mairix(1) index and search mail folders

SYNOPSIS

Indexing

mairix [ -v|--verbose ] [ -p|--purge ] [ -f|--rcfile mairixrc ] [ -F|--fast-index ] [ --force-hash-key-new-database hash ]

Searching

mairix [ -v|--verbose ] [ -f|--rcfile mairixrc ] [ -r|--raw-output ] [ -x|--excerpt-output ] [ -H|--force-hardlinks ] [ -o|--mfolder mfolder ] [ -a|--augment ] [ -t|--threads ] search-patterns

Other

mairix [ -h|--help ]

mairix [ -V|--version ]

mairix [ -d|--dump ]

DESCRIPTION

mairix indexes and searches a collection of email messages. The folders containing the messages for indexing are defined in the configuration file. The indexing stage produces a database file. The database file provides rapid access to details of the indexed messages during searching operations. A search normally produces a folder (so-called mfolder) containing the matched messages. However, a raw mode (-r) exists which just lists the matched messages instead.

It can operate with the following folder types

*
maildir
*
MH (compatible with the MH folder formats used by xmh, sylpheed, claws-mail, nnml (Gnus) and evolution)
*
mbox (including mboxes that have been compressed with gzip or bzip2)

If maildir or MH source folders are used, and a search outputs its matches to an mfolder in maildir or MH format, symbolic links are used to reference the original messages inside the mfolder. However, if mbox folders are involved, copies of messages are made instead.

OPTIONS

mairix decides whether indexing or searching is required by looking for the presence of any search-patterns on the command line.

Special modes

-h, --help

Show usage summary and exit

-V, --version
Show program version and exit

-d

Dump the database's contents in human-readable form to stdout.

General options

-f mairixrc

--rcfile mairixrc

Specify an alternative configuration file to use. The default configuration file is ~/.mairixrc.

-v, --verbose

Make the output more verbose

-Q, --no-integrity-checks

Normally mairix will do some internal integrity tests on the database. The -Q option removes these checks, making mairix run faster, but it will be less likely to detect internal problems if any bugs creep in.

The nochecks directive in the rc file has the same effect.

--unlock

mairix locks its database file during any indexing or searching operation to prevent multiple indexing runs interfering with each other, or an indexing run interfering with search runs. The --unlock option removes the lockfile before doing the requested indexing or searching operation. This is a convenient way of cleaning up a stale lockfile if an earlier run crashed for some reason or was aborted.

Indexing options

-p, --purge

Cause stale (dead) messages to be purged from the database during an indexing run. (Normally, stale messages are left in the database because of the additional cost of compacting away the storage that they take up.)

-F, --fast-index

When processing maildir and MH folders, mairix normally compares the mtime and size of each message against the values stored in the database. If they have changed, the message will be rescanned. This check requires each message file to be stat'ed. For large numbers of messages in these folder types, this can be a sizeable overhead.

This option tells mairix to assume that when a message currently on-disc has a name matching one already in the database, it should assume the message is unchanged.

A later indexing run without using this option will fix up any rescans that were missed due to its use.

--force-hash-key-new-database hash

This option should only be used for debugging.
If a new database is created, hash is used as hash key, instead of a random hash.

Search options

-a, --augment

Append newly matches messages to the current mfolder instead of creating the mfolder from scratch.

-t, --threads

As well as returning the matched messages, also return every message in the same thread as one of the real matches.

-r, --raw-output

Instead of creating an mfolder containing the matched messages, just show their paths on stdout.

-x, --excerpt-output

Instead of creating an mfolder containing the matched messages, display an excerpt from their headers on stdout. The excerpt shows To, Cc, From, Subject and Date.

-H, --force-hardlinks

Instead of creating symbolic links, force the use of hardlinks. This helps mailers such as alpine to realize that there are new mails in the search folder.

-o mfolder

--mfolder mfolder

Specify a temporary alternative path for the mfolder to use, overriding the mfolder directive in the rc file.

mairix will refuse to output search results into any folder that appears to be amongst those that are indexed. This is to prevent accidental deletion of emails.

Search patterns

t:word

Match word in the To: header.

c:word

Match word in the Cc: header.

f:word

Match word in the From: header.

s:word

Match word in the Subject: header.

m:word

Match word in the Message-ID: header.

b:word

Match word in the message body.

Message body is taken to mean any body part of type text/plain or text/html. For text/html, text within meta tags is ignored. In particular, the URLs inside <A HREF="..."> tags are not currently indexed. Non-text attachments are ignored. If there's an attachment of type message/rfc822, this is parsed and the match is performed on this sub-message too. If a hit occurs, the enclosing message is treated as having a hit.

d:[start-datespec]-[end-datespec]

Match messages with Date: headers lying in the specific range.

z:[low-size]-[high-size]

Match messages whose size lies in the specified range. If the low-size argument is omitted it defaults to zero. If the high-size argument is omitted it defaults to infinite size.

For example, to match messages between 10kilobytes and 20kilobytes in size, the following search term can be used:

mairix z:10k-20k

The suffix 'k' on a number means multiply by 1024, and the suffix 'M' on a number means multiply by 1024*1024.

n:word

Match word occurring as the name of an attachment in the message. Since attachment names are usually long, this option would usually be used in the substring form. So

mairix n:mairix=

would match all messages which have attachments whose names contain the substring mairix.

The attachment name is determined from the name=xxx or filename=xxx qualifiers on the Content-Type: and Content-Disposition: headers respectively.

F:flags

Match messages with particular flag settings. The available flags are 's' meaning seen, 'r' meaning replied, and 'f' meaning flagged. The flags are case-insensitive. A flag letter may be prefixed by a '-' to negate its sense. Thus

mairix F:-s d:1w-

would match any unread message less than a week old, and

mairix F:f-r d:-1m

would match any flagged message older than a month which you haven't replied to yet.

Note that the flag characters and their meanings agree with those used as the suffix letters on message filenames in maildir folders.

Searching for a match amongst more than one part of a message

Multiple body parts may be grouped together, if a match in any of them is sought. Common examples follow.

tc:word

Match word in either the To: or Cc: headers (or both).

bs:word

Match word in either the Subject: header or the message body (or both).

The a: search pattern is an abbreviation for tcf:; i.e. match the word in the To:, Cc: or From: headers. ("a" stands for "address" in this case.)

Match words

The word argument to the search strings can take various forms.

~word

Match messages not containing the word.

word1,word2

This matches if both the words are matched in the specified message part.

word1/word2

This matches if either of the words are matched in the specified message part.

substring=

Match any word containing substring as a substring

substring=N

Match any word containing substring, allowing up to N errors in the match. For example, if N is 1, a single error is allowed, where an error can be
*
a missing letter
*
an extra letter
*
a different letter.

^substring=

Match any word containing substring as a substring, with the requirement that substring occurs at the beginning of the matched word.

Precedence matters

The binding order of the constructions is:

1.
Individual command line arguments define separate conditions which are AND-ed together

2.
Within a single argument, the letters before the colon define which message parts the expression applies to. If there is no colon, the expression applies to all the headers listed earlier and the body.

3.
After the colon, commas delineate separate disjuncts, which are OR-ed together.

4.
Each disjunct may contain separate conjuncts, which are separated by plus signs. These conditions are AND-ed together.

5.
Each conjunct may start with a tilde to negate it, and may be followed by a slash to indicate a substring match, optionally followed by an integer to define the maximum number of errors allowed.

Date specification

This section describes the syntax used for specifying dates when searching using the `d:' option.

Dates are specified as a range. The start and end of the range can both be specified. Alternatively, if the start is omitted, it is treated as being the beginning of time. If the end is omitted, it is treated as the current time.

There are 4 basic formats:

d:start-end

Specify both start and end explicitly
d:start-
Specify start, end is the current time
d:-end
Specify end, start is 'a long time ago' (i.e. early enough to include any message).
d:period
Specify start and end implicitly, as the start and end of the period given.

The start and end can be specified either absolute or relative. A relative endpoint is given as a number followed by a single letter defining the scaling:

letter short for example meaning


d days 3d 3 days
w weeks 2w 2 weeks (14 days)
m months 5m 5 months (150 days)
y years 4y 4 years (4*365 days)

Months are always treated as 30 days, and years as 365 days, for this purpose.

Absolute times can be specified in many forms. Some forms have different meanings when they define a start date from that when they define an end date. Where a single expression specifies both the start and end (i.e. where the argument to d: doesn't contain a `-'), it will usually have different interpretations in the two cases.

In the examples below, suppose the current date is Sunday May 18th, 2003 (when I started to write this material.)

Example Start date End date Notes


d:20030301-20030425 March 1st, 2003 25th April, 2003
d:030301-030425 March 1st, 2003 April 25th, 2003 century assumed
d:mar1-apr25 March 1st, 2003 April 25th, 2003
d:Mar1-Apr25 March 1st, 2003 April 25th, 2003 case insensitive
d:MAR1-APR25 March 1st, 2003 April 25th, 2003 case insensitive
d:1mar-25apr March 1st, 2003 April 25th, 2003 date and month in either order
d:2002 January 1st, 2002 December 31st, 2002 whole year
d:mar March 1st, 2003 March 31st, 2003 most recent March
d:oct October 1st, 2002 October 31st, 2002 most recent October
d:21oct-mar October 21st, 2002 March 31st, 2003 start before end
d:21apr-mar April 21st, 2002 March 31st, 2003 start before end
d:21apr- April 21st, 2003 May 18th, 2003 end omitted
d:-21apr January 1st, 1900 April 21st, 2003 start omitted
d:6w-2w April 6th, 2003 May 4th, 2003 both dates relative
d:21apr-1w April 21st, 2003 May 11th, 2003 one date relative
d:21apr-2y April 21st, 2001 May 11th, 2001 start before end
d:99-11 January 1st, 1999 May 11th, 2003 2 digits are a day of the month if possible, otherwise a year
d:99oct-1oct October 1st, 1999 October 1st, 2002 end before now, single digit is a day of the month
d:99oct-01oct October 1st, 1999 October 31st, 2001 2 digits starting with zero treated as a year
d:oct99-oct1 October 1st, 1999 October 1st, 2002 day and month in either order
d:oct99-oct01 October 1st, 1999 October 31st, 2001 year and month in either order

The principles in the table work as follows.

  • When the expression defines a period of more than a day (i.e. if a month or year is specified), the earliest day in the period is taken when the start date is defined, and the last day in the period if the end of the range is being defined.
  • The end date is always taken to be on or before the current date.
  • The start date is always taken to be on or before the end date.

SETTING UP THE MATCH FOLDER

If the match folder does not exist when running in search mode, it is automatically created. For 'mformat=maildir' (the default), this should be all you need to do. If you use 'mformat=mh', you may have to run some commands before your mailer will recognize the folder. e.g. for mutt, you could do

mkdir -p /home/richard/Mail/mfolder touch /home/richard/Mail/mfolder/.mh_sequences

which seems to work. Alternatively, within mutt, you could set MBOX_TYPE to in advance.

If you use Sylpheed, the best way seems to be to create the new folder from within Sylpheed before letting mairix write into it.

EXAMPLES

Suppose my email address is <[email protected]>.

Either of the following will match all messages newer than 3 months from me with the word 'chrony' in the subject line:

mairix d:3m- f:richard+doesnt+exist s:chrony mairix d:3m- f:[email protected] s:chrony

Suppose I don't mind a few spurious matches on the address, I want a wider date range, and I suspect that some messages I replied to might have had the subject keyword spelt wrongly (let's allow up to 2 errors):

mairix d:6m- f:richard s:chrony=2

NOTES

mairix works exclusively in terms of words. The index that's built in indexing mode contains a table of which words occur in which messages. Hence, the search capability is based on finding messages that contain particular words. mairix defines a word as any string of alphanumeric characters + underscore. Any whitespace, punctuation, hyphens etc are treated as word boundaries.

mairix has special handling for the To:, Cc: and From: headers. Besides the normal word scan, these headers are scanned a second time, where the characters '@', '-' and '.' are also treated as word characters. This allows most (if not all) email addresses to appear in the database as single words. So if you have a mail from [email protected], it will match on both these searches

mairix f:foobar mairix f:[email protected]

It should be clear by now that the searching cannot be used to find messages matching general regular expressions. This has never been much of a limitation. Most searches are for particular keywords that were in the messages, or details of the recipients, or the approximate date.

It's also worth pointing out that there is no 'locality' information stored, so you can't search for messages that have one words 'close' to some other word. For every message and every word, there is a simple yes/no condition stored - whether the message contains the word in a particular header or in the body. So far this has proved to be adequate. mairix has a similar feel to using an Internet search engine.

FILES

~/.mairixrc

AUTHOR

Copyright (C) 2002-2006 Richard P. Curnow <[email protected]>

BUGS

We need a plugin scheme to allow more types of attachment to be scanned and indexed.