dash.texi

\input texinfo    @c -*- texinfo -*-
@c %**start of header
@setfilename dash.info
@set DASHVER 2.19.1
@settitle Dash: A modern list library for GNU Emacs.
@documentencoding UTF-8
@documentlanguage en
@c %**end of header

@copying
This manual is for Dash version @value{DASHVER}.

Copyright @copyright{} 2012--2024 Free Software Foundation, Inc.

@quotation
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.3 or
any later version published by the Free Software Foundation; with the
Invariant Sections being ``GNU General Public License,'' and no
Front-Cover Texts or Back-Cover Texts.  A copy of the license is
included in the section entitled ``GNU Free Documentation License''.
@end quotation
@end copying

@dircategory Emacs
@direntry
* Dash: (dash.info).    A modern list library for GNU Emacs.
@end direntry

@titlepage
@title Dash Manual
@subtitle For Dash Version @value{DASHVER}.
@author Magnar Sveen
@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage

@contents

@ifnottex
@node Top
@top Dash

@insertcopying
@end ifnottex

@menu
* Installation::        Installing and configuring Dash.
* Functions::           Dash API reference.
* Development::         Contributing to Dash development.

Appendices

* FDL::                 The license for this documentation.
* GPL::                 Conditions for copying and changing Dash.
* Index::               Index including functions and macros.

@detailmenu
 --- The Detailed Node Listing ---

Installation

* Using in a package::  Listing Dash as a package dependency.
* Fontification of special variables::  Font Lock of anaphoric macro variables.
* Info symbol lookup::  Looking up Dash symbols in this manual.

Functions

* Maps::
* Sublist selection::
* List to list::
* Reductions::
* Unfolding::
* Predicates::
* Partitioning::
* Indexing::
* Set operations::
* Other list operations::
* Tree operations::
* Threading macros::
* Binding::
* Side effects::
* Destructive operations::
* Function combinators::

Development

* Contribute::          How to contribute.
* Contributors::        List of contributors.
@end detailmenu
@end menu

@node Installation
@chapter Installation

Dash is available on @url{https://elpa.gnu.org/, GNU ELPA},
@url{https://elpa.gnu.org/devel/, GNU-devel ELPA}, and
@url{https://melpa.org/, MELPA}, and can be installed with the
standard command @code{package-install} (@pxref{Package
Installation,,, emacs, The GNU Emacs Manual}).

@table @kbd
@item M-x package-install @key{RET} dash @key{RET}
Install the Dash library.
@end table

Alternatively, you can just dump @file{dash.el} in your
@code{load-path} somewhere (@pxref{Lisp Libraries,,, emacs, The GNU
Emacs Manual}).

@menu
* Using in a package::  Listing Dash as a package dependency.
* Fontification of special variables::  Font Lock of anaphoric macro variables.
* Info symbol lookup::  Looking up Dash symbols in this manual.
@end menu

@node Using in a package
@section Using in a package

If you use Dash in your own package, be sure to list it as a
dependency in the library's headers as follows (@pxref{Library
Headers,,, elisp, The Emacs Lisp Reference Manual}).

@lisp
;; Package-Requires: ((dash "@value{DASHVER}"))
@end lisp

@node Fontification of special variables
@section Fontification of special variables

@findex dash-fontify-mode
The autoloaded minor mode @code{dash-fontify-mode} is provided for
optional fontification of anaphoric Dash variables (@code{it},
@code{acc}, etc.@:) in Emacs Lisp buffers using search-based Font Lock
(@pxref{Font Lock,,, emacs, The GNU Emacs Manual}).  In older Emacs
versions which do not dynamically detect macros, the minor mode also
fontifies calls to Dash macros.

@findex global-dash-fontify-mode
To automatically enable the minor mode in all Emacs Lisp buffers, just
call its autoloaded global counterpart
@code{global-dash-fontify-mode}, either interactively or from your
@code{user-init-file}:

@lisp
(global-dash-fontify-mode)
@end lisp

@node Info symbol lookup
@section Info symbol lookup

@findex dash-register-info-lookup
While editing Elisp files, you can use @kbd{C-h S}
(@code{info-lookup-symbol}) to look up Elisp symbols in the relevant
Info manuals (@pxref{Info Lookup,,, emacs, The GNU Emacs Manual}).  To
enable the same for Dash symbols, use the command
@code{dash-register-info-lookup}.  It can be called directly when
needed, or automatically from your @code{user-init-file}.  For
example:

@lisp
(with-eval-after-load 'info-look
  (dash-register-info-lookup))
@end lisp

@node Functions
@chapter Functions

This chapter contains reference documentation for the Dash
@acronym{API, Application Programming Interface}.  The names of all
public functions defined in the library are prefixed with a dash
character (@samp{-}).

The library also provides anaphoric macro versions of functions where
that makes sense.  The names of these macros are prefixed with two
dashes (@samp{--}) instead of one.

For instance, while the function @code{-map} applies a function to
each element of a list, its anaphoric counterpart @code{--map}
evaluates a form with the local variable @code{it} temporarily bound
to the current list element instead.

@lisp
@group
;; Normal version.
(-map (lambda (n) (* n n)) '(1 2 3 4))
    @result{} (1 4 9 16)
@end group

@group
;; Anaphoric version.
(--map (* it it) '(1 2 3 4))
    @result{} (1 4 9 16)
@end group
@end lisp

The normal version can, of course, also be written as in the following
example, which demonstrates the utility of both versions.

@lisp
@group
(defun my-square (n)
  "Return N multiplied by itself."
  (* n n))

(-map #'my-square '(1 2 3 4))
    @result{} (1 4 9 16)
@end group
@end lisp

@menu
* Maps::
* Sublist selection::
* List to list::
* Reductions::
* Unfolding::
* Predicates::
* Partitioning::
* Indexing::
* Set operations::
* Other list operations::
* Tree operations::
* Threading macros::
* Binding::
* Side effects::
* Destructive operations::
* Function combinators::
@end menu

@node Maps
@section Maps

Functions in this category take a transforming function, which
is then applied sequentially to each or selected elements of the
input list.  The results are collected in order and returned as a
new list.

@anchor{-map}
@defun -map (fn list)
Apply @var{fn} to each item in @var{list} and return the list of results.

This function's anaphoric counterpart is @code{--map}.

@example
@group
(-map (lambda (num) (* num num)) '(1 2 3 4))
    @result{} (1 4 9 16)
@end group
@group
(-map #'1+ '(1 2 3 4))
    @result{} (2 3 4 5)
@end group
@group
(--map (* it it) '(1 2 3 4))
    @result{} (1 4 9 16)
@end group
@end example
@end defun

@anchor{-map-when}
@defun -map-when (pred rep list)
Use @var{pred} to conditionally apply @var{rep} to each item in @var{list}.
Return a copy of @var{list} where the items for which @var{pred} returns @code{nil}
are unchanged, and the rest are mapped through the @var{rep} function.

Alias: @code{-replace-where}

See also: @code{-update-at} (@pxref{-update-at})

@example
@group
(-map-when 'even? 'square '(1 2 3 4))
    @result{} (1 4 3 16)
@end group
@group
(--map-when (> it 2) (* it it) '(1 2 3 4))
    @result{} (1 2 9 16)
@end group
@group
(--map-when (= it 2) 17 '(1 2 3 4))
    @result{} (1 17 3 4)
@end group
@end example
@end defun

@anchor{-map-first}
@defun -map-first (pred rep list)
Use @var{pred} to determine the first item in @var{list} to call @var{rep} on.
Return a copy of @var{list} where the first item for which @var{pred} returns
non-@code{nil} is replaced with the result of calling @var{rep} on that item.

See also: @code{-map-when} (@pxref{-map-when}), @code{-replace-first} (@pxref{-replace-first})

@example
@group
(-map-first 'even? 'square '(1 2 3 4))
    @result{} (1 4 3 4)
@end group
@group
(--map-first (> it 2) (* it it) '(1 2 3 4))
    @result{} (1 2 9 4)
@end group
@group
(--map-first (= it 2) 17 '(1 2 3 2))
    @result{} (1 17 3 2)
@end group
@end example
@end defun

@anchor{-map-last}
@defun -map-last (pred rep list)
Use @var{pred} to determine the last item in @var{list} to call @var{rep} on.
Return a copy of @var{list} where the last item for which @var{pred} returns
non-@code{nil} is replaced with the result of calling @var{rep} on that item.

See also: @code{-map-when} (@pxref{-map-when}), @code{-replace-last} (@pxref{-replace-last})

@example
@group
(-map-last 'even? 'square '(1 2 3 4))
    @result{} (1 2 3 16)
@end group
@group
(--map-last (> it 2) (* it it) '(1 2 3 4))
    @result{} (1 2 3 16)
@end group
@group
(--map-last (= it 2) 17 '(1 2 3 2))
    @result{} (1 2 3 17)
@end group
@end example
@end defun

@anchor{-map-indexed}
@defun -map-indexed (fn list)
Apply @var{fn} to each index and item in @var{list} and return the list of results.
This is like @code{-map} (@pxref{-map}), but @var{fn} takes two arguments: the index of the
current element within @var{list}, and the element itself.

This function's anaphoric counterpart is @code{--map-indexed}.

For a side-effecting variant, see also @code{-each-indexed} (@pxref{-each-indexed}).

@example
@group
(-map-indexed (lambda (index item) (- item index)) '(1 2 3 4))
    @result{} (1 1 1 1)
@end group
@group
(--map-indexed (- it it-index) '(1 2 3 4))
    @result{} (1 1 1 1)
@end group
@group
(-map-indexed #'* '(1 2 3 4))
    @result{} (0 2 6 12)
@end group
@end example
@end defun

@anchor{-annotate}
@defun -annotate (fn list)
Pair each item in @var{list} with the result of passing it to @var{fn}.

Return an alist of (@var{result} . @var{item}), where each @var{item} is the
corresponding element of @var{list}, and @var{result} is the value obtained
by calling @var{fn} on @var{item}.

This function's anaphoric counterpart is @code{--annotate}.

@example
@group
(-annotate #'1+ '(1 2 3))
    @result{} ((2 . 1) (3 . 2) (4 . 3))
@end group
@group
(-annotate #'length '((f o o) (bar baz)))
    @result{} ((3 f o o) (2 bar baz))
@end group
@group
(--annotate (> it 1) '(0 1 2 3))
    @result{} ((nil . 0) (nil . 1) (t . 2) (t . 3))
@end group
@end example
@end defun

@anchor{-splice}
@defun -splice (pred fun list)
Splice lists generated by @var{fun} in place of items satisfying @var{pred} in @var{list}.

Call @var{pred} on each element of @var{list}.  Whenever the result of @var{pred}
is @code{nil}, leave that @code{it} as-is.  Otherwise, call @var{fun} on the same
@code{it} that satisfied @var{pred}.  The result should be a (possibly
empty) list of items to splice in place of @code{it} in @var{list}.

This can be useful as an alternative to the @code{,@@} construct in a
@code{`} structure, in case you need to splice several lists at
marked positions (for example with keywords).

This function's anaphoric counterpart is @code{--splice}.

See also: @code{-splice-list} (@pxref{-splice-list}), @code{-insert-at} (@pxref{-insert-at}).

@example
@group
(-splice #'numberp (lambda (n) (list n n)) '(a 1 b 2))
    @result{} (a 1 1 b 2 2)
@end group
@group
(--splice t (list it it) '(1 2 3 4))
    @result{} (1 1 2 2 3 3 4 4)
@end group
@group
(--splice (eq it :magic) '((magical) (code)) '((foo) :magic (bar)))
    @result{} ((foo) (magical) (code) (bar))
@end group
@end example
@end defun

@anchor{-splice-list}
@defun -splice-list (pred new-list list)
Splice @var{new-list} in place of elements matching @var{pred} in @var{list}.

See also: @code{-splice} (@pxref{-splice}), @code{-insert-at} (@pxref{-insert-at})

@example
@group
(-splice-list 'keywordp '(a b c) '(1 :foo 2))
    @result{} (1 a b c 2)
@end group
@group
(-splice-list 'keywordp nil '(1 :foo 2))
    @result{} (1 2)
@end group
@group
(--splice-list (keywordp it) '(a b c) '(1 :foo 2))
    @result{} (1 a b c 2)
@end group
@end example
@end defun

@anchor{-mapcat}
@defun -mapcat (fn list)
Return the concatenation of the result of mapping @var{fn} over @var{list}.
Thus function @var{fn} should return a list.

@example
@group
(-mapcat 'list '(1 2 3))
    @result{} (1 2 3)
@end group
@group
(-mapcat (lambda (item) (list 0 item)) '(1 2 3))
    @result{} (0 1 0 2 0 3)
@end group
@group
(--mapcat (list 0 it) '(1 2 3))
    @result{} (0 1 0 2 0 3)
@end group
@end example
@end defun

@anchor{-copy}
@defun -copy (list)
Create a shallow copy of @var{list}.

@example
@group
(-copy '(1 2 3))
    @result{} (1 2 3)
@end group
@group
(let ((a '(1 2 3))) (eq a (-copy a)))
    @result{} nil
@end group
@end example
@end defun

@node Sublist selection
@section Sublist selection

Functions returning a sublist of the original list.

@anchor{-filter}
@defun -filter (pred list)
Return a new list of the items in @var{list} for which @var{pred} returns non-@code{nil}.

Alias: @code{-select}.

This function's anaphoric counterpart is @code{--filter}.

For similar operations, see also @code{-keep} (@pxref{-keep}) and @code{-remove} (@pxref{-remove}).

@example
@group
(-filter (lambda (num) (= 0 (% num 2))) '(1 2 3 4))
    @result{} (2 4)
@end group
@group
(-filter #'natnump '(-2 -1 0 1 2))
    @result{} (0 1 2)
@end group
@group
(--filter (= 0 (% it 2)) '(1 2 3 4))
    @result{} (2 4)
@end group
@end example
@end defun

@anchor{-remove}
@defun -remove (pred list)
Return a new list of the items in @var{list} for which @var{pred} returns @code{nil}.

Alias: @code{-reject}.

This function's anaphoric counterpart is @code{--remove}.

For similar operations, see also @code{-keep} (@pxref{-keep}) and @code{-filter} (@pxref{-filter}).

@example
@group
(-remove (lambda (num) (= 0 (% num 2))) '(1 2 3 4))
    @result{} (1 3)
@end group
@group
(-remove #'natnump '(-2 -1 0 1 2))
    @result{} (-2 -1)
@end group
@group
(--remove (= 0 (% it 2)) '(1 2 3 4))
    @result{} (1 3)
@end group
@end example
@end defun

@anchor{-remove-first}
@defun -remove-first (pred list)
Remove the first item from @var{list} for which @var{pred} returns non-@code{nil}.
This is a non-destructive operation, but only the front of @var{list}
leading up to the removed item is a copy; the rest is @var{list}'s
original tail.  If no item is removed, then the result is a
complete copy.

Alias: @code{-reject-first}.

This function's anaphoric counterpart is @code{--remove-first}.

See also @code{-map-first} (@pxref{-map-first}), @code{-remove-item} (@pxref{-remove-item}), and @code{-remove-last} (@pxref{-remove-last}).

@example
@group
(-remove-first #'natnump '(-2 -1 0 1 2))
    @result{} (-2 -1 1 2)
@end group
@group
(-remove-first #'stringp '(1 2 "first" "second"))
    @result{} (1 2 "second")
@end group
@group
(--remove-first (> it 3) '(1 2 3 4 5 6))
    @result{} (1 2 3 5 6)
@end group
@end example
@end defun

@anchor{-remove-last}
@defun -remove-last (pred list)
Remove the last item from @var{list} for which @var{pred} returns non-@code{nil}.
The result is a copy of @var{list} regardless of whether an element is
removed.

Alias: @code{-reject-last}.

This function's anaphoric counterpart is @code{--remove-last}.

See also @code{-map-last} (@pxref{-map-last}), @code{-remove-item} (@pxref{-remove-item}), and @code{-remove-first} (@pxref{-remove-first}).

@example
@group
(-remove-last #'natnump '(1 3 5 4 7 8 10 -11))
    @result{} (1 3 5 4 7 8 -11)
@end group
@group
(-remove-last #'stringp '(1 2 "last" "second"))
    @result{} (1 2 "last")
@end group
@group
(--remove-last (> it 3) '(1 2 3 4 5 6 7 8 9 10))
    @result{} (1 2 3 4 5 6 7 8 9)
@end group
@end example
@end defun

@anchor{-remove-item}
@defun -remove-item (item list)
Return a copy of @var{list} with all occurrences of @var{item} removed.
The comparison is done with @code{equal}.

@example
@group
(-remove-item 3 '(1 2 3 2 3 4 5 3))
    @result{} (1 2 2 4 5)
@end group
@group
(-remove-item 'foo '(foo bar baz foo))
    @result{} (bar baz)
@end group
@group
(-remove-item "bob" '("alice" "bob" "eve" "bob"))
    @result{} ("alice" "eve")
@end group
@end example
@end defun

@anchor{-non-nil}
@defun -non-nil (list)
Return a copy of @var{list} with all @code{nil} items removed.

@example
@group
(-non-nil '(nil 1 nil 2 nil nil 3 4 nil 5 nil))
    @result{} (1 2 3 4 5)
@end group
@group
(-non-nil '((nil)))
    @result{} ((nil))
@end group
@group
(-non-nil ())
    @result{} ()
@end group
@end example
@end defun

@anchor{-slice}
@defun -slice (list from &optional to step)
Return copy of @var{list}, starting from index @var{from} to index @var{to}.

@var{from} or @var{to} may be negative.  These values are then interpreted
modulo the length of the list.

If @var{step} is a number, only each STEPth item in the resulting
section is returned.  Defaults to 1.

@example
@group
(-slice '(1 2 3 4 5) 1)
    @result{} (2 3 4 5)
@end group
@group
(-slice '(1 2 3 4 5) 0 3)
    @result{} (1 2 3)
@end group
@group
(-slice '(1 2 3 4 5 6 7 8 9) 1 -1 2)
    @result{} (2 4 6 8)
@end group
@end example
@end defun

@anchor{-take}
@defun -take (n list)
Return a copy of the first @var{n} items in @var{list}.
Return a copy of @var{list} if it contains @var{n} items or fewer.
Return @code{nil} if @var{n} is zero or less.

See also: @code{-take-last} (@pxref{-take-last}).

@example
@group
(-take 3 '(1 2 3 4 5))
    @result{} (1 2 3)
@end group
@group
(-take 17 '(1 2 3 4 5))
    @result{} (1 2 3 4 5)
@end group
@group
(-take 0 '(1 2 3 4 5))
    @result{} ()
@end group
@end example
@end defun

@anchor{-take-last}
@defun -take-last (n list)
Return a copy of the last @var{n} items of @var{list} in order.
Return a copy of @var{list} if it contains @var{n} items or fewer.
Return @code{nil} if @var{n} is zero or less.

See also: @code{-take} (@pxref{-take}).

@example
@group
(-take-last 3 '(1 2 3 4 5))
    @result{} (3 4 5)
@end group
@group
(-take-last 17 '(1 2 3 4 5))
    @result{} (1 2 3 4 5)
@end group
@group
(-take-last 1 '(1 2 3 4 5))
    @result{} (5)
@end group
@end example
@end defun

@anchor{-drop}
@defun -drop (n list)
Return the tail (not a copy) of @var{list} without the first @var{n} items.
Return @code{nil} if @var{list} contains @var{n} items or fewer.
Return @var{list} if @var{n} is zero or less.

For another variant, see also @code{-drop-last} (@pxref{-drop-last}).

@example
@group
(-drop 3 '(1 2 3 4 5))
    @result{} (4 5)
@end group
@group
(-drop 17 '(1 2 3 4 5))
    @result{} ()
@end group
@group
(-drop 0 '(1 2 3 4 5))
    @result{} (1 2 3 4 5)
@end group
@end example
@end defun

@anchor{-drop-last}
@defun -drop-last (n list)
Return a copy of @var{list} without its last @var{n} items.
Return a copy of @var{list} if @var{n} is zero or less.
Return @code{nil} if @var{list} contains @var{n} items or fewer.

See also: @code{-drop} (@pxref{-drop}).

@example
@group
(-drop-last 3 '(1 2 3 4 5))
    @result{} (1 2)
@end group
@group
(-drop-last 17 '(1 2 3 4 5))
    @result{} ()
@end group
@group
(-drop-last 0 '(1 2 3 4 5))
    @result{} (1 2 3 4 5)
@end group
@end example
@end defun

@anchor{-take-while}
@defun -take-while (pred list)
Take successive items from @var{list} for which @var{pred} returns non-@code{nil}.
@var{pred} is a function of one argument.  Return a new list of the
successive elements from the start of @var{list} for which @var{pred} returns
non-@code{nil}.

This function's anaphoric counterpart is @code{--take-while}.

For another variant, see also @code{-drop-while} (@pxref{-drop-while}).

@example
@group
(-take-while #'even? '(1 2 3 4))
    @result{} ()
@end group
@group
(-take-while #'even? '(2 4 5 6))
    @result{} (2 4)
@end group
@group
(--take-while (< it 4) '(1 2 3 4 3 2 1))
    @result{} (1 2 3)
@end group
@end example
@end defun

@anchor{-drop-while}
@defun -drop-while (pred list)
Drop successive items from @var{list} for which @var{pred} returns non-@code{nil}.
@var{pred} is a function of one argument.  Return the tail (not a copy)
of @var{list} starting from its first element for which @var{pred} returns
@code{nil}.

This function's anaphoric counterpart is @code{--drop-while}.

For another variant, see also @code{-take-while} (@pxref{-take-while}).

@example
@group
(-drop-while #'even? '(1 2 3 4))
    @result{} (1 2 3 4)
@end group
@group
(-drop-while #'even? '(2 4 5 6))
    @result{} (5 6)
@end group
@group
(--drop-while (< it 4) '(1 2 3 4 3 2 1))
    @result{} (4 3 2 1)
@end group
@end example
@end defun

@anchor{-select-by-indices}
@defun -select-by-indices (indices list)
Return a list whose elements are elements from @var{list} selected
as `(nth i list)` for all i from @var{indices}.

@example
@group
(-select-by-indices '(4 10 2 3 6) '("v" "e" "l" "o" "c" "i" "r" "a" "p" "t" "o" "r"))
    @result{} ("c" "o" "l" "o" "r")
@end group
@group
(-select-by-indices '(2 1 0) '("a" "b" "c"))
    @result{} ("c" "b" "a")
@end group
@group
(-select-by-indices '(0 1 2 0 1 3 3 1) '("f" "a" "r" "l"))
    @result{} ("f" "a" "r" "f" "a" "l" "l" "a")
@end group
@end example
@end defun

@anchor{-select-columns}
@defun -select-columns (columns table)
Select @var{columns} from @var{table}.

@var{table} is a list of lists where each element represents one row.
It is assumed each row has the same length.

Each row is transformed such that only the specified @var{columns} are
selected.

See also: @code{-select-column} (@pxref{-select-column}), @code{-select-by-indices} (@pxref{-select-by-indices})

@example
@group
(-select-columns '(0 2) '((1 2 3) (a b c) (:a :b :c)))
    @result{} ((1 3) (a c) (:a :c))
@end group
@group
(-select-columns '(1) '((1 2 3) (a b c) (:a :b :c)))
    @result{} ((2) (b) (:b))
@end group
@group
(-select-columns nil '((1 2 3) (a b c) (:a :b :c)))
    @result{} (nil nil nil)
@end group
@end example
@end defun

@anchor{-select-column}
@defun -select-column (column table)
Select @var{column} from @var{table}.

@var{table} is a list of lists where each element represents one row.
It is assumed each row has the same length.

The single selected column is returned as a list.

See also: @code{-select-columns} (@pxref{-select-columns}), @code{-select-by-indices} (@pxref{-select-by-indices})

@example
@group
(-select-column 1 '((1 2 3) (a b c) (:a :b :c)))
    @result{} (2 b :b)
@end group
@end example
@end defun

@node List to list
@section List to list

Functions returning a modified copy of the input list.

@anchor{-keep}
@defun -keep (fn list)
Return a new list of the non-@code{nil} results of applying @var{fn} to each item in @var{list}.
Like @code{-filter} (@pxref{-filter}), but returns the non-@code{nil} results of @var{fn} instead of
the corresponding elements of @var{list}.

Its anaphoric counterpart is @code{--keep}.

@example
@group
(-keep #'cdr '((1 2 3) (4 5) (6)))
    @result{} ((2 3) (5))
@end group
@group
(-keep (lambda (n) (and (> n 3) (* 10 n))) '(1 2 3 4 5 6))
    @result{} (40 50 60)
@end group
@group
(--keep (and (> it 3) (* 10 it)) '(1 2 3 4 5 6))
    @result{} (40 50 60)
@end group
@end example
@end defun

@anchor{-concat}
@defun -concat (&rest sequences)
Concatenate all the arguments and make the result a list.
The result is a list whose elements are the elements of all the arguments.
Each argument may be a list, vector or string.

All arguments except the last argument are copied.  The last argument
is just used as the tail of the new list.

@example
@group
(-concat '(1))
    @result{} (1)
@end group
@group
(-concat '(1) '(2))
    @result{} (1 2)
@end group
@group
(-concat '(1) '(2 3) '(4))
    @result{} (1 2 3 4)
@end group
@end example
@end defun

@anchor{-flatten}
@defun -flatten (l)
Take a nested list @var{l} and return its contents as a single, flat list.

Note that because @code{nil} represents a list of zero elements (an
empty list), any mention of @code{nil} in @var{l} will disappear after
flattening.  If you need to preserve nils, consider @code{-flatten-n} (@pxref{-flatten-n})
or map them to some unique symbol and then map them back.

Conses of two atoms are considered "terminals", that is, they
aren't flattened further.

See also: @code{-flatten-n} (@pxref{-flatten-n})

@example
@group
(-flatten '((1)))
    @result{} (1)
@end group
@group
(-flatten '((1 (2 3) (((4 (5)))))))
    @result{} (1 2 3 4 5)
@end group
@group
(-flatten '(1 2 (3 . 4)))
    @result{} (1 2 (3 . 4))
@end group
@end example
@end defun

@anchor{-flatten-n}
@defun -flatten-n (num list)
Flatten @var{num} levels of a nested @var{list}.

See also: @code{-flatten} (@pxref{-flatten})

@example
@group
(-flatten-n 1 '((1 2) ((3 4) ((5 6)))))
    @result{} (1 2 (3 4) ((5 6)))
@end group
@group
(-flatten-n 2 '((1 2) ((3 4) ((5 6)))))
    @result{} (1 2 3 4 (5 6))
@end group
@group
(-flatten-n 3 '((1 2) ((3 4) ((5 6)))))
    @result{} (1 2 3 4 5 6)
@end group
@end example
@end defun

@anchor{-replace}
@defun -replace (old new list)
Replace all @var{old} items in @var{list} with @var{new}.

Elements are compared using @code{equal}.

See also: @code{-replace-at} (@pxref{-replace-at})

@example
@group
(-replace 1 "1" '(1 2 3 4 3 2 1))
    @result{} ("1" 2 3 4 3 2 "1")
@end group
@group
(-replace "foo" "bar" '("a" "nice" "foo" "sentence" "about" "foo"))
    @result{} ("a" "nice" "bar" "sentence" "about" "bar")
@end group
@group
(-replace 1 2 nil)
    @result{} nil
@end group
@end example
@end defun

@anchor{-replace-first}
@defun -replace-first (old new list)
Replace the first occurrence of @var{old} with @var{new} in @var{list}.

Elements are compared using @code{equal}.

See also: @code{-map-first} (@pxref{-map-first})

@example
@group
(-replace-first 1 "1" '(1 2 3 4 3 2 1))
    @result{} ("1" 2 3 4 3 2 1)
@end group
@group
(-replace-first "foo" "bar" '("a" "nice" "foo" "sentence" "about" "foo"))
    @result{} ("a" "nice" "bar" "sentence" "about" "foo")
@end group
@group
(-replace-first 1 2 nil)
    @result{} nil
@end group
@end example
@end defun

@anchor{-replace-last}
@defun -replace-last (old new list)
Replace the last occurrence of @var{old} with @var{new} in @var{list}.

Elements are compared using @code{equal}.

See also: @code{-map-last} (@pxref{-map-last})

@example
@group
(-replace-last 1 "1" '(1 2 3 4 3 2 1))
    @result{} (1 2 3 4 3 2 "1")
@end group
@group
(-replace-last "foo" "bar" '("a" "nice" "foo" "sentence" "about" "foo"))
    @result{} ("a" "nice" "foo" "sentence" "about" "bar")
@end group
@group
(-replace-last 1 2 nil)
    @result{} nil
@end group
@end example
@end defun

@anchor{-insert-at}
@defun -insert-at (n x list)
Return a list with @var{x} inserted into @var{list} at position @var{n}.

See also: @code{-splice} (@pxref{-splice}), @code{-splice-list} (@pxref{-splice-list})

@example
@group
(-insert-at 1 'x '(a b c))
    @result{} (a x b c)
@end group
@group
(-insert-at 12 'x '(a b c))
    @result{} (a b c x)
@end group
@end example
@end defun

@anchor{-replace-at}
@defun -replace-at (n x list)
Return a list with element at Nth position in @var{list} replaced with @var{x}.

See also: @code{-replace} (@pxref{-replace})

@example
@group
(-replace-at 0 9 '(0 1 2 3 4 5))
    @result{} (9 1 2 3 4 5)
@end group
@group
(-replace-at 1 9 '(0 1 2 3 4 5))
    @result{} (0 9 2 3 4 5)
@end group
@group
(-replace-at 4 9 '(0 1 2 3 4 5))
    @result{} (0 1 2 3 9 5)
@end group
@end example
@end defun

@anchor{-update-at}
@defun -update-at (n func list)
Use @var{func} to update the Nth element of @var{list}.
Return a copy of @var{list} where the Nth element is replaced with the
result of calling @var{func} on it.

See also: @code{-map-when} (@pxref{-map-when})

@example
@group
(-update-at 0 (lambda (x) (+ x 9)) '(0 1 2 3 4 5))
    @result{} (9 1 2 3 4 5)
@end group
@group
(-update-at 1 (lambda (x) (+ x 8)) '(0 1 2 3 4 5))
    @result{} (0 9 2 3 4 5)
@end group
@group
(--update-at 2 (length it) '("foo" "bar" "baz" "quux"))
    @result{} ("foo" "bar" 3 "quux")
@end group
@end example
@end defun

@anchor{-remove-at}
@defun -remove-at (n list)
Return @var{list} with its element at index @var{n} removed.
That is, remove any element selected as (nth @var{n} @var{list}) from @var{list}
and return the result.

This is a non-destructive operation: parts of @var{list} (but not
necessarily all of it) are copied as needed to avoid
destructively modifying it.

See also: @code{-remove-at-indices} (@pxref{-remove-at-indices}), @code{-remove} (@pxref{-remove}).

@example
@group
(-remove-at 0 '(a b c))
    @result{} (b c)
@end group
@group
(-remove-at 1 '(a b c))
    @result{} (a c)
@end group
@group
(-remove-at 2 '(a b c))
    @result{} (a b)
@end group
@end example
@end defun

@anchor{-remove-at-indices}
@defun -remove-at-indices (indices list)
Return @var{list} with its elements at @var{indices} removed.
That is, for each index @var{i} in @var{indices}, remove any element selected
as (nth @var{i} @var{list}) from @var{list}.

This is a non-destructive operation: parts of @var{list} (but not
necessarily all of it) are copied as needed to avoid
destructively modifying it.

See also: @code{-remove-at} (@pxref{-remove-at}), @code{-remove} (@pxref{-remove}).

@example
@group
(-remove-at-indices '(0) '(a b c d e))
    @result{} (b c d e)
@end group
@group
(-remove-at-indices '(1 3) '(a b c d e))
    @result{} (a c e)
@end group
@group
(-remove-at-indices '(4 0 2) '(a b c d e))
    @result{} (b d)
@end group
@end example
@end defun

@node Reductions
@section Reductions

Functions reducing lists to a single value (which may also be a list).

@anchor{-reduce-from}
@defun -reduce-from (fn init list)
Reduce the function @var{fn} across @var{list}, starting with @var{init}.
Return the result of applying @var{fn} to @var{init} and the first element of
@var{list}, then applying @var{fn} to that result and the second element,
etc.  If @var{list} is empty, return @var{init} without calling @var{fn}.

This function's anaphoric counterpart is @code{--reduce-from}.

For other folds, see also @code{-reduce} (@pxref{-reduce}) and @code{-reduce-r} (@pxref{-reduce-r}).

@example
@group
(-reduce-from #'- 10 '(1 2 3))
    @result{} 4
@end group
@group
(-reduce-from #'list 10 '(1 2 3))
    @result{} (((10 1) 2) 3)
@end group
@group
(--reduce-from (concat acc " " it) "START" '("a" "b" "c"))
    @result{} "START a b c"
@end group
@end example
@end defun

@anchor{-reduce-r-from}
@defun -reduce-r-from (fn init list)
Reduce the function @var{fn} across @var{list} in reverse, starting with @var{init}.
Return the result of applying @var{fn} to the last element of @var{list} and
@var{init}, then applying @var{fn} to the second-to-last element and the
previous result of @var{fn}, etc.  That is, the first argument of @var{fn} is
the current element, and its second argument the accumulated
value.  If @var{list} is empty, return @var{init} without calling @var{fn}.

This function is like @code{-reduce-from} (@pxref{-reduce-from}) but the operation associates
from the right rather than left.  In other words, it starts from
the end of @var{list} and flips the arguments to @var{fn}.  Conceptually, it
is like replacing the conses in @var{list} with applications of @var{fn}, and
its last link with @var{init}, and evaluating the resulting expression.

This function's anaphoric counterpart is @code{--reduce-r-from}.

For other folds, see also @code{-reduce-r} (@pxref{-reduce-r}) and @code{-reduce} (@pxref{-reduce}).

@example
@group
(-reduce-r-from #'- 10 '(1 2 3))
    @result{} -8
@end group
@group
(-reduce-r-from #'list 10 '(1 2 3))
    @result{} (1 (2 (3 10)))
@end group
@group
(--reduce-r-from (concat it " " acc) "END" '("a" "b" "c"))
    @result{} "a b c END"
@end group
@end example
@end defun

@anchor{-reduce}
@defun -reduce (fn list)
Reduce the function @var{fn} across @var{list}.
Return the result of applying @var{fn} to the first two elements of
@var{list}, then applying @var{fn} to that result and the third element, etc.
If @var{list} contains a single element, return it without calling @var{fn}.
If @var{list} is empty, return the result of calling @var{fn} with no
arguments.

This function's anaphoric counterpart is @code{--reduce}.

For other folds, see also @code{-reduce-from} (@pxref{-reduce-from}) and @code{-reduce-r} (@pxref{-reduce-r}).

@example
@group
(-reduce #'- '(1 2 3 4))
    @result{} -8
@end group
@group
(-reduce #'list '(1 2 3 4))
    @result{} (((1 2) 3) 4)
@end group
@group
(--reduce (format "%s-%d" acc it) '(1 2 3))
    @result{} "1-2-3"
@end group
@end example
@end defun

@anchor{-reduce-r}
@defun -reduce-r (fn list)
Reduce the function @var{fn} across @var{list} in reverse.
Return the result of applying @var{fn} to the last two elements of
@var{list}, then applying @var{fn} to the third-to-last element and the
previous result of @var{fn}, etc.  That is, the first argument of @var{fn} is
the current element, and its second argument the accumulated
value.  If @var{list} contains a single element, return it without
calling @var{fn}.  If @var{list} is empty, return the result of calling @var{fn}
with no arguments.

This function is like @code{-reduce} (@pxref{-reduce}) but the operation associates from
the right rather than left.  In other words, it starts from the
end of @var{list} and flips the arguments to @var{fn}.  Conceptually, it is
like replacing the conses in @var{list} with applications of @var{fn},
ignoring its last link, and evaluating the resulting expression.

This function's anaphoric counterpart is @code{--reduce-r}.

For other folds, see also @code{-reduce-r-from} (@pxref{-reduce-r-from}) and @code{-reduce} (@pxref{-reduce}).

@example
@group
(-reduce-r #'- '(1 2 3 4))
    @result{} -2
@end group
@group
(-reduce-r #'list '(1 2 3 4))
    @result{} (1 (2 (3 4)))
@end group
@group
(--reduce-r (format "%s-%d" acc it) '(1 2 3))
    @result{} "3-2-1"
@end group
@end example
@end defun

@anchor{-reductions-from}
@defun -reductions-from (fn init list)
Return a list of @var{fn}'s intermediate reductions across @var{list}.
That is, a list of the intermediate values of the accumulator
when @code{-reduce-from} (@pxref{-reduce-from}) (which see) is called with the same
arguments.

This function's anaphoric counterpart is @code{--reductions-from}.

For other folds, see also @code{-reductions} (@pxref{-reductions}) and @code{-reductions-r} (@pxref{-reductions-r}).

@example
@group
(-reductions-from #'max 0 '(2 1 4 3))
    @result{} (0 2 2 4 4)
@end group
@group
(-reductions-from #'* 1 '(1 2 3 4))
    @result{} (1 1 2 6 24)
@end group
@group
(--reductions-from (format "(FN %s %d)" acc it) "INIT" '(1 2 3))
    @result{} ("INIT" "(FN INIT 1)" "(FN (FN INIT 1) 2)" "(FN (FN (FN INIT 1) 2) 3)")
@end group
@end example
@end defun

@anchor{-reductions-r-from}
@defun -reductions-r-from (fn init list)
Return a list of @var{fn}'s intermediate reductions across reversed @var{list}.
That is, a list of the intermediate values of the accumulator
when @code{-reduce-r-from} (@pxref{-reduce-r-from}) (which see) is called with the same
arguments.

This function's anaphoric counterpart is @code{--reductions-r-from}.

For other folds, see also @code{-reductions} (@pxref{-reductions}) and @code{-reductions-r} (@pxref{-reductions-r}).

@example
@group
(-reductions-r-from #'max 0 '(2 1 4 3))
    @result{} (4 4 4 3 0)
@end group
@group
(-reductions-r-from #'* 1 '(1 2 3 4))
    @result{} (24 24 12 4 1)
@end group
@group
(--reductions-r-from (format "(FN %d %s)" it acc) "INIT" '(1 2 3))
    @result{} ("(FN 1 (FN 2 (FN 3 INIT)))" "(FN 2 (FN 3 INIT))" "(FN 3 INIT)" "INIT")
@end group
@end example
@end defun

@anchor{-reductions}
@defun -reductions (fn list)
Return a list of @var{fn}'s intermediate reductions across @var{list}.
That is, a list of the intermediate values of the accumulator
when @code{-reduce} (@pxref{-reduce}) (which see) is called with the same arguments.

This function's anaphoric counterpart is @code{--reductions}.

For other folds, see also @code{-reductions} (@pxref{-reductions}) and @code{-reductions-r} (@pxref{-reductions-r}).

@example
@group
(-reductions #'+ '(1 2 3 4))
    @result{} (1 3 6 10)
@end group
@group
(-reductions #'* '(1 2 3 4))
    @result{} (1 2 6 24)
@end group
@group
(--reductions (format "(FN %s %d)" acc it) '(1 2 3))
    @result{} (1 "(FN 1 2)" "(FN (FN 1 2) 3)")
@end group
@end example
@end defun

@anchor{-reductions-r}
@defun -reductions-r (fn list)
Return a list of @var{fn}'s intermediate reductions across reversed @var{list}.
That is, a list of the intermediate values of the accumulator
when @code{-reduce-r} (@pxref{-reduce-r}) (which see) is called with the same arguments.

This function's anaphoric counterpart is @code{--reductions-r}.

For other folds, see also @code{-reductions-r-from} (@pxref{-reductions-r-from}) and
@code{-reductions} (@pxref{-reductions}).

@example
@group
(-reductions-r #'+ '(1 2 3 4))
    @result{} (10 9 7 4)
@end group
@group
(-reductions-r #'* '(1 2 3 4))
    @result{} (24 24 12 4)
@end group
@group
(--reductions-r (format "(FN %d %s)" it acc) '(1 2 3))
    @result{} ("(FN 1 (FN 2 3))" "(FN 2 3)" 3)
@end group
@end example
@end defun

@anchor{-count}
@defun -count (pred list)
Counts the number of items in @var{list} where (@var{pred} item) is non-@code{nil}.

@example
@group
(-count 'even? '(1 2 3 4 5))
    @result{} 2
@end group
@group
(--count (< it 4) '(1 2 3 4))
    @result{} 3
@end group
@end example
@end defun

@anchor{-sum}
@defun -sum (list)
Return the sum of @var{list}.

@example
@group
(-sum ())
    @result{} 0
@end group
@group
(-sum '(1))
    @result{} 1
@end group
@group
(-sum '(1 2 3 4))
    @result{} 10
@end group
@end example
@end defun

@anchor{-running-sum}
@defun -running-sum (list)
Return a list with running sums of items in @var{list}.
@var{list} must be non-empty.

@example
@group
(-running-sum '(1 2 3 4))
    @result{} (1 3 6 10)
@end group
@group
(-running-sum '(1))
    @result{} (1)
@end group
@group
(-running-sum ())
    @error{} Wrong type argument: consp, nil
@end group
@end example
@end defun

@anchor{-product}
@defun -product (list)
Return the product of @var{list}.

@example
@group
(-product ())
    @result{} 1
@end group
@group
(-product '(1))
    @result{} 1
@end group
@group
(-product '(1 2 3 4))
    @result{} 24
@end group
@end example
@end defun

@anchor{-running-product}
@defun -running-product (list)
Return a list with running products of items in @var{list}.
@var{list} must be non-empty.

@example
@group
(-running-product '(1 2 3 4))
    @result{} (1 2 6 24)
@end group
@group
(-running-product '(1))
    @result{} (1)
@end group
@group
(-running-product ())
    @error{} Wrong type argument: consp, nil
@end group
@end example
@end defun

@anchor{-inits}
@defun -inits (list)
Return all prefixes of @var{list}.

@example
@group
(-inits '(1 2 3 4))
    @result{} (nil (1) (1 2) (1 2 3) (1 2 3 4))
@end group
@group
(-inits nil)
    @result{} (nil)
@end group
@group
(-inits '(1))
    @result{} (nil (1))
@end group
@end example
@end defun

@anchor{-tails}
@defun -tails (list)
Return all suffixes of @var{list}.

@example
@group
(-tails '(1 2 3 4))
    @result{} ((1 2 3 4) (2 3 4) (3 4) (4) nil)
@end group
@group
(-tails nil)
    @result{} (nil)
@end group
@group
(-tails '(1))
    @result{} ((1) nil)
@end group
@end example
@end defun

@anchor{-common-prefix}
@defun -common-prefix (&rest lists)
Return the longest common prefix of @var{lists}.

@example
@group
(-common-prefix '(1))
    @result{} (1)
@end group
@group
(-common-prefix '(1 2) '(3 4) '(1 2))
    @result{} ()
@end group
@group
(-common-prefix '(1 2) '(1 2 3) '(1 2 3 4))
    @result{} (1 2)
@end group
@end example
@end defun

@anchor{-common-suffix}
@defun -common-suffix (&rest lists)
Return the longest common suffix of @var{lists}.

@example
@group
(-common-suffix '(1))
    @result{} (1)
@end group
@group
(-common-suffix '(1 2) '(3 4) '(1 2))
    @result{} ()
@end group
@group
(-common-suffix '(1 2 3 4) '(2 3 4) '(3 4))
    @result{} (3 4)
@end group
@end example
@end defun

@anchor{-min}
@defun -min (list)
Return the smallest value from @var{list} of numbers or markers.

@example
@group
(-min '(0))
    @result{} 0
@end group
@group
(-min '(3 2 1))
    @result{} 1
@end group
@group
(-min '(1 2 3))
    @result{} 1
@end group
@end example
@end defun

@anchor{-min-by}
@defun -min-by (comparator list)
Take a comparison function @var{comparator} and a @var{list} and return
the least element of the list by the comparison function.

See also combinator @code{-on} (@pxref{-on}) which can transform the values before
comparing them.

@example
@group
(-min-by '> '(4 3 6 1))
    @result{} 1
@end group
@group
(--min-by (> (car it) (car other)) '((1 2 3) (2) (3 2)))
    @result{} (1 2 3)
@end group
@group
(--min-by (> (length it) (length other)) '((1 2 3) (2) (3 2)))
    @result{} (2)
@end group
@end example
@end defun

@anchor{-max}
@defun -max (list)
Return the largest value from @var{list} of numbers or markers.

@example
@group
(-max '(0))
    @result{} 0
@end group
@group
(-max '(3 2 1))
    @result{} 3
@end group
@group
(-max '(1 2 3))
    @result{} 3
@end group
@end example
@end defun

@anchor{-max-by}
@defun -max-by (comparator list)
Take a comparison function @var{comparator} and a @var{list} and return
the greatest element of the list by the comparison function.

See also combinator @code{-on} (@pxref{-on}) which can transform the values before
comparing them.

@example
@group
(-max-by '> '(4 3 6 1))
    @result{} 6
@end group
@group
(--max-by (> (car it) (car other)) '((1 2 3) (2) (3 2)))
    @result{} (3 2)
@end group
@group
(--max-by (> (length it) (length other)) '((1 2 3) (2) (3 2)))
    @result{} (1 2 3)
@end group
@end example
@end defun

@anchor{-frequencies}
@defun -frequencies (list)
Count the occurrences of each distinct element of @var{list}.

Return an alist of (@var{element} . @var{n}), where each @var{element} occurs @var{n}
times in @var{list}.

The test for equality is done with @code{equal}, or with @code{-compare-fn}
if that is non-@code{nil}.

See also @code{-count} (@pxref{-count}) and @code{-group-by} (@pxref{-group-by}).

@example
@group
(-frequencies ())
    @result{} ()
@end group
@group
(-frequencies '(1 2 3 1 2 1))
    @result{} ((1 . 3) (2 . 2) (3 . 1))
@end group
@group
(let ((-compare-fn #'string=)) (-frequencies '(a "a")))
    @result{} ((a . 2))
@end group
@end example
@end defun

@node Unfolding
@section Unfolding

Operations dual to reductions, building lists from a seed
value rather than consuming a list to produce a single value.

@anchor{-iterate}
@defun -iterate (fun init n)
Return a list of iterated applications of @var{fun} to @var{init}.

This means a list of the form:

  (@var{init} (@var{fun} @var{init}) (@var{fun} (@var{fun} @var{init})) @dots{})

@var{n} is the length of the returned list.

@example
@group
(-iterate #'1+ 1 10)
    @result{} (1 2 3 4 5 6 7 8 9 10)
@end group
@group
(-iterate (lambda (x) (+ x x)) 2 5)
    @result{} (2 4 8 16 32)
@end group
@group
(--iterate (* it it) 2 5)
    @result{} (2 4 16 256 65536)
@end group
@end example
@end defun

@anchor{-unfold}
@defun -unfold (fun seed)
Build a list from @var{seed} using @var{fun}.

This is "dual" operation to @code{-reduce-r} (@pxref{-reduce-r}): while -reduce-r
consumes a list to produce a single value, @code{-unfold} (@pxref{-unfold}) takes a
seed value and builds a (potentially infinite!) list.

@var{fun} should return @code{nil} to stop the generating process, or a
cons (@var{a} . @var{b}), where @var{a} will be prepended to the result and @var{b} is
the new seed.

@example
@group
(-unfold (lambda (x) (unless (= x 0) (cons x (1- x)))) 10)
    @result{} (10 9 8 7 6 5 4 3 2 1)
@end group
@group
(--unfold (when it (cons it (cdr it))) '(1 2 3 4))
    @result{} ((1 2 3 4) (2 3 4) (3 4) (4))
@end group
@group
(--unfold (when it (cons it (butlast it))) '(1 2 3 4))
    @result{} ((1 2 3 4) (1 2 3) (1 2) (1))
@end group
@end example
@end defun

@anchor{-repeat}
@defun -repeat (n x)
Return a new list of length @var{n} with each element being @var{x}.
Return @code{nil} if @var{n} is less than 1.

@example
@group
(-repeat 3 :a)
    @result{} (:a :a :a)
@end group
@group
(-repeat 1 :a)
    @result{} (:a)
@end group
@group
(-repeat 0 :a)
    @result{} ()
@end group
@end example
@end defun

@anchor{-cycle}
@defun -cycle (list)
Return an infinite circular copy of @var{list}.
The returned list cycles through the elements of @var{list} and repeats
from the beginning.

@example
@group
(-take 5 (-cycle '(1 2 3)))
    @result{} (1 2 3 1 2)
@end group
@group
(-take 7 (-cycle '(1 "and" 3)))
    @result{} (1 "and" 3 1 "and" 3 1)
@end group
@group
(-zip-lists (-cycle '(3)) '(1 2))
    @result{} ((3 1) (3 2))
@end group
@end example
@end defun

@node Predicates
@section Predicates

Reductions of one or more lists to a boolean value.

@anchor{-some}
@defun -some (pred list)
Return (@var{pred} x) for the first @var{list} item where (@var{pred} x) is non-@code{nil}, else @code{nil}.

Alias: @code{-any}.

This function's anaphoric counterpart is @code{--some}.

@example
@group
(-some #'stringp '(1 "2" 3))
    @result{} t
@end group
@group
(--some (string-match-p "x" it) '("foo" "axe" "xor"))
    @result{} 1
@end group
@group
(--some (= it-index 3) '(0 1 2))
    @result{} nil
@end group
@end example
@end defun

@anchor{-every}
@defun -every (pred list)
Return non-@code{nil} if @var{pred} returns non-@code{nil} for all items in @var{list}.
If so, return the last such result of @var{pred}.  Otherwise, once an
item is reached for which @var{pred} returns @code{nil}, return @code{nil} without
calling @var{pred} on any further @var{list} elements.

This function is like @code{-every-p}, but on success returns the last
non-@code{nil} result of @var{pred} instead of just @code{t}.

This function's anaphoric counterpart is @code{--every}.

@example
@group
(-every #'numberp '(1 2 3))
    @result{} t
@end group
@group
(--every (string-match-p "x" it) '("axe" "xor"))
    @result{} 0
@end group
@group
(--every (= it it-index) '(0 1 3))
    @result{} nil
@end group
@end example
@end defun

@anchor{-any?}
@defun -any? (pred list)
Return @code{t} if (@var{pred} @var{x}) is non-@code{nil} for any @var{x} in @var{list}, else @code{nil}.

Alias: @code{-any-p}, @code{-some?}, @code{-some-p}

@example
@group
(-any? #'numberp '(nil 0 t))
    @result{} t
@end group
@group
(-any? #'numberp '(nil t t))
    @result{} nil
@end group
@group
(-any? #'null '(1 3 5))
    @result{} nil
@end group
@end example
@end defun

@anchor{-all?}
@defun -all? (pred list)
Return @code{t} if (@var{pred} @var{x}) is non-@code{nil} for all @var{x} in @var{list}, else @code{nil}.
In the latter case, stop after the first @var{x} for which (@var{pred} @var{x}) is
@code{nil}, without calling @var{pred} on any subsequent elements of @var{list}.

The similar function @code{-every} (@pxref{-every}) is more widely useful, since it
returns the last non-@code{nil} result of @var{pred} instead of just @code{t} on
success.

Alias: @code{-all-p}, @code{-every-p}, @code{-every?}.

This function's anaphoric counterpart is @code{--all?}.

@example
@group
(-all? #'numberp '(1 2 3))
    @result{} t
@end group
@group
(-all? #'numberp '(2 t 6))
    @result{} nil
@end group
@group
(--all? (= 0 (% it 2)) '(2 4 6))
    @result{} t
@end group
@end example
@end defun

@anchor{-none?}
@defun -none? (pred list)
Return @code{t} if (@var{pred} @var{x}) is @code{nil} for all @var{x} in @var{list}, else @code{nil}.

Alias: @code{-none-p}

@example
@group
(-none? 'even? '(1 2 3))
    @result{} nil
@end group
@group
(-none? 'even? '(1 3 5))
    @result{} t
@end group
@group
(--none? (= 0 (% it 2)) '(1 2 3))
    @result{} nil
@end group
@end example
@end defun

@anchor{-only-some?}
@defun -only-some? (pred list)
Return @code{t} if different @var{list} items both satisfy and do not satisfy @var{pred}.
That is, if @var{pred} returns both @code{nil} for at least one item, and
non-@code{nil} for at least one other item in @var{list}.  Return @code{nil} if all
items satisfy the predicate or none of them do.

Alias: @code{-only-some-p}

@example
@group
(-only-some? 'even? '(1 2 3))
    @result{} t
@end group
@group
(-only-some? 'even? '(1 3 5))
    @result{} nil
@end group
@group
(-only-some? 'even? '(2 4 6))
    @result{} nil
@end group
@end example
@end defun

@anchor{-contains?}
@defun -contains? (list element)
Return non-@code{nil} if @var{list} contains @var{element}.

The test for equality is done with @code{equal}, or with @code{-compare-fn}
if that is non-@code{nil}.  As with @code{member}, the return value is
actually the tail of @var{list} whose car is @var{element}.

Alias: @code{-contains-p}.

@example
@group
(-contains? '(1 2 3) 1)
    @result{} (1 2 3)
@end group
@group
(-contains? '(1 2 3) 2)
    @result{} (2 3)
@end group
@group
(-contains? '(1 2 3) 4)
    @result{} ()
@end group
@end example
@end defun

@anchor{-is-prefix?}
@defun -is-prefix? (prefix list)
Return non-@code{nil} if @var{prefix} is a prefix of @var{list}.

Alias: @code{-is-prefix-p}.

@example
@group
(-is-prefix? '(1 2 3) '(1 2 3 4 5))
    @result{} t
@end group
@group
(-is-prefix? '(1 2 3 4 5) '(1 2 3))
    @result{} nil
@end group
@group
(-is-prefix? '(1 3) '(1 2 3 4 5))
    @result{} nil
@end group
@end example
@end defun

@anchor{-is-suffix?}
@defun -is-suffix? (suffix list)
Return non-@code{nil} if @var{suffix} is a suffix of @var{list}.

Alias: @code{-is-suffix-p}.

@example
@group
(-is-suffix? '(3 4 5) '(1 2 3 4 5))
    @result{} t
@end group
@group
(-is-suffix? '(1 2 3 4 5) '(3 4 5))
    @result{} nil
@end group
@group
(-is-suffix? '(3 5) '(1 2 3 4 5))
    @result{} nil
@end group
@end example
@end defun

@anchor{-is-infix?}
@defun -is-infix? (infix list)
Return non-@code{nil} if @var{infix} is infix of @var{list}.

This operation runs in @var{o}(n^2) time

Alias: @code{-is-infix-p}

@example
@group
(-is-infix? '(1 2 3) '(1 2 3 4 5))
    @result{} t
@end group
@group
(-is-infix? '(2 3 4) '(1 2 3 4 5))
    @result{} t
@end group
@group
(-is-infix? '(3 4 5) '(1 2 3 4 5))
    @result{} t
@end group
@end example
@end defun

@anchor{-cons-pair?}
@defun -cons-pair? (obj)
Return non-@code{nil} if @var{obj} is a true cons pair.
That is, a cons (@var{a} . @var{b}) where @var{b} is not a list.

Alias: @code{-cons-pair-p}.

@example
@group
(-cons-pair? '(1 . 2))
    @result{} t
@end group
@group
(-cons-pair? '(1 2))
    @result{} nil
@end group
@group
(-cons-pair? '(1))
    @result{} nil
@end group
@end example
@end defun

@node Partitioning
@section Partitioning

Functions partitioning the input list into a list of lists.

@anchor{-split-at}
@defun -split-at (n list)
Split @var{list} into two sublists after the Nth element.
The result is a list of two elements (@var{take} @var{drop}) where @var{take} is a
new list of the first @var{n} elements of @var{list}, and @var{drop} is the
remaining elements of @var{list} (not a copy).  @var{take} and @var{drop} are like
the results of @code{-take} (@pxref{-take}) and @code{-drop} (@pxref{-drop}), respectively, but the split
is done in a single list traversal.

@example
@group
(-split-at 3 '(1 2 3 4 5))
    @result{} ((1 2 3) (4 5))
@end group
@group
(-split-at 17 '(1 2 3 4 5))
    @result{} ((1 2 3 4 5) nil)
@end group
@group
(-split-at 0 '(1 2 3 4 5))
    @result{} (nil (1 2 3 4 5))
@end group
@end example
@end defun

@anchor{-split-with}
@defun -split-with (pred list)
Split @var{list} into a prefix satisfying @var{pred}, and the rest.
The first sublist is the prefix of @var{list} with successive elements
satisfying @var{pred}, and the second sublist is the remaining elements
that do not.  The result is like performing

  ((-take-while @var{pred} @var{list}) (-drop-while @var{pred} @var{list}))

but in no more than a single pass through @var{list}.

@example
@group
(-split-with 'even? '(1 2 3 4))
    @result{} (nil (1 2 3 4))
@end group
@group
(-split-with 'even? '(2 4 5 6))
    @result{} ((2 4) (5 6))
@end group
@group
(--split-with (< it 4) '(1 2 3 4 3 2 1))
    @result{} ((1 2 3) (4 3 2 1))
@end group
@end example
@end defun

@anchor{-split-on}
@defmac -split-on (item list)
Split the @var{list} each time @var{item} is found.

Unlike @code{-partition-by} (@pxref{-partition-by}), the @var{item} is discarded from the results.
Empty lists are also removed from the result.

Comparison is done by @code{equal}.

See also @code{-split-when} (@pxref{-split-when})

@example
@group
(-split-on '| '(Nil | Leaf a | Node [Tree a]))
    @result{} ((Nil) (Leaf a) (Node [Tree a]))
@end group
@group
(-split-on :endgroup '("a" "b" :endgroup "c" :endgroup "d" "e"))
    @result{} (("a" "b") ("c") ("d" "e"))
@end group
@group
(-split-on :endgroup '("a" "b" :endgroup :endgroup "d" "e"))
    @result{} (("a" "b") ("d" "e"))
@end group
@end example
@end defmac

@anchor{-split-when}
@defun -split-when (fn list)
Split the @var{list} on each element where @var{fn} returns non-@code{nil}.

Unlike @code{-partition-by} (@pxref{-partition-by}), the "matched" element is discarded from
the results.  Empty lists are also removed from the result.

This function can be thought of as a generalization of
@code{split-string}.

@example
@group
(-split-when 'even? '(1 2 3 4 5 6))
    @result{} ((1) (3) (5))
@end group
@group
(-split-when 'even? '(1 2 3 4 6 8 9))
    @result{} ((1) (3) (9))
@end group
@group
(--split-when (memq it '(&optional &rest)) '(a b &optional c d &rest args))
    @result{} ((a b) (c d) (args))
@end group
@end example
@end defun

@anchor{-separate}
@defun -separate (pred list)
Split @var{list} into two sublists based on whether items satisfy @var{pred}.
The result is like performing

  ((-filter @var{pred} @var{list}) (-remove @var{pred} @var{list}))

but in a single pass through @var{list}.

@example
@group
(-separate (lambda (num) (= 0 (% num 2))) '(1 2 3 4 5 6 7))
    @result{} ((2 4 6) (1 3 5 7))
@end group
@group
(--separate (< it 5) '(3 7 5 9 3 2 1 4 6))
    @result{} ((3 3 2 1 4) (7 5 9 6))
@end group
@group
(-separate 'cdr '((1 2) (1) (1 2 3) (4)))
    @result{} (((1 2) (1 2 3)) ((1) (4)))
@end group
@end example
@end defun

@anchor{-partition}
@defun -partition (n list)
Return a new list with the items in @var{list} grouped into @var{n}-sized sublists.
If there are not enough items to make the last group @var{n}-sized,
those items are discarded.

@example
@group
(-partition 2 '(1 2 3 4 5 6))
    @result{} ((1 2) (3 4) (5 6))
@end group
@group
(-partition 2 '(1 2 3 4 5 6 7))
    @result{} ((1 2) (3 4) (5 6))
@end group
@group
(-partition 3 '(1 2 3 4 5 6 7))
    @result{} ((1 2 3) (4 5 6))
@end group
@end example
@end defun

@anchor{-partition-all}
@defun -partition-all (n list)
Return a new list with the items in @var{list} grouped into @var{n}-sized sublists.
The last group may contain less than @var{n} items.

@example
@group
(-partition-all 2 '(1 2 3 4 5 6))
    @result{} ((1 2) (3 4) (5 6))
@end group
@group
(-partition-all 2 '(1 2 3 4 5 6 7))
    @result{} ((1 2) (3 4) (5 6) (7))
@end group
@group
(-partition-all 3 '(1 2 3 4 5 6 7))
    @result{} ((1 2 3) (4 5 6) (7))
@end group
@end example
@end defun

@anchor{-partition-in-steps}
@defun -partition-in-steps (n step list)
Partition @var{list} into sublists of length @var{n} that are @var{step} items apart.
Like @code{-partition-all-in-steps} (@pxref{-partition-all-in-steps}), but if there are not enough items
to make the last group @var{n}-sized, those items are discarded.

@example
@group
(-partition-in-steps 2 1 '(1 2 3 4))
    @result{} ((1 2) (2 3) (3 4))
@end group
@group
(-partition-in-steps 3 2 '(1 2 3 4))
    @result{} ((1 2 3))
@end group
@group
(-partition-in-steps 3 2 '(1 2 3 4 5))
    @result{} ((1 2 3) (3 4 5))
@end group
@end example
@end defun

@anchor{-partition-all-in-steps}
@defun -partition-all-in-steps (n step list)
Partition @var{list} into sublists of length @var{n} that are @var{step} items apart.
Adjacent groups may overlap if @var{n} exceeds the @var{step} stride.
Trailing groups may contain less than @var{n} items.

@example
@group
(-partition-all-in-steps 2 1 '(1 2 3 4))
    @result{} ((1 2) (2 3) (3 4) (4))
@end group
@group
(-partition-all-in-steps 3 2 '(1 2 3 4))
    @result{} ((1 2 3) (3 4))
@end group
@group
(-partition-all-in-steps 3 2 '(1 2 3 4 5))
    @result{} ((1 2 3) (3 4 5) (5))
@end group
@end example
@end defun

@anchor{-partition-by}
@defun -partition-by (fn list)
Apply @var{fn} to each item in @var{list}, splitting it each time @var{fn} returns a new value.

@example
@group
(-partition-by 'even? ())
    @result{} ()
@end group
@group
(-partition-by 'even? '(1 1 2 2 2 3 4 6 8))
    @result{} ((1 1) (2 2 2) (3) (4 6 8))
@end group
@group
(--partition-by (< it 3) '(1 2 3 4 3 2 1))
    @result{} ((1 2) (3 4 3) (2 1))
@end group
@end example
@end defun

@anchor{-partition-by-header}
@defun -partition-by-header (fn list)
Apply @var{fn} to the first item in @var{list}. That is the header
value. Apply @var{fn} to each item in @var{list}, splitting it each time @var{fn}
returns the header value, but only after seeing at least one
other value (the body).

@example
@group
(--partition-by-header (= it 1) '(1 2 3 1 2 1 2 3 4))
    @result{} ((1 2 3) (1 2) (1 2 3 4))
@end group
@group
(--partition-by-header (> it 0) '(1 2 0 1 0 1 2 3 0))
    @result{} ((1 2 0) (1 0) (1 2 3 0))
@end group
@group
(-partition-by-header 'even? '(2 1 1 1 4 1 3 5 6 6 1))
    @result{} ((2 1 1 1) (4 1 3 5) (6 6 1))
@end group
@end example
@end defun

@anchor{-partition-after-pred}
@defun -partition-after-pred (pred list)
Partition @var{list} after each element for which @var{pred} returns non-@code{nil}.

This function's anaphoric counterpart is @code{--partition-after-pred}.

@example
@group
(-partition-after-pred #'booleanp ())
    @result{} ()
@end group
@group
(-partition-after-pred #'booleanp '(t t))
    @result{} ((t) (t))
@end group
@group
(-partition-after-pred #'booleanp '(0 0 t t 0 t))
    @result{} ((0 0 t) (t) (0 t))
@end group
@end example
@end defun

@anchor{-partition-before-pred}
@defun -partition-before-pred (pred list)
Partition directly before each time @var{pred} is true on an element of @var{list}.

@example
@group
(-partition-before-pred #'booleanp ())
    @result{} ()
@end group
@group
(-partition-before-pred #'booleanp '(0 t))
    @result{} ((0) (t))
@end group
@group
(-partition-before-pred #'booleanp '(0 0 t 0 t t))
    @result{} ((0 0) (t 0) (t) (t))
@end group
@end example
@end defun

@anchor{-partition-before-item}
@defun -partition-before-item (item list)
Partition directly before each time @var{item} appears in @var{list}.

@example
@group
(-partition-before-item 3 ())
    @result{} ()
@end group
@group
(-partition-before-item 3 '(1))
    @result{} ((1))
@end group
@group
(-partition-before-item 3 '(3))
    @result{} ((3))
@end group
@end example
@end defun

@anchor{-partition-after-item}
@defun -partition-after-item (item list)
Partition directly after each time @var{item} appears in @var{list}.

@example
@group
(-partition-after-item 3 ())
    @result{} ()
@end group
@group
(-partition-after-item 3 '(1))
    @result{} ((1))
@end group
@group
(-partition-after-item 3 '(3))
    @result{} ((3))
@end group
@end example
@end defun

@anchor{-group-by}
@defun -group-by (fn list)
Separate @var{list} into an alist whose keys are @var{fn} applied to the
elements of @var{list}.  Keys are compared by @code{equal}.

@example
@group
(-group-by 'even? ())
    @result{} ()
@end group
@group
(-group-by 'even? '(1 1 2 2 2 3 4 6 8))
    @result{} ((nil 1 1 3) (t 2 2 2 4 6 8))
@end group
@group
(--group-by (car (split-string it "/")) '("a/b" "c/d" "a/e"))
    @result{} (("a" "a/b" "a/e") ("c" "c/d"))
@end group
@end example
@end defun

@node Indexing
@section Indexing

Functions retrieving or sorting based on list indices and
related predicates.

@anchor{-elem-index}
@defun -elem-index (elem list)
Return the first index of @var{elem} in @var{list}.
That is, the index within @var{list} of the first element that is
@code{equal} to @var{elem}.  Return @code{nil} if there is no such element.

See also: @code{-find-index} (@pxref{-find-index}).

@example
@group
(-elem-index 2 '(6 7 8 3 4))
    @result{} nil
@end group
@group
(-elem-index "bar" '("foo" "bar" "baz"))
    @result{} 1
@end group
@group
(-elem-index '(1 2) '((3) (5 6) (1 2) nil))
    @result{} 2
@end group
@end example
@end defun

@anchor{-elem-indices}
@defun -elem-indices (elem list)
Return the list of indices at which @var{elem} appears in @var{list}.
That is, the indices of all elements of @var{list} @code{equal} to @var{elem}, in
the same ascending order as they appear in @var{list}.

@example
@group
(-elem-indices 2 '(6 7 8 3 4 1))
    @result{} ()
@end group
@group
(-elem-indices "bar" '("foo" "bar" "baz"))
    @result{} (1)
@end group
@group
(-elem-indices '(1 2) '((3) (1 2) (5 6) (1 2) nil))
    @result{} (1 3)
@end group
@end example
@end defun

@anchor{-find-index}
@defun -find-index (pred list)
Return the index of the first item satisfying @var{pred} in @var{list}.
Return @code{nil} if no such item is found.

@var{pred} is called with one argument, the current list element, until
it returns non-@code{nil}, at which point the search terminates.

This function's anaphoric counterpart is @code{--find-index}.

See also: @code{-first} (@pxref{-first}), @code{-find-last-index} (@pxref{-find-last-index}).

@example
@group
(-find-index #'numberp '(a b c))
    @result{} nil
@end group
@group
(-find-index #'natnump '(1 0 -1))
    @result{} 0
@end group
@group
(--find-index (> it 5) '(2 4 1 6 3 3 5 8))
    @result{} 3
@end group
@end example
@end defun

@anchor{-find-last-index}
@defun -find-last-index (pred list)
Return the index of the last item satisfying @var{pred} in @var{list}.
Return @code{nil} if no such item is found.

Predicate @var{pred} is called with one argument each time, namely the
current list element.

This function's anaphoric counterpart is @code{--find-last-index}.

See also: @code{-last} (@pxref{-last}), @code{-find-index} (@pxref{-find-index}).

@example
@group
(-find-last-index #'numberp '(a b c))
    @result{} nil
@end group
@group
(--find-last-index (> it 5) '(2 7 1 6 3 8 5 2))
    @result{} 5
@end group
@group
(-find-last-index (-partial #'string< 'a) '(c b a))
    @result{} 1
@end group
@end example
@end defun

@anchor{-find-indices}
@defun -find-indices (pred list)
Return the list of indices in @var{list} satisfying @var{pred}.

Each element of @var{list} in turn is passed to @var{pred}.  If the result is
non-@code{nil}, the index of that element in @var{list} is included in the
result.  The returned indices are in ascending order, i.e., in
the same order as they appear in @var{list}.

This function's anaphoric counterpart is @code{--find-indices}.

See also: @code{-find-index} (@pxref{-find-index}), @code{-elem-indices} (@pxref{-elem-indices}).

@example
@group
(-find-indices #'numberp '(a b c))
    @result{} ()
@end group
@group
(-find-indices #'numberp '(8 1 d 2 b c a 3))
    @result{} (0 1 3 7)
@end group
@group
(--find-indices (> it 5) '(2 4 1 6 3 3 5 8))
    @result{} (3 7)
@end group
@end example
@end defun

@anchor{-grade-up}
@defun -grade-up (comparator list)
Grade elements of @var{list} using @var{comparator} relation.
This yields a permutation vector such that applying this
permutation to @var{list} sorts it in ascending order.

@example
@group
(-grade-up #'< '(3 1 4 2 1 3 3))
    @result{} (1 4 3 0 5 6 2)
@end group
@group
(let ((l '(3 1 4 2 1 3 3))) (-select-by-indices (-grade-up #'< l) l))
    @result{} (1 1 2 3 3 3 4)
@end group
@end example
@end defun

@anchor{-grade-down}
@defun -grade-down (comparator list)
Grade elements of @var{list} using @var{comparator} relation.
This yields a permutation vector such that applying this
permutation to @var{list} sorts it in descending order.

@example
@group
(-grade-down #'< '(3 1 4 2 1 3 3))
    @result{} (2 0 5 6 3 1 4)
@end group
@group
(let ((l '(3 1 4 2 1 3 3))) (-select-by-indices (-grade-down #'< l) l))
    @result{} (4 3 3 3 2 1 1)
@end group
@end example
@end defun

@node Set operations
@section Set operations

Operations pretending lists are sets.

@anchor{-union}
@defun -union (list1 list2)
Return a new list of distinct elements appearing in either @var{list1} or @var{list2}.

The test for equality is done with @code{equal}, or with @code{-compare-fn}
if that is non-@code{nil}.

@example
@group
(-union '(1 2 3) '(3 4 5))
    @result{} (1 2 3 4 5)
@end group
@group
(-union '(1 2 2 4) ())
    @result{} (1 2 4)
@end group
@group
(-union '(1 1 2 2) '(4 4 3 2 1))
    @result{} (1 2 4 3)
@end group
@end example
@end defun

@anchor{-difference}
@defun -difference (list1 list2)
Return a new list with the distinct members of @var{list1} that are not in @var{list2}.

The test for equality is done with @code{equal}, or with @code{-compare-fn}
if that is non-@code{nil}.

@example
@group
(-difference () ())
    @result{} ()
@end group
@group
(-difference '(1 2 3) '(4 5 6))
    @result{} (1 2 3)
@end group
@group
(-difference '(1 2 3 4) '(3 4 5 6))
    @result{} (1 2)
@end group
@end example
@end defun

@anchor{-intersection}
@defun -intersection (list1 list2)
Return a new list of distinct elements appearing in both @var{list1} and @var{list2}.

The test for equality is done with @code{equal}, or with @code{-compare-fn}
if that is non-@code{nil}.

@example
@group
(-intersection () ())
    @result{} ()
@end group
@group
(-intersection '(1 2 3) '(4 5 6))
    @result{} ()
@end group
@group
(-intersection '(1 2 2 3) '(4 3 3 2))
    @result{} (2 3)
@end group
@end example
@end defun

@anchor{-powerset}
@defun -powerset (list)
Return the power set of @var{list}.

@example
@group
(-powerset ())
    @result{} (nil)
@end group
@group
(-powerset '(x y))
    @result{} ((x y) (x) (y) nil)
@end group
@group
(-powerset '(x y z))
    @result{} ((x y z) (x y) (x z) (x) (y z) (y) (z) nil)
@end group
@end example
@end defun

@anchor{-permutations}
@defun -permutations (list)
Return the distinct permutations of @var{list}.

Duplicate elements of @var{list} are determined by @code{equal}, or by
@code{-compare-fn} if that is non-@code{nil}.

@example
@group
(-permutations ())
    @result{} (nil)
@end group
@group
(-permutations '(a a b))
    @result{} ((a a b) (a b a) (b a a))
@end group
@group
(-permutations '(a b c))
    @result{} ((a b c) (a c b) (b a c) (b c a) (c a b) (c b a))
@end group
@end example
@end defun

@anchor{-distinct}
@defun -distinct (list)
Return a copy of @var{list} with all duplicate elements removed.

The test for equality is done with @code{equal}, or with @code{-compare-fn}
if that is non-@code{nil}.

Alias: @code{-uniq}.

@example
@group
(-distinct ())
    @result{} ()
@end group
@group
(-distinct '(1 1 2 3 3))
    @result{} (1 2 3)
@end group
@group
(-distinct '(t t t))
    @result{} (t)
@end group
@end example
@end defun

@anchor{-same-items?}
@defun -same-items? (list1 list2)
Return non-@code{nil} if @var{list1} and @var{list2} have the same distinct elements.

The order of the elements in the lists does not matter.  The
lists may be of different lengths, i.e., contain duplicate
elements.  The test for equality is done with @code{equal}, or with
@code{-compare-fn} if that is non-@code{nil}.

Alias: @code{-same-items-p}.

@example
@group
(-same-items? '(1 2 3) '(1 2 3))
    @result{} t
@end group
@group
(-same-items? '(1 1 2 3) '(3 3 2 1))
    @result{} t
@end group
@group
(-same-items? '(1 2 3) '(1 2 3 4))
    @result{} nil
@end group
@end example
@end defun

@node Other list operations
@section Other list operations

Other list functions not fit to be classified elsewhere.

@anchor{-rotate}
@defun -rotate (n list)
Rotate @var{list} @var{n} places to the right (left if @var{n} is negative).
The time complexity is @var{o}(n).

@example
@group
(-rotate 3 '(1 2 3 4 5 6 7))
    @result{} (5 6 7 1 2 3 4)
@end group
@group
(-rotate -3 '(1 2 3 4 5 6 7))
    @result{} (4 5 6 7 1 2 3)
@end group
@group
(-rotate 16 '(1 2 3 4 5 6 7))
    @result{} (6 7 1 2 3 4 5)
@end group
@end example
@end defun

@anchor{-cons*}
@defun -cons* (&rest args)
Make a new list from the elements of @var{args}.
The last 2 elements of @var{args} are used as the final cons of the
result, so if the final element of @var{args} is not a list, the result
is a dotted list.  With no @var{args}, return @code{nil}.

@example
@group
(-cons* 1 2)
    @result{} (1 . 2)
@end group
@group
(-cons* 1 2 3)
    @result{} (1 2 . 3)
@end group
@group
(-cons* 1)
    @result{} 1
@end group
@end example
@end defun

@anchor{-snoc}
@defun -snoc (list elem &rest elements)
Append @var{elem} to the end of the list.

This is like @code{cons}, but operates on the end of list.

If any @var{elements} are given, append them to the list as well.

@example
@group
(-snoc '(1 2 3) 4)
    @result{} (1 2 3 4)
@end group
@group
(-snoc '(1 2 3) 4 5 6)
    @result{} (1 2 3 4 5 6)
@end group
@group
(-snoc '(1 2 3) '(4 5 6))
    @result{} (1 2 3 (4 5 6))
@end group
@end example
@end defun

@anchor{-interpose}
@defun -interpose (sep list)
Return a new list of all elements in @var{list} separated by @var{sep}.

@example
@group
(-interpose "-" ())
    @result{} ()
@end group
@group
(-interpose "-" '("a"))
    @result{} ("a")
@end group
@group
(-interpose "-" '("a" "b" "c"))
    @result{} ("a" "-" "b" "-" "c")
@end group
@end example
@end defun

@anchor{-interleave}
@defun -interleave (&rest lists)
Return a new list of the first item in each list, then the second etc.

@example
@group
(-interleave '(1 2) '("a" "b"))
    @result{} (1 "a" 2 "b")
@end group
@group
(-interleave '(1 2) '("a" "b") '("A" "B"))
    @result{} (1 "a" "A" 2 "b" "B")
@end group
@group
(-interleave '(1 2 3) '("a" "b"))
    @result{} (1 "a" 2 "b")
@end group
@end example
@end defun

@anchor{-iota}
@defun -iota (count &optional start step)
Return a list containing @var{count} numbers.
Starts from @var{start} and adds @var{step} each time.  The default @var{start} is
zero, the default @var{step} is 1.
This function takes its name from the corresponding primitive in
the @var{apl} language.

@example
@group
(-iota 6)
    @result{} (0 1 2 3 4 5)
@end group
@group
(-iota 4 2.5 -2)
    @result{} (2.5 0.5 -1.5 -3.5)
@end group
@group
(-iota -1)
    @error{} Wrong type argument: natnump, -1
@end group
@end example
@end defun

@anchor{-zip-with}
@defun -zip-with (fn list1 list2)
Zip @var{list1} and @var{list2} into a new list using the function @var{fn}.
That is, apply @var{fn} pairwise taking as first argument the next
element of @var{list1} and as second argument the next element of @var{list2}
at the corresponding position.  The result is as long as the
shorter list.

This function's anaphoric counterpart is @code{--zip-with}.

For other zips, see also @code{-zip-lists} (@pxref{-zip-lists}) and @code{-zip-fill} (@pxref{-zip-fill}).

@example
@group
(-zip-with #'+ '(1 2 3 4) '(5 6 7))
    @result{} (6 8 10)
@end group
@group
(-zip-with #'cons '(1 2 3) '(4 5 6 7))
    @result{} ((1 . 4) (2 . 5) (3 . 6))
@end group
@group
(--zip-with (format "%s & %s" it other) '(Batman Jekyll) '(Robin Hyde))
    @result{} ("Batman & Robin" "Jekyll & Hyde")
@end group
@end example
@end defun

@anchor{-zip-pair}
@defun -zip-pair (list1 list2)
Zip @var{list1} and @var{list2} together.

Make a pair with the head of each list, followed by a pair with
the second element of each list, and so on.  The number of pairs
returned is equal to the length of the shorter input list.

See also: @code{-zip-lists} (@pxref{-zip-lists}).

@example
@group
(-zip-pair '(1 2 3 4) '(5 6 7))
    @result{} ((1 . 5) (2 . 6) (3 . 7))
@end group
@group
(-zip-pair '(1 2 3) '(4 5 6))
    @result{} ((1 . 4) (2 . 5) (3 . 6))
@end group
@group
(-zip-pair '(1 2) '(3))
    @result{} ((1 . 3))
@end group
@end example
@end defun

@anchor{-zip-lists}
@defun -zip-lists (&rest lists)
Zip @var{lists} together.

Group the head of each list, followed by the second element of
each list, and so on.  The number of returned groupings is equal
to the length of the shortest input list, and the length of each
grouping is equal to the number of input @var{lists}.

The return value is always a list of proper lists, in contrast to
@code{-zip} (@pxref{-zip}) which returns a list of dotted pairs when only two input
@var{lists} are provided.

See also: @code{-zip-pair} (@pxref{-zip-pair}).

@example
@group
(-zip-lists '(1 2 3) '(4 5 6))
    @result{} ((1 4) (2 5) (3 6))
@end group
@group
(-zip-lists '(1 2 3) '(4 5 6 7))
    @result{} ((1 4) (2 5) (3 6))
@end group
@group
(-zip-lists '(1 2) '(3 4 5) '(6))
    @result{} ((1 3 6))
@end group
@end example
@end defun

@anchor{-zip-lists-fill}
@defun -zip-lists-fill (fill-value &rest lists)
Zip @var{lists} together, padding shorter lists with @var{fill-value}.
This is like @code{-zip-lists} (@pxref{-zip-lists}) (which see), except it retains all
elements at positions beyond the end of the shortest list.  The
number of returned groupings is equal to the length of the
longest input list, and the length of each grouping is equal to
the number of input @var{lists}.

@example
@group
(-zip-lists-fill 0 '(1 2) '(3 4 5) '(6))
    @result{} ((1 3 6) (2 4 0) (0 5 0))
@end group
@group
(-zip-lists-fill 0 '(1 2) '(3 4) '(5 6))
    @result{} ((1 3 5) (2 4 6))
@end group
@group
(-zip-lists-fill 0 '(1 2 3) nil)
    @result{} ((1 0) (2 0) (3 0))
@end group
@end example
@end defun

@anchor{-zip}
@defun -zip (&rest lists)
Zip @var{lists} together.

Group the head of each list, followed by the second element of
each list, and so on.  The number of returned groupings is equal
to the length of the shortest input list, and the number of items
in each grouping is equal to the number of input @var{lists}.

If only two @var{lists} are provided as arguments, return the groupings
as a list of dotted pairs.  Otherwise, return the groupings as a
list of proper lists.

Since the return value changes form depending on the number of
arguments, it is generally recommended to use @code{-zip-lists} (@pxref{-zip-lists})
instead, or @code{-zip-pair} (@pxref{-zip-pair}) if a list of dotted pairs is desired.

See also: @code{-unzip} (@pxref{-unzip}).

@example
@group
(-zip '(1 2 3 4) '(5 6 7) '(8 9))
    @result{} ((1 5 8) (2 6 9))
@end group
@group
(-zip '(1 2 3) '(4 5 6) '(7 8 9))
    @result{} ((1 4 7) (2 5 8) (3 6 9))
@end group
@group
(-zip '(1 2 3))
    @result{} ((1) (2) (3))
@end group
@end example
@end defun

@anchor{-zip-fill}
@defun -zip-fill (fill-value &rest lists)
Zip @var{lists} together, padding shorter lists with @var{fill-value}.
This is like @code{-zip} (@pxref{-zip}) (which see), except it retains all elements
at positions beyond the end of the shortest list.  The number of
returned groupings is equal to the length of the longest input
list, and the length of each grouping is equal to the number of
input @var{lists}.

Since the return value changes form depending on the number of
arguments, it is generally recommended to use @code{-zip-lists-fill} (@pxref{-zip-lists-fill})
instead, unless a list of dotted pairs is explicitly desired.

@example
@group
(-zip-fill 0 '(1 2 3) '(4 5))
    @result{} ((1 . 4) (2 . 5) (3 . 0))
@end group
@group
(-zip-fill 0 () '(1 2 3))
    @result{} ((0 . 1) (0 . 2) (0 . 3))
@end group
@group
(-zip-fill 0 '(1 2) '(3 4) '(5 6))
    @result{} ((1 3 5) (2 4 6))
@end group
@end example
@end defun

@anchor{-unzip-lists}
@defun -unzip-lists (lists)
Unzip @var{lists}.

This works just like @code{-zip-lists} (@pxref{-zip-lists}) (which see), but takes a list
of lists instead of a variable number of arguments, such that

  (-unzip-lists (-zip-lists @var{args}@dots{}))

is identity (given that the lists comprising @var{args} are of the same
length).

@example
@group
(-unzip-lists (-zip-lists '(1 2) '(3 4) '(5 6)))
    @result{} ((1 2) (3 4) (5 6))
@end group
@group
(-unzip-lists '((1 2 3) (4 5) (6 7) (8 9)))
    @result{} ((1 4 6 8) (2 5 7 9))
@end group
@group
(-unzip-lists '((1 2 3) (4 5 6)))
    @result{} ((1 4) (2 5) (3 6))
@end group
@end example
@end defun

@anchor{-unzip}
@defun -unzip (lists)
Unzip @var{lists}.

This works just like @code{-zip} (@pxref{-zip}) (which see), but takes a list of
lists instead of a variable number of arguments, such that

  (-unzip (-zip @var{l1} @var{l2} @var{l3} @dots{}))

is identity (given that the lists are of the same length, and
that @code{-zip} (@pxref{-zip}) is not called with two arguments, because of the
caveat described in its docstring).

Note in particular that calling @code{-unzip} (@pxref{-unzip}) on a list of two lists
will return a list of dotted pairs.

Since the return value changes form depending on the number of
@var{lists}, it is generally recommended to use @code{-unzip-lists} (@pxref{-unzip-lists}) instead.

@example
@group
(-unzip (-zip '(1 2) '(3 4) '(5 6)))
    @result{} ((1 . 2) (3 . 4) (5 . 6))
@end group
@group
(-unzip '((1 2 3) (4 5 6)))
    @result{} ((1 . 4) (2 . 5) (3 . 6))
@end group
@group
(-unzip '((1 2 3) (4 5) (6 7) (8 9)))
    @result{} ((1 4 6 8) (2 5 7 9))
@end group
@end example
@end defun

@anchor{-pad}
@defun -pad (fill-value &rest lists)
Pad each of @var{lists} with @var{fill-value} until they all have equal lengths.

Ensure all @var{lists} are as long as the longest one by repeatedly
appending @var{fill-value} to the shorter lists, and return the
resulting @var{lists}.

@example
@group
(-pad 0 ())
    @result{} (nil)
@end group
@group
(-pad 0 '(1 2) '(3 4))
    @result{} ((1 2) (3 4))
@end group
@group
(-pad 0 '(1 2) '(3 4 5 6) '(7 8 9))
    @result{} ((1 2 0 0) (3 4 5 6) (7 8 9 0))
@end group
@end example
@end defun

@anchor{-table}
@defun -table (fn &rest lists)
Compute outer product of @var{lists} using function @var{fn}.

The function @var{fn} should have the same arity as the number of
supplied lists.

The outer product is computed by applying fn to all possible
combinations created by taking one element from each list in
order.  The dimension of the result is (length lists).

See also: @code{-table-flat} (@pxref{-table-flat})

@example
@group
(-table '* '(1 2 3) '(1 2 3))
    @result{} ((1 2 3) (2 4 6) (3 6 9))
@end group
@group
(-table (lambda (a b) (-sum (-zip-with '* a b))) '((1 2) (3 4)) '((1 3) (2 4)))
    @result{} ((7 15) (10 22))
@end group
@group
(apply '-table 'list (-repeat 3 '(1 2)))
    @result{} ((((1 1 1) (2 1 1)) ((1 2 1) (2 2 1))) (((1 1 2) (2 1 2)) ((1 2 2) (2 2 2))))
@end group
@end example
@end defun

@anchor{-table-flat}
@defun -table-flat (fn &rest lists)
Compute flat outer product of @var{lists} using function @var{fn}.

The function @var{fn} should have the same arity as the number of
supplied lists.

The outer product is computed by applying fn to all possible
combinations created by taking one element from each list in
order.  The results are flattened, ignoring the tensor structure
of the result.  This is equivalent to calling:

  (-flatten-n (1- (length lists)) (apply '-table fn lists))

but the implementation here is much more efficient.

See also: @code{-flatten-n} (@pxref{-flatten-n}), @code{-table} (@pxref{-table})

@example
@group
(-table-flat 'list '(1 2 3) '(a b c))
    @result{} ((1 a) (2 a) (3 a) (1 b) (2 b) (3 b) (1 c) (2 c) (3 c))
@end group
@group
(-table-flat '* '(1 2 3) '(1 2 3))
    @result{} (1 2 3 2 4 6 3 6 9)
@end group
@group
(apply '-table-flat 'list (-repeat 3 '(1 2)))
    @result{} ((1 1 1) (2 1 1) (1 2 1) (2 2 1) (1 1 2) (2 1 2) (1 2 2) (2 2 2))
@end group
@end example
@end defun

@anchor{-first}
@defun -first (pred list)
Return the first item in @var{list} for which @var{pred} returns non-@code{nil}.
Return @code{nil} if no such element is found.

To get the first item in the list no questions asked,
use @code{-first-item} (@pxref{-first-item}).

Alias: @code{-find}.

This function's anaphoric counterpart is @code{--first}.

@example
@group
(-first #'natnump '(-1 0 1))
    @result{} 0
@end group
@group
(-first #'null '(1 2 3))
    @result{} nil
@end group
@group
(--first (> it 2) '(1 2 3))
    @result{} 3
@end group
@end example
@end defun

@anchor{-last}
@defun -last (pred list)
Return the last x in @var{list} where (@var{pred} x) is non-@code{nil}, else @code{nil}.

@example
@group
(-last 'even? '(1 2 3 4 5 6 3 3 3))
    @result{} 6
@end group
@group
(-last 'even? '(1 3 7 5 9))
    @result{} nil
@end group
@group
(--last (> (length it) 3) '("a" "looong" "word" "and" "short" "one"))
    @result{} "short"
@end group
@end example
@end defun

@anchor{-first-item}
@defun -first-item (list)
Return the first item of @var{list}, or @code{nil} on an empty list.

See also: @code{-second-item} (@pxref{-second-item}), @code{-last-item} (@pxref{-last-item}), etc.

@example
@group
(-first-item ())
    @result{} ()
@end group
@group
(-first-item '(1 2 3 4 5))
    @result{} 1
@end group
@group
(let ((list (list 1 2 3))) (setf (-first-item list) 5) list)
    @result{} (5 2 3)
@end group
@end example
@end defun

@anchor{-second-item}
@defun -second-item (list)
Return the second item of @var{list}, or @code{nil} if @var{list} is too short.

See also: @code{-first-item} (@pxref{-first-item}), @code{-third-item} (@pxref{-third-item}), etc.

@example
@group
(-second-item ())
    @result{} ()
@end group
@group
(-second-item '(1 2 3 4 5))
    @result{} 2
@end group
@group
(let ((list (list 1 2))) (setf (-second-item list) 5) list)
    @result{} (1 5)
@end group
@end example
@end defun

@anchor{-third-item}
@defun -third-item (list)
Return the third item of @var{list}, or @code{nil} if @var{list} is too short.

See also: @code{-second-item} (@pxref{-second-item}), @code{-fourth-item} (@pxref{-fourth-item}), etc.

@example
@group
(-third-item ())
    @result{} ()
@end group
@group
(-third-item '(1 2))
    @result{} ()
@end group
@group
(-third-item '(1 2 3 4 5))
    @result{} 3
@end group
@end example
@end defun

@anchor{-fourth-item}
@defun -fourth-item (list)
Return the fourth item of @var{list}, or @code{nil} if @var{list} is too short.

See also: @code{-third-item} (@pxref{-third-item}), @code{-fifth-item} (@pxref{-fifth-item}), etc.

@example
@group
(-fourth-item ())
    @result{} ()
@end group
@group
(-fourth-item '(1 2 3))
    @result{} ()
@end group
@group
(-fourth-item '(1 2 3 4 5))
    @result{} 4
@end group
@end example
@end defun

@anchor{-fifth-item}
@defun -fifth-item (list)
Return the fifth item of @var{list}, or @code{nil} if @var{list} is too short.

See also: @code{-fourth-item} (@pxref{-fourth-item}), @code{-last-item} (@pxref{-last-item}), etc.

@example
@group
(-fifth-item ())
    @result{} ()
@end group
@group
(-fifth-item '(1 2 3 4))
    @result{} ()
@end group
@group
(-fifth-item '(1 2 3 4 5))
    @result{} 5
@end group
@end example
@end defun

@anchor{-last-item}
@defun -last-item (list)
Return the last item of @var{list}, or @code{nil} on an empty list.

See also: @code{-first-item} (@pxref{-first-item}), etc.

@example
@group
(-last-item ())
    @result{} ()
@end group
@group
(-last-item '(1 2 3 4 5))
    @result{} 5
@end group
@group
(let ((list (list 1 2 3))) (setf (-last-item list) 5) list)
    @result{} (1 2 5)
@end group
@end example
@end defun

@anchor{-butlast}
@defun -butlast (list)
Return a list of all items in list except for the last.

@example
@group
(-butlast '(1 2 3))
    @result{} (1 2)
@end group
@group
(-butlast '(1 2))
    @result{} (1)
@end group
@group
(-butlast '(1))
    @result{} nil
@end group
@end example
@end defun

@anchor{-sort}
@defun -sort (comparator list)
Sort @var{list}, stably, comparing elements using @var{comparator}.
Return the sorted list.  @var{list} is @var{not} modified by side effects.
@var{comparator} is called with two elements of @var{list}, and should return non-@code{nil}
if the first element should sort before the second.

@example
@group
(-sort #'< '(3 1 2))
    @result{} (1 2 3)
@end group
@group
(-sort #'> '(3 1 2))
    @result{} (3 2 1)
@end group
@group
(--sort (< it other) '(3 1 2))
    @result{} (1 2 3)
@end group
@end example
@end defun

@anchor{-list}
@defun -list (arg)
Ensure @var{arg} is a list.
If @var{arg} is already a list, return it as is (not a copy).
Otherwise, return a new list with @var{arg} as its only element.

Another supported calling convention is (-list &rest @var{args}).
In this case, if @var{arg} is not a list, a new list with all of
@var{args} as elements is returned.  This use is supported for
backward compatibility and is otherwise deprecated.

@example
@group
(-list 1)
    @result{} (1)
@end group
@group
(-list ())
    @result{} ()
@end group
@group
(-list '(1 2 3))
    @result{} (1 2 3)
@end group
@end example
@end defun

@anchor{-fix}
@defun -fix (fn list)
Compute the (least) fixpoint of @var{fn} with initial input @var{list}.

@var{fn} is called at least once, results are compared with @code{equal}.

@example
@group
(-fix (lambda (l) (-non-nil (--mapcat (-split-at (/ (length it) 2) it) l))) '((1 2 3)))
    @result{} ((1) (2) (3))
@end group
@group
(let ((l '((starwars scifi) (jedi starwars warrior)))) (--fix (-uniq (--mapcat (cons it (cdr (assq it l))) it)) '(jedi book)))
    @result{} (jedi starwars warrior scifi book)
@end group
@end example
@end defun

@node Tree operations
@section Tree operations

Functions pretending lists are trees.

@anchor{-tree-seq}
@defun -tree-seq (branch children tree)
Return a sequence of the nodes in @var{tree}, in depth-first search order.

@var{branch} is a predicate of one argument that returns non-@code{nil} if the
passed argument is a branch, that is, a node that can have children.

@var{children} is a function of one argument that returns the children
of the passed branch node.

Non-branch nodes are simply copied.

@example
@group
(-tree-seq 'listp 'identity '(1 (2 3) 4 (5 (6 7))))
    @result{} ((1 (2 3) 4 (5 (6 7))) 1 (2 3) 2 3 4 (5 (6 7)) 5 (6 7) 6 7)
@end group
@group
(-tree-seq 'listp 'reverse '(1 (2 3) 4 (5 (6 7))))
    @result{} ((1 (2 3) 4 (5 (6 7))) (5 (6 7)) (6 7) 7 6 5 4 (2 3) 3 2 1)
@end group
@group
(--tree-seq (vectorp it) (append it nil) [1 [2 3] 4 [5 [6 7]]])
    @result{} ([1 [2 3] 4 [5 [6 7]]] 1 [2 3] 2 3 4 [5 [6 7]] 5 [6 7] 6 7)
@end group
@end example
@end defun

@anchor{-tree-map}
@defun -tree-map (fn tree)
Apply @var{fn} to each element of @var{tree} while preserving the tree structure.

@example
@group
(-tree-map '1+ '(1 (2 3) (4 (5 6) 7)))
    @result{} (2 (3 4) (5 (6 7) 8))
@end group
@group
(-tree-map '(lambda (x) (cons x (expt 2 x))) '(1 (2 3) 4))
    @result{} ((1 . 2) ((2 . 4) (3 . 8)) (4 . 16))
@end group
@group
(--tree-map (length it) '("<body>" ("<p>" "text" "</p>") "</body>"))
    @result{} (6 (3 4 4) 7)
@end group
@end example
@end defun

@anchor{-tree-map-nodes}
@defun -tree-map-nodes (pred fun tree)
Call @var{fun} on each node of @var{tree} that satisfies @var{pred}.

If @var{pred} returns @code{nil}, continue descending down this node.  If @var{pred}
returns non-@code{nil}, apply @var{fun} to this node and do not descend
further.

@example
@group
(-tree-map-nodes 'vectorp (lambda (x) (-sum (append x nil))) '(1 [2 3] 4 (5 [6 7] 8)))
    @result{} (1 5 4 (5 13 8))
@end group
@group
(-tree-map-nodes 'keywordp (lambda (x) (symbol-name x)) '(1 :foo 4 ((5 6 :bar) :baz 8)))
    @result{} (1 ":foo" 4 ((5 6 ":bar") ":baz" 8))
@end group
@group
(--tree-map-nodes (eq (car-safe it) 'add-mode) (-concat it (list :mode 'emacs-lisp-mode)) '(with-mode emacs-lisp-mode (foo bar) (add-mode a b) (baz (add-mode c d))))
    @result{} (with-mode emacs-lisp-mode (foo bar) (add-mode a b :mode emacs-lisp-mode) (baz (add-mode c d :mode emacs-lisp-mode)))
@end group
@end example
@end defun

@anchor{-tree-reduce}
@defun -tree-reduce (fn tree)
Use @var{fn} to reduce elements of list @var{tree}.
If elements of @var{tree} are lists themselves, apply the reduction recursively.

@var{fn} is first applied to first element of the list and second
element, then on this result and third element from the list etc.

See @code{-reduce-r} (@pxref{-reduce-r}) for how exactly are lists of zero or one element handled.

@example
@group
(-tree-reduce '+ '(1 (2 3) (4 5)))
    @result{} 15
@end group
@group
(-tree-reduce 'concat '("strings" (" on" " various") ((" levels"))))
    @result{} "strings on various levels"
@end group
@group
(--tree-reduce (cond ((stringp it) (concat it " " acc)) (t (let ((sn (symbol-name it))) (concat "<" sn ">" acc "</" sn ">")))) '(body (p "some words") (div "more" (b "bold") "words")))
    @result{} "<body><p>some words</p> <div>more <b>bold</b> words</div></body>"
@end group
@end example
@end defun

@anchor{-tree-reduce-from}
@defun -tree-reduce-from (fn init-value tree)
Use @var{fn} to reduce elements of list @var{tree}.
If elements of @var{tree} are lists themselves, apply the reduction recursively.

@var{fn} is first applied to @var{init-value} and first element of the list,
then on this result and second element from the list etc.

The initial value is ignored on cons pairs as they always contain
two elements.

@example
@group
(-tree-reduce-from '+ 1 '(1 (1 1) ((1))))
    @result{} 8
@end group
@group
(--tree-reduce-from (-concat acc (list it)) nil '(1 (2 3 (4 5)) (6 7)))
    @result{} ((7 6) ((5 4) 3 2) 1)
@end group
@end example
@end defun

@anchor{-tree-mapreduce}
@defun -tree-mapreduce (fn folder tree)
Apply @var{fn} to each element of @var{tree}, and make a list of the results.
If elements of @var{tree} are lists themselves, apply @var{fn} recursively to
elements of these nested lists.

Then reduce the resulting lists using @var{folder} and initial value
@var{init-value}. See @code{-reduce-r-from} (@pxref{-reduce-r-from}).

This is the same as calling @code{-tree-reduce} (@pxref{-tree-reduce}) after @code{-tree-map} (@pxref{-tree-map})
but is twice as fast as it only traverse the structure once.

@example
@group
(-tree-mapreduce 'list 'append '(1 (2 (3 4) (5 6)) (7 (8 9))))
    @result{} (1 2 3 4 5 6 7 8 9)
@end group
@group
(--tree-mapreduce 1 (+ it acc) '(1 (2 (4 9) (2 1)) (7 (4 3))))
    @result{} 9
@end group
@group
(--tree-mapreduce 0 (max acc (1+ it)) '(1 (2 (4 9) (2 1)) (7 (4 3))))
    @result{} 3
@end group
@end example
@end defun

@anchor{-tree-mapreduce-from}
@defun -tree-mapreduce-from (fn folder init-value tree)
Apply @var{fn} to each element of @var{tree}, and make a list of the results.
If elements of @var{tree} are lists themselves, apply @var{fn} recursively to
elements of these nested lists.

Then reduce the resulting lists using @var{folder} and initial value
@var{init-value}. See @code{-reduce-r-from} (@pxref{-reduce-r-from}).

This is the same as calling @code{-tree-reduce-from} (@pxref{-tree-reduce-from}) after @code{-tree-map} (@pxref{-tree-map})
but is twice as fast as it only traverse the structure once.

@example
@group
(-tree-mapreduce-from 'identity '* 1 '(1 (2 (3 4) (5 6)) (7 (8 9))))
    @result{} 362880
@end group
@group
(--tree-mapreduce-from (+ it it) (cons it acc) nil '(1 (2 (4 9) (2 1)) (7 (4 3))))
    @result{} (2 (4 (8 18) (4 2)) (14 (8 6)))
@end group
@group
(concat "@{" (--tree-mapreduce-from (cond ((-cons-pair? it) (concat (symbol-name (car it)) " -> " (symbol-name (cdr it)))) (t (concat (symbol-name it) " : @{"))) (concat it (unless (or (equal acc "@}") (equal (substring it (1- (length it))) "@{")) ", ") acc) "@}" '((elisp-mode (foo (bar . booze)) (baz . qux)) (c-mode (foo . bla) (bum . bam)))))
    @result{} "@{elisp-mode : @{foo : @{bar -> booze@}, baz -> qux@}, c-mode : @{foo -> bla, bum -> bam@}@}"
@end group
@end example
@end defun

@anchor{-clone}
@defun -clone (list)
Create a deep copy of @var{list}.
The new list has the same elements and structure but all cons are
replaced with new ones.  This is useful when you need to clone a
structure such as plist or alist.

@example
@group
(let* ((a (list (list 1))) (b (-clone a))) (setcar (car a) 2) b)
    @result{} ((1))
@end group
@end example
@end defun

@node Threading macros
@section Threading macros

Macros that conditionally combine sequential forms for brevity
or readability.

@anchor{->}
@defmac -> (x &optional form &rest more)
Thread the expr through the forms. Insert @var{x} as the second item
in the first form, making a list of it if it is not a list
already. If there are more forms, insert the first form as the
second item in second form, etc.

@example
@group
(-> '(2 3 5))
    @result{} (2 3 5)
@end group
@group
(-> '(2 3 5) (append '(8 13)))
    @result{} (2 3 5 8 13)
@end group
@group
(-> '(2 3 5) (append '(8 13)) (-slice 1 -1))
    @result{} (3 5 8)
@end group
@end example
@end defmac

@anchor{->>}
@defmac ->> (x &optional form &rest more)
Thread the expr through the forms. Insert @var{x} as the last item
in the first form, making a list of it if it is not a list
already. If there are more forms, insert the first form as the
last item in second form, etc.

@example
@group
(->> '(1 2 3) (-map 'square))
    @result{} (1 4 9)
@end group
@group
(->> '(1 2 3) (-map 'square) (-remove 'even?))
    @result{} (1 9)
@end group
@group
(->> '(1 2 3) (-map 'square) (-reduce '+))
    @result{} 14
@end group
@end example
@end defmac

@anchor{-->}
@defmac --> (x &rest forms)
Starting with the value of @var{x}, thread each expression through @var{forms}.

Insert @var{x} at the position signified by the symbol @code{it} in the first
form.  If there are more forms, insert the first form at the position
signified by @code{it} in the second form, etc.

@example
@group
(--> "def" (concat "abc" it "ghi"))
    @result{} "abcdefghi"
@end group
@group
(--> "def" (concat "abc" it "ghi") (upcase it))
    @result{} "ABCDEFGHI"
@end group
@group
(--> "def" (concat "abc" it "ghi") upcase)
    @result{} "ABCDEFGHI"
@end group
@end example
@end defmac

@anchor{-as->}
@defmac -as-> (value variable &rest forms)
Starting with @var{value}, thread @var{variable} through @var{forms}.

In the first form, bind @var{variable} to @var{value}.  In the second form, bind
@var{variable} to the result of the first form, and so forth.

@example
@group
(-as-> 3 my-var (1+ my-var) (list my-var) (mapcar (lambda (ele) (* 2 ele)) my-var))
    @result{} (8)
@end group
@group
(-as-> 3 my-var 1+)
    @result{} 4
@end group
@group
(-as-> 3 my-var)
    @result{} 3
@end group
@end example
@end defmac

@anchor{-some->}
@defmac -some-> (x &optional form &rest more)
When expr is non-@code{nil}, thread it through the first form (via @code{->} (@pxref{->})),
and when that result is non-@code{nil}, through the next form, etc.

@example
@group
(-some-> '(2 3 5))
    @result{} (2 3 5)
@end group
@group
(-some-> 5 square)
    @result{} 25
@end group
@group
(-some-> 5 even? square)
    @result{} nil
@end group
@end example
@end defmac

@anchor{-some->>}
@defmac -some->> (x &optional form &rest more)
When expr is non-@code{nil}, thread it through the first form (via @code{->>} (@pxref{->>})),
and when that result is non-@code{nil}, through the next form, etc.

@example
@group
(-some->> '(1 2 3) (-map 'square))
    @result{} (1 4 9)
@end group
@group
(-some->> '(1 3 5) (-last 'even?) (+ 100))
    @result{} nil
@end group
@group
(-some->> '(2 4 6) (-last 'even?) (+ 100))
    @result{} 106
@end group
@end example
@end defmac

@anchor{-some-->}
@defmac -some--> (expr &rest forms)
Thread @var{expr} through @var{forms} via @code{-->} (@pxref{-->}), while the result is non-@code{nil}.
When @var{expr} evaluates to non-@code{nil}, thread the result through the
first of @var{forms}, and when that result is non-@code{nil}, thread it
through the next form, etc.

@example
@group
(-some--> "def" (concat "abc" it "ghi"))
    @result{} "abcdefghi"
@end group
@group
(-some--> nil (concat "abc" it "ghi"))
    @result{} nil
@end group
@group
(-some--> '(0 1) (-remove #'natnump it) (append it it) (-map #'1+ it))
    @result{} ()
@end group
@end example
@end defmac

@anchor{-doto}
@defmac -doto (init &rest forms)
Evaluate @var{init} and pass it as argument to @var{forms} with @code{->} (@pxref{->}).
The @var{result} of evaluating @var{init} is threaded through each of @var{forms}
individually using @code{->} (@pxref{->}), which see.  The return value is @var{result},
which @var{forms} may have modified by side effect.

@example
@group
(-doto (list 1 2 3) pop pop)
    @result{} (3)
@end group
@group
(-doto (cons 1 2) (setcar 3) (setcdr 4))
    @result{} (3 . 4)
@end group
@group
(gethash 'k (--doto (make-hash-table) (puthash 'k 'v it)))
    @result{} v
@end group
@end example
@end defmac

@node Binding
@section Binding

Macros that combine @code{let} and @code{let*} with destructuring and flow control.

@anchor{-when-let}
@defmac -when-let ((var val) &rest body)
If @var{val} evaluates to non-@code{nil}, bind it to @var{var} and execute body.

Note: binding is done according to @code{-let} (@pxref{-let}).

@example
@group
(-when-let (match-index (string-match "d" "abcd")) (+ match-index 2))
    @result{} 5
@end group
@group
(-when-let ((&plist :foo foo) (list :foo "foo")) foo)
    @result{} "foo"
@end group
@group
(-when-let ((&plist :foo foo) (list :bar "bar")) foo)
    @result{} nil
@end group
@end example
@end defmac

@anchor{-when-let*}
@defmac -when-let* (vars-vals &rest body)
If all @var{vals} evaluate to true, bind them to their corresponding
@var{vars} and execute body. @var{vars-vals} should be a list of (@var{var} @var{val})
pairs.

Note: binding is done according to @code{-let*} (@pxref{-let*}).  @var{vals} are evaluated
sequentially, and evaluation stops after the first @code{nil} @var{val} is
encountered.

@example
@group
(-when-let* ((x 5) (y 3) (z (+ y 4))) (+ x y z))
    @result{} 15
@end group
@group
(-when-let* ((x 5) (y nil) (z 7)) (+ x y z))
    @result{} nil
@end group
@end example
@end defmac

@anchor{-if-let}
@defmac -if-let ((var val) then &rest else)
If @var{val} evaluates to non-@code{nil}, bind it to @var{var} and do @var{then},
otherwise do @var{else}.

Note: binding is done according to @code{-let} (@pxref{-let}).

@example
@group
(-if-let (match-index (string-match "d" "abc")) (+ match-index 3) 7)
    @result{} 7
@end group
@group
(--if-let (even? 4) it nil)
    @result{} t
@end group
@end example
@end defmac

@anchor{-if-let*}
@defmac -if-let* (vars-vals then &rest else)
If all @var{vals} evaluate to true, bind them to their corresponding
@var{vars} and do @var{then}, otherwise do @var{else}. @var{vars-vals} should be a list
of (@var{var} @var{val}) pairs.

Note: binding is done according to @code{-let*} (@pxref{-let*}).  @var{vals} are evaluated
sequentially, and evaluation stops after the first @code{nil} @var{val} is
encountered.

@example
@group
(-if-let* ((x 5) (y 3) (z 7)) (+ x y z) "foo")
    @result{} 15
@end group
@group
(-if-let* ((x 5) (y nil) (z 7)) (+ x y z) "foo")
    @result{} "foo"
@end group
@group
(-if-let* (((_ _ x) '(nil nil 7))) x)
    @result{} 7
@end group
@end example
@end defmac

@anchor{-let}
@defmac -let (varlist &rest body)
Bind variables according to @var{varlist} then eval @var{body}.

@var{varlist} is a list of lists of the form (@var{pattern} @var{source}).  Each
@var{pattern} is matched against the @var{source} "structurally".  @var{source}
is only evaluated once for each @var{pattern}.  Each @var{pattern} is matched
recursively, and can therefore contain sub-patterns which are
matched against corresponding sub-expressions of @var{source}.

All the SOURCEs are evalled before any symbols are
bound (i.e. "in parallel").

If @var{varlist} only contains one (@var{pattern} @var{source}) element, you can
optionally specify it using a vector and discarding the
outer-most parens.  Thus

  (-let ((@var{pattern} @var{source})) @dots{})

becomes

  (-let [@var{pattern} @var{source}] @dots{}).

@code{-let} (@pxref{-let}) uses a convention of not binding places (symbols) starting
with _ whenever it's possible.  You can use this to skip over
entries you don't care about.  However, this is not *always*
possible (as a result of implementation) and these symbols might
get bound to undefined values.

Following is the overview of supported patterns.  Remember that
patterns can be matched recursively, so every a, b, aK in the
following can be a matching construct and not necessarily a
symbol/variable.

Symbol:

  a - bind the @var{source} to @var{a}.  This is just like regular @code{let}.

Conses and lists:

  (a) - bind @code{car} of cons/list to @var{a}

  (a . b) - bind car of cons to @var{a} and @code{cdr} to @var{b}

  (a b) - bind car of list to @var{a} and @code{cadr} to @var{b}

  (a1 a2 a3 @dots{}) - bind 0th car of list to @var{a1}, 1st to @var{a2}, 2nd to @var{a3}@enddots{}

  (a1 a2 a3 @dots{} aN . rest) - as above, but bind the Nth cdr to @var{rest}.

Vectors:

  [a] - bind 0th element of a non-list sequence to @var{a} (works with
        vectors, strings, bit arrays@dots{})

  [a1 a2 a3 @dots{}] - bind 0th element of non-list sequence to @var{a0}, 1st to
                   @var{a1}, 2nd to @var{a2}, @enddots{}
                   If the @var{pattern} is shorter than @var{source}, the values at
                   places not in @var{pattern} are ignored.
                   If the @var{pattern} is longer than @var{source}, an @code{error} is
                   thrown.

  [a1 a2 a3 @dots{} &rest rest] - as above, but bind the rest of
                              the sequence to @var{rest}.  This is
                              conceptually the same as improper list
                              matching (a1 a2 @dots{} aN . rest)

Key/value stores:

  (&plist key0 a0 @dots{} keyN aN) - bind value mapped by keyK in the
                                 @var{source} plist to aK.  If the
                                 value is not found, aK is @code{nil}.
                                 Uses @code{plist-get} to fetch values.

  (&alist key0 a0 @dots{} keyN aN) - bind value mapped by keyK in the
                                 @var{source} alist to aK.  If the
                                 value is not found, aK is @code{nil}.
                                 Uses @code{assoc} to fetch values.

  (&hash key0 a0 @dots{} keyN aN) - bind value mapped by keyK in the
                                @var{source} hash table to aK.  If the
                                value is not found, aK is @code{nil}.
                                Uses @code{gethash} to fetch values.

Further, special keyword &keys supports "inline" matching of
plist-like key-value pairs, similarly to &keys keyword of
@code{cl-defun}.

  (a1 a2 @dots{} aN &keys key1 b1 @dots{} keyN bK)

This binds @var{n} values from the list to a1 @dots{} aN, then interprets
the cdr as a plist (see key/value matching above).

@var{a} shorthand notation for kv-destructuring exists which allows the
patterns be optionally left out and derived from the key name in
the following fashion:

- a key :foo is converted into @code{foo} pattern,
- a key 'bar is converted into @code{bar} pattern,
- a key "baz" is converted into @code{baz} pattern.

That is, the entire value under the key is bound to the derived
variable without any further destructuring.

This is possible only when the form following the key is not a
valid pattern (i.e. not a symbol, a cons cell or a vector).
Otherwise the matching proceeds as usual and in case of an
invalid spec fails with an error.

Thus the patterns are normalized as follows:

   ;; derive all the missing patterns
   (&plist :foo 'bar "baz") => (&plist :foo foo 'bar bar "baz" baz)

   ;; we can specify some but not others
   (&plist :foo 'bar explicit-bar) => (&plist :foo foo 'bar explicit-bar)

   ;; nothing happens, we store :foo in x
   (&plist :foo x) => (&plist :foo x)

   ;; nothing happens, we match recursively
   (&plist :foo (a b c)) => (&plist :foo (a b c))

You can name the source using the syntax @var{symbol} &as @var{pattern}.
This syntax works with lists (proper or improper), vectors and
all types of maps.

  (list &as a b c) (list 1 2 3)

binds @var{a} to 1, @var{b} to 2, @var{c} to 3 and @var{list} to (1 2 3).

Similarly:

  (bounds &as beg . end) (cons 1 2)

binds @var{beg} to 1, @var{end} to 2 and @var{bounds} to (1 . 2).

  (items &as first . rest) (list 1 2 3)

binds @var{first} to 1, @var{rest} to (2 3) and @var{items} to (1 2 3)

  [vect &as _ b c] [1 2 3]

binds @var{b} to 2, @var{c} to 3 and @var{vect} to [1 2 3] (_ avoids binding as usual).

  (plist &as &plist :b b) (list :a 1 :b 2 :c 3)

binds @var{b} to 2 and @var{plist} to (:a 1 :b 2 :c 3).  Same for &alist and &hash.

This is especially useful when we want to capture the result of a
computation and destructure at the same time.  Consider the
form (function-returning-complex-structure) returning a list of
two vectors with two items each.  We want to capture this entire
result and pass it to another computation, but at the same time
we want to get the second item from each vector.  We can achieve
it with pattern

  (result &as [_ a] [_ b]) (function-returning-complex-structure)

Note: Clojure programmers may know this feature as the ":as
binding".  The difference is that we put the &as at the front
because we need to support improper list binding.

@example
@group
(-let (([a (b c) d] [1 (2 3) 4])) (list a b c d))
    @result{} (1 2 3 4)
@end group
@group
(-let [(a b c . d) (list 1 2 3 4 5 6)] (list a b c d))
    @result{} (1 2 3 (4 5 6))
@end group
@group
(-let [(&plist :foo foo :bar bar) (list :baz 3 :foo 1 :qux 4 :bar 2)] (list foo bar))
    @result{} (1 2)
@end group
@end example
@end defmac

@anchor{-let*}
@defmac -let* (varlist &rest body)
Bind variables according to @var{varlist} then eval @var{body}.

@var{varlist} is a list of lists of the form (@var{pattern} @var{source}).  Each
@var{pattern} is matched against the @var{source} structurally.  @var{source} is
only evaluated once for each @var{pattern}.

Each @var{source} can refer to the symbols already bound by this
@var{varlist}.  This is useful if you want to destructure @var{source}
recursively but also want to name the intermediate structures.

See @code{-let} (@pxref{-let}) for the list of all possible patterns.

@example
@group
(-let* (((a . b) (cons 1 2)) ((c . d) (cons 3 4))) (list a b c d))
    @result{} (1 2 3 4)
@end group
@group
(-let* (((a . b) (cons 1 (cons 2 3))) ((c . d) b)) (list a b c d))
    @result{} (1 (2 . 3) 2 3)
@end group
@group
(-let* (((&alist "foo" foo "bar" bar) (list (cons "foo" 1) (cons "bar" (list 'a 'b 'c)))) ((a b c) bar)) (list foo a b c bar))
    @result{} (1 a b c (a b c))
@end group
@end example
@end defmac

@anchor{-lambda}
@defmac -lambda (match-form &rest body)
Return a lambda which destructures its input as @var{match-form} and executes @var{body}.

Note that you have to enclose the @var{match-form} in a pair of parens,
such that:

  (-lambda (x) body)
  (-lambda (x y @dots{}) body)

has the usual semantics of @code{lambda}.  Furthermore, these get
translated into normal @code{lambda}, so there is no performance
penalty.

See @code{-let} (@pxref{-let}) for a description of the destructuring mechanism.

@example
@group
(-map (-lambda ((x y)) (+ x y)) '((1 2) (3 4) (5 6)))
    @result{} (3 7 11)
@end group
@group
(-map (-lambda ([x y]) (+ x y)) '([1 2] [3 4] [5 6]))
    @result{} (3 7 11)
@end group
@group
(funcall (-lambda ((_ . a) (_ . b)) (-concat a b)) '(1 2 3) '(4 5 6))
    @result{} (2 3 5 6)
@end group
@end example
@end defmac

@anchor{-setq}
@defmac -setq ([match-form val] ...)
Bind each @var{match-form} to the value of its @var{val}.

@var{match-form} destructuring is done according to the rules of @code{-let} (@pxref{-let}).

This macro allows you to bind multiple variables by destructuring
the value, so for example:

  (-setq (a b) x
         (&plist :c c) plist)

expands roughly speaking to the following code

  (setq a (car x)
        b (cadr x)
        c (plist-get plist :c))

Care is taken to only evaluate each @var{val} once so that in case of
multiple assignments it does not cause unexpected side effects.

@example
@group
(let (a) (-setq a 1) a)
    @result{} 1
@end group
@group
(let (a b) (-setq (a b) (list 1 2)) (list a b))
    @result{} (1 2)
@end group
@group
(let (c) (-setq (&plist :c c) (list :c "c")) c)
    @result{} "c"
@end group
@end example
@end defmac

@node Side effects
@section Side effects

Functions iterating over lists for side effect only.

@anchor{-each}
@defun -each (list fn)
Call @var{fn} on each element of @var{list}.
Return @code{nil}; this function is intended for side effects.

Its anaphoric counterpart is @code{--each}.

For access to the current element's index in @var{list}, see
@code{-each-indexed} (@pxref{-each-indexed}).

@example
@group
(let (l) (-each '(1 2 3) (lambda (x) (push x l))) l)
    @result{} (3 2 1)
@end group
@group
(let (l) (--each '(1 2 3) (push it l)) l)
    @result{} (3 2 1)
@end group
@group
(-each '(1 2 3) #'identity)
    @result{} nil
@end group
@end example
@end defun

@anchor{-each-while}
@defun -each-while (list pred fn)
Call @var{fn} on each @var{item} in @var{list}, while (@var{pred} @var{item}) is non-@code{nil}.
Once an @var{item} is reached for which @var{pred} returns @code{nil}, @var{fn} is no
longer called.  Return @code{nil}; this function is intended for side
effects.

Its anaphoric counterpart is @code{--each-while}.

@example
@group
(let (l) (-each-while '(2 4 5 6) #'even? (lambda (x) (push x l))) l)
    @result{} (4 2)
@end group
@group
(let (l) (--each-while '(1 2 3 4) (< it 3) (push it l)) l)
    @result{} (2 1)
@end group
@group
(let ((s 0)) (--each-while '(1 3 4 5) (< it 5) (setq s (+ s it))) s)
    @result{} 8
@end group
@end example
@end defun

@anchor{-each-indexed}
@defun -each-indexed (list fn)
Call @var{fn} on each index and element of @var{list}.
For each @var{item} at @var{index} in @var{list}, call (funcall @var{fn} @var{index} @var{item}).
Return @code{nil}; this function is intended for side effects.

See also: @code{-map-indexed} (@pxref{-map-indexed}).

@example
@group
(let (l) (-each-indexed '(a b c) (lambda (i x) (push (list x i) l))) l)
    @result{} ((c 2) (b 1) (a 0))
@end group
@group
(let (l) (--each-indexed '(a b c) (push (list it it-index) l)) l)
    @result{} ((c 2) (b 1) (a 0))
@end group
@group
(let (l) (--each-indexed () (push it l)) l)
    @result{} ()
@end group
@end example
@end defun

@anchor{-each-r}
@defun -each-r (list fn)
Call @var{fn} on each element of @var{list} in reversed order.
Return @code{nil}; this function is intended for side effects.

Its anaphoric counterpart is @code{--each-r}.

@example
@group
(let (l) (-each-r '(1 2 3) (lambda (x) (push x l))) l)
    @result{} (1 2 3)
@end group
@group
(let (l) (--each-r '(1 2 3) (push it l)) l)
    @result{} (1 2 3)
@end group
@group
(-each-r '(1 2 3) #'identity)
    @result{} nil
@end group
@end example
@end defun

@anchor{-each-r-while}
@defun -each-r-while (list pred fn)
Call @var{fn} on each @var{item} in reversed @var{list}, while (@var{pred} @var{item}) is non-@code{nil}.
Once an @var{item} is reached for which @var{pred} returns @code{nil}, @var{fn} is no
longer called.  Return @code{nil}; this function is intended for side
effects.

Its anaphoric counterpart is @code{--each-r-while}.

@example
@group
(let (l) (-each-r-while '(2 4 5 6) #'even? (lambda (x) (push x l))) l)
    @result{} (6)
@end group
@group
(let (l) (--each-r-while '(1 2 3 4) (>= it 3) (push it l)) l)
    @result{} (3 4)
@end group
@group
(let ((s 0)) (--each-r-while '(1 2 3 5) (> it 1) (setq s (+ s it))) s)
    @result{} 10
@end group
@end example
@end defun

@anchor{-dotimes}
@defun -dotimes (num fn)
Call @var{fn} @var{num} times, presumably for side effects.
@var{fn} is called with a single argument on successive integers
running from 0, inclusive, to @var{num}, exclusive.  @var{fn} is not called
if @var{num} is less than 1.

This function's anaphoric counterpart is @code{--dotimes}.

@example
@group
(let (s) (-dotimes 3 (lambda (n) (push n s))) s)
    @result{} (2 1 0)
@end group
@group
(let (s) (-dotimes 0 (lambda (n) (push n s))) s)
    @result{} ()
@end group
@group
(let (s) (--dotimes 5 (push it s)) s)
    @result{} (4 3 2 1 0)
@end group
@end example
@end defun

@node Destructive operations
@section Destructive operations

Macros that modify variables holding lists.

@anchor{!cons}
@defmac !cons (car cdr)
Destructive: Set @var{cdr} to the cons of @var{car} and @var{cdr}.

@example
@group
(let (l) (!cons 5 l) l)
    @result{} (5)
@end group
@group
(let ((l '(3))) (!cons 5 l) l)
    @result{} (5 3)
@end group
@end example
@end defmac

@anchor{!cdr}
@defmac !cdr (list)
Destructive: Set @var{list} to the cdr of @var{list}.

@example
@group
(let ((l '(3))) (!cdr l) l)
    @result{} ()
@end group
@group
(let ((l '(3 5))) (!cdr l) l)
    @result{} (5)
@end group
@end example
@end defmac

@node Function combinators
@section Function combinators

Functions that manipulate and compose other functions.

@anchor{-partial}
@defun -partial (fun &rest args)
Return a function that is a partial application of @var{fun} to @var{args}.
@var{args} is a list of the first @var{n} arguments to pass to @var{fun}.
The result is a new function which does the same as @var{fun}, except that
the first @var{n} arguments are fixed at the values with which this function
was called.

@example
@group
(funcall (-partial #'+ 5))
    @result{} 5
@end group
@group
(funcall (-partial #'- 5) 3)
    @result{} 2
@end group
@group
(funcall (-partial #'+ 5 2) 3)
    @result{} 10
@end group
@end example
@end defun

@anchor{-rpartial}
@defun -rpartial (fn &rest args)
Return a function that is a partial application of @var{fn} to @var{args}.
@var{args} is a list of the last @var{n} arguments to pass to @var{fn}.  The result
is a new function which does the same as @var{fn}, except that the last
@var{n} arguments are fixed at the values with which this function was
called.  This is like @code{-partial} (@pxref{-partial}), except the arguments are fixed
starting from the right rather than the left.

@example
@group
(funcall (-rpartial #'- 5))
    @result{} -5
@end group
@group
(funcall (-rpartial #'- 5) 8)
    @result{} 3
@end group
@group
(funcall (-rpartial #'- 5 2) 10)
    @result{} 3
@end group
@end example
@end defun

@anchor{-juxt}
@defun -juxt (&rest fns)
Return a function that is the juxtaposition of @var{fns}.
The returned function takes a variable number of @var{args}, applies
each of @var{fns} in turn to @var{args}, and returns the list of results.

@example
@group
(funcall (-juxt) 1 2)
    @result{} ()
@end group
@group
(funcall (-juxt #'+ #'- #'* #'/) 7 5)
    @result{} (12 2 35 1)
@end group
@group
(mapcar (-juxt #'number-to-string #'1+) '(1 2))
    @result{} (("1" 2) ("2" 3))
@end group
@end example
@end defun

@anchor{-compose}
@defun -compose (&rest fns)
Compose @var{fns} into a single composite function.
Return a function that takes a variable number of @var{args}, applies
the last function in @var{fns} to @var{args}, and returns the result of
calling each remaining function on the result of the previous
function, right-to-left.  If no @var{fns} are given, return a variadic
@code{identity} function.

@example
@group
(funcall (-compose #'- #'1+ #'+) 1 2 3)
    @result{} -7
@end group
@group
(funcall (-compose #'identity #'1+) 3)
    @result{} 4
@end group
@group
(mapcar (-compose #'not #'stringp) '(nil ""))
    @result{} (t nil)
@end group
@end example
@end defun

@anchor{-applify}
@defun -applify (fn)
Return a function that applies @var{fn} to a single list of args.
This changes the arity of @var{fn} from taking @var{n} distinct arguments to
taking 1 argument which is a list of @var{n} arguments.

@example
@group
(funcall (-applify #'+) nil)
    @result{} 0
@end group
@group
(mapcar (-applify #'+) '((1 1 1) (1 2 3) (5 5 5)))
    @result{} (3 6 15)
@end group
@group
(funcall (-applify #'<) '(3 6))
    @result{} t
@end group
@end example
@end defun

@anchor{-on}
@defun -on (op trans)
Return a function that calls @var{trans} on each arg and @var{op} on the results.
The returned function takes a variable number of arguments, calls
the function @var{trans} on each one in turn, and then passes those
results as the list of arguments to @var{op}, in the same order.

For example, the following pairs of expressions are morally
equivalent:

  (funcall (-on #'+ #'1+) 1 2 3) = (+ (1+ 1) (1+ 2) (1+ 3))
  (funcall (-on #'+ #'1+))       = (+)

@example
@group
(-sort (-on #'< #'length) '((1 2 3) (1) (1 2)))
    @result{} ((1) (1 2) (1 2 3))
@end group
@group
(funcall (-on #'min #'string-to-number) "22" "2" "1" "12")
    @result{} 1
@end group
@group
(-min-by (-on #'> #'length) '((1 2 3) (4) (1 2)))
    @result{} (4)
@end group
@end example
@end defun

@anchor{-flip}
@defun -flip (fn)
Return a function that calls @var{fn} with its arguments reversed.
The returned function takes the same number of arguments as @var{fn}.

For example, the following two expressions are morally
equivalent:

  (funcall (-flip #'-) 1 2) = (- 2 1)

See also: @code{-rotate-args} (@pxref{-rotate-args}).

@example
@group
(-sort (-flip #'<) '(4 3 6 1))
    @result{} (6 4 3 1)
@end group
@group
(funcall (-flip #'-) 3 2 1 10)
    @result{} 4
@end group
@group
(funcall (-flip #'1+) 1)
    @result{} 2
@end group
@end example
@end defun

@anchor{-rotate-args}
@defun -rotate-args (n fn)
Return a function that calls @var{fn} with args rotated @var{n} places to the right.
The returned function takes the same number of arguments as @var{fn},
rotates the list of arguments @var{n} places to the right (left if @var{n} is
negative) just like @code{-rotate} (@pxref{-rotate}), and applies @var{fn} to the result.

See also: @code{-flip} (@pxref{-flip}).

@example
@group
(funcall (-rotate-args -1 #'list) 1 2 3 4)
    @result{} (2 3 4 1)
@end group
@group
(funcall (-rotate-args 1 #'-) 1 10 100)
    @result{} 89
@end group
@group
(funcall (-rotate-args 2 #'list) 3 4 5 1 2)
    @result{} (1 2 3 4 5)
@end group
@end example
@end defun

@anchor{-const}
@defun -const (c)
Return a function that returns @var{c} ignoring any additional arguments.

In types: a -> b -> a

@example
@group
(funcall (-const 2) 1 3 "foo")
    @result{} 2
@end group
@group
(mapcar (-const 1) '("a" "b" "c" "d"))
    @result{} (1 1 1 1)
@end group
@group
(-sum (mapcar (-const 1) '("a" "b" "c" "d")))
    @result{} 4
@end group
@end example
@end defun

@anchor{-cut}
@defmac -cut (&rest params)
Take n-ary function and n arguments and specialize some of them.
Arguments denoted by <> will be left unspecialized.

See @var{srfi-26} for detailed description.

@example
@group
(funcall (-cut list 1 <> 3 <> 5) 2 4)
    @result{} (1 2 3 4 5)
@end group
@group
(-map (-cut funcall <> 5) `(1+ 1- ,(lambda (x) (/ 1.0 x))))
    @result{} (6 4 0.2)
@end group
@group
(-map (-cut <> 1 2 3) '(list vector string))
    @result{} ((1 2 3) [1 2 3] "\1\2\3")
@end group
@end example
@end defmac

@anchor{-not}
@defun -not (pred)
Return a predicate that negates the result of @var{pred}.
The returned predicate passes its arguments to @var{pred}.  If @var{pred}
returns @code{nil}, the result is non-@code{nil}; otherwise the result is @code{nil}.

See also: @code{-andfn} (@pxref{-andfn}) and @code{-orfn} (@pxref{-orfn}).

@example
@group
(funcall (-not #'numberp) "5")
    @result{} t
@end group
@group
(-sort (-not #'<) '(5 2 1 0 6))
    @result{} (6 5 2 1 0)
@end group
@group
(-filter (-not (-partial #'< 4)) '(1 2 3 4 5 6 7 8))
    @result{} (1 2 3 4)
@end group
@end example
@end defun

@anchor{-orfn}
@defun -orfn (&rest preds)
Return a predicate that returns the first non-@code{nil} result of @var{preds}.
The returned predicate takes a variable number of arguments,
passes them to each predicate in @var{preds} in turn until one of them
returns non-@code{nil}, and returns that non-@code{nil} result without calling
the remaining @var{preds}.  If all @var{preds} return @code{nil}, or if no @var{preds} are
given, the returned predicate returns @code{nil}.

See also: @code{-andfn} (@pxref{-andfn}) and @code{-not} (@pxref{-not}).

@example
@group
(-filter (-orfn #'natnump #'booleanp) '(1 nil "a" -4 b c t))
    @result{} (1 nil t)
@end group
@group
(funcall (-orfn #'symbolp (-cut string-match-p "x" <>)) "axe")
    @result{} 1
@end group
@group
(funcall (-orfn #'= #'+) 1 1)
    @result{} t
@end group
@end example
@end defun

@anchor{-andfn}
@defun -andfn (&rest preds)
Return a predicate that returns non-@code{nil} if all @var{preds} do so.
The returned predicate @var{p} takes a variable number of arguments and
passes them to each predicate in @var{preds} in turn.  If any one of
@var{preds} returns @code{nil}, @var{p} also returns @code{nil} without calling the
remaining @var{preds}.  If all @var{preds} return non-@code{nil}, @var{p} returns the last
such value.  If no @var{preds} are given, @var{p} always returns non-@code{nil}.

See also: @code{-orfn} (@pxref{-orfn}) and @code{-not} (@pxref{-not}).

@example
@group
(-filter (-andfn #'numberp (-cut < <> 5)) '(a 1 b 6 c 2))
    @result{} (1 2)
@end group
@group
(mapcar (-andfn #'numberp #'1+) '(a 1 b 6))
    @result{} (nil 2 nil 7)
@end group
@group
(funcall (-andfn #'= #'+) 1 1)
    @result{} 2
@end group
@end example
@end defun

@anchor{-iteratefn}
@defun -iteratefn (fn n)
Return a function @var{fn} composed @var{n} times with itself.

@var{fn} is a unary function.  If you need to use a function of higher
arity, use @code{-applify} (@pxref{-applify}) first to turn it into a unary function.

With n = 0, this acts as identity function.

In types: (a -> a) -> Int -> a -> a.

This function satisfies the following law:

  (funcall (-iteratefn fn n) init) = (-last-item (-iterate fn init (1+ n))).

@example
@group
(funcall (-iteratefn (lambda (x) (* x x)) 3) 2)
    @result{} 256
@end group
@group
(funcall (-iteratefn '1+ 3) 1)
    @result{} 4
@end group
@group
(funcall (-iteratefn 'cdr 3) '(1 2 3 4 5))
    @result{} (4 5)
@end group
@end example
@end defun

@anchor{-fixfn}
@defun -fixfn (fn &optional equal-test halt-test)
Return a function that computes the (least) fixpoint of @var{fn}.

@var{fn} must be a unary function. The returned lambda takes a single
argument, @var{x}, the initial value for the fixpoint iteration. The
iteration halts when either of the following conditions is satisfied:

 1. Iteration converges to the fixpoint, with equality being
    tested using @var{equal-test}. If @var{equal-test} is not specified,
    @code{equal} is used. For functions over the floating point
    numbers, it may be necessary to provide an appropriate
    approximate comparison test.

 2. @var{halt-test} returns a non-@code{nil} value. @var{halt-test} defaults to a
    simple counter that returns @code{t} after @code{-fixfn-max-iterations},
    to guard against infinite iteration. Otherwise, @var{halt-test}
    must be a function that accepts a single argument, the
    current value of @var{x}, and returns non-@code{nil} as long as iteration
    should continue. In this way, a more sophisticated
    convergence test may be supplied by the caller.

The return value of the lambda is either the fixpoint or, if
iteration halted before converging, a cons with car @code{halted} and
cdr the final output from @var{halt-test}.

In types: (a -> a) -> a -> a.

@example
@group
(funcall (-fixfn #'cos #'approx=) 0.7)
    @result{} 0.7390851332151607
@end group
@group
(funcall (-fixfn (lambda (x) (expt (+ x 10) 0.25))) 2.0)
    @result{} 1.8555845286409378
@end group
@group
(funcall (-fixfn #'sin #'approx=) 0.1)
    @result{} (halted . t)
@end group
@end example
@end defun

@anchor{-prodfn}
@defun -prodfn (&rest fns)
Return a function that applies each of @var{fns} to each of a list of arguments.

Takes a list of @var{n} functions and returns a function that takes a
list of length @var{n}, applying Ith function to Ith element of the
input list.  Returns a list of length @var{n}.

In types (for @var{n}=2): ((a -> b), (c -> d)) -> (a, c) -> (b, d)

This function satisfies the following laws:

    (-compose (-prodfn f g @dots{})
              (-prodfn f' g' @dots{}))
  = (-prodfn (-compose f f')
             (-compose g g')
             @dots{})

    (-prodfn f g @dots{})
  = (-juxt (-compose f (-partial #'nth 0))
           (-compose g (-partial #'nth 1))
           @dots{})

    (-compose (-prodfn f g @dots{})
              (-juxt f' g' @dots{}))
  = (-juxt (-compose f f')
           (-compose g g')
           @dots{})

    (-compose (-partial #'nth n)
              (-prod f1 f2 @dots{}))
  = (-compose fn (-partial #'nth n))

@example
@group
(funcall (-prodfn #'1+ #'1- #'number-to-string) '(1 2 3))
    @result{} (2 1 "3")
@end group
@group
(-map (-prodfn #'1- #'1+) '((1 2) (3 4) (5 6)))
    @result{} ((0 3) (2 5) (4 7))
@end group
@group
(apply #'+ (funcall (-prodfn #'length #'string-to-number) '((t) "5")))
    @result{} 6
@end group
@end example
@end defun

@node Development
@chapter Development

The Dash repository is hosted on GitHub at
@url{https://github.com/magnars/dash.el}.

@menu
* Contribute::          How to contribute.
* Contributors::        List of contributors.
@end menu

@node Contribute
@section Contribute

Yes, please do.  Pure functions in the list manipulation realm only,
please.  There's a suite of examples/tests in @file{dev/examples.el},
so remember to add tests for your additions, or they may get broken
later.

Run the tests with @samp{make check}.  Regenerate the docs with
@samp{make docs}.  Contributors are encouraged to install these
commands as a Git pre-commit hook, so that the tests are always
running and the docs are always in sync:

@example
$ cp dev/pre-commit.sh .git/hooks/pre-commit
@end example

Oh, and don't edit @file{README.md} or @file{dash.texi} directly, as
they are auto-generated.  Instead, change their respective templates
@file{readme-template.md} or @file{dash-template.texi}.

To ensure that Dash can be distributed with GNU ELPA or Emacs, we
require that all contributors assign copyright to the Free Software
Foundation.  For more on this, @pxref{Copyright Assignment,,, emacs,
The GNU Emacs Manual}.

@node Contributors
@section Contributors

@itemize
@item
@url{https://github.com/Fuco1, Matus Goljer} contributed lots of
features and functions.
@item
@url{https://github.com/tkf, Takafumi Arakaki} contributed
@code{-group-by}.
@item
@url{https://github.com/tali713, tali713} is the author of
@code{-applify}.
@item
@url{https://github.com/vemv, V@'{i}ctor M. Valenzuela} contributed
@code{-repeat}.
@item
@url{https://github.com/nicferrier, Nic Ferrier} contributed
@code{-cons*}.
@item
@url{https://github.com/Wilfred, Wilfred Hughes} contributed
@code{-slice}, @code{-first-item}, and @code{-last-item}.
@item
@url{https://github.com/shosti, Emanuel Evans} contributed
@code{-if-let}, @code{-when-let}, and @code{-insert-at}.
@item
@url{https://github.com/rejeep, Johan Andersson} contributed
@code{-sum}, @code{-product}, and @code{-same-items?}.
@item
@url{https://github.com/kurisuwhyte, Christina Whyte} contributed
@code{-compose}.
@item
@url{https://github.com/steventlamb, Steve Lamb} contributed
@code{-cycle}, @code{-pad}, @code{-annotate}, @code{-zip-fill}, and a
variadic version of @code{-zip}.
@item
@url{https://github.com/fbergroth, Fredrik Bergroth} made the
@code{-if-let} family use @code{-let} destructuring and improved the
script for generating documentation.
@item
@url{https://github.com/holomorph, Mark Oteiza} contributed
@code{-iota} and the script to create an Info manual.
@item
@url{https://github.com/wasamasa, Vasilij Schneidermann} contributed
@code{-some}.
@item
@url{https://github.com/occidens, William West} made @code{-fixfn}
more robust at handling floats.
@item
@url{https://github.com/camsaul, Cam Saul} contributed @code{-some->},
@code{-some->>}, and @code{-some-->}.
@item
@url{https://github.com/basil-conto, Basil L. Contovounesios}
contributed @code{-common-prefix}, @code{-common-suffix}, and various
other improvements.
@item
@url{https://github.com/doublep, Paul Pogonyshev} contributed
@code{-each-r} and @code{-each-r-while}.
@end itemize

Thanks!

New contributors are very welcome.  @xref{Contribute}.

@c Appendices.

@node FDL
@appendix GNU Free Documentation License
@include doc/fdl.texi

@node GPL
@appendix GNU General Public License
@include doc/gpl.texi

@node Index
@unnumbered Index
@printindex fn

@bye