1NOTMUCH-SEARCH-TERMS(7) notmuch NOTMUCH-SEARCH-TERMS(7)
2
3
4
6 notmuch-search-terms - syntax for notmuch queries
7
9 notmuch count [option ...] <search-term> ...
10
11 notmuch dump [--gzip] [--format=(batch-tag|sup)] [--output=<file>] [--]
12 [<search-term> ...]
13
14 notmuch reindex [option ...] <search-term> ...
15
16 notmuch search [option ...] <search-term> ...
17
18 notmuch show [option ...] <search-term> ...
19
20 notmuch tag +<tag> ... -<tag> [--] <search-term> ...
21
23 Several notmuch commands accept a common syntax for search terms.
24
25 The search terms can consist of free-form text (and quoted phrases)
26 which will match all messages that contain all of the given
27 terms/phrases in the body, the subject, or any of the sender or recipi‐
28 ent headers.
29
30 As a special case, a search string consisting of exactly a single as‐
31 terisk ("*") will match all messages.
32
33 Search prefixes
34 In addition to free text, the following prefixes can be used to force
35 terms to match against specific portions of an email, (where <brackets>
36 indicate user-supplied values).
37
38 Some of the prefixes with <regex> forms can be also used to restrict
39 the results to those whose value matches a regular expression (see
40 regex(7)) delimited with //, for example:
41
42 notmuch search 'from:"/bob@.*[.]example[.]com/"'
43
44 body:<word-or-quoted-phrase>
45 Match terms in the body of messages.
46
47 from:<name-or-address> or from:/<regex>/
48 The from: prefix is used to match the name or address of the
49 sender of an email message.
50
51 to:<name-or-address>
52 The to: prefix is used to match the names or addresses of any
53 recipient of an email message, (whether To, Cc, or Bcc).
54
55 subject:<word-or-quoted-phrase> or subject:/<regex>/
56 Any term prefixed with subject: will match only text from the
57 subject of an email. Searching for a phrase in the subject is
58 supported by including quotation marks around the phrase, imme‐
59 diately following subject:.
60
61 attachment:<word>
62 The attachment: prefix can be used to search for specific file‐
63 names (or extensions) of attachments to email messages.
64
65 mimetype:<word>
66 The mimetype: prefix will be used to match text from the con‐
67 tent-types of MIME parts within email messages (as specified by
68 the sender).
69
70 tag:<tag> or tag:/<regex>/ or is:<tag> or is:/<regex>/
71 For tag: and is: valid tag values include inbox and unread by
72 default for new messages added by notmuch-new(1) as well as any
73 other tag values added manually with notmuch-tag(1).
74
75 id:<message-id> or mid:<message-id> or mid:/<regex>/
76 For id: and mid:, message ID values are the literal contents of
77 the Message-ID: header of email messages, but without the '<',
78 '>' delimiters.
79
80 thread:<thread-id>
81 The thread: prefix can be used with the thread ID values that
82 are generated internally by notmuch (and do not appear in email
83 messages). These thread ID values can be seen in the first col‐
84 umn of output from notmuch-search(1)
85
86 thread:{<notmuch query>}
87 Threads may be searched for indirectly by providing an arbitrary
88 notmuch query in {}. For example, the following returns threads
89 containing a message from mallory and one (not necessarily the
90 same message) with Subject containing the word "crypto".
91
92 % notmuch search 'thread:"{from:mallory}" and thread:"{subject:crypto}"'
93
94 The performance of such queries can vary wildly. To understand
95 this, the user should think of the query thread:{<something>} as
96 expanding to all of the thread IDs which match <something>; not‐
97 much then performs a second search using the expanded query.
98
99 path:<directory-path> or path:<directory-path>/** or path:/<regex>/
100 The path: prefix searches for email messages that are in partic‐
101 ular directories within the mail store. The directory must be
102 specified relative to the top-level maildir (and without the
103 leading slash). By default, path: matches messages in the speci‐
104 fied directory only. The "/**" suffix can be used to match mes‐
105 sages in the specified directory and all its subdirectories re‐
106 cursively. path:"" matches messages in the root of the mail
107 store and, likewise, path:** matches all messages.
108
109 path: will find a message if any copy of that message is in the
110 specific directory.
111
112 folder:<maildir-folder> or folder:/<regex>/
113 The folder: prefix searches for email messages by maildir or MH
114 folder. For MH-style folders, this is equivalent to path:. For
115 maildir, this includes messages in the "new" and "cur" subdirec‐
116 tories. The exact syntax for maildir folders depends on your
117 mail configuration. For maildir++, folder:"" matches the inbox
118 folder (which is the root in maildir++), other folder names al‐
119 ways start with ".", and nested folders are separated by "."s,
120 such as folder:.classes.topology. For "file system" maildir, the
121 inbox is typically folder:INBOX and nested folders are separated
122 by slashes, such as folder:classes/topology.
123
124 folder: will find a message if any copy of that message is in
125 the specific folder.
126
127 date:<since>..<until> or date:<date>
128 The date: prefix can be used to restrict the results to only
129 messages within a particular time range (based on the Date:
130 header).
131
132 See DATE AND TIME SEARCH below for details on the range expres‐
133 sion, and supported syntax for <since> and <until> date and time
134 expressions.
135
136 The time range can also be specified using timestamps without
137 including the date prefix using a syntax of:
138
139 <initial-timestamp>..<final-timestamp>
140
141 Each timestamp is a number representing the number of seconds
142 since 1970-01-01 00:00:00 UTC. Specifying a time range this way
143 is considered legacy and predates the date prefix.
144
145 lastmod:<initial-revision>..<final-revision>
146 The lastmod: prefix can be used to restrict the result by the
147 database revision number of when messages were last modified
148 (tags were added/removed or filenames changed). This is usually
149 used in conjunction with the --uuid argument to not‐
150 much-search(1) to find messages that have changed since an ear‐
151 lier query.
152
153 query:<name>
154 The query: prefix allows queries to refer to previously saved
155 queries added with notmuch-config(1).
156
157 property:<key>=<value>
158 The property: prefix searches for messages with a particular
159 <key>=<value> property pair. Properties are used internally by
160 notmuch (and extensions) to add metadata to messages. A given
161 key can be present on a given message with several different
162 values. See notmuch-properties(7) for more details.
163
164 sexp:<subquery>
165 The sexp: prefix allows subqueries in the format documented in
166 notmuch-sexp-queries(7). Note that subqueries containing spaces
167 must be quoted, and any embedded double quotes must be escaped
168 (see Quoting).
169
170 User defined prefixes are also supported, see notmuch-config(1) for de‐
171 tails.
172
173 Operators
174 In addition to individual terms, multiple terms can be combined with
175 Boolean operators (and, or, not, and xor). Each term in the query will
176 be implicitly connected by a logical AND if no explicit operator is
177 provided (except that terms with a common prefix will be implicitly
178 combined with OR). The shorthand '-<term>' can be used for 'not
179 <term>' but unfortunately this does not work at the start of an expres‐
180 sion. Parentheses can also be used to control the combination of the
181 Boolean operators, but will have to be protected from interpretation by
182 the shell, (such as by putting quotation marks around any parenthesized
183 expression).
184
185 In addition to the standard boolean operators, Xapian provides several
186 operators specific to text searching.
187
188 notmuch search term1 NEAR term2
189
190 will return results where term1 is within 10 words of term2. The
191 threshold can be set like this:
192
193 notmuch search term1 NEAR/2 term2
194
195 The search
196
197 notmuch search term1 ADJ term2
198
199 will return results where term1 is within 10 words of term2, but in the
200 same order as in the query. The threshold can be set the same as with
201 NEAR:
202
203 notmuch search term1 ADJ/7 term2
204
205 Stemming
206 Stemming in notmuch means that these searches
207
208 notmuch search detailed
209 notmuch search details
210 notmuch search detail
211
212 will all return identical results, because Xapian first "reduces" the
213 term to the common stem (here 'detail') and then performs the search.
214
215 There are two ways to turn this off: a search for a capitalized word
216 will be performed unstemmed, so that one can search for "John" and not
217 get results for "Johnson"; phrase searches are also unstemmed (see be‐
218 low for details). Stemming is currently only supported for English.
219 Searches for words in other languages will be performed unstemmed.
220
221 Wildcards
222 It is possible to use a trailing '*' as a wildcard. A search for
223 'wildc*' will match 'wildcard', 'wildcat', etc.
224
225 Boolean and Probabilistic Prefixes
226 Xapian (and hence notmuch) prefixes are either boolean, supporting ex‐
227 act matches like "tag:inbox" or probabilistic, supporting a more flexi‐
228 ble term based searching. Certain special prefixes are processed by
229 notmuch in a way not strictly fitting either of Xapian's built in
230 styles. The prefixes currently supported by notmuch are as follows.
231
232 Boolean
233 tag:, id:, thread:, folder:, path:, property:
234
235 Probabilistic
236 body:, to:, attachment:, mimetype:
237
238 Special
239 from:, query:, subject:, sexp:
240
241 Terms and phrases
242 In general Xapian distinguishes between lists of terms and phrases.
243 Phrases are indicated by double quotes (but beware you probably need to
244 protect those from your shell) and insist that those unstemmed words
245 occur in that order. One useful, but initially surprising feature is
246 that the following are equivalent ways to write the same phrase.
247
248 • "a list of words"
249
250 • a-list-of-words
251
252 • a/list/of/words
253
254 • a.list.of.words
255
256 Both parenthesised lists of terms and quoted phrases are ok with proba‐
257 bilistic prefixes such as to:, from:, and subject:. For prefixes sup‐
258 porting regex search, the parenthesised list should be quoted. In par‐
259 ticular
260
261 subject:"(pizza free)"
262
263 is equivalent to
264
265 subject:pizza and subject:free
266
267 Both of these will match a subject "Free Delicious Pizza" while
268
269 subject:"pizza free"
270
271 will not.
272
273 Quoting
274 Double quotes are also used by the notmuch query parser to protect
275 boolean terms, regular expressions, or subqueries containing spaces or
276 other special characters, e.g.
277
278 tag:"a tag"
279
280 folder:"/^.*/(Junk|Spam)$/"
281
282 thread:"{from:mallory and date:2009}"
283
284 As with phrases, you need to protect the double quotes from the shell
285 e.g.
286
287 % notmuch search 'folder:"/^.*/(Junk|Spam)$/"'
288 % notmuch search 'thread:"{from:mallory and date:2009}" and thread:{to:mallory}'
289
290 Double quotes within query strings need to be doubled to escape them.
291
292 % notmuch search 'tag:"""quoted tag"""'
293 % notmuch search 'sexp:"(or ""wizard"" ""php"")"'
294
296 notmuch understands a variety of standard and natural ways of express‐
297 ing dates and times, both in absolute terms ("2012-10-24") and in rela‐
298 tive terms ("yesterday"). Any number of relative terms can be combined
299 ("1 hour 25 minutes") and an absolute date/time can be combined with
300 relative terms to further adjust it. A non-exhaustive description of
301 the syntax supported for absolute and relative terms is given below.
302
303 The range expression
304 date:<since>..<until>
305
306 The above expression restricts the results to only messages from
307 <since> to <until>, based on the Date: header.
308
309 <since> and <until> can describe imprecise times, such as "yesterday".
310 In this case, <since> is taken as the earliest time it could describe
311 (the beginning of yesterday) and <until> is taken as the latest time it
312 could describe (the end of yesterday). Similarly, date:january..febru‐
313 ary matches from the beginning of January to the end of February.
314
315 If specifying a time range using timestamps in conjunction with the
316 date prefix, each timestamp must be preceded by @ (ASCII hex 40). As
317 above, each timestamp is a number representing the number of seconds
318 since 1970-01-01 00:00:00 UTC. For example:
319 date:@<initial-timestamp>..@<final-timestamp>
320
321 Currently, spaces in range expressions are not supported. You can re‐
322 place the spaces with '_', or (in most cases) '-', or (in some cases)
323 leave the spaces out altogether. Examples in this man page use spaces
324 for clarity.
325
326 Open-ended ranges are supported. I.e. it's possible to specify
327 date:..<until> or date:<since>.. to not limit the start or end time,
328 respectively.
329
330 Single expression
331 date:<expr> works as a shorthand for date:<expr>..<expr>. For example,
332 date:monday matches from the beginning of Monday until the end of Mon‐
333 day.
334
335 Relative date and time
336 [N|number] (years|months|weeks|days|hours|hrs|minutes|mins|sec‐
337 onds|secs) [...]
338
339 All refer to past, can be repeated and will be accumulated.
340
341 Units can be abbreviated to any length, with the otherwise ambiguous
342 single m being m for minutes and M for months.
343
344 Number can also be written out one, two, ..., ten, dozen, hundred. Ad‐
345 ditionally, the unit may be preceded by "last" or "this" (e.g., "last
346 week" or "this month").
347
348 When combined with absolute date and time, the relative date and time
349 specification will be relative from the specified absolute date and
350 time.
351
352 Examples: 5M2d, two weeks
353
354 Supported absolute time formats
355 • H[H]:MM[:SS] [(am|a.m.|pm|p.m.)]
356
357 • H[H] (am|a.m.|pm|p.m.)
358
359 • HHMMSS
360
361 • now
362
363 • noon
364
365 • midnight
366
367 • Examples: 17:05, 5pm
368
369 Supported absolute date formats
370 • YYYY-MM[-DD]
371
372 • DD-MM[-[YY]YY]
373
374 • MM-YYYY
375
376 • M[M]/D[D][/[YY]YY]
377
378 • M[M]/YYYY
379
380 • D[D].M[M][.[YY]YY]
381
382 • D[D][(st|nd|rd|th)] Mon[thname] [YYYY]
383
384 • Mon[thname] D[D][(st|nd|rd|th)] [YYYY]
385
386 • Wee[kday]
387
388 Month names can be abbreviated at three or more characters.
389
390 Weekday names can be abbreviated at three or more characters.
391
392 Examples: 2012-07-31, 31-07-2012, 7/31/2012, August 3
393
394 Time zones
395 • (+|-)HH:MM
396
397 • (+|-)HH[MM]
398
399 Some time zone codes, e.g. UTC, EET.
400
402 notmuch(1), notmuch-config(1), notmuch-count(1), notmuch-dump(1), not‐
403 much-hooks(5), notmuch-insert(1), notmuch-new(1), notmuch-proper‐
404 ties(7), notmuch-reindex(1), notmuch-reply(1), notmuch-restore(1), not‐
405 much-search(1), notmuch-show(1), notmuch-tag(1)
406
408 Carl Worth and many others
409
411 2009-2022, Carl Worth and many others
412
413
414
415
4160.36 Jun 25, 2022 NOTMUCH-SEARCH-TERMS(7)