<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<link rel="stylesheet" href="style.css" type="text/css">
<meta content="text/html; charset=iso-8859-1" http-equiv="Content-Type">
<link rel="Start" href="index.html">
<link rel="previous" href="Ratio.html">
<link rel="next" href="Set.html">
<link rel="Up" href="index.html">
<link title="Index of types" rel=Appendix href="index_types.html">
<link title="Index of exceptions" rel=Appendix href="index_exceptions.html">
<link title="Index of values" rel=Appendix href="index_values.html">
<link title="Index of modules" rel=Appendix href="index_modules.html">
<link title="Index of module types" rel=Appendix href="index_module_types.html">
<link title="Arg" rel="Chapter" href="Arg.html">
<link title="Arith_status" rel="Chapter" href="Arith_status.html">
<link title="Array" rel="Chapter" href="Array.html">
<link title="ArrayLabels" rel="Chapter" href="ArrayLabels.html">
<link title="Big_int" rel="Chapter" href="Big_int.html">
<link title="Bigarray" rel="Chapter" href="Bigarray.html">
<link title="Buffer" rel="Chapter" href="Buffer.html">
<link title="Callback" rel="Chapter" href="Callback.html">
<link title="CamlinternalLazy" rel="Chapter" href="CamlinternalLazy.html">
<link title="CamlinternalMod" rel="Chapter" href="CamlinternalMod.html">
<link title="CamlinternalOO" rel="Chapter" href="CamlinternalOO.html">
<link title="Char" rel="Chapter" href="Char.html">
<link title="Complex" rel="Chapter" href="Complex.html">
<link title="Condition" rel="Chapter" href="Condition.html">
<link title="Digest" rel="Chapter" href="Digest.html">
<link title="Dynlink" rel="Chapter" href="Dynlink.html">
<link title="Event" rel="Chapter" href="Event.html">
<link title="Filename" rel="Chapter" href="Filename.html">
<link title="Format" rel="Chapter" href="Format.html">
<link title="Gc" rel="Chapter" href="Gc.html">
<link title="Genlex" rel="Chapter" href="Genlex.html">
<link title="Graphics" rel="Chapter" href="Graphics.html">
<link title="GraphicsX11" rel="Chapter" href="GraphicsX11.html">
<link title="Hashtbl" rel="Chapter" href="Hashtbl.html">
<link title="Int32" rel="Chapter" href="Int32.html">
<link title="Int64" rel="Chapter" href="Int64.html">
<link title="Lazy" rel="Chapter" href="Lazy.html">
<link title="Lexing" rel="Chapter" href="Lexing.html">
<link title="List" rel="Chapter" href="List.html">
<link title="ListLabels" rel="Chapter" href="ListLabels.html">
<link title="Map" rel="Chapter" href="Map.html">
<link title="Marshal" rel="Chapter" href="Marshal.html">
<link title="MoreLabels" rel="Chapter" href="MoreLabels.html">
<link title="Mutex" rel="Chapter" href="Mutex.html">
<link title="Nativeint" rel="Chapter" href="Nativeint.html">
<link title="Num" rel="Chapter" href="Num.html">
<link title="Obj" rel="Chapter" href="Obj.html">
<link title="Oo" rel="Chapter" href="Oo.html">
<link title="Parsing" rel="Chapter" href="Parsing.html">
<link title="Pervasives" rel="Chapter" href="Pervasives.html">
<link title="Printexc" rel="Chapter" href="Printexc.html">
<link title="Printf" rel="Chapter" href="Printf.html">
<link title="Queue" rel="Chapter" href="Queue.html">
<link title="Random" rel="Chapter" href="Random.html">
<link title="Ratio" rel="Chapter" href="Ratio.html">
<link title="Scanf" rel="Chapter" href="Scanf.html">
<link title="Set" rel="Chapter" href="Set.html">
<link title="Sort" rel="Chapter" href="Sort.html">
<link title="Stack" rel="Chapter" href="Stack.html">
<link title="StdLabels" rel="Chapter" href="StdLabels.html">
<link title="Str" rel="Chapter" href="Str.html">
<link title="Stream" rel="Chapter" href="Stream.html">
<link title="String" rel="Chapter" href="String.html">
<link title="StringLabels" rel="Chapter" href="StringLabels.html">
<link title="Sys" rel="Chapter" href="Sys.html">
<link title="Thread" rel="Chapter" href="Thread.html">
<link title="ThreadUnix" rel="Chapter" href="ThreadUnix.html">
<link title="Tk" rel="Chapter" href="Tk.html">
<link title="Unix" rel="Chapter" href="Unix.html">
<link title="UnixLabels" rel="Chapter" href="UnixLabels.html">
<link title="Weak" rel="Chapter" href="Weak.html"><link title="Introduction" rel="Section" href="#6_Introduction">
<link title="Formatted input channel" rel="Section" href="#6_Formattedinputchannel">
<link title="Type of formatted input functions" rel="Section" href="#6_Typeofformattedinputfunctions">
<link title="The general formatted input function" rel="Section" href="#6_Thegeneralformattedinputfunction">
<link title="Format string description" rel="Section" href="#6_Formatstringdescription">
<link title="Specialised formatted input functions" rel="Section" href="#6_Specialisedformattedinputfunctions">
<link title="Reading format strings from input" rel="Section" href="#6_Readingformatstringsfrominput">
<link title="Functional input with format strings" rel="Subsection" href="#7_Functionalinputwithformatstrings">
<link title="A simple example" rel="Subsection" href="#7_Asimpleexample">
<link title="Formatted input as a functional feature" rel="Subsection" href="#7_Formattedinputasafunctionalfeature">
<link title="The space character in format strings" rel="Subsection" href="#space">
<link title="Conversion specifications in format strings" rel="Subsection" href="#conversion">
<link title="Scanning indications in format strings" rel="Subsection" href="#indication">
<link title="Exceptions during scanning" rel="Subsection" href="#7_Exceptionsduringscanning">
<title>Scanf</title>
</head>
<body>
<div class="navbar"><a class="pre" href="Ratio.html" title="Ratio">Previous</a>
<a class="up" href="index.html" title="Index">Up</a>
<a class="post" href="Set.html" title="Set">Next</a>
</div>
<h1>Module <a href="type_Scanf.html">Scanf</a></h1>
<pre><span class="keyword">module</span> Scanf: <code class="code"><span class="keyword">sig</span></code> <a href="Scanf.html">..</a> <code class="code"><span class="keyword">end</span></code></pre><div class="info module top">
Formatted input functions.<br>
</div>
<hr width="100%">
<br>
<h6 id="6_Introduction">Introduction</h6><br>
<br>
<div class="h7" id="7_Functionalinputwithformatstrings">Functional input with format strings</div><br>
<br>
The module <code class="code"><span class="constructor">Scanf</span></code> provides formatted input functions or <em>scanners</em>.
<p>
The formatted input functions can read from any kind of input, including
strings, files, or anything that can return characters. The more general
source of characters is named a <em>formatted input channel</em> (or <em> scanning buffer</em>) and has type <a href="Scanf.Scanning.html#TYPEin_channel"><code class="code"><span class="constructor">Scanf</span>.<span class="constructor">Scanning</span>.in_channel</code></a>. The more general
formatted input function reads from any scanning buffer and is named
<code class="code">bscanf</code>.
<p>
Generally speaking, the formatted input functions have 3 arguments:<ul>
<li>the first argument is a source of characters for the input,</li>
<li>the second argument is a format string that specifies the values to
read,</li>
<li>the third argument is a <em>receiver function</em> that is applied to the
values read.</li>
</ul>
Hence, a typical call to the formatted input function <a href="Scanf.html#VALbscanf"><code class="code"><span class="constructor">Scanf</span>.bscanf</code></a> is
<code class="code">bscanf ic fmt f</code>, where:
<p>
<ul>
<li><code class="code">ic</code> is a source of characters (typically a <em> formatted input channel</em> with type <a href="Scanf.Scanning.html#TYPEin_channel"><code class="code"><span class="constructor">Scanf</span>.<span class="constructor">Scanning</span>.in_channel</code></a>),</li>
</ul>
<ul>
<li><code class="code">fmt</code> is a format string (the same format strings as those used to print
material with module <a href="Printf.html"><code class="code"><span class="constructor">Printf</span></code></a> or <a href="Format.html"><code class="code"><span class="constructor">Format</span></code></a>),</li>
</ul>
<ul>
<li><code class="code">f</code> is a function that has as many arguments as the number of values to
read in the input.</li>
</ul>
<br>
<br>
<div class="h7" id="7_Asimpleexample">A simple example</div><br>
<br>
As suggested above, the expression <code class="code">bscanf ic <span class="string">"%d"</span> f</code> reads a decimal
integer <code class="code">n</code> from the source of characters <code class="code">ic</code> and returns <code class="code">f n</code>.
<p>
For instance,
<p>
<ul>
<li>if we use <code class="code">stdin</code> as the source of characters (<a href="Scanf.Scanning.html#VALstdin"><code class="code"><span class="constructor">Scanf</span>.<span class="constructor">Scanning</span>.stdin</code></a> is
the predefined formatted input channel that reads from standard input),</li>
</ul>
<ul>
<li>if we define the receiver <code class="code">f</code> as <code class="code"><span class="keyword">let</span> f x = x + 1</code>,</li>
</ul>
then <code class="code">bscanf <span class="constructor">Scanning</span>.stdin <span class="string">"%d"</span> f</code> reads an integer <code class="code">n</code> from the
standard input and returns <code class="code">f n</code> (that is <code class="code">n + 1</code>). Thus, if we
evaluate <code class="code">bscanf stdin <span class="string">"%d"</span> f</code>, and then enter <code class="code">41</code> at the
keyboard, we get <code class="code">42</code> as the final result.<br>
<br>
<div class="h7" id="7_Formattedinputasafunctionalfeature">Formatted input as a functional feature</div><br>
<br>
The OCaml scanning facility is reminiscent of the corresponding C feature.
However, it is also largely different, simpler, and yet more powerful:
the formatted input functions are higher-order functionals and the
parameter passing mechanism is just the regular function application not
the variable assignment based mechanism which is typical for formatted
input in imperative languages; the OCaml format strings also feature
useful additions to easily define complex tokens; as expected within a
functional programming language, the formatted input functions also
support polymorphism, in particular arbitrary interaction with
polymorphic user-defined scanners. Furthermore, the OCaml formatted input
facility is fully type-checked at compile time.<br>
<br>
<h6 id="6_Formattedinputchannel">Formatted input channel</h6><br>
<pre><span class="keyword">module</span> <a href="Scanf.Scanning.html">Scanning</a>: <code class="code"><span class="keyword">sig</span></code> <a href="Scanf.Scanning.html">..</a> <code class="code"><span class="keyword">end</span></code></pre><br>
<h6 id="6_Typeofformattedinputfunctions">Type of formatted input functions</h6><br>
<pre><span id="TYPEscanner"><span class="keyword">type</span> <code class="type">('a, 'b, 'c, 'd)</code> scanner</span> = <code class="type">('a, <a href="Scanf.Scanning.html#TYPEin_channel">Scanning.in_channel</a>, 'b, 'c, 'a -> 'd, 'd) format6 -> 'c</code> </pre>
<div class="info ">
The type of formatted input scanners: <code class="code">(<span class="keywordsign">'</span>a, <span class="keywordsign">'</span>b, <span class="keywordsign">'</span>c, <span class="keywordsign">'</span>d) scanner</code>
is the type of a formatted input function that reads from some
formatted input channel according to some format string; more
precisely, if <code class="code">scan</code> is some formatted input function, then <code class="code">scan
ic fmt f</code> applies <code class="code">f</code> to the arguments specified by the format
string <code class="code">fmt</code>, when <code class="code">scan</code> has read those arguments from the
formatted input channel <code class="code">ic</code>.
<p>
For instance, the <code class="code">scanf</code> function below has type <code class="code">(<span class="keywordsign">'</span>a, <span class="keywordsign">'</span>b, <span class="keywordsign">'</span>c, <span class="keywordsign">'</span>d)
scanner</code>, since it is a formatted input function that reads from
<code class="code"><span class="constructor">Scanning</span>.stdin</code>: <code class="code">scanf fmt f</code> applies <code class="code">f</code> to the arguments specified by
<code class="code">fmt</code>, reading those arguments from <code class="code"><span class="constructor">Pervasives</span>.stdin</code> as expected.
<p>
If the format <code class="code">fmt</code> has some <code class="code">%r</code> indications, the corresponding input
functions must be provided before the receiver <code class="code">f</code> argument. For
instance, if <code class="code">read_elem</code> is an input function for values of type <code class="code">t</code>,
then <code class="code">bscanf ic <span class="string">"%r;"</span> read_elem f</code> reads a value <code class="code">v</code> of type <code class="code">t</code> followed
by a <code class="code"><span class="string">';'</span></code> character, and returns <code class="code">f v</code>.<br>
<b>Since</b> 3.10.0<br>
</div>
<pre><span id="EXCEPTIONScan_failure"><span class="keyword">exception</span> Scan_failure</span> <span class="keyword">of</span> <code class="type">string</code></pre>
<div class="info ">
The exception that formatted input functions raise when the input cannot
be read according to the given format.<br>
</div>
<br>
<h6 id="6_Thegeneralformattedinputfunction">The general formatted input function</h6><br>
<pre><span id="VALbscanf"><span class="keyword">val</span> bscanf</span> : <code class="type"><a href="Scanf.Scanning.html#TYPEin_channel">Scanning.in_channel</a> -> ('a, 'b, 'c, 'd) <a href="Scanf.html#TYPEscanner">scanner</a></code></pre><div class="info ">
<code class="code">bscanf ic fmt r1 ... rN f</code> reads arguments for the function <code class="code">f</code>, from the
formatted input channel <code class="code">ic</code>, according to the format string <code class="code">fmt</code>, and
applies <code class="code">f</code> to these values.
The result of this call to <code class="code">f</code> is returned as the result of the entire
<code class="code">bscanf</code> call.
For instance, if <code class="code">f</code> is the function <code class="code"><span class="keyword">fun</span> s i <span class="keywordsign">-></span> i + 1</code>, then
<code class="code"><span class="constructor">Scanf</span>.sscanf <span class="string">"x= 1"</span> <span class="string">"%s = %i"</span> f</code> returns <code class="code">2</code>.
<p>
Arguments <code class="code">r1</code> to <code class="code">rN</code> are user-defined input functions that read the
argument corresponding to the <code class="code">%r</code> conversions specified in the format
string.<br>
</div>
<br>
<h6 id="6_Formatstringdescription">Format string description</h6><br>
<br>
The format string is a character string which contains three types of
objects:<ul>
<li>plain characters, which are simply matched with the characters of the
input (with a special case for space and line feed, see <a href="Scanf.html#space"><i>The space character in format strings</i></a>),</li>
<li>conversion specifications, each of which causes reading and conversion of
one argument for the function <code class="code">f</code> (see <a href="Scanf.html#conversion"><i>Conversion specifications in format strings</i></a>),</li>
<li>scanning indications to specify boundaries of tokens
(see scanning <a href="Scanf.html#indication"><i>Scanning indications in format strings</i></a>).</li>
</ul>
<br>
<br>
<div class="h7" id="space">The space character in format strings</div><br>
<br>
As mentioned above, a plain character in the format string is just
matched with the next character of the input; however, two characters are
special exceptions to this rule: the space character (<code class="code"><span class="string">' '</span></code> or ASCII code
32) and the line feed character (<code class="code"><span class="string">'\n'</span></code> or ASCII code 10).
A space does not match a single space character, but any amount of
'whitespace' in the input. More precisely, a space inside the format
string matches <em>any number</em> of tab, space, line feed and carriage
return characters. Similarly, a line feed character in the format string
matches either a single line feed or a carriage return followed by a line
feed.
<p>
Matching <em>any</em> amount of whitespace, a space in the format string
also matches no amount of whitespace at all; hence, the call <code class="code">bscanf ib
<span class="string">"Price = %d $"</span> (<span class="keyword">fun</span> p <span class="keywordsign">-></span> p)</code> succeeds and returns <code class="code">1</code> when reading an
input with various whitespace in it, such as <code class="code"><span class="constructor">Price</span> = 1 $</code>,
<code class="code"><span class="constructor">Price</span> = 1 $</code>, or even <code class="code"><span class="constructor">Price</span>=1$</code>.<br>
<br>
<div class="h7" id="conversion">Conversion specifications in format strings</div><br>
<br>
Conversion specifications consist in the <code class="code">%</code> character, followed by
an optional flag, an optional field width, and followed by one or
two conversion characters. The conversion characters and their
meanings are:
<p>
<ul>
<li><code class="code">d</code>: reads an optionally signed decimal integer.</li>
<li><code class="code">i</code>: reads an optionally signed integer
(usual input conventions for decimal (<code class="code">0-9</code>+), hexadecimal
(<code class="code">0x[0-9a-f]+</code> and <code class="code">0<span class="constructor">X</span>[0-9<span class="constructor">A</span>-<span class="constructor">F</span>]+</code>), octal (<code class="code">0o[0-7]+</code>), and binary
(<code class="code">0b[0-1]+</code>) notations are understood).</li>
<li><code class="code">u</code>: reads an unsigned decimal integer.</li>
<li><code class="code">x</code> or <code class="code"><span class="constructor">X</span></code>: reads an unsigned hexadecimal integer (<code class="code">[0-9a-fA-<span class="constructor">F</span>]+</code>).</li>
<li><code class="code">o</code>: reads an unsigned octal integer (<code class="code">[0-7]+</code>).</li>
<li><code class="code">s</code>: reads a string argument that spreads as much as possible, until the
following bounding condition holds: <ul>
<li>a whitespace has been found (see <a href="Scanf.html#space"><i>The space character in format strings</i></a>),</li>
<li>a scanning indication (see scanning <a href="Scanf.html#indication"><i>Scanning indications in format strings</i></a>) has been
encountered,</li>
<li>the end-of-input has been reached.</li>
</ul>
Hence, this conversion always succeeds: it returns an empty
string if the bounding condition holds when the scan begins.</li>
<li><code class="code"><span class="constructor">S</span></code>: reads a delimited string argument (delimiters and special
escaped characters follow the lexical conventions of OCaml).</li>
<li><code class="code">c</code>: reads a single character. To test the current input character
without reading it, specify a null field width, i.e. use
specification <code class="code">%0c</code>. Raise <code class="code"><span class="constructor">Invalid_argument</span></code>, if the field width
specification is greater than 1.</li>
<li><code class="code"><span class="constructor">C</span></code>: reads a single delimited character (delimiters and special
escaped characters follow the lexical conventions of OCaml).</li>
<li><code class="code">f</code>, <code class="code">e</code>, <code class="code"><span class="constructor">E</span></code>, <code class="code">g</code>, <code class="code"><span class="constructor">G</span></code>: reads an optionally signed
floating-point number in decimal notation, in the style <code class="code">dddd.ddd
e/<span class="constructor">E</span>+-dd</code>.</li>
<li><code class="code"><span class="constructor">F</span></code>: reads a floating point number according to the lexical
conventions of OCaml (hence the decimal point is mandatory if the
exponent part is not mentioned).</li>
<li><code class="code"><span class="constructor">B</span></code>: reads a boolean argument (<code class="code"><span class="keyword">true</span></code> or <code class="code"><span class="keyword">false</span></code>).</li>
<li><code class="code">b</code>: reads a boolean argument (for backward compatibility; do not use
in new programs).</li>
<li><code class="code">ld</code>, <code class="code">li</code>, <code class="code">lu</code>, <code class="code">lx</code>, <code class="code">lX</code>, <code class="code">lo</code>: reads an <code class="code">int32</code> argument to
the format specified by the second letter for regular integers.</li>
<li><code class="code">nd</code>, <code class="code">ni</code>, <code class="code">nu</code>, <code class="code">nx</code>, <code class="code">nX</code>, <code class="code">no</code>: reads a <code class="code">nativeint</code> argument to
the format specified by the second letter for regular integers.</li>
<li><code class="code"><span class="constructor">Ld</span></code>, <code class="code"><span class="constructor">Li</span></code>, <code class="code"><span class="constructor">Lu</span></code>, <code class="code"><span class="constructor">Lx</span></code>, <code class="code"><span class="constructor">LX</span></code>, <code class="code"><span class="constructor">Lo</span></code>: reads an <code class="code">int64</code> argument to
the format specified by the second letter for regular integers.</li>
<li><code class="code">[ range ]</code>: reads characters that matches one of the characters
mentioned in the range of characters <code class="code">range</code> (or not mentioned in
it, if the range starts with <code class="code">^</code>). Reads a <code class="code">string</code> that can be
empty, if the next input character does not match the range. The set of
characters from <code class="code">c1</code> to <code class="code">c2</code> (inclusively) is denoted by <code class="code">c1-c2</code>.
Hence, <code class="code">%[0-9]</code> returns a string representing a decimal number
or an empty string if no decimal digit is found; similarly,
<code class="code">%[\\048-\\057\\065-\\070]</code> returns a string of hexadecimal digits.
If a closing bracket appears in a range, it must occur as the
first character of the range (or just after the <code class="code">^</code> in case of
range negation); hence <code class="code">[]]</code> matches a <code class="code">]</code> character and
<code class="code">[^]]</code> matches any character that is not <code class="code">]</code>.
Use <code class="code">%%</code> and <code class="code">%@</code> to include a <code class="code">%</code> or a <code class="code">@</code> in a range.</li>
<li><code class="code">r</code>: user-defined reader. Takes the next <code class="code">ri</code> formatted input
function and applies it to the scanning buffer <code class="code">ib</code> to read the
next argument. The input function <code class="code">ri</code> must therefore have type
<code class="code"><span class="constructor">Scanning</span>.in_channel <span class="keywordsign">-></span> <span class="keywordsign">'</span>a</code> and the argument read has type <code class="code"><span class="keywordsign">'</span>a</code>.</li>
<li><code class="code">{ fmt %}</code>: reads a format string argument. The format string
read must have the same type as the format string specification
<code class="code">fmt</code>. For instance, <code class="code"><span class="string">"%{ %i %}"</span></code> reads any format string that
can read a value of type <code class="code">int</code>; hence, if <code class="code">s</code> is the string
<code class="code"><span class="string">"fmt:\"number is %u\""</span></code>, then <code class="code"><span class="constructor">Scanf</span>.sscanf s <span class="string">"fmt: %{%i%}"</span></code>
succeeds and returns the format string <code class="code"><span class="string">"number is %u"</span></code>.</li>
<li><code class="code">\( fmt %\)</code>: scanning sub-format substitution.
Reads a format string <code class="code">rf</code> in the input, then goes on scanning with
<code class="code">rf</code> instead of scanning with <code class="code">fmt</code>.
The format string <code class="code">rf</code> must have the same type as the format string
specification <code class="code">fmt</code> that it replaces.
For instance, <code class="code"><span class="string">"%( %i %)"</span></code> reads any format string that can read a value
of type <code class="code">int</code>.
The conversion returns the format string read <code class="code">rf</code>, and then a value
read using <code class="code">rf</code>.
Hence, if <code class="code">s</code> is the string <code class="code"><span class="string">"\"%4d\"1234.00"</span></code>, then
<code class="code"><span class="constructor">Scanf</span>.sscanf s <span class="string">"%(%i%)"</span> (<span class="keyword">fun</span> fmt i <span class="keywordsign">-></span> fmt, i)</code> evaluates to
<code class="code">(<span class="string">"%4d"</span>, 1234)</code>.</li>
</ul>
This behaviour is not mere format substitution, since the conversion
returns the format string read as additional argument. If you need
pure format substitution, use special flag <code class="code">_</code> to discard the
extraneous argument: conversion <code class="code">%_\( fmt %\)</code> reads a format string
<code class="code">rf</code> and then behaves the same as format string <code class="code">rf</code>. Hence, if <code class="code">s</code> is
the string <code class="code"><span class="string">"\"%4d\"1234.00"</span></code>, then <code class="code"><span class="constructor">Scanf</span>.sscanf s <span class="string">"%_(%i%)"</span></code> is
simply equivalent to <code class="code"><span class="constructor">Scanf</span>.sscanf <span class="string">"1234.00"</span> <span class="string">"%4d"</span></code>.
<p>
<ul>
<li><code class="code">l</code>: returns the number of lines read so far.</li>
<li><code class="code">n</code>: returns the number of characters read so far.</li>
<li><code class="code"><span class="constructor">N</span></code> or <code class="code"><span class="constructor">L</span></code>: returns the number of tokens read so far.</li>
<li><code class="code">!</code>: matches the end of input condition.</li>
<li><code class="code">%</code>: matches one <code class="code">%</code> character in the input.</li>
<li><code class="code">@</code>: matches one <code class="code">@</code> character in the input.</li>
<li><code class="code">,</code>: does nothing.</li>
</ul>
Following the <code class="code">%</code> character that introduces a conversion, there may be
the special flag <code class="code">_</code>: the conversion that follows occurs as usual,
but the resulting value is discarded.
For instance, if <code class="code">f</code> is the function <code class="code"><span class="keyword">fun</span> i <span class="keywordsign">-></span> i + 1</code>, and <code class="code">s</code> is the
string <code class="code"><span class="string">"x = 1"</span></code>, then <code class="code"><span class="constructor">Scanf</span>.sscanf s <span class="string">"%_s = %i"</span> f</code> returns <code class="code">2</code>.
<p>
The field width is composed of an optional integer literal
indicating the maximal width of the token to read.
For instance, <code class="code">%6d</code> reads an integer, having at most 6 decimal digits;
<code class="code">%4f</code> reads a float with at most 4 characters; and <code class="code">%8[\\000-\\255]</code>
returns the next 8 characters (or all the characters still available,
if fewer than 8 characters are available in the input).
<p>
Notes:
<p>
<ul>
<li>as mentioned above, a <code class="code">%s</code> conversion always succeeds, even if there is
nothing to read in the input: in this case, it simply returns <code class="code"><span class="string">""</span></code>.</li>
</ul>
<ul>
<li>in addition to the relevant digits, <code class="code"><span class="string">'_'</span></code> characters may appear
inside numbers (this is reminiscent to the usual OCaml lexical
conventions). If stricter scanning is desired, use the range
conversion facility instead of the number conversions.</li>
</ul>
<ul>
<li>the <code class="code">scanf</code> facility is not intended for heavy duty lexical
analysis and parsing. If it appears not expressive enough for your
needs, several alternative exists: regular expressions (module
<code class="code"><span class="constructor">Str</span></code>), stream parsers, <code class="code">ocamllex</code>-generated lexers,
<code class="code">ocamlyacc</code>-generated parsers.</li>
</ul>
<br>
<br>
<div class="h7" id="indication">Scanning indications in format strings</div><br>
<br>
Scanning indications appear just after the string conversions <code class="code">%s</code>
and <code class="code">%[ range ]</code> to delimit the end of the token. A scanning
indication is introduced by a <code class="code">@</code> character, followed by some
plain character <code class="code">c</code>. It means that the string token should end
just before the next matching <code class="code">c</code> (which is skipped). If no <code class="code">c</code>
character is encountered, the string token spreads as much as
possible. For instance, <code class="code"><span class="string">"%s@\t"</span></code> reads a string up to the next
tab character or to the end of input. If a <code class="code">@</code> character appears
anywhere else in the format string, it is treated as a plain character.
<p>
Note:
<p>
<ul>
<li>As usual in format strings, <code class="code">%</code> and <code class="code">@</code> characters must be escaped
using <code class="code">%%</code> and <code class="code">%@</code>; this rule still holds within range specifications
and scanning indications.
For instance, <code class="code"><span class="string">"%s@%%"</span></code> reads a string up to the next <code class="code">%</code> character.</li>
<li>The scanning indications introduce slight differences in the syntax of
<code class="code"><span class="constructor">Scanf</span></code> format strings, compared to those used for the <code class="code"><span class="constructor">Printf</span></code>
module. However, the scanning indications are similar to those used in
the <code class="code"><span class="constructor">Format</span></code> module; hence, when producing formatted text to be scanned
by <code class="code">!<span class="constructor">Scanf</span>.bscanf</code>, it is wise to use printing functions from the
<code class="code"><span class="constructor">Format</span></code> module (or, if you need to use functions from <code class="code"><span class="constructor">Printf</span></code>, banish
or carefully double check the format strings that contain <code class="code"><span class="string">'@'</span></code>
characters).</li>
</ul>
<br>
<br>
<div class="h7" id="7_Exceptionsduringscanning">Exceptions during scanning</div><br>
<br>
Scanners may raise the following exceptions when the input cannot be read
according to the format string:
<p>
<ul>
<li>Raise <code class="code"><span class="constructor">Scanf</span>.<span class="constructor">Scan_failure</span></code> if the input does not match the format.</li>
</ul>
<ul>
<li>Raise <code class="code"><span class="constructor">Failure</span></code> if a conversion to a number is not possible.</li>
</ul>
<ul>
<li>Raise <code class="code"><span class="constructor">End_of_file</span></code> if the end of input is encountered while some more
characters are needed to read the current conversion specification.</li>
</ul>
<ul>
<li>Raise <code class="code"><span class="constructor">Invalid_argument</span></code> if the format string is invalid.</li>
</ul>
Note:
<p>
<ul>
<li>as a consequence, scanning a <code class="code">%s</code> conversion never raises exception
<code class="code"><span class="constructor">End_of_file</span></code>: if the end of input is reached the conversion succeeds and
simply returns the characters read so far, or <code class="code"><span class="string">""</span></code> if none were ever read.</li>
</ul>
<br>
<br>
<h6 id="6_Specialisedformattedinputfunctions">Specialised formatted input functions</h6><br>
<pre><span id="VALfscanf"><span class="keyword">val</span> fscanf</span> : <code class="type"><a href="Pervasives.html#TYPEin_channel">in_channel</a> -> ('a, 'b, 'c, 'd) <a href="Scanf.html#TYPEscanner">scanner</a></code></pre><div class="info ">
Same as <a href="Scanf.html#VALbscanf"><code class="code"><span class="constructor">Scanf</span>.bscanf</code></a>, but reads from the given regular input channel.
<p>
Warning: since all formatted input functions operate from a <em>formatted
input channel</em>, be aware that each <code class="code">fscanf</code> invocation will operate with a
formatted input channel reading from the given channel. This extra level
of bufferization can lead to a strange scanning behaviour if you use low
level primitives on the channel (reading characters, seeking the reading
position, and so on).
<p>
As a consequence, never mix direct low level reading and high level
scanning from the same regular input channel.<br>
</div>
<pre><span id="VALsscanf"><span class="keyword">val</span> sscanf</span> : <code class="type">string -> ('a, 'b, 'c, 'd) <a href="Scanf.html#TYPEscanner">scanner</a></code></pre><div class="info ">
Same as <a href="Scanf.html#VALbscanf"><code class="code"><span class="constructor">Scanf</span>.bscanf</code></a>, but reads from the given string.<br>
</div>
<pre><span id="VALscanf"><span class="keyword">val</span> scanf</span> : <code class="type">('a, 'b, 'c, 'd) <a href="Scanf.html#TYPEscanner">scanner</a></code></pre><div class="info ">
Same as <a href="Scanf.html#VALbscanf"><code class="code"><span class="constructor">Scanf</span>.bscanf</code></a>, but reads from the predefined formatted input
channel <a href="Scanf.Scanning.html#VALstdin"><code class="code"><span class="constructor">Scanf</span>.<span class="constructor">Scanning</span>.stdin</code></a> that is connected to <code class="code"><span class="constructor">Pervasives</span>.stdin</code>.<br>
</div>
<pre><span id="VALkscanf"><span class="keyword">val</span> kscanf</span> : <code class="type"><a href="Scanf.Scanning.html#TYPEin_channel">Scanning.in_channel</a> -><br> (<a href="Scanf.Scanning.html#TYPEin_channel">Scanning.in_channel</a> -> exn -> 'd) -> ('a, 'b, 'c, 'd) <a href="Scanf.html#TYPEscanner">scanner</a></code></pre><div class="info ">
Same as <a href="Scanf.html#VALbscanf"><code class="code"><span class="constructor">Scanf</span>.bscanf</code></a>, but takes an additional function argument
<code class="code">ef</code> that is called in case of error: if the scanning process or
some conversion fails, the scanning function aborts and calls the
error handling function <code class="code">ef</code> with the formatted input channel and the
exception that aborted the scanning process as arguments.<br>
</div>
<br>
<h6 id="6_Readingformatstringsfrominput">Reading format strings from input</h6><br>
<pre><span id="VALbscanf_format"><span class="keyword">val</span> bscanf_format</span> : <code class="type"><a href="Scanf.Scanning.html#TYPEin_channel">Scanning.in_channel</a> -><br> ('a, 'b, 'c, 'd, 'e, 'f) format6 -><br> (('a, 'b, 'c, 'd, 'e, 'f) format6 -> 'g) -> 'g</code></pre><div class="info ">
<code class="code">bscanf_format ic fmt f</code> reads a format string token from the formatted
input channel <code class="code">ic</code>, according to the given format string <code class="code">fmt</code>, and
applies <code class="code">f</code> to the resulting format string value.
Raise <code class="code"><span class="constructor">Scan_failure</span></code> if the format string value read does not have the
same type as <code class="code">fmt</code>.<br>
<b>Since</b> 3.09.0<br>
</div>
<pre><span id="VALsscanf_format"><span class="keyword">val</span> sscanf_format</span> : <code class="type">string -><br> ('a, 'b, 'c, 'd, 'e, 'f) format6 -><br> (('a, 'b, 'c, 'd, 'e, 'f) format6 -> 'g) -> 'g</code></pre><div class="info ">
Same as <a href="Scanf.html#VALbscanf_format"><code class="code"><span class="constructor">Scanf</span>.bscanf_format</code></a>, but reads from the given string.<br>
<b>Since</b> 3.09.0<br>
</div>
<pre><span id="VALformat_from_string"><span class="keyword">val</span> format_from_string</span> : <code class="type">string -><br> ('a, 'b, 'c, 'd, 'e, 'f) format6 -> ('a, 'b, 'c, 'd, 'e, 'f) format6</code></pre><div class="info ">
<code class="code">format_from_string s fmt</code> converts a string argument to a format string,
according to the given format string <code class="code">fmt</code>.
Raise <code class="code"><span class="constructor">Scan_failure</span></code> if <code class="code">s</code>, considered as a format string, does not
have the same type as <code class="code">fmt</code>.<br>
<b>Since</b> 3.10.0<br>
</div>
<pre><span id="VALunescaped"><span class="keyword">val</span> unescaped</span> : <code class="type">string -> string</code></pre><div class="info ">
Return a copy of the argument with escape sequences, following the
lexical conventions of OCaml, replaced by their corresponding
special characters. If there is no escape sequence in the
argument, still return a copy, contrary to String.escaped.<br>
<b>Since</b> 4.00.0<br>
</div>
</body></html>
