<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=US-ASCII">
<meta name="generator" content="hevea 2.32">
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1">
<link rel="stylesheet" type="text/css" href="manual.css">
<title>Chapter 4  Labels and variants</title>
</head>
<body>
<a href="objectexamples.html"><img src="previous_motif.svg" alt="Previous"></a>
<a href="index.html"><img src="contents_motif.svg" alt="Up"></a>
<a href="polymorphism.html"><img src="next_motif.svg" alt="Next"></a>
<hr>
<h1 class="chapter" id="sec42">Chapter 4  Labels and variants</h1>
<ul>
<li><a href="lablexamples.html#sec43">4.1  Labels</a>
</li><li><a href="lablexamples.html#sec47">4.2  Polymorphic variants</a>
</li></ul>
<p> <a id="c:labl-examples"></a>
</p><p>
<span class="c009">(Chapter written by Jacques Garrigue)</span></p><p><br>
<br>
</p><p>This chapter gives an overview of the new features in
OCaml 3: labels, and polymorphic variants.</p>
<h2 class="section" id="sec43">4.1  Labels</h2>
<p>If you have a look at modules ending in <span class="c003">Labels</span> in the standard
library, you will see that function types have annotations you did not
have in the functions you defined yourself.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> ListLabels.map;;
</div><div class="caml-output ok">- : f:('a -> 'b) -> 'a list -> 'b list = <fun>
</div></pre>
<pre><div class="caml-input"> StringLabels.sub;;
</div><div class="caml-output ok">- : string -> pos:int -> len:int -> string = <fun>
</div></pre>
</div><p>Such annotations of the form <span class="c003">name:</span> are called <em>labels</em>. They are
meant to document the code, allow more checking, and give more
flexibility to function application.
You can give such names to arguments in your programs, by prefixing them
with a tilde <span class="c003">~</span>.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f ~x ~y = x - y;;
</div><div class="caml-output ok">val f : x:int -> y:int -> int = <fun>
</div></pre>
<pre><div class="caml-input"> let x = 3 and y = 2 in f ~x ~y;;
</div><div class="caml-output ok">- : int = 1
</div></pre>
</div><p>When you want to use distinct names for the variable and the label
appearing in the type, you can use a naming label of the form
<span class="c003">~name:</span>. This also applies when the argument is not a variable.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f ~x:x1 ~y:y1 = x1 - y1;;
</div><div class="caml-output ok">val f : x:int -> y:int -> int = <fun>
</div></pre>
<pre><div class="caml-input"> f ~x:3 ~y:2;;
</div><div class="caml-output ok">- : int = 1
</div></pre>
</div><p>Labels obey the same rules as other identifiers in OCaml, that is you
cannot use a reserved keyword (like <span class="c003">in</span> or <span class="c003">to</span>) as label.</p><p>Formal parameters and arguments are matched according to their
respective labels<sup><a id="text1" href="#note1">1</a></sup>, the absence of label
being interpreted as the empty label.
This allows commuting arguments in applications. One can also
partially apply a function on any argument, creating a new function of
the remaining parameters.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f ~x ~y = x - y;;
</div><div class="caml-output ok">val f : x:int -> y:int -> int = <fun>
</div></pre>
<pre><div class="caml-input"> f ~y:2 ~x:3;;
</div><div class="caml-output ok">- : int = 1
</div></pre>
<pre><div class="caml-input"> ListLabels.fold_left;;
</div><div class="caml-output ok">- : f:('a -> 'b -> 'a) -> init:'a -> 'b list -> 'a = <fun>
</div></pre>
<pre><div class="caml-input"> ListLabels.fold_left [1;2;3] ~init:0 ~f:( + );;
</div><div class="caml-output ok">- : int = 6
</div></pre>
<pre><div class="caml-input"> ListLabels.fold_left ~init:0;;
</div><div class="caml-output ok">- : f:(int -> 'a -> int) -> 'a list -> int = <fun>
</div></pre>
</div><p>If several arguments of a function bear the same label (or no label),
they will not commute among themselves, and order matters. But they
can still commute with other arguments.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let hline ~x:x1 ~x:x2 ~y = (x1, x2, y);;
</div><div class="caml-output ok">val hline : x:'a -> x:'b -> y:'c -> 'a * 'b * 'c = <fun>
</div></pre>
<pre><div class="caml-input"> hline ~x:3 ~y:2 ~x:5;;
</div><div class="caml-output ok">- : int * int * int = (3, 5, 2)
</div></pre>
</div><p>As an exception to the above parameter matching rules, if an
application is total (omitting all optional arguments), labels may be
omitted.
In practice, many applications are total, so that labels can often be
omitted.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> f 3 2;;
</div><div class="caml-output ok">- : int = 1
</div></pre>
<pre><div class="caml-input"> ListLabels.map succ [1;2;3];;
</div><div class="caml-output ok">- : int list = [2; 3; 4]
</div></pre>
</div><p>
But beware that functions like <span class="c003">ListLabels.fold_left</span> whose result
type is a type variable will never be considered as totally applied.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> ListLabels.fold_left <U>( + )</U> 0 [1;2;3];;
</div><div class="caml-output error">Error: This expression has type int -> int -> int
but an expression was expected of type 'a list
</div></pre>
</div><p>When a function is passed as an argument to a higher-order function,
labels must match in both types. Neither adding nor removing labels
are allowed.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let h g = g ~x:3 ~y:2;;
</div><div class="caml-output ok">val h : (x:int -> y:int -> 'a) -> 'a = <fun>
</div></pre>
<pre><div class="caml-input"> h f;;
</div><div class="caml-output ok">- : int = 1
</div></pre>
<pre><div class="caml-input"> h <U>( + )</U> ;;
</div><div class="caml-output error">Error: This expression has type int -> int -> int
but an expression was expected of type x:int -> y:int -> 'a
</div></pre>
</div><p>
Note that when you don’t need an argument, you can still use a wildcard
pattern, but you must prefix it with the label.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> h (fun ~x:_ ~y -> y+1);;
</div><div class="caml-output ok">- : int = 3
</div></pre>
</div>
<h3 class="subsection" id="sec44">4.1.1  Optional arguments</h3>
<p>An interesting feature of labeled arguments is that they can be made
optional. For optional parameters, the question mark <span class="c003">?</span> replaces the
tilde <span class="c003">~</span> of non-optional ones, and the label is also prefixed by <span class="c003">?</span>
in the function type.
Default values may be given for such optional parameters.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let bump ?(step = 1) x = x + step;;
</div><div class="caml-output ok">val bump : ?step:int -> int -> int = <fun>
</div></pre>
<pre><div class="caml-input"> bump 2;;
</div><div class="caml-output ok">- : int = 3
</div></pre>
<pre><div class="caml-input"> bump ~step:3 2;;
</div><div class="caml-output ok">- : int = 5
</div></pre>
</div><p>A function taking some optional arguments must also take at least one
non-optional argument. The criterion for deciding whether an optional
argument has been omitted is the non-labeled application of an
argument appearing after this optional argument in the function type.
Note that if that argument is labeled, you will only be able to
eliminate optional arguments through the special case for total
applications.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let test ?(x = 0) ?(y = 0) () ?(z = 0) () = (x, y, z);;
</div><div class="caml-output ok">val test : ?x:int -> ?y:int -> unit -> ?z:int -> unit -> int * int * int =
<fun>
</div></pre>
<pre><div class="caml-input"> test ();;
</div><div class="caml-output ok">- : ?z:int -> unit -> int * int * int = <fun>
</div></pre>
<pre><div class="caml-input"> test ~x:2 () ~z:3 ();;
</div><div class="caml-output ok">- : int * int * int = (2, 0, 3)
</div></pre>
</div><p>Optional parameters may also commute with non-optional or unlabeled
ones, as long as they are applied simultaneously. By nature, optional
arguments do not commute with unlabeled arguments applied
independently.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> test ~y:2 ~x:3 () ();;
</div><div class="caml-output ok">- : int * int * int = (3, 2, 0)
</div></pre>
<pre><div class="caml-input"> test () () ~z:1 ~y:2 ~x:3;;
</div><div class="caml-output ok">- : int * int * int = (3, 2, 1)
</div></pre>
<pre><div class="caml-input"> <U>(test () ())</U> ~z:1 ;;
</div><div class="caml-output error">Error: This expression has type int * int * int
This is not a function; it cannot be applied.
</div></pre>
</div><p>
Here <span class="c003">(test () ())</span> is already <span class="c003">(0,0,0)</span> and cannot be further
applied.</p><p>Optional arguments are actually implemented as option types. If
you do not give a default value, you have access to their internal
representation, <span class="c003">type 'a option = None | Some of 'a</span>. You can then
provide different behaviors when an argument is present or not.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let bump ?step x =
match step with
| None -> x * 2
| Some y -> x + y
;;
</div><div class="caml-output ok">val bump : ?step:int -> int -> int = <fun>
</div></pre>
</div><p>It may also be useful to relay an optional argument from a function
call to another. This can be done by prefixing the applied argument
with <span class="c003">?</span>. This question mark disables the wrapping of optional
argument in an option type.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let test2 ?x ?y () = test ?x ?y () ();;
</div><div class="caml-output ok">val test2 : ?x:int -> ?y:int -> unit -> int * int * int = <fun>
</div></pre>
<pre><div class="caml-input"> test2 ?x:None;;
</div><div class="caml-output ok">- : ?y:int -> unit -> int * int * int = <fun>
</div></pre>
</div>
<h3 class="subsection" id="sec45">4.1.2  Labels and type inference</h3>
<p>
<a id="ss:label-inference"></a></p><p>While they provide an increased comfort for writing function
applications, labels and optional arguments have the pitfall that they
cannot be inferred as completely as the rest of the language.</p><p>You can see it in the following two examples.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let h' g = g ~y:2 ~x:3;;
</div><div class="caml-output ok">val h' : (y:int -> x:int -> 'a) -> 'a = <fun>
</div></pre>
<pre><div class="caml-input"> h' <U>f</U> ;;
</div><div class="caml-output error">Error: This expression has type x:int -> y:int -> int
but an expression was expected of type y:int -> x:int -> 'a
</div></pre>
<pre><div class="caml-input"> let bump_it bump x =
bump ~step:2 x;;
</div><div class="caml-output ok">val bump_it : (step:int -> 'a -> 'b) -> 'a -> 'b = <fun>
</div></pre>
<pre><div class="caml-input"> bump_it <U>bump</U> 1 ;;
</div><div class="caml-output error">Error: This expression has type ?step:int -> int -> int
but an expression was expected of type step:int -> 'a -> 'b
</div></pre>
</div><p>
The first case is simple: <span class="c003">g</span> is passed <span class="c003">~y</span> and then <span class="c003">~x</span>, but <span class="c003">f</span>
expects <span class="c003">~x</span> and then <span class="c003">~y</span>. This is correctly handled if we know the
type of <span class="c003">g</span> to be <span class="c003">x:int -> y:int -> int</span> in advance, but otherwise
this causes the above type clash. The simplest workaround is to apply
formal parameters in a standard order.</p><p>The second example is more subtle: while we intended the argument
<span class="c003">bump</span> to be of type <span class="c003">?step:int -> int -> int</span>, it is inferred as
<span class="c003">step:int -> int -> 'a</span>.
These two types being incompatible (internally normal and optional
arguments are different), a type error occurs when applying <span class="c003">bump_it</span>
to the real <span class="c003">bump</span>.</p><p>We will not try here to explain in detail how type inference works.
One must just understand that there is not enough information in the
above program to deduce the correct type of <span class="c003">g</span> or <span class="c003">bump</span>. That is,
there is no way to know whether an argument is optional or not, or
which is the correct order, by looking only at how a function is
applied. The strategy used by the compiler is to assume that there are
no optional arguments, and that applications are done in the right
order.</p><p>The right way to solve this problem for optional parameters is to add
a type annotation to the argument <span class="c003">bump</span>.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let bump_it (bump : ?step:int -> int -> int) x =
bump ~step:2 x;;
</div><div class="caml-output ok">val bump_it : (?step:int -> int -> int) -> int -> int = <fun>
</div></pre>
<pre><div class="caml-input"> bump_it bump 1;;
</div><div class="caml-output ok">- : int = 3
</div></pre>
</div><p>
In practice, such problems appear mostly when using objects whose
methods have optional arguments, so that writing the type of object
arguments is often a good idea.</p><p>Normally the compiler generates a type error if you attempt to pass to
a function a parameter whose type is different from the expected one.
However, in the specific case where the expected type is a non-labeled
function type, and the argument is a function expecting optional
parameters, the compiler will attempt to transform the argument to
have it match the expected type, by passing <span class="c003">None</span> for all optional
parameters.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let twice f (x : int) = f(f x);;
</div><div class="caml-output ok">val twice : (int -> int) -> int -> int = <fun>
</div></pre>
<pre><div class="caml-input"> twice bump 2;;
</div><div class="caml-output ok">- : int = 8
</div></pre>
</div><p>This transformation is coherent with the intended semantics,
including side-effects. That is, if the application of optional
parameters shall produce side-effects, these are delayed until the
received function is really applied to an argument.</p>
<h3 class="subsection" id="sec46">4.1.3  Suggestions for labeling</h3>
<p>Like for names, choosing labels for functions is not an easy task. A
good labeling is a labeling which</p><ul class="itemize"><li class="li-itemize">
makes programs more readable,
</li><li class="li-itemize">is easy to remember,
</li><li class="li-itemize">when possible, allows useful partial applications.
</li></ul><p>We explain here the rules we applied when labeling OCaml
libraries.</p><p>To speak in an “object-oriented” way, one can consider that each
function has a main argument, its <em>object</em>, and other arguments
related with its action, the <em>parameters</em>. To permit the
combination of functions through functionals in commuting label mode, the
object will not be labeled. Its role is clear from the function
itself. The parameters are labeled with names reminding of
their nature or their role. The best labels combine nature and
role. When this is not possible the role is to be preferred, since the
nature will
often be given by the type itself. Obscure abbreviations should be
avoided.
</p><pre>
<span class="c003">ListLabels.map : f:('a -> 'b) -> 'a list -> 'b list</span>
UnixLabels.write : file_descr -> buf:bytes -> pos:int -> len:int -> unit
</pre><p>When there are several objects of same nature and role, they are all
left unlabeled.
</p><pre>
<span class="c003">ListLabels.iter2 : f:('a -> 'b -> 'c) -> 'a list -> 'b list -> unit</span>
</pre><p>When there is no preferable object, all arguments are labeled.
</p><pre>
BytesLabels.blit :
src:bytes -> src_pos:int -> dst:bytes -> dst_pos:int -> len:int -> unit
</pre><p>However, when there is only one argument, it is often left unlabeled.
</p><pre>
BytesLabels.create : int -> bytes
</pre><p>
This principle also applies to functions of several arguments whose
return type is a type variable, as long as the role of each argument
is not ambiguous. Labeling such functions may lead to awkward error
messages when one attempts to omit labels in an application, as we
have seen with <span class="c003">ListLabels.fold_left</span>.</p><p>Here are some of the label names you will find throughout the
libraries.</p><div class="tableau">
<div class="center"><table class="c000 cellpadding1" border=1><tr><td class="c014"><span class="c013">Label</span></td><td class="c014"><span class="c013">Meaning</span> </td></tr>
<tr><td class="c016">
<span class="c003">f:</span></td><td class="c016">a function to be applied </td></tr>
<tr><td class="c016"><span class="c003">pos:</span></td><td class="c016">a position in a string, array or byte sequence </td></tr>
<tr><td class="c016"><span class="c003">len:</span></td><td class="c016">a length </td></tr>
<tr><td class="c016"><span class="c003">buf:</span></td><td class="c016">a byte sequence or string used as buffer </td></tr>
<tr><td class="c016"><span class="c003">src:</span></td><td class="c016">the source of an operation </td></tr>
<tr><td class="c016"><span class="c003">dst:</span></td><td class="c016">the destination of an operation </td></tr>
<tr><td class="c016"><span class="c003">init:</span></td><td class="c016">the initial value for an iterator </td></tr>
<tr><td class="c016"><span class="c003">cmp:</span></td><td class="c016">a comparison function, <span class="c009">e.g.</span> <span class="c003">Pervasives.compare</span> </td></tr>
<tr><td class="c016"><span class="c003">mode:</span></td><td class="c016">an operation mode or a flag list </td></tr>
</table></div></div><p>All these are only suggestions, but keep in mind that the
choice of labels is essential for readability. Bizarre choices will
make the program harder to maintain.</p><p>In the ideal, the right function name with right labels should be
enough to understand the function’s meaning. Since one can get this
information with OCamlBrowser or the <span class="c003">ocaml</span> toplevel, the documentation
is only used when a more detailed specification is needed.</p>
<h2 class="section" id="sec47">4.2  Polymorphic variants</h2>
<p>Variants as presented in section <a href="coreexamples.html#s%3Atut-recvariants">1.4</a> are a
powerful tool to build data structures and algorithms. However they
sometimes lack flexibility when used in modular programming. This is
due to the fact that every constructor is assigned to a unique type
when defined and used. Even if the same name appears in the definition
of multiple types, the constructor itself belongs to only one type.
Therefore, one cannot decide that a given constructor belongs to
multiple types, or consider a value of some type to belong to some
other type with more constructors.</p><p>With polymorphic variants, this original assumption is removed. That
is, a variant tag does not belong to any type in particular, the type
system will just check that it is an admissible value according to its
use. You need not define a type before using a variant tag. A variant
type will be inferred independently for each of its uses.</p><h3 class="subsection" id="sec48">Basic use</h3>
<p>In programs, polymorphic variants work like usual ones. You just have
to prefix their names with a backquote character <span class="c003">`</span>.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> [`On; `Off];;
</div><div class="caml-output ok">- : [> `Off | `On ] list = [`On; `Off]
</div></pre>
<pre><div class="caml-input"> `Number 1;;
</div><div class="caml-output ok">- : [> `Number of int ] = `Number 1
</div></pre>
<pre><div class="caml-input"> let f = function `On -> 1 | `Off -> 0 | `Number n -> n;;
</div><div class="caml-output ok">val f : [< `Number of int | `Off | `On ] -> int = <fun>
</div></pre>
<pre><div class="caml-input"> List.map f [`On; `Off];;
</div><div class="caml-output ok">- : int list = [1; 0]
</div></pre>
</div><p>
<span class="c003">[>`Off|`On] list</span> means that to match this list, you should at
least be able to match <span class="c003">`Off</span> and <span class="c003">`On</span>, without argument.
<span class="c003">[<`On|`Off|`Number of int]</span> means that <span class="c003">f</span> may be applied to <span class="c003">`Off</span>,
<span class="c003">`On</span> (both without argument), or <span class="c003">`Number</span> <span class="c009">n</span> where
<span class="c009">n</span> is an integer.
The <span class="c003">></span> and <span class="c003"><</span> inside the variant types show that they may still be
refined, either by defining more tags or by allowing less. As such, they
contain an implicit type variable. Because each of the variant types
appears only once in the whole type, their implicit type variables are
not shown.</p><p>The above variant types were polymorphic, allowing further refinement.
When writing type annotations, one will most often describe fixed
variant types, that is types that cannot be refined. This is
also the case for type abbreviations. Such types do not contain <span class="c003"><</span> or
<span class="c003">></span>, but just an enumeration of the tags and their associated types,
just like in a normal datatype definition.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> type 'a vlist = [`Nil | `Cons of 'a * 'a vlist];;
</div><div class="caml-output ok">type 'a vlist = [ `Cons of 'a * 'a vlist | `Nil ]
</div></pre>
<pre><div class="caml-input"> let rec map f : 'a vlist -> 'b vlist = function
| `Nil -> `Nil
| `Cons(a, l) -> `Cons(f a, map f l)
;;
</div><div class="caml-output ok">val map : ('a -> 'b) -> 'a vlist -> 'b vlist = <fun>
</div></pre>
</div><h3 class="subsection" id="sec49">Advanced use</h3>
<p>Type-checking polymorphic variants is a subtle thing, and some
expressions may result in more complex type information.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f = function `A -> `C | `B -> `D | x -> x;;
</div><div class="caml-output ok">val f : ([> `A | `B | `C | `D ] as 'a) -> 'a = <fun>
</div></pre>
<pre><div class="caml-input"> f `E;;
</div><div class="caml-output ok">- : [> `A | `B | `C | `D | `E ] = `E
</div></pre>
</div><p>
Here we are seeing two phenomena. First, since this matching is open
(the last case catches any tag), we obtain the type <span class="c003">[> `A | `B]</span>
rather than <span class="c003">[< `A | `B]</span> in a closed matching. Then, since <span class="c003">x</span> is
returned as is, input and return types are identical. The notation <span class="c003">as 'a</span> denotes such type sharing. If we apply <span class="c003">f</span> to yet another tag
<span class="c003">`E</span>, it gets added to the list.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f1 = function `A x -> x = 1 | `B -> true | `C -> false
let f2 = function `A x -> x = "a" | `B -> true ;;
</div><div class="caml-output ok">val f1 : [< `A of int | `B | `C ] -> bool = <fun>
val f2 : [< `A of string | `B ] -> bool = <fun>
</div></pre>
<pre><div class="caml-input"> let f x = f1 x && f2 x;;
</div><div class="caml-output ok">val f : [< `A of string & int | `B ] -> bool = <fun>
</div></pre>
</div><p>
Here <span class="c003">f1</span> and <span class="c003">f2</span> both accept the variant tags <span class="c003">`A</span> and <span class="c003">`B</span>, but the
argument of <span class="c003">`A</span> is <span class="c003">int</span> for <span class="c003">f1</span> and <span class="c003">string</span> for <span class="c003">f2</span>. In <span class="c003">f</span>’s
type <span class="c003">`C</span>, only accepted by <span class="c003">f1</span>, disappears, but both argument types
appear for <span class="c003">`A</span> as <span class="c003">int & string</span>. This means that if we
pass the variant tag <span class="c003">`A</span> to <span class="c003">f</span>, its argument should be <em>both</em>
<span class="c003">int</span> and <span class="c003">string</span>. Since there is no such value, <span class="c003">f</span> cannot be
applied to <span class="c003">`A</span>, and <span class="c003">`B</span> is the only accepted input.</p><p>Even if a value has a fixed variant type, one can still give it a
larger type through coercions. Coercions are normally written with
both the source type and the destination type, but in simple cases the
source type may be omitted.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> type 'a wlist = [`Nil | `Cons of 'a * 'a wlist | `Snoc of 'a wlist * 'a];;
</div><div class="caml-output ok">type 'a wlist = [ `Cons of 'a * 'a wlist | `Nil | `Snoc of 'a wlist * 'a ]
</div></pre>
<pre><div class="caml-input"> let wlist_of_vlist l = (l : 'a vlist :> 'a wlist);;
</div><div class="caml-output ok">val wlist_of_vlist : 'a vlist -> 'a wlist = <fun>
</div></pre>
<pre><div class="caml-input"> let open_vlist l = (l : 'a vlist :> [> 'a vlist]);;
</div><div class="caml-output ok">val open_vlist : 'a vlist -> [> 'a vlist ] = <fun>
</div></pre>
<pre><div class="caml-input"> fun x -> (x :> [`A|`B|`C]);;
</div><div class="caml-output ok">- : [< `A | `B | `C ] -> [ `A | `B | `C ] = <fun>
</div></pre>
</div><p>You may also selectively coerce values through pattern matching.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let split_cases = function
| `Nil | `Cons _ as x -> `A x
| `Snoc _ as x -> `B x
;;
</div><div class="caml-output ok">val split_cases :
[< `Cons of 'a | `Nil | `Snoc of 'b ] ->
[> `A of [> `Cons of 'a | `Nil ] | `B of [> `Snoc of 'b ] ] = <fun>
</div></pre>
</div><p>
When an or-pattern composed of variant tags is wrapped inside an
alias-pattern, the alias is given a type containing only the tags
enumerated in the or-pattern. This allows for many useful idioms, like
incremental definition of functions.</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let num x = `Num x
let eval1 eval (`Num x) = x
let rec eval x = eval1 eval x ;;
</div><div class="caml-output ok">val num : 'a -> [> `Num of 'a ] = <fun>
val eval1 : 'a -> [< `Num of 'b ] -> 'b = <fun>
val eval : [< `Num of 'a ] -> 'a = <fun>
</div></pre>
<pre><div class="caml-input"> let plus x y = `Plus(x,y)
let eval2 eval = function
| `Plus(x,y) -> eval x + eval y
| `Num _ as x -> eval1 eval x
let rec eval x = eval2 eval x ;;
</div><div class="caml-output ok">val plus : 'a -> 'b -> [> `Plus of 'a * 'b ] = <fun>
val eval2 : ('a -> int) -> [< `Num of int | `Plus of 'a * 'a ] -> int = <fun>
val eval : ([< `Num of int | `Plus of 'a * 'a ] as 'a) -> int = <fun>
</div></pre>
</div><p>To make this even more comfortable, you may use type definitions as
abbreviations for or-patterns. That is, if you have defined <span class="c003">type myvariant = [`Tag1 of int | `Tag2 of bool]</span>, then the pattern <span class="c003">#myvariant</span> is
equivalent to writing <span class="c003">(`Tag1(_ : int) | `Tag2(_ : bool))</span>.</p><p>Such abbreviations may be used alone,
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f = function
| #myvariant -> "myvariant"
| `Tag3 -> "Tag3";;
</div><div class="caml-output ok">val f : [< `Tag1 of int | `Tag2 of bool | `Tag3 ] -> string = <fun>
</div></pre>
</div><p>
or combined with with aliases.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let g1 = function `Tag1 _ -> "Tag1" | `Tag2 _ -> "Tag2";;
</div><div class="caml-output ok">val g1 : [< `Tag1 of 'a | `Tag2 of 'b ] -> string = <fun>
</div></pre>
<pre><div class="caml-input"> let g = function
| #myvariant as x -> g1 x
| `Tag3 -> "Tag3";;
</div><div class="caml-output ok">val g : [< `Tag1 of int | `Tag2 of bool | `Tag3 ] -> string = <fun>
</div></pre>
</div>
<h3 class="subsection" id="sec50">4.2.1  Weaknesses of polymorphic variants</h3>
<p>After seeing the power of polymorphic variants, one may wonder why
they were added to core language variants, rather than replacing them.</p><p>The answer is twofold. One first aspect is that while being pretty
efficient, the lack of static type information allows for less
optimizations, and makes polymorphic variants slightly heavier than
core language ones. However noticeable differences would only
appear on huge data structures.</p><p>More important is the fact that polymorphic variants, while being
type-safe, result in a weaker type discipline. That is, core language
variants do actually much more than ensuring type-safety, they also
check that you use only declared constructors, that all constructors
present in a data-structure are compatible, and they enforce typing
constraints to their parameters.</p><p>For this reason, you must be more careful about making types explicit
when you use polymorphic variants. When you write a library, this is
easy since you can describe exact types in interfaces, but for simple
programs you are probably better off with core language variants.</p><p>Beware also that some idioms make trivial errors very hard to find.
For instance, the following code is probably wrong but the compiler
has no way to see it.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> type abc = [`A | `B | `C] ;;
</div><div class="caml-output ok">type abc = [ `A | `B | `C ]
</div></pre>
<pre><div class="caml-input"> let f = function
| `As -> "A"
| #abc -> "other" ;;
</div><div class="caml-output ok">val f : [< `A | `As | `B | `C ] -> string = <fun>
</div></pre>
<pre><div class="caml-input"> let f : abc -> string = f ;;
</div><div class="caml-output ok">val f : abc -> string = <fun>
</div></pre>
</div><p>
You can avoid such risks by annotating the definition itself.
</p><div class="caml-example toplevel">
<pre><div class="caml-input"> let f : abc -> string = function
| <U>`As</U> -> "A"
| #abc -> "other" ;;
</div><div class="caml-output error">Error: This pattern matches values of type [? `As ]
but a pattern was expected which matches values of type abc
The second variant type does not allow tag(s) `As
</div></pre>
</div>
<hr class="footnoterule"><dl class="thefootnotes"><dt class="dt-thefootnotes">
<a id="note1" href="#text1">1</a></dt><dd class="dd-thefootnotes"><div class="footnotetext">This correspond to the commuting label mode
of Objective Caml 3.00 through 3.02, with some additional flexibility
on total applications. The so-called classic mode (<span class="c003">-nolabels</span>
options) is now deprecated for normal use.</div></dd></dl>
<hr>
<a href="objectexamples.html"><img src="previous_motif.svg" alt="Previous"></a>
<a href="index.html"><img src="contents_motif.svg" alt="Up"></a>
<a href="polymorphism.html"><img src="next_motif.svg" alt="Next"></a>
</body>
</html>
distrib
>
Mageia
>
7
>
armv7hl
>
media
>
core-release
>
by-pkgid
>
fb18813323b88f9a6e869238ab603257
>
files
>
21
