6.1 Sequences

This section describes functions that accept any kind of sequence.

Function: sequencep object ¶

This function returns t if object is a list, vector, string, bool-vector, or char-table, nil otherwise. See also seqp below.

Function: length sequence ¶

This function returns the number of elements in sequence. The function signals the wrong-type-argument error if the argument is not a sequence or is a dotted list; it signals the circular-list error if the argument is a circular list. For a char-table, the value returned is always one more than the maximum Emacs character code.

See Definition of safe-length, for the related function safe-length.

(length '(1 2 3))
    ⇒ 3

(length ())
    ⇒ 0

(length "foobar")
    ⇒ 6

(length [1 2 3])
    ⇒ 3

(length (make-bool-vector 5 nil))
    ⇒ 5

See also string-bytes, in Text Representations.

If you need to compute the width of a string on display, you should use string-width (see Size of Displayed Text), not length, since length only counts the number of characters, but does not account for the display width of each character.

Function: length< sequence length ¶

Return non-nil if sequence is shorter than length. This may be more efficient than computing the length of sequence if sequence is a long list.

Function: length> sequence length ¶

Return non-nil if sequence is longer than length.

Function: length= sequence length ¶

Return non-nil if the length of sequence is equal to length.

Function: elt sequence index ¶

This function returns the element of sequence indexed by index. Legitimate values of index are integers ranging from 0 up to one less than the length of sequence. If sequence is a list, out-of-range values behave as for nth. See Definition of nth. Otherwise, out-of-range values trigger an args-out-of-range error.

(elt [1 2 3 4] 2)
     ⇒ 3

(elt '(1 2 3 4) 2)
     ⇒ 3

;; We use string to show clearly which character elt returns.
(string (elt "1234" 2))
     ⇒ "3"

(elt [1 2 3 4] 4)
     error→ Args out of range: [1 2 3 4], 4

(elt [1 2 3 4] -1)
     error→ Args out of range: [1 2 3 4], -1

This function generalizes aref (see Functions that Operate on Arrays) and nth (see Definition of nth).

Function: copy-sequence seqr ¶

This function returns a copy of seqr, which should be either a sequence or a record. The copy is the same type of object as the original, and it has the same elements in the same order. However, if seqr is empty, like a string or a vector of zero length, the value returned by this function might not be a copy, but an empty object of the same type and identical to seqr.

Storing a new element into the copy does not affect the original seqr, and vice versa. However, the elements of the copy are not copies; they are identical (eq) to the elements of the original. Therefore, changes made within these elements, as found via the copy, are also visible in the original.

If the argument is a string with text properties, the property list in the copy is itself a copy, not shared with the original’s property list. However, the actual values of the properties are shared. See Text Properties.

This function does not work for dotted lists. Trying to copy a circular list may cause an infinite loop.

See also append in Building Cons Cells and Lists, concat in Creating Strings, and vconcat in Functions for Vectors, for other ways to copy sequences.

(setq bar (list 1 2))
     ⇒ (1 2)

(setq x (vector 'foo bar))
     ⇒ [foo (1 2)]

(setq y (copy-sequence x))
     ⇒ [foo (1 2)]

(eq x y)
     ⇒ nil

(equal x y)
     ⇒ t

(eq (elt x 1) (elt y 1))
     ⇒ t

;; Replacing an element of one sequence.
(aset x 0 'quux)
x ⇒ [quux (1 2)]
y ⇒ [foo (1 2)]

;; Modifying the inside of a shared element.
(setcar (aref x 1) 69)
x ⇒ [quux (69 2)]
y ⇒ [foo (69 2)]

Function: reverse sequence ¶

This function creates a new sequence whose elements are the elements of sequence, but in reverse order. The original argument sequence is not altered. Note that char-tables cannot be reversed.

(setq x '(1 2 3 4))
     ⇒ (1 2 3 4)

(reverse x)
     ⇒ (4 3 2 1)
x
     ⇒ (1 2 3 4)

(setq x [1 2 3 4])
     ⇒ [1 2 3 4]

(reverse x)
     ⇒ [4 3 2 1]
x
     ⇒ [1 2 3 4]

(setq x "xyzzy")
     ⇒ "xyzzy"

(reverse x)
     ⇒ "yzzyx"
x
     ⇒ "xyzzy"

Function: nreverse sequence ¶

This function reverses the order of the elements of sequence. Unlike reverse the original sequence may be modified.

For example:

(setq x (list 'a 'b 'c))
     ⇒ (a b c)

x
     ⇒ (a b c)
(nreverse x)
     ⇒ (c b a)

;; The cons cell that was first is now last.
x
     ⇒ (a)

To avoid confusion, we usually store the result of nreverse back in the same variable which held the original list:

(setq x (nreverse x))

Here is the nreverse of our favorite example, (a b c), presented graphically:

Original list head:                       Reversed list:
 -------------        -------------        ------------
| car  | cdr  |      | car  | cdr  |      | car | cdr  |
|   a  |  nil |<--   |   b  |   o  |<--   |   c |   o  |
|      |      |   |  |      |   |  |   |  |     |   |  |
 -------------    |   --------- | -    |   -------- | -
                  |             |      |            |
                   -------------        ------------

For the vector, it is even simpler because you don’t need setq:

(setq x (copy-sequence [1 2 3 4]))
     ⇒ [1 2 3 4]
(nreverse x)
     ⇒ [4 3 2 1]
x
     ⇒ [4 3 2 1]

Note that unlike reverse, this function doesn’t work with strings. Although you can alter string data by using aset, it is strongly encouraged to treat strings as immutable even when they are mutable. See Mutability.

Function: sort sequence &rest keyword-args ¶

This function sorts sequence, which must be a list or vector, and returns a sorted sequence of the same type. The sort is stable, which means that elements with equal sort keys maintain their relative order. It takes the following optional keyword arguments:

:key keyfunc

Use keyfunc, a function that takes a single element from sequence and returns its key value, to generate the keys used in comparison. If this argument is absent or if keyfunc is nil then identity is assumed; that is, the elements themselves are used as sorting keys.

:lessp predicate

Use predicate to order the keys. predicate is a function that takes two sort keys as arguments and returns non-nil if the first should come before the second. If this argument is absent or predicate is nil, then value< is used, which is applicable to many different Lisp types and generally sorts in ascending order (see definition of value<, below).

For consistency, any predicate must obey the following rules:

• It must be antisymmetric: it cannot both order a before b and b before a.
• It must be transitive: if it orders a before b and b before c, then it must also order a before c.

:reverse flag

If flag is non-nil, the sorting order is reversed. With the default :lessp predicate this means sorting in descending order.

:in-place flag

If flag is non-nil, then sequence is sorted in-place (destructively) and returned. If nil, or if this argument is not given, a sorted copy of the input is returned and sequence itself remains unmodified. In-place sorting is slightly faster, but the original sequence is lost.

If the default behavior is not suitable for your needs, it is usually easier and faster to supply a new :key function than a different :lessp predicate. For example, consider sorting these strings:

(setq numbers '("one" "two" "three" "four" "five" "six"))
(sort numbers)
     ⇒ ("five" "four" "one" "six" "three" "two")

You can sort the strings by length instead by supplying a different key function:

(sort numbers :key #'length)
     ⇒ ("one" "two" "six" "four" "five" "three")

Note how strings of the same length keep their original order, thanks to the sorting stability. Now suppose you want to sort by length, but use the string contents to break ties. The easiest way is to specify a key function that transforms an element to a value that is sorted this way. Since value< orders compound objects (conses, lists, vectors and records) lexicographically, you could do:

(sort numbers :key (lambda (x) (cons (length x) x)))
     ⇒ ("one" "six" "two" "five" "four" "three")

because (3 . "six") is ordered before (3 . "two") and so on.

For compatibility with previous versions of Emacs, the sort function can also be called using the fixed two-argument form:

(sort sequence predicate)

where predicate is the :lessp argument. When using this form, sorting is always done in-place.

See Sorting Text, for more functions that perform sorting. See documentation in Access to Documentation Strings, for a useful example of sort.

Function: value< a b ¶

This function returns non-nil if a comes before b in the standard sorting order; this means that it returns nil when b comes before a, or if they are equal or unordered.

The arguments a and b must have the same type. Specifically:

• Numbers are compared using < (see definition of <).
• Strings are compared using string-lessp (see definition of string-lessp) and symbols are compared by comparing their names as strings.
• Conses, lists, vectors and records are compared lexicographically. This means that the two sequences are compared element-wise from left to right until they differ, and the result is then that of value< on the first pair of differing elements. If one sequence runs out of elements before the other, the shorter sequence comes before the longer.
• Markers are compared first by buffer, then by position.
• Buffers and processes are compared by comparing their names as strings. Dead buffers (whose name is nil) will compare before any live buffer.
• Other types are considered unordered and the return value will be nil.

Examples:

(value< -4 3.5) ⇒ t
(value< "dog" "cat") ⇒ nil
(value< 'yip 'yip) ⇒ nil
(value< '(3 2) '(3 2 0)) ⇒ t
(value< [3 2 "a"] [3 2 "b"]) ⇒ t

Note that nil is treated as either a symbol or an empty list, depending on what it is compared against:

(value< nil '(0)) ⇒ t
(value< 'nib nil) ⇒ t

There is no limit to the length of sequences (lists, vectors and so on) that can be compared, but value< may fail with an error if used to compare circular or deeply nested data structures.

The seq.el library provides the following additional sequence manipulation macros and functions, prefixed with seq-.

All functions defined in this library are free of side-effects; i.e., they do not modify any sequence (list, vector, or string) that you pass as an argument. Unless otherwise stated, the result is a sequence of the same type as the input. For those functions that take a predicate, this should be a function of one argument.

The seq.el library can be extended to work with additional types of sequential data-structures. For that purpose, all functions are defined using cl-defgeneric. See Generic Functions, for more details about using cl-defgeneric for adding extensions.

Function: seq-elt sequence index ¶

This function returns the element of sequence at the specified index, which is an integer whose valid value range is zero to one less than the length of sequence. For out-of-range values on built-in sequence types, seq-elt behaves like elt. For the details, see Definition of elt.

(seq-elt [1 2 3 4] 2)
⇒ 3

seq-elt returns places settable using setf (see The setf Macro).

(setq vec [1 2 3 4])
(setf (seq-elt vec 2) 5)
vec
⇒ [1 2 5 4]

Function: seq-length sequence ¶

This function returns the number of elements in sequence. For built-in sequence types, seq-length behaves like length. See Definition of length.

Function: seqp object ¶

This function returns non-nil if object is a sequence (a list or array), or any additional type of sequence defined via seq.el generic functions. This is an extensible variant of sequencep.

(seqp [1 2])
⇒ t

(seqp 2)
⇒ nil

Function: seq-drop sequence n ¶

This function returns all but the first n (an integer) elements of sequence. If n is negative or zero, the result is sequence.

(seq-drop [1 2 3 4 5 6] 3)
⇒ [4 5 6]

(seq-drop "hello world" -4)
⇒ "hello world"

Function: seq-take sequence n ¶

This function returns the first n (an integer) elements of sequence. If n is negative or zero, the result is nil.

(seq-take '(1 2 3 4) 3)
⇒ (1 2 3)

(seq-take [1 2 3 4] 0)
⇒ []

Function: seq-take-while predicate sequence ¶

This function returns the members of sequence in order, stopping before the first one for which predicate returns nil.

(seq-take-while (lambda (elt) (> elt 0)) '(1 2 3 -1 -2))
⇒ (1 2 3)

(seq-take-while (lambda (elt) (> elt 0)) [-1 4 6])
⇒ []

Function: seq-drop-while predicate sequence ¶

This function returns the members of sequence in order, starting from the first one for which predicate returns nil.

(seq-drop-while (lambda (elt) (> elt 0)) '(1 2 3 -1 -2))
⇒ (-1 -2)

(seq-drop-while (lambda (elt) (< elt 0)) [1 4 6])
⇒ [1 4 6]

Function: seq-split sequence length ¶

This function returns a list consisting of sub-sequences of sequence of (at most) length length. (The final element may be shorter than length if the length of sequence isn’t a multiple of length.

(seq-split [0 1 2 3 4] 2)
⇒ ([0 1] [2 3] [4])

Function: seq-do function sequence ¶

This function applies function to each element of sequence in turn (presumably for side effects), and returns sequence.

Function: seq-map function sequence ¶

This function returns the result of applying function to each element of sequence. The returned value is a list.

(seq-map #'1+ '(2 4 6))
⇒ (3 5 7)

(seq-map #'symbol-name [foo bar])
⇒ ("foo" "bar")

Function: seq-map-indexed function sequence ¶

This function returns the result of applying function to each element of sequence and its index within seq. The returned value is a list.

(seq-map-indexed (lambda (elt idx)
                   (list idx elt))
                 '(a b c))
⇒ ((0 a) (1 b) (2 c))

Function: seq-mapn function &rest sequences ¶

This function returns the result of applying function to each element of sequences. The arity (see subr-arity) of function must match the number of sequences. Mapping stops at the end of the shortest sequence, and the returned value is a list.

(seq-mapn #'+ '(2 4 6) '(20 40 60))
⇒ (22 44 66)

(seq-mapn #'concat '("moskito" "bite") ["bee" "sting"])
⇒ ("moskitobee" "bitesting")

Function: seq-filter predicate sequence ¶

This function returns a list of all the elements in sequence for which predicate returns non-nil.

(seq-filter (lambda (elt) (> elt 0)) [1 -1 3 -3 5])
⇒ (1 3 5)

(seq-filter (lambda (elt) (> elt 0)) '(-1 -3 -5))
⇒ nil

Function: seq-remove predicate sequence ¶

This function returns a list of all the elements in sequence for which predicate returns nil.

(seq-remove (lambda (elt) (> elt 0)) [1 -1 3 -3 5])
⇒ (-1 -3)

(seq-remove (lambda (elt) (< elt 0)) '(-1 -3 -5))
⇒ nil

Function: seq-remove-at-position sequence n ¶

This function returns a copy of sequence where the element at (zero-based) index n got removed. The result is a sequence of the same type as sequence.

(seq-remove-at-position [1 -1 3 -3 5] 0)
⇒ [-1 3 -3 5]

(seq-remove-at-position [1 -1 3 -3 5] 3)
⇒ [1 -1 3 5]

Function: seq-keep function sequence ¶

This function returns a list of all non-nil results from calling function on the elements in sequence.

(seq-keep #'cl-digit-char-p '(?6 ?a ?7))
⇒ (6 7)

Function: seq-reduce function sequence initial-value ¶

This function returns the result of calling function with initial-value and the first element of sequence, then calling function with that result and the second element of sequence, then with that result and the third element of sequence, etc. function should be a function of two arguments.

function is called with two arguments. initial-value (and then the accumulated value) is used as the first argument, and the elements in sequence are used for the second argument.

If sequence is empty, this returns initial-value without calling function.

(seq-reduce #'+ [1 2 3 4] 0)
⇒ 10

(seq-reduce #'+ '(1 2 3 4) 5)
⇒ 15

(seq-reduce #'+ '() 3)
⇒ 3

Function: seq-some predicate sequence ¶

This function returns the first non-nil value returned by applying predicate to each element of sequence in turn.

(seq-some #'numberp ["abc" 1 nil])
⇒ t

(seq-some #'numberp ["abc" "def"])
⇒ nil

(seq-some #'null ["abc" 1 nil])
⇒ t

(seq-some #'1+ [2 4 6])
⇒ 3

Function: seq-find predicate sequence &optional default ¶

This function returns the first element in sequence for which predicate returns non-nil. If no element matches predicate, the function returns default.

Note that this function has an ambiguity if the found element is identical to default, as in that case it cannot be known whether an element was found or not.

(seq-find #'numberp ["abc" 1 nil])
⇒ 1

(seq-find #'numberp ["abc" "def"])
⇒ nil

Function: seq-every-p predicate sequence ¶

This function returns non-nil if applying predicate to every element of sequence returns non-nil.

(seq-every-p #'numberp [2 4 6])
⇒ t

(seq-every-p #'numberp [2 4 "6"])
⇒ nil

Function: seq-empty-p sequence ¶

This function returns non-nil if sequence is empty.

(seq-empty-p "not empty")
⇒ nil

(seq-empty-p "")
⇒ t

Function: seq-count predicate sequence ¶

This function returns the number of elements in sequence for which predicate returns non-nil.

(seq-count (lambda (elt) (> elt 0)) [-1 2 0 3 -2])
⇒ 2

Function: seq-sort function sequence ¶

This function returns a copy of sequence that is sorted according to function, a function of two arguments that returns non-nil if the first argument should sort before the second.

Function: seq-sort-by function predicate sequence ¶

This function is similar to seq-sort, but the elements of sequence are transformed by applying function on them before being sorted. function is a function of one argument.

(seq-sort-by #'seq-length #'> ["a" "ab" "abc"])
⇒ ["abc" "ab" "a"]

Function: seq-contains-p sequence elt &optional function ¶

This function returns non-nil if at least one element in sequence is equal to elt. If the optional argument function is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-contains-p '(symbol1 symbol2) 'symbol1)
⇒ t

(seq-contains-p '(symbol1 symbol2) 'symbol3)
⇒ nil

Function: seq-set-equal-p sequence1 sequence2 &optional testfn ¶

This function checks whether sequence1 and sequence2 contain the same elements, regardless of the order. If the optional argument testfn is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-set-equal-p '(a b c) '(c b a))
⇒ t

(seq-set-equal-p '(a b c) '(c b))
⇒ nil

(seq-set-equal-p '("a" "b" "c") '("c" "b" "a"))
⇒ t

(seq-set-equal-p '("a" "b" "c") '("c" "b" "a") #'eq)
⇒ nil

Function: seq-position sequence elt &optional function ¶

This function returns the (zero-based) index of the first element in sequence that is equal to elt. If the optional argument function is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-position '(a b c) 'b)
⇒ 1

(seq-position '(a b c) 'd)
⇒ nil

Function: seq-positions sequence elt &optional testfn ¶

This function returns a list of the (zero-based) indices of the elements in sequence for which testfn returns non-nil when passed the element and elt as arguments. testfn defaults to equal.

(seq-positions '(a b c a d) 'a)
⇒ (0 3)

(seq-positions '(a b c a d) 'z)
⇒ nil

(seq-positions '(11 5 7 12 9 15) 10 #'>=)
⇒ (0 3 5)

Function: seq-uniq sequence &optional function ¶

This function returns a list of the elements of sequence with duplicates removed. If the optional argument function is non-nil, it is a function of two arguments to use instead of the default equal.

(seq-uniq '(1 2 2 1 3))
⇒ (1 2 3)

(seq-uniq '(1 2 2.0 1.0) #'=)
⇒ (1 2)

Function: seq-subseq sequence start &optional end ¶

This function returns a subset of sequence from start to end, both integers (end defaults to the last element). If start or end is negative, it counts from the end of sequence.

(seq-subseq '(1 2 3 4 5) 1)
⇒ (2 3 4 5)

(seq-subseq '[1 2 3 4 5] 1 3)
⇒ [2 3]

(seq-subseq '[1 2 3 4 5] -3 -1)
⇒ [3 4]

Function: seq-concatenate type &rest sequences ¶

This function returns a sequence of type type made of the concatenation of sequences. type may be: vector, list or string.

(seq-concatenate 'list '(1 2) '(3 4) [5 6])
⇒ (1 2 3 4 5 6)

(seq-concatenate 'string "Hello " "world")
⇒ "Hello world"

Function: seq-mapcat function sequence &optional type ¶

This function returns the result of applying seq-concatenate to the result of applying function to each element of sequence. The result is a sequence of type type, or a list if type is nil.

(seq-mapcat #'seq-reverse '((3 2 1) (6 5 4)))
⇒ (1 2 3 4 5 6)

Function: seq-partition sequence n ¶

This function returns a list of the elements of sequence grouped into sub-sequences of length n. The last sequence may contain less elements than n. n must be an integer. If n is a negative integer or 0, the return value is nil.

(seq-partition '(0 1 2 3 4 5 6 7) 3)
⇒ ((0 1 2) (3 4 5) (6 7))

Function: seq-union sequence1 sequence2 &optional function ¶

This function returns a list of the elements that appear either in sequence1 or sequence2. The elements of the returned list are all unique, in the sense that no two elements there will compare equal. If the optional argument function is non-nil, it should be a function of two arguments to use to compare elements, instead of the default equal.

(seq-union [1 2 3] [3 5])
⇒ (1 2 3 5)

Function: seq-intersection sequence1 sequence2 &optional function ¶

This function returns a list of the elements that appear both in sequence1 and sequence2. If the optional argument function is non-nil, it is a function of two arguments to use to compare elements instead of the default equal.

(seq-intersection [2 3 4 5] [1 3 5 6 7])
⇒ (3 5)

Function: seq-difference sequence1 sequence2 &optional function ¶

This function returns a list of the elements that appear in sequence1 but not in sequence2. If the optional argument function is non-nil, it is a function of two arguments to use to compare elements instead of the default equal.

(seq-difference '(2 3 4 5) [1 3 5 6 7])
⇒ (2 4)

Function: seq-group-by function sequence ¶

This function separates the elements of sequence into an alist whose keys are the result of applying function to each element of sequence. Keys are compared using equal.

(seq-group-by #'integerp '(1 2.1 3 2 3.2))
⇒ ((t 1 3 2) (nil 2.1 3.2))

(seq-group-by #'car '((a 1) (b 2) (a 3) (c 4)))
⇒ ((b (b 2)) (a (a 1) (a 3)) (c (c 4)))

Function: seq-into sequence type ¶

This function converts the sequence sequence into a sequence of type type. type can be one of the following symbols: vector, string or list.

(seq-into [1 2 3] 'list)
⇒ (1 2 3)

(seq-into nil 'vector)
⇒ []

(seq-into "hello" 'vector)
⇒ [104 101 108 108 111]

Function: seq-min sequence ¶

This function returns the smallest element of sequence. The elements of sequence must be numbers or markers (see Markers).

(seq-min [3 1 2])
⇒ 1

(seq-min "Hello")
⇒ 72

Function: seq-max sequence ¶

This function returns the largest element of sequence. The elements of sequence must be numbers or markers.

(seq-max [1 3 2])
⇒ 3

(seq-max "Hello")
⇒ 111

Macro: seq-doseq (var sequence) body… ¶

This macro is like dolist (see dolist), except that sequence can be a list, vector or string. This is primarily useful for side-effects.

Macro: seq-let var-sequence val-sequence body… ¶

This macro binds the variables defined in var-sequence to the values that are the corresponding elements of val-sequence. This is known as destructuring binding. The elements of var-sequence can themselves include sequences, allowing for nested destructuring.

The var-sequence sequence can also include the &rest marker followed by a variable name to be bound to the rest of val-sequence.

(seq-let [first second] [1 2 3 4]
  (list first second))
⇒ (1 2)

(seq-let (_ a _ b) '(1 2 3 4)
  (list a b))
⇒ (2 4)

(seq-let [a [b [c]]] [1 [2 [3]]]
  (list a b c))
⇒ (1 2 3)

(seq-let [a b &rest others] [1 2 3 4]
  others)

⇒ [3 4]

The pcase patterns provide an alternative facility for destructuring binding, see Destructuring with pcase Patterns.

Macro: seq-setq var-sequence val-sequence ¶

This macro works similarly to seq-let, except that values are assigned to variables as if by setq instead of as in a let binding.

(let ((a nil)
      (b nil))
  (seq-setq (_ a _ b) '(1 2 3 4))
  (list a b))
⇒ (2 4)

Function: seq-random-elt sequence ¶

This function returns an element of sequence taken at random.

(seq-random-elt [1 2 3 4])
⇒ 3
(seq-random-elt [1 2 3 4])
⇒ 2
(seq-random-elt [1 2 3 4])
⇒ 4
(seq-random-elt [1 2 3 4])
⇒ 2
(seq-random-elt [1 2 3 4])
⇒ 1

If sequence is empty, this function signals an error.