
1072because Exim keywords may contain underscores, but not hyphens. If the \"from"\
1073keyword is present and the given address does not match the user who owns the
1074forward file, Exim normally adds a \"Sender:"\ header to the message,
1075though it can be configured not to do this.
1077The \extra@_headers\ keyword allows you to add custom header lines to the
1078message. The text supplied must be one or more syntactically valid RFC 2882
1079header lines. You can use `@\n' within quoted text to specify newlines between
1080headers, and also to define continued header lines. For example:
1081.display asis
1082extra_headers "h1: first\nh2: second\n continued\nh3: third"
1084No newline should appear at the end of the final header line.
1086If no \"to"\ argument appears, the message is sent to the address in the
1087\"@$reply@_address"\ variable (see section ~~SECTfilterstringexpansion above).
1088An \"In-Reply-To:"\ header is automatically included in the created message,
1089giving a reference to the message identification of the incoming message.
1091If \"return message"\ is specified, the incoming message that caused the filter
1092file to be run is added to the end of the message, subject to a maximum size
1095If a log file is specified, a line is added to it for each message sent.
1097If a \"once"\ file is specified, it is used to hold a database for remembering
1098who has received a message, and no more than one message is ever sent to any
1099particular address, unless \"once@_repeat"\ is set. This specifies a time
1100interval after which another copy of the message is sent. The interval is
1101specified as a sequence of numbers, each followed by the initial letter of one
1102of `seconds', `minutes', `hours', `days', or `weeks'. For example,
1103.display asis
1104once_repeat 5d4h
1106causes a new message to be sent if 5 days and 4 hours have elapsed since the
1107last one was sent. There must be no white space in a time interval.
1109Commonly, the file name specified for \"once"\ is used as the base name for
1110direct-access (DBM) file operations. There are a number of different DBM
1111libraries in existence. Some operating systems provide one as a default, but
1112even in this case a different one may have been used when building Exim. With
1113some DBM libraries, specifying \"once"\ results in two files being created,
1114with the suffixes \".dir"\ and \".pag"\ being added to the given name. With
1115some others a single file with the suffix \".db"\ is used, or the name is used
1118Using a DBM file for implementing the \"once"\ feature means that the file
1119grows as large as necessary. This is not usually a problem, but some system
1120administrators want to put a limit on it. The facility can be configured not to
1121use a DBM file, but instead, to use a regular file with a maximum size. The
1122data in such a file is searched sequentially, and if the file fills up, the
1123oldest entry is deleted to make way for a new one. This means that some
1124correspondents may receive a second copy of the message after an unpredictable
1125interval. Consult your local information to see if your system is configured
1126this way.
1128More than one \"mail"\ or \"vacation"\ command may be obeyed in a single filter
1129run; they are all honoured, even when they are to the same recipient.
1132.section Logging commands
1133.rset SECTlog "~~chapter.~~section"
1134A log can be kept of actions taken by a filter file. This facility is normally
1135available in conventional configurations, but there are some situations where
1136it might not be. Also, the system administrator may choose to disable it. Check
1137your local information if in doubt.
1139Logging takes place while the filter file is being interpreted. It does not
1140queue up for later like the delivery commands. The reason for this is so that a
1141log file need be opened only once for several write operations. There are two
1142commands, neither of which constitutes a significant delivery. The first
1143defines a file to which logging output is subsequently written:
1145 logfile <<file name>>
1146e.g. logfile @$home/filter.log
1148The file name must be fully qualified. You can use \$home$\, as in this
1149example, to refer to your home directory. The file name may optionally be
1150followed by a mode for the file, which is used if the file has to be created.
1151For example,
1153 logfile @$home/filter.log 0644
1155The number is interpreted as octal, even if it does not begin with a zero.
1156The default for the mode is 600. It is suggested that the \"logfile"\ command
1157normally appear as the first command in a filter file. Once \"logfile"\ has
1158been obeyed, the \"logwrite"\ command can be used to write to the log file:
1160 logwrite "<<some text string>>"
1161e.g. logwrite "@$tod@_log @$message@_id processed"
1163It is possible to have more than one \"logfile"\ command, to specify writing to
1164different log files in different circumstances. Writing takes place at the end
1165of the file, and a newline character is added to the end of each string if
1166there isn't one already there. Newlines can be put in the middle of the string
1167by using the `@\n' escape sequence. Lines from simultaneous deliveries may get
1168interleaved in the file, as there is no interlocking, so you should plan your
1169logging with this in mind. However, data should not get lost.
1172.section The finish command
1173.rset SECTfinish "~~chapter.~~section"
1174The command \"finish"\, which has no arguments, causes Exim to stop
1175interpreting the filter file. This is not a significant action unless preceded
1176by \"seen"\. A filter file containing only \"seen finish"\ is a black hole.
1178.section The testprint command
1179.rset SECTtestprint "~~chapter.~~section"
1180It is sometimes helpful to be able to print out the values of variables when
1181testing filter files. The command
1183 testprint <<text>>
1184e.g. testprint "home=@$home reply@_address=@$reply@_address"
1186does nothing when mail is being delivered. However, when the filtering code is
1187being tested by means of the \-bf-\ option (see section ~~SECTtesting above),
1188the value of the string is written to the standard output.
1190.section The fail command
1191.rset SECTfail "~~chapter.~~section"
1192When Exim's filtering facilities are being used as a system filter, the
1193\"fail"\ command is available, to force delivery failure. Because this command
1194is normally usable only by the system administrator, and not enabled for use by
1195ordinary users, it is described in more detail in the main Exim specification
1196rather than in this document.
1198.section The freeze command
1199.rset SECTfreeze "~~chapter.~~section"
1200When Exim's filtering facilities are being used as a system filter, the
1201\"freeze"\ command is available, to freeze a message on the queue. Because this
1202command is normally usable only by the system administrator, and not enabled
1203for use by ordinary users, it is described in more detail in the main Exim
1204specification rather than in this document.
1207.section The headers command
1208.rset SECTheaders "~~chapter.~~section"
1209The \"headers"\ command can be used to change the target character set which is
1210used when translating the contents of encoded header lines for insertion by the
1211\"@$header@_"\ mechanism (see section ~~SECTheadervariables above). The default
1212can be set in the Exim configuration; if not specified, ISO-8859-1 is used. The
1213only currently supported format for the \"headers"\ command is as in this
1215.display asis
1216headers charset "UTF-8"
1218That is, \"headers"\ is followed by the word \"charset"\ and then the name of a
1219character set. This particular example would be useful if you wanted to compare
1220the contents of a header to a UTF-8 string.
1223.section Obeying commands conditionally
1224.rset SECTif "~~chapter.~~section"
1225Most of the power of filtering comes from the ability to test conditions and
1226obey different commands depending on the outcome. The \"if"\ command is used to
1227specify conditional execution, and its general form is
1229if <<condition>>
1230then <<commands>>
1231elif <<condition>>
1232then <<commands>>
1233else <<commands>>
1236There may be any number of \"elif"\ and \"then"\ sections (including none) and
1237the \"else"\ section is also optional. Any number of commands, including nested
1238\"if"\ commands, may appear in any of the <<commands>> sections.
1240Conditions can be combined by using the words \"and"\ and \"or"\, and round
1241brackets (parentheses) can be used to specify how several conditions are to
1242combine. Without brackets, \"and"\ is more binding than \"or"\.
1243For example,
1244.display asis
1246 $h_subject: contains "Make money" or
1247 $h_precedence: is "junk" or
1248 ($h_sender: matches ^\\d{8}@ and not personal) or
1249 $message_body contains "this is spam"
1251 seen finish
1254A condition can be preceded by \"not"\ to negate it, and there are also some
1255negative forms of condition that are more English-like.
1259.section String testing conditions
1260There are a number of conditions that operate on text strings, using the words
1261`begins', `ends', `is', `contains' and `matches'.
1263Note that if you want to apply the same test to more than one header line, you
1264can easily concatenate them into a single string for testing, as in this
1266.display asis
1267if "$h_to:, $h_cc:" contains me@domain.example then ...
1270If any of the condition names are written in lower case, the testing of letters
1271is done without regard to case; if they are written in upper case (for example,
1272`CONTAINS') then the case of letters is significant.
1274 <<text1>> begins <<text2>>
1275 <<text1>> does not begin <<text2>>
1276e.g. @$header@_from: begins "Friend@@"
1278A `begins' test checks for the presence of the second string at the start of
1279the first, both strings having been expanded.
1281 <<text1>> ends <<text2>>
1282 <<text1>> does not end <<text2>>
1283e.g. @$header@_from: ends ""
1285An `ends' test checks for the presence of the second string at the end of
1286the first, both strings having been expanded.
1288 <<text1>> is <<text2>>
1289 <<text1>> is not <<text2>>
1290e.g. @$local@_part@_suffix is "-foo"
1292An `is' test does an exact match between the strings, having first expanded
1293both strings.
1295 <<text1>> contains <<text2>>
1296 <<text1>> does not contain <<text2>>
1297e.g. @$header@_subject: contains "evolution"
1299A `contains' test does a partial string match, having expanded both strings.
1301 <<text1>> matches <<text2>>
1302 <<text1>> does not match <<text2>>
1303e.g. @$sender@_address matches "(bill|john)@@"
1305For a `matches' test, after expansion of both strings, the second one is
1306interpreted as a regular expression. Exim uses the PCRE regular expression
1307library, which provides regular expressions that are compatible with Perl.
1310The match succeeds if the regular expression matches any part of the first
1311string. If you want a regular expression to match only at the start or end of
1312the subject string, you must encode that requirement explicitly, using the @^
1313or @$ metacharacters. The above example, which is not so constrained, matches
1314all these addresses:
1315.display asis
1321To match only the first two, you could use this:
1322.display asis
1323if $sender_address matches "^(bill|john)@" then ...
1327Care must be taken if you need a backslash in a regular expression, because
1328backslashes are interpreted as escape characters both by the string expansion
1329code and by Exim's normal processing of strings in quotes. For example, if you
1330want to test the sender address for a domain ending in \".com"\ the regular
1331expression is
1332.display asis
1335The backslash and dollar sign in that expression have to be escaped when used
1336in a filter command, as otherwise they would be interpreted by the expansion
1337code. Thus what you actually write is
1338.display asis
1339if $sender_address matches \\.com\$
1341An alternative way of handling this is to make use of the \"@\N"\ expansion
1342flag for suppressing expansion:
1343.display asis
1344if $sender_address matches \N\.com$\N
1346Everything between the two occurrences of \"@\N"\ is copied without change by
1347the string expander (and in fact you do not need the final one, because it is
1348at the end of the string).
1350If the regular expression is given in quotes (mandatory only if it contains
1351white space) you have to write either
1352.display asis
1353if $sender_address matches "\\\\.com\\$"
1356.display asis
1357if $sender_address matches "\\N\\.com$\\N"
1361If the regular expression contains bracketed sub-expressions, numeric
1362variable substitutions such as \$1$\ can be used in the subsequent actions
1363after a successful match. If the match fails, the values of the numeric
1364variables remain unchanged. Previous values are not restored after \"endif"\ --
1365in other words, only one set of values is ever available. If the condition
1366contains several sub-conditions connected by \"and"\ or \"or"\, it is the
1367strings extracted from the last successful match that are available in
1368subsequent actions. Numeric variables from any one sub-condition are also
1369available for use in subsequent sub-conditions, since string expansion of a
1370condition occurs just before it is tested.
1372.section Numeric testing conditions
1373The following conditions are available for performing numerical tests:
1375 <<number1>> is above <<number2>>
1376 <<number1>> is not above <<number2>>
1377 <<number1>> is below <<number2>>
1378 <<number1>> is not below <<number2>>
1379e.g. @$message@_size is not above 10k
1381The <<number>> arguments must expand to strings of digits, optionally followed
1382by one of the letters K or M (upper case or lower case) which cause
1383multiplication by 1024 and 1024x1024 respectively.
1385.section Testing for significant deliveries
1386Whether or not any previously obeyed filter commands have resulted in a
1387significant delivery can be tested by the condition \"delivered"\, for example:
1388.display asis
1389if not delivered then save mail/anomalous endif
1392.section Testing for error messages
1393The condition \"error@_message"\ is true if the incoming message is a mail
1394delivery error message (bounce message). Putting the command
1395.display asis
1396if error_message then finish endif
1398at the head of your filter file is a useful insurance against things going
1399wrong in such a way that you cannot receive delivery error reports, and is
1400highly recommended. Note that \"error@_message"\ is a condition, not an
1401expansion variable, and therefore is not preceded by \@$\.
1403.section Testing for personal mail
1404.rset SECTpersonal "~~chapter.~~section"
1405A common requirement is to distinguish between incoming personal mail and mail
1406from a mailing list,
1407or from a robot or other automatic process (for example, a bounce message).
1408In particular, this test is normally required for so-called `vacation
1409messages'. The condition
1411 personal
1413is a shorthand for
1415 not error@_message and
1416 @$header@_to: contains "@$local@_part@@@$domain" and
1417 @$header@_from: does not contain "@$local@_part@@@$domain" and
1418 @$header@_from: does not contain "server@@" and
1419 @$header@_from: does not contain "daemon@@" and
1420 @$header@_from: does not contain "root@@" and
1421 @$header@_auto-submitted: does not contain "auto-" and
1422 @$header@_subject: does not contain "circular" and
1423 @$header@_precedence: does not contain "bulk" and
1424 @$header@_precedence: does not contain "list" and
1425 @$header@_precedence: does not contain "junk"
1427The variable \$local@_part$\ contains the local part of the mail address of
1428the user whose filter file is being run -- it is normally your login id. The
1429\$domain$\ variable contains the mail domain. This condition tests for the
1430appearance of the current user in the \"To:"\ header, checks that the sender is
1431not the current user or one of a number of common daemons, and checks the
1432content of the \"Subject:"\ and \"Precedence:"\ headers.
1434If prefixes or suffixes are in use for local parts -- something which depends
1435on the configuration of Exim (see section ~~SECTmbox below) -- the first two
1436tests above are also done with
1437.display asis
1440instead of just \$local@_part$\. If the system is configured to rewrite local
1441parts of mail addresses, for example, to rewrite `dag46' as `Dirk.Gently',
1442the rewritten form of the address is also used in the tests.
1444This example shows the use of \"personal"\ in a filter file that is sending out
1445vacation messages:
1446.display asis
1447if personal then
1448 mail
1449 to $reply_address
1450 subject "Re: $h_subject:"
1451 file $home/vacation/message
1452 once $home/vacation/once
1453 once_repeat 10d
1457It is quite common for people who have mail accounts on a number of different
1458systems to forward all their mail to one system, and in this case a check for
1459personal mail should test all their various mail addresses. To allow for this,
1460the \"personal"\ condition keyword can be followed by
1462alias <<address>>
1464any number of times, for example
1465.display asis
1466if personal alias smith@else.where.example
1467 alias
1468then ...
1470This causes messages containing the alias addresses in any places where the
1471local address is tested to be treated as personal.
1473.section Testing delivery status
1474There are two conditions which are intended mainly for use in system filter
1475files, but which are available in users' filter files as well. The condition
1476\"first@_delivery"\ is true if this is the first attempt to deliver the
1477message, and false otherwise.
1479This indicator is not reset until the first delivery process successfully
1480terminates; if there is a crash or a power failure (for example), the next
1481delivery attempt is also a `first delivery'.
1484In a user filter file it will be false only if
1485there was previously an error in the filter, or if a delivery for the user
1486failed owing to, for example, a quota error, or forwarding to a remote
1487address that was deferred for some reason.
1489The condition \"manually@_thawed"\ is true only if the message was `frozen' for
1490some reason, and was subsequently released by the system administrator. It is
1491unlikely to be of use in users' filter files.
1493.section Testing a list of addresses
1494There is a facility for looping through a list of addresses and applying a
1495condition to each of them. It takes the form
1497foranyaddress <<string>> (<<condition>>)
1499where <<string>> is interpreted as a list of RFC 2822 addresses, as in a
1500typical header line, and <<condition>> is any valid filter condition or
1501combination of conditions. The `group' syntax that is defined for certain
1502header lines that contain addresses is supported.
1504The parentheses surrounding the condition are mandatory, to delimit it from
1505possible further sub-conditions of the enclosing \"if"\ command. Within the
1506condition, the expansion variable \$thisaddress$\ is set to the non-comment
1507portion of each of the addresses in the string in turn. For example, if the
1508string is
1509.display asis
1510B.Simpson <bart@sfld.example>, lisa@sfld.example (his sister)
1512then \$thisaddress$\ would take on the values \"bart@@sfld.example"\ and
1513\"lisa@@sfld.example"\ in turn.
1515If there are no valid addresses in the list, the whole condition is false. If
1516the internal condition is true for any one address, the overall condition is
1517true and the loop ends. If the internal condition is false for all addresses in
1518the list, the overall condition is false. This example tests for the presence
1519of an eight-digit local part in any address in a \To:\ header:
1520.display asis
1521if foranyaddress $h_to: ( $thisaddress matches ^\\d{8}@ ) then ...
1523When the overall condition is true, the value of \$thisaddress$\ in the
1524commands that follow \"then"\ is the last value it took on inside the loop. At
1525the end of the \"if"\ command, the value of \$thisaddress$\ is reset to what it
1526was before. It is best to avoid the use of multiple occurrences of
1527\"foranyaddress"\, nested or otherwise, in a single \"if"\ command, if the
1528value of \$thisaddress$\ is to be used afterwards, because it isn't always
1529clear what the value will be. Nested \"if"\ commands should be used instead.
1531Header lines can be joined together if a check is to be applied to more than
1532one of them. For example:
1533.display asis
1534if foranyaddress $h_to:,$h_cc: ....
1536scans through the addresses in both the \To:\ and the \Cc:\ headers.
1538.section Multiple personal mailboxes
1539.rset SECTmbox "~~chapter.~~section"
1540The system administrator can configure Exim so that users can set up variants
1541on their email addresses and handle them separately. Consult your system
1542administrator or local documentation to see if this facility is enabled on your
1543system, and if so, what the details are.
1545The facility involves the use of a prefix or a suffix on an email address. For
1546example, all mail addressed to \lg103-<<something>>\ would be the property of
1547user \lg103\, who could determine how it was to be handled, depending on the
1548value of <<something>>.
1550There are two possible ways in which this can be set up. The first possibility
1551is the use of multiple \(.forward)\ files. In this case, mail to \lg103-foo\,
1552for example, is handled by looking for a file called \.forward-foo\ in
1553\lg103's\ home directory. If such a file does not exist, delivery fails and the
1554message is returned to its sender.
1556The alternative approach is to pass all messages through a single \(.forward)\
1557file, which must be a filter file in order to distinguish between the different
1558cases by referencing the variables \$local@_part@_prefix$\ or
1559\$local@_part@_suffix$\, as in the final example in section ~~SECTex below. If
1560the filter file does not handle a prefixed or suffixed address, delivery fails
1561and the message is returned to its sender.
1563It is possible to configure Exim to support both schemes at once. In this case,
1564a specific \.forward-foo\ file is first sought; if it is not found, the basic
1565\(.forward)\ file is used.
1567The \"personal"\ test (see section ~~SECTpersonal) includes prefixes and
1568suffixes in its checking.
1571.section Ignoring delivery errors
1572As was explained above, filtering just sets up addresses for delivery -- no
1573deliveries are actually done while a filter file is active. If any of the
1574generated addresses subsequently suffers a delivery failure, an error message
1575is generated in the normal way. However, if the filter command which sets up a
1576delivery is preceded by the word \"noerror"\, errors for that delivery,
1577$it{and any deliveries consequent on it} (that is, from alias, forwarding, or
1578filter files it invokes) are ignored.
1581.section Examples of Exim filter commands
1582.rset SECTex "~~chapter.~~section"
1583Simple forwarding:
1584.display asis
1585# Exim filter
1586deliver baggins@rivendell.middle-earth.example
1588Vacation handling using traditional means, assuming that the \.vacation.msg\
1589and other files have been set up in your home directory:
1590.display asis
1591# Exim filter
1592unseen pipe "/usr/ucb/vacation \"$local_part\""
1594Vacation handling inside Exim, having first created a file called
1595\.vacation.msg\ in your home directory:
1596.display asis
1597# Exim filter
1598if personal then vacation endif
1600File some messages by subject:
1601.display asis
1602# Exim filter
1603if $header_subject: contains "empire" or
1604 $header_subject: contains "foundation"
1606 save $home/mail/f+e
1609Save all non-urgent messages by weekday:
1610.display asis
1611# Exim filter
1612if $header_subject: does not contain "urgent" and
1613 $tod_full matches "^(...),"
1615 save $home/mail/$1
1618Throw away all mail from one site, except from postmaster:
1619.display asis
1620# Exim filter
1621if $reply_address contains "" and
1622 $reply_address does not contain "postmaster@"
1624 seen finish
1627.if ~~sgcal
1628.if ~~sys.leftonpage < 6ld
1632Handle multiple personal mailboxes
1633.display asis
1634# Exim filter
1635if $local_part_suffix is "-foo"
1637 save $home/mail/foo
1638elif $local_part_suffix is "-bar"
1640 save $home/mail/bar
1644. End of filter