thirdparty/scummvm
2
0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
scummvm/engine/lua/doc/manual.html

2908 lines
111 KiB
HTML
Raw Blame History

<HEAD>
<TITLE>Lua 3.1 Reference Manual</TITLE>
</HEAD>
<BODY>
<h1>Lua 3.1 Reference Manual</h1>
<!-- ====================================================================== -->
<HR>
<A NAME="1."></A>
<H1>1 - Introduction</H1>
<P>
Lua is an extension programming language designed to support
general procedural programming with data description
facilities.
Lua is intended to be used as a light-weight, but powerful,
configuration language for any program that needs one.
Lua has been designed and implemented by
W.&nbsp;Celes,
R.&nbsp;Ierusalimschy and
L.&nbsp;H.&nbsp;de Figueiredo.
<P>
Lua is implemented as a library, written in C.
Being an extension language, Lua has no notion of a ``main'' program:
it only works <EM>embedded</EM> in a host client,
called the <EM>embedding</EM> program.
This host program can invoke functions to execute a piece of
code in Lua, can write and read Lua variables,
and can register C functions to be called by Lua code.
Through the use of C functions, Lua can be augmented to cope with
a wide range of different domains,
thus creating customized programming languages sharing a syntactical framework.
<P>
Lua is free-distribution software,
and provided as usual with no guarantees,
as stated in the copyright notice.
The implementation described in this manual is available
at the following URL's:
<PRE>
<A HREF="http://www.tecgraf.puc-rio.br/lua/">http://www.tecgraf.puc-rio.br/lua/</A>
<A HREF="ftp://ftp.tecgraf.puc-rio.br/pub/lua/lua.tar.gz">ftp://ftp.tecgraf.puc-rio.br/pub/lua/lua.tar.gz</A>
</PRE>
<P>
<P>
<!-- ====================================================================== -->
<HR>
<A NAME="2."></A>
<H1>2 - Environment and Chunks</H1>
<P>
All statements in Lua are executed in a <A NAME="global environment"><EM>global environment</EM></A>.
This environment, which keeps all global variables,
is initialized at the beginning of the embedding program and
persists until its end.
Optionally, a user can create multiple independent global
environments (see Section&nbsp;<A HREF="#mangstate">5.1</A>).
<P>
The global environment can be manipulated by Lua code or
by the embedding program,
which can read and write global variables
using functions from the API library that implements Lua.
<P>
<A NAME="Global variables">Global variables</A> do not need declaration.
Any variable is assumed to be global unless explicitly declared local
(see Section&nbsp;<A HREF="#localvar">4.5.5</A>).
Before the first assignment, the value of a global variable is <B>nil</B>;
this default can be changed (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
<P>
The unit of execution of Lua is called a <A NAME="chunk"><EM>chunk</EM></A>.
A chunk is simply a sequence of statements:
<PRE>
chunk ::= {stat} [ret]
</PRE>
Statements are described in Section&nbsp;<A HREF="#stats">4.5</A>.
(As usual, {<EM>a</EM>} means 0 or more <EM>a</EM>'s,
[<EM>a</EM>] means an optional <EM>a</EM> and ('<EM>a</EM>)+ means
one or more <EM>a</EM>'s.)
<P>
A chunk may be in a file or in a string inside the host program.
A chunk may optionally end with a <CODE>return</CODE> statement (see Section&nbsp;<A HREF="#return">4.5.3</A>).
When a chunk is executed, first all its code is pre-compiled,
then the statements are executed in sequential order.
All modifications a chunk effects on the global environment persist
after its end.
<P>
Chunks may also be pre-compiled into binary form;
see program <A NAME="luac"><TT><A HREF="luac.html">luac</A></TT></A> for details.
Text files with chunks and their binary pre-compiled forms
are interchangeable.
Lua automatically detects the file type and acts accordingly.
<A NAME="pre-compilation"></A>
<P>
<A NAME="TypesSec"></A>
<!-- ====================================================================== -->
<HR>
<A NAME="3."></A>
<H1>3 - Types and Tags</H1>
<P>
Lua is a dynamically typed language.
Variables do not have types; only values do.
Therefore, there are no type definitions in the language.
All values carry their own type.
Besides a type, all values also have a <A NAME="tag">tag</A>.
<P>
There are six <A NAME="basic types">basic types</A> in Lua: <A NAME="nil"><EM>nil</EM></A>, <A NAME="number"><EM>number</EM></A>,
<A NAME="string"><EM>string</EM></A>, <A NAME="function"><EM>function</EM></A>, <A NAME="userdata"><EM>userdata</EM></A>, and <A NAME="table"><EM>table</EM></A>.
<EM>Nil</EM> is the type of the value <B>nil</B>,
whose main property is to be different from any other value.
<EM>Number</EM> represents real (double precision floating point) numbers,
while <EM>string</EM> has the usual meaning.
Lua is <A NAME="eight-bit clean">eight-bit clean</A>,
and so strings may contain any 8-bit character,
<EM>including</EM> embedded zeros (<CODE>'\0'</CODE>).
The function <CODE>type</CODE> returns a string describing the type
of a given value (see Section&nbsp;<A HREF="#pdf-type">6.1</A>).
<P>
Functions are considered first-class values in Lua.
This means that functions can be stored in variables,
passed as arguments to other functions, and returned as results.
Lua can call (and manipulate) functions written in Lua and
functions written in C.
They can be distinguished by their tags:
all Lua functions have the same tag,
and all C functions have the same tag,
which is different from the tag of Lua functions.
<P>
The type <EM>userdata</EM> is provided to allow
arbitrary <A NAME="C pointers">C pointers</A> to be stored in Lua variables.
It corresponds to a <CODE>void*</CODE> and has no pre-defined operations in Lua,
besides assignment and equality test.
However, by using <EM>tag methods</EM>,
the programmer can define operations for <EM>userdata</EM> values
(see Section&nbsp;<A HREF="#tag-method">4.8</A>).
<P>
The type <EM>table</EM> implements <A NAME="associative arrays">associative arrays</A>,
that is, <A NAME="arrays">arrays</A> that can be indexed not only with numbers,
but with any value (except <B>nil</B>).
Therefore, this type may be used not only to represent ordinary arrays,
but also symbol tables, sets, records, etc.
Tables are the main data structuring mechanism in Lua.
To represent <A NAME="records">records</A>, Lua uses the field name as an index.
The language supports this representation by
providing <CODE>a.name</CODE> as syntactic sugar for <CODE>a["name"]</CODE>.
Tables may also carry methods.
Because functions are first class values,
table fields may contain functions.
The form <CODE>t:f(x)</CODE> is syntactic sugar for <CODE>t.f(t,x)</CODE>,
which calls the method <CODE>f</CODE> from the table <CODE>t</CODE> passing
itself as the first parameter (see Section&nbsp;<A HREF="#func-def">4.6.9</A>).
<P>
Note that tables are <EM>objects</EM>, and not values.
Variables cannot contain tables, only <EM>references</EM> to them.
Assignment, parameter passing, and returns always manipulate references
to tables, and do not imply any kind of copy.
Moreover, tables must be explicitly created before used
(see Section&nbsp;<A HREF="#tableconstructor">4.6.7</A>).
<P>
Tags are mainly used to select tag methods when
some events occur.
Tag methods are the main mechanism for extending the
semantics of Lua (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
Each of the types <EM>nil</EM>, <EM>number</EM> and <EM>string</EM> has a different tag.
All values of each of these types have this same pre-defined tag.
Values of type <EM>function</EM> can have two different tags,
depending on whether they are Lua or C functions.
Finally,
values of type <EM>userdata</EM> and <EM>table</EM> can have
as many different tags as needed (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
Tags are created with the function <CODE>newtag</CODE>,
and the function <CODE>tag</CODE> returns the tag of a given value.
To change the tag of a given table,
there is the function <CODE>settag</CODE> (see Section&nbsp;<A HREF="#pdf-newtag">6.1</A>).
<P>
<P>
<!-- ====================================================================== -->
<HR>
<A NAME="4."></A>
<H1>4 - The Language</H1>
<P>
This section describes the lexis, the syntax and the semantics of Lua.
<P>
<P>
<A NAME="lexical"></A>
<A NAME="4.1"></A>
<H2>4.1 - Lexical Conventions</H2>
<P>
<A NAME="Identifiers">Identifiers</A> in Lua can be any string of letters,
digits, and underscores,
not beginning with a digit.
The definition of letter depends on the current locale:
Any character considered alphabetic by the current locale
can be used in an identifier.
The following words are reserved, and cannot be used as identifiers:
<A NAME="reserved words"></A>
<PRE>
and do else elseif
end function if local
nil not or repeat
return then until while
</PRE>
Lua is a case-sensitive language:
<TT>and</TT> is a reserved word, but <TT>And</TT> and <TT>\'and</TT>
(if the locale permits) are two other different identifiers.
As a convention, identifiers starting with underscore followed by
uppercase letters should not be used in regular programs.
<P>
The following strings denote other <A NAME="tokens">tokens</A>:
<PRE>
~= &lt;= &gt;= &lt; &gt; == = + - * /
( ) { } [ ] ; , . .. ...
</PRE>
<P>
<A NAME="Literal strings">Literal strings</A> can be delimited by matching single or double quotes,
and can contain the C-like escape sequences
<CODE>'\a'</CODE> (bell),
<CODE>'\b'</CODE> (back space),
<CODE>'\f'</CODE> (form feed),
<CODE>'\n'</CODE> (new line),
<CODE>'\r'</CODE> (carriage return),
<CODE>'\t'</CODE> (horizontal tab),
<CODE>'\v'</CODE> (vertical tab),
<CODE>'\\'</CODE>, (backslash),
<CODE>'\"'</CODE>, (double quote),
and <CODE>'\''</CODE> (single quote).
A character in a string may also be specified by its numerical value,
through the escape sequence <CODE>'\ddd'</CODE>,
where <CODE>ddd</CODE> is a sequence of up to three <EM>decimal</EM> digits.
Strings in Lua may contain any 8-bit value, including embedded 0.
<P>
Literal strings can also be delimited by matching <CODE>[[ ... ]]</CODE>.
Literals in this bracketed form may run for several lines,
may contain nested <CODE>[[ ... ]]</CODE> pairs,
and do not interpret escape sequences.
This form is specially convenient for
writing strings that contain program pieces or
other quoted strings.
As an example, in a system using ASCII,
the following three literals are equivalent:
<PRE>
1) "alo\n123\""
2) '\97lo\10\04923"'
3) [[alo
123"]]
</PRE>
<P>
<P>
<A NAME="Comments">Comments</A> start anywhere outside a string with a
double hyphen (<CODE>--</CODE>) and run until the end of the line.
Moreover,
the first line of a chunk is skipped if it starts with <CODE>#</CODE>.
This facility allows the use of Lua as a script interpreter
in Unix systems (see Section&nbsp;<A HREF="#lua-sa">8</A>).
<P>
<A NAME="Numerical constants">Numerical constants</A> may be written with an optional decimal part,
and an optional decimal exponent.
Examples of valid numerical constants are:
<PRE>
4 4.0 0.4 4.57e-3 0.3e12
</PRE>
<P>
<A NAME="pre-processor"></A>
<A NAME="4.2"></A>
<H2>4.2 - The Pre-processor</H2>
<P>
All lines that start with a <CODE>$</CODE> sign are handled by a pre-processor.
The <CODE>$</CODE> sign must be immediately
followed by one of the following directives:
<DL>
<DT><B><TT>debug</TT></B><DD> - turn on debugging facilities (see Section&nbsp;<A HREF="#pragma">4.9</A>).
<DT><B><TT>nodebug</TT></B><DD> - turn off debugging facilities (see Section&nbsp;<A HREF="#pragma">4.9</A>).
<DT><B><TT>if <EM>cond</TT></EM></B><DD> - starts a conditional part.
If <EM>cond</EM> is false, then this part is skipped by the lexical analyzer.
<DT><B><TT>ifnot <EM>cond</TT></EM></B><DD> - starts a conditional part.
If <EM>cond</EM> is true, then this part is skipped by the lexical analyzer.
<DT><B><TT>end</TT></B><DD> - ends a conditional part.
<DT><B><TT>else</TT></B><DD> - starts an ``else'' conditional part,
flipping the ``skip'' status.
<DT><B><TT>endinput</TT></B><DD> - ends the lexical parse of the file.
</DL>
<P>
Directives may be freely nested.
Particularly, a <CODE>$endinput</CODE> may occur inside a <CODE>$if</CODE>;
in that case, even the matching <CODE>$end</CODE> is not parsed.
<P>
A <EM>cond</EM> part may be:
<DL>
<DT><B><TT>nil</TT></B><DD> - always false.
<DT><B><TT>1</TT></B><DD> - always true.
<DT><B><EM>name</EM></B><DD> - true if the value of the
global variable <EM>name</EM> is different from <B>nil</B>.
Note that <EM>name</EM> is evaluated <EM>before</EM> the chunk starts its execution.
Therefore, actions in a chunk do not affect its own conditional directives.
</DL>
<P>
<A NAME="coercion"></A>
<A NAME="4.3"></A>
<H2>4.3 - Coercion</H2>
<P>
Lua provides some automatic conversions between values at run time.
Any arithmetic operation applied to a string tries to convert
that string to a number, following the usual rules.
Conversely, whenever a number is used when a string is expected,
that number is converted to a string, according to the following rule:
if the number is an integer, it is written without exponent or decimal point;
otherwise, it is formatted following the <CODE>%g</CODE>
conversion specification of the <CODE>printf</CODE> function in the
standard C library.
For complete control on how numbers are converted to strings,
use the <CODE>format</CODE> function (see Section&nbsp;<A HREF="#format">6.2</A>).
<P>
<P>
<A NAME="adjust"></A>
<A NAME="4.4"></A>
<H2>4.4 - Adjustment</H2>
<P>
Functions in Lua can return many values.
Because there are no type declarations,
the system does not know how many values a function will return,
or how many parameters it needs.
Therefore, sometimes, a list of values must be <EM>adjusted</EM>, at run time,
to a given length.
If there are more values than are needed,
then the last values are thrown away.
If there are more needs than values,
then the list is extended with as many <B>nil</B>'s as needed.
Adjustment occurs in multiple assignment (see Section&nbsp;<A HREF="#assignment">4.5.2</A>)
and function calls (see Section&nbsp;<A HREF="#functioncall">4.6.8</A>).
<P>
<P>
<A NAME="stats"></A>
<A NAME="4.5"></A>
<H2>4.5 - Statements</H2>
<P>
Lua supports an almost conventional set of <A NAME="statements">statements</A>,
similar to those in Pascal or C.
The conventional commands include
assignment, control structures and procedure calls.
Non-conventional commands include table constructors
(see Section&nbsp;<A HREF="#tableconstructor">4.6.7</A>),
and local variable declarations (see Section&nbsp;<A HREF="#localvar">4.5.5</A>).
<P>
<H3>4.5.1 - Blocks</H3>
A <A NAME="block">block</A> is a list of statements, which are executed sequentially.
A statement may be optionally followed by a semicolon:
<PRE>
block ::= {stat sc} [ret]
sc ::= ['<B>;</B>']
</PRE>
For syntactic reasons, a <A NAME="return"><TT>return</TT></A> statement can only be written
as the last statement of a block.
This restriction also avoids some ``statement not reached'' conditions.
<P>
A block may be explicitly delimited:
<PRE>
stat ::= <B>do</B> block <B>end</B>
</PRE>
This is useful to control the scope of local variables (see Section&nbsp;<A HREF="#localvar">4.5.5</A>).
<P>
<A NAME="assignment"></A>
<H3>4.5.2 - <A NAME="Assignment</H3>">Assignment</H3></A>
The language allows <A NAME="multiple assignment">multiple assignment</A>.
Therefore, the syntax for assignment
defines a list of variables on the left side,
and a list of expressions on the right side.
Both lists have their elements separated by commas:
<PRE>
stat ::= varlist1 '<B>=</B>' explist1
varlist1 ::= var {'<B>,</B>' var}
</PRE>
This statement first evaluates all values on the right side
and eventual indices on the left side,
and then makes the assignments.
Therefore, it can be used to exchange two values, as in
<PRE>
x, y = y, x
</PRE>
The two lists may have different lengths.
Before the assignment, the list of values is <EM>adjusted</EM> to
the length of the list of variables (see Section&nbsp;<A HREF="#adjust">4.4</A>).
<P>
A single name can denote a global or a local variable,
or a formal parameter:
<PRE>
var ::= name
</PRE>
Square brackets are used to index a table:
<PRE>
var ::= simpleexp '<B>[</B>' exp1 '<B>]</B>'
</PRE>
The <EM>simpleexp</EM> should result in a table value,
from where the field indexed by the expression
value gets the assigned value.
<P>
The syntax <CODE>var.NAME</CODE> is just syntactic sugar for
<CODE>var["NAME"]</CODE>:
<PRE>
var ::= simpleexp '<B>.</B>' name
</PRE>
<P>
The meaning of assignments and evaluations of global variables and
indexed variables can be changed by tag methods (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
Actually,
an assignment <CODE>x = val</CODE>, where <CODE>x</CODE> is a global variable,
is equivalent to a call <CODE>setglobal('x', val)</CODE>;
an assignment <CODE>t[i] = val</CODE> is equivalent to
<CODE>settable_event(t, i, val)</CODE>.
See Section&nbsp;<A HREF="#tag-method">4.8</A> for a complete description of these functions.
(Function <CODE>setglobal</CODE> is pre-defined in Lua.
Function <TT>settable_event</TT> is used only for explanatory purposes.)
<P>
<H3>4.5.3 - Control Structures</H3>
The <A NAME="condition expression">condition expression</A> of a control structure may return any value.
All values different from <B>nil</B> are considered true;
only <B>nil</B> is considered false.
<TT>if</TT>'s, <TT>while</TT>'s and <TT>repeat</TT>'s have the usual meaning.
<P>
<A NAME="while-do"></A><A NAME="repeat-until"></A><A NAME="if-then-else"></A>
<PRE>
stat ::= <B>while</B> exp1 <B>do</B> block <B>end</B> <BR> | <B>repeat</B> block <B>until</B> exp1 <BR> | <B>if</B> exp1 <B>then</B> block {elseif} [<B>else</B> block] <B>end</B>
elseif ::= <B>elseif</B> exp1 <B>then</B> block
</PRE>
<P>
A <TT>return</TT> is used to return values from a function or from a chunk.
<A NAME="return"></A>
Because they may return more than one value,
the syntax for a <A NAME="return statement">return statement</A> is:
<PRE>
ret ::= <B>return</B> [explist1] [sc]
</PRE>
<P>
<A NAME="funcstat"></A>
<H3>4.5.4 - Function Calls as Statements</H3>
Because of possible side-effects,
function calls can be executed as statements:
<PRE>
stat ::= functioncall
</PRE>
In this case, all returned values are thrown away.
Function calls are explained in Section&nbsp;<A HREF="#functioncall">4.6.8</A>.
<P>
<A NAME="localvar"></A>
<H3>4.5.5 - Local Declarations</H3>
<A NAME="Local variables">Local variables</A> may be declared anywhere inside a block.
Their scope begins after the declaration and lasts until the
end of the block.
The declaration may include an initial assignment:
<PRE>
stat ::= <B>local</B> declist [init]
declist ::= name {'<B>,</B>' name}
init ::= '<B>=</B>' explist1
</PRE>
If present, an initial assignment has the same semantics
of a multiple assignment.
Otherwise, all variables are initialized with <B>nil</B>.
<P>
<P>
<A NAME="4.6"></A>
<H2>4.6 - Expressions</H2>
<P>
<H3>4.6.1 - <A NAME="Basic Expressions</H3>">Basic Expressions</H3></A>
Basic expressions are:
<PRE>
exp ::= '<B>(</B>' exp '<B>)</B>'
exp ::= <B>nil</B>
exp ::= '<B>number</B>'
exp ::= '<B>literal</B>'
exp ::= function
exp ::= simpleexp
</PRE>
<PRE>
simpleexp ::= var
simpleexp ::= upvalue
simpleexp ::= functioncall
</PRE>
<P>
Numbers (numerical constants) and
string literals are explained in Section&nbsp;<A HREF="#lexical">4.1</A>;
variables are explained in Section&nbsp;<A HREF="#assignment">4.5.2</A>;
upvalues are explained in Section&nbsp;<A HREF="#upvalue">4.7</A>;
function definitions (<EM>function</EM>) are explained in Section&nbsp;<A HREF="#func-def">4.6.9</A>;
function call are explained in Section&nbsp;<A HREF="#functioncall">4.6.8</A>.
<P>
An access to a global variable <CODE>x</CODE> is equivalent to a
call <CODE>getglobal('x')</CODE>;
an access to an indexed variable <CODE>t[i]</CODE> is equivalent to
a call <CODE>gettable_event(t, i)</CODE>.
See Section&nbsp;<A HREF="#tag-method">4.8</A> for a description of these functions.
(Function <CODE>getglobal</CODE> is pre-defined in Lua.
Function <TT>gettable_event</TT> is used only for explanatory purposes.)
<P>
The non-terminal <EM>exp1</EM> is used to indicate that the values
returned by an expression must be adjusted to one single value:
<PRE>
exp1 ::= exp
</PRE>
<P>
<H3>4.6.2 - Arithmetic Operators</H3>
Lua supports the usual <A NAME="arithmetic operators">arithmetic operators</A>:
the binary <CODE>+</CODE> (addition),
<CODE>-</CODE> (subtraction), <CODE>*</CODE> (multiplication),
<CODE>/</CODE> (division) and <CODE>^</CODE> (exponentiation),
and unary <CODE>-</CODE> (negation).
If the operands are numbers, or strings that can be converted to
numbers (according to the rules given in Section&nbsp;<A HREF="#coercion">4.3</A>),
then all operations except exponentiation have the usual meaning.
Otherwise, an appropriate tag method is called (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
An exponentiation always calls a tag method.
The standard mathematical library redefines this method for numbers,
giving the expected meaning to <A NAME="exponentiation">exponentiation</A>
(see Section&nbsp;<A HREF="#mathlib">6.3</A>).
<P>
<H3>4.6.3 - Relational Operators</H3>
Lua provides the following <A NAME="relational operators">relational operators</A>:
<PRE>
&lt; &gt; &lt;= &gt;= ~= ==
</PRE>
All these return <B>nil</B> as false and a value different from <B>nil</B> as true.
<P>
Equality first compares the tags of its operands.
If they are different, then the result is <B>nil</B>.
Otherwise, their values are compared.
Numbers and strings are compared in the usual way.
Tables, userdata and functions are compared by reference,
that is, two tables are considered equal only if they are the same table.
The operator <CODE>~=</CODE> is exactly the negation of equality (<CODE>==</CODE>).
Note that the conversion rules of Section&nbsp;<A HREF="#coercion">4.3</A>
<EM>do not</EM> apply to equality comparisons.
Thus, <CODE>"0"==0</CODE> evaluates to false,
and <CODE>t[0]</CODE> and <CODE>t["0"]</CODE> denote different
entries in a table.
<P>
The other operators work as follows.
If both arguments are numbers, then they are compared as such.
Otherwise, if both arguments are strings,
then their values are compared using lexicographical order.
Otherwise, the ``order'' tag method is called (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
<P>
<H3>4.6.4 - Logical Operators</H3>
The <A NAME="logical operators">logical operators</A> are:
<A NAME="and"></A><A NAME="or"></A><A NAME="not"></A>
<PRE>
and or not
</PRE>
Like control structures, all logical operators
consider <B>nil</B> as false and anything else as true.
The operator <CODE>and</CODE> returns <B>nil</B> if its first argument is <B>nil</B>;
otherwise, it returns its second argument.
The operator <CODE>or</CODE> returns its first argument
if it is different from <B>nil</B>;
otherwise, it returns its second argument.
Both <CODE>and</CODE> and <CODE>or</CODE> use <A NAME="short-cut evaluation">short-cut evaluation</A>,
that is,
the second operand is evaluated only when necessary.
<P>
A useful Lua idiom is <CODE>x = x or v</CODE>,
which is equivalent to
<PRE>
if x == nil then x = v end
</PRE>
i.e., it sets <CODE>x</CODE> to a default value <CODE>v</CODE> when
<CODE>x</CODE> is not set.
<P>
<H3>4.6.5 - Concatenation</H3>
The string <A NAME="concatenation">concatenation</A> operator in Lua is
denoted by ``<A NAME=".."><TT>..</TT></A>''.
If both operands are strings or numbers, they are converted to
strings according to the rules in Section&nbsp;<A HREF="#coercion">4.3</A>.
Otherwise, the ``concat'' tag method is called (see Section&nbsp;<A HREF="#tag-method">4.8</A>).
<P>
<H3>4.6.6 - Precedence</H3>
<A NAME="Operator precedence">Operator precedence</A> follows the table below,
from the lower to the higher priority:
<PRE>
and or
&lt; &gt; &lt;= &gt;= ~= ==
..
+ -
* /
not - (unary)
^
</PRE>
All binary operators are left associative,
except for <CODE>^</CODE> (exponentiation),
which is right associative.
<P>
<A NAME="tableconstructor"></A>
<H3>4.6.7 - Table Constructors</H3>
Table <A NAME="constructors">constructors</A> are expressions that create tables;
every time a constructor is evaluated, a new table is created.
Constructors can be used to create empty tables,
or to create a table and initialize some fields.
<P>
The general syntax for constructors is:
<PRE>
tableconstructor ::= '<B>{</B>' fieldlist '<B>}</B>'
fieldlist ::= lfieldlist | ffieldlist | lfieldlist '<B>;</B>' ffieldlist | ffieldlist '<B>;</B>' lfieldlist
lfieldlist ::= [lfieldlist1]
ffieldlist ::= [ffieldlist1]
</PRE>
<P>
The form <EM>lfieldlist1</EM> is used to initialize lists.
<PRE>
lfieldlist1 ::= exp {'<B>,</B>' exp} ['<B>,</B>']
</PRE>
The expressions in the list are assigned to consecutive numerical indices,
starting with 1.
For example:
<PRE>
a = {"v1", "v2", 34}
</PRE>
is equivalent to:
<PRE>
do
local temp = {}
temp[1] = "v1"
temp[2] = "v2"
temp[3] = 34
a = temp
end
</PRE>
<P>
The form <EM>ffieldlist1</EM> initializes other fields in a table:
<PRE>
ffieldlist1 ::= ffield {'<B>,</B>' ffield} ['<B>,</B>']
ffield ::= '<B>[</B>' exp '<B>]</B>' '<B>=</B>' exp | name '<B>=</B>' exp
</PRE>
For example:
<PRE>
a = {[f(k)] = g(y), x = 1, y = 3, [0] = b+c}
</PRE>
is equivalent to:
<PRE>
do
local temp = {}
temp[f(k)] = g(y)
temp.x = 1 -- or temp["x"] = 1
temp.y = 3 -- or temp["y"] = 3
temp[0] = b+c
a = temp
end
</PRE>
An expression like <CODE>{x = 1, y = 4}</CODE> is
in fact syntactic sugar for <CODE>{["x"] = 1, ["y"] = 4}</CODE>.
<P>
Both forms may have an optional trailing comma,
and can be used in the same constructor separated by
a semi-collon.
For example, all forms below are correct:
<PRE>
x = {;}
x = {'a', 'b',}
x = {type='list'; 'a', 'b'}
x = {f(0), f(1), f(2),; n=3}
</PRE>
<P>
<A NAME="functioncall"></A>
<H3>4.6.8 - Function Calls</H3>
A <A NAME="function call">function call</A> has the following syntax:
<PRE>
functioncall ::= simpleexp args
</PRE>
First, <EM>simpleexp</EM> is evaluated.
If its value has type <EM>function</EM>,
then this function is called,
with the given arguments.
Otherwise, the ``function'' tag method is called,
having as first parameter the value of <EM>simpleexp</EM>,
and then the original call parameters.
<P>
The form:
<PRE>
functioncall ::= simpleexp '<B>:</B>' name args
</PRE>
can be used to call ``methods''.
A call <CODE>simpleexp:name(...)</CODE>
is syntactic sugar for
<PRE>
simpleexp.name(simpleexp, ...)
</PRE>
except that <CODE>simpleexp</CODE> is evaluated only once.
<P>
<PRE>
args ::= '<B>(</B>' [explist1] '<B>)</B>'
args ::= tableconstructor
args ::= '<B>literal</B>'
explist1 ::= exp1 {'<B>,</B>' exp1}
</PRE>
All argument expressions are evaluated before the call.
A call of the form <CODE>f{...}</CODE> is syntactic sugar for
<CODE>f({...})</CODE>, that is,
the parameter list is a single new table.
A call of the form <CODE>f'...'</CODE>
(or <CODE>f"..."</CODE> or <CODE>f[[...]]</CODE>) is syntactic sugar for
<CODE>f('...')</CODE>, that is,
the parameter list is a single literal string.
<P>
Because a function can return any number of results
(see Section&nbsp;<A HREF="#return">4.5.3</A>),
the number of results must be adjusted before used.
If the function is called as a statement (see Section&nbsp;<A HREF="#funcstat">4.5.4</A>),
then its return list is adjusted to&nbsp;0,
thus discarding all returned values.
If the function is called in a place that needs a single value
(syntactically denoted by the non-terminal <EM>exp1</EM>),
then its return list is adjusted to&nbsp;1,
thus discarding all returned values but the first one.
If the function is called in a place that can hold many values
(syntactically denoted by the non-terminal <EM>exp</EM>),
then no adjustment is made.
Note that the only place that can hold many values
is the last expression (or the only one) in an assignment
or in a return statement; see examples below.
<PRE>
f(); -- adjusted to 0
g(x, f()); -- f() is adjusted to 1
a,b,c = f(), x; -- f() is adjusted to 1 result (and c gets nil)
a,b,c = x, f(); -- f() is adjusted to 2
a,b,c = f(); -- f() is adjusted to 3
return f(); -- returns all values returned by f()
</PRE>
<P>
<A NAME="func-def"></A>
<H3>4.6.9 - <A NAME="Function Definitions</H3>">Function Definitions</H3></A>
<P>
The syntax for function definition is:
<PRE>
function ::= <B>function</B> '<B>(</B>' [parlist1] '<B>)</B>' block <B>end</B>
stat ::= <B>function</B> funcname '<B>(</B>' [parlist1] '<B>)</B>' block <B>end</B>
funcname ::= name | name '<B>.</B>' name
</PRE>
The statement:
<PRE>
function f (...)
...
end
</PRE>
is just syntactic sugar for:
<PRE>
f = function (...)
...
end
</PRE>
<P>
A function definition is an executable expression,
whose value has type <EM>function</EM>.
When Lua pre-compiles a chunk,
all its function bodies are pre-compiled, too.
Then, whenever Lua executes the function definition,
its upvalues are fixed (see Section&nbsp;<A HREF="#upvalue">4.7</A>),
and the function is <EM>instantiated</EM> (or ``closed'').
This function instance (or ``closure'')
is the final value of the expression.
Different instances of a same function
may have different upvalues.
<P>
Parameters act as local variables,
initialized with the argument values:
<PRE>
parlist1 ::= '<B>...</B>'
parlist1 ::= name {'<B>,</B>' name} ['<B>,</B>' '<B>...</B>']
</PRE>
<A NAME="vararg"></A>
When a function is called,
the list of <A NAME="arguments">arguments</A> is adjusted to
the length of the list of parameters (see Section&nbsp;<A HREF="#adjust">4.4</A>),
unless the function is a <A NAME="vararg"><EM>vararg</EM></A> function,
indicated by the dots (...) at the end of its parameter list.
A vararg function does not adjust its argument list;
instead, it collects any extra arguments into an implicit parameter,
called <A NAME="arg"><TT>arg</TT></A>.
This parameter is always initialized as a table,
with a field <CODE>n</CODE> with the number of extra arguments,
and the extra arguments at positions 1, 2, ...
<P>
As an example, suppose definitions like:
<PRE>
function f(a, b) end
function g(a, b, ...) end
</PRE>
Then, we have the following mapping from arguments to parameters:
<PRE>
CALL PARAMETERS
<P>
f(3) a=3, b=nil
f(3, 4) a=3, b=4
f(3, 4, 5) a=3, b=4
<P>
g(3) a=3, b=nil, arg={n=0}
g(3, 4) a=3, b=4, arg={n=0}
g(3, 4, 5, 8) a=3, b=4, arg={5, 8; n=2}
</PRE>
<P>
Results are returned using the <CODE>return</CODE> statement (see Section&nbsp;<A HREF="#return">4.5.3</A>).
If control reaches the end of a function without a return instruction,
then the function returns with no results.
<P>
There is a special syntax for defining <A NAME="methods">methods</A>,
that is, functions that have an implicit extra parameter <A NAME="self"><TT>self</TT></A>.
<PRE>
function ::= <B>function</B> name '<B>:</B>' name '<B>(</B>' [parlist1] '<B>)</B>' block <B>end</B>
</PRE>
Thus, a declaration like
<PRE>
function v:f (...)
...
end
</PRE>
is equivalent to
<PRE>
v.f = function (self, ...)
...
end
</PRE>
that is, the function gets an extra formal parameter called <CODE>self</CODE>.
Note that the variable <CODE>v</CODE> must have been
previously initialized with a table value.
<P>
<P>
<A NAME="upvalue"></A>
<A NAME="4.7"></A>
<H2>4.7 - Visibility and Upvalues</H2>
<A NAME="Visibility"></A> <A NAME="Upvalues"></A>
<P>
A function body may refer to its own local variables
(which includes its parameters) and to global variables,
as long as they are not shadowed by local
variables from enclosing functions.
A function <EM>cannot</EM> access a local
variable from an enclosing function,
since such variables may no longer exist when the function is called.
However, a function may access the <EM>value</EM> of a local variable
from an enclosing function, using <EM>upvalues</EM>.
<P>
<PRE>
upvalue ::= '<B>%</B>' name
</PRE>
An upvalue is somewhat similar to a variable expression,
but whose value is frozen when the function wherein it
appears is instantiated.
The name used in an upvalue may be the name of any variable visible
at the point where the function is defined.
<P>
Here are some examples:
<PRE>
a,b,c = 1,2,3 -- global variables
function f (x)
local b -- x and b are local to f
local g = function (a)
local y -- a and y are local to g
p = a -- OK, access local 'a'
p = c -- OK, access global 'c'
p = b -- ERROR: cannot access a variable in outer scope
p = %b -- OK, access frozen value of 'b' (local to 'f')
p = %c -- OK, access frozen value of global 'c'
p = %y -- ERROR: 'y' is not visible where 'g' is defined
end -- g
end -- f
</PRE>
<P>
<P>
<A NAME="tag-method"></A>
<A NAME="4.8"></A>
<H2>4.8 - Tag Methods</H2>
<P>
Lua provides a powerful mechanism to extend its semantics,
called <A NAME="Tag Methods"><EM>Tag Methods</EM></A>.
A tag method is a programmer-defined function
that is called at specific key points during the evaluation of a program,
allowing the programmer to change the standard Lua behavior at these points.
Each of these points is called an <A NAME="event"><EM>event</EM></A>.
<P>
The tag method called for any specific event is selected
according to the tag of the values involved
in the event (see Section&nbsp;<A HREF="#TypesSec">3</A>).
The function <A NAME="settagmethod"><TT>settagmethod</TT></A> changes the tag method
associated with a given pair <EM>(tag, event)</EM>.
Its first parameter is the tag, the second is the event name
(a string, see below),
and the third parameter is the new method (a function),
or <B>nil</B> to restore the default behavior.
The function returns the previous tag method for that pair.
Another function, <A NAME="gettagmethod"><TT>gettagmethod</TT></A>,
receives a tag and an event name and returns the
current method associated with the pair.
<P>
Tag methods are called in the following events,
identified by the given names.
The semantics of tag methods is better explained by a Lua function
describing the behavior of the interpreter at each event.
The function not only shows when a tag method is called,
but also its arguments, its results and the default behavior.
Please notice that the code shown here is only illustrative;
the real behavior is hard coded in the interpreter,
and it is much more efficient than this simulation.
All functions used in these descriptions
(<CODE>rawgetglobal</CODE>, <CODE>tonumber</CODE>, <CODE>call</CODE>, etc)
are described in Section&nbsp;<A HREF="#predefined">6.1</A>.
<P>
<DL>
<P>
<DT><B>``add'':</B><DD><A NAME="add event"></A>
called when a <CODE>+</CODE> operation is applied to non numerical operands.
<P>
The function <CODE>getbinmethod</CODE> defines how Lua chooses a tag method
for a binary operation.
First, Lua tries the first operand.
If its tag does not define a tag method for the operation,
then Lua tries the second operand.
If it also fails, then it gets a tag method from tag&nbsp;0:
<PRE>
function getbinmethod (op1, op2, event)
return gettagmethod(tag(op1), event) or
gettagmethod(tag(op2), event) or
gettagmethod(0, event)
end
</PRE>
<PRE>
function add_event (op1, op2)
local o1, o2 = tonumber(op1), tonumber(op2)
if o1 and o2 then -- both operands are numeric
return o1+o2 -- '+' here is the primitive 'add'
else -- at least one of the operands is not numeric
local tm = getbinmethod(op1, op2, "add")
if tm then
-- call the method with both operands and an extra
-- argument with the event name
return tm(op1, op2, "add")
else -- no tag method available: default behavior
error("unexpected type at arithmetic operation")
end
end
end
</PRE>
<P>
<DT><B>``sub'':</B><DD><A NAME="sub event"></A>
called when a <CODE>-</CODE> operation is applied to non numerical operands.
Behavior similar to the <CODE>"add"</CODE> event.
<P>
<DT><B>``mul'':</B><DD><A NAME="mul event"></A>
called when a <CODE>*</CODE> operation is applied to non numerical operands.
Behavior similar to the <CODE>"add"</CODE> event.
<P>
<DT><B>``div'':</B><DD><A NAME="div event"></A>
called when a <CODE>/</CODE> operation is applied to non numerical operands.
Behavior similar to the <CODE>"add"</CODE> event.
<P>
<DT><B>``pow'':</B><DD><A NAME="pow event"></A>
called when a <CODE>^</CODE> operation is applied.
<PRE>
function pow_event (op1, op2)
local tm = getbinmethod(op1, op2, "pow")
if tm then
-- call the method with both operands and an extra
-- argument with the event name
return tm(op1, op2, "pow")
else -- no tag method available: default behavior
error("unexpected type at arithmetic operation")
end
end
</PRE>
<P>
<DT><B>``unm'':</B><DD><A NAME="unm event"></A>
called when an unary <CODE>-</CODE> operation is applied to a non numerical operand.
<PRE>
function unm_event (op)
local o = tonumber(op)
if o then -- operand is numeric
return -o -- '-' here is the primitive 'unm'
else -- the operand is not numeric.
-- Try to get a tag method from the operand;
-- if it does not have one, try a "global" one (tag 0)
local tm = gettagmethod(tag(op), "unm") or
gettagmethod(0, "unm")
if tm then
-- call the method with the operand, nil, and an extra
-- argument with the event name
return tm(op, nil, "unm")
else -- no tag method available: default behavior
error("unexpected type at arithmetic operation")
end
end
end
</PRE>
<P>
<DT><B>``lt'':</B><DD><A NAME="lt event"></A>
called when a <CODE>&lt;</CODE> operation is applied to non numerical
or non string operands.
<PRE>
function lt_event (op1, op2)
if type(op1) == "number" and type(op2) == "number" then
return op1 &lt; op2 -- numeric comparison
elseif type(op1) == "string" and type(op2) == "string" then
return op1 &lt; op2 -- lexicographic comparison
else
local tm = getbinmethod(op1, op2, "lt")
if tm then
return tm(op1, op2, "lt")
else
error("unexpected type at comparison");
end
end
end
</PRE>
<P>
<DT><B>``gt'':</B><DD><A NAME="gt event"></A>
called when a <CODE>&gt;</CODE> operation is applied to non numerical
or non string operands.
Behavior similar to the <CODE>"lt"</CODE> event.
<P>
<DT><B>``le'':</B><DD><A NAME="le event"></A>
called when a <CODE>&lt;=</CODE> operation is applied to non numerical
or non string operands.
Behavior similar to the <CODE>"lt"</CODE> event.
<P>
<DT><B>``ge'':</B><DD><A NAME="ge event"></A>
called when a <CODE>&gt;=</CODE> operation is applied to non numerical
or non string operands.
Behavior similar to the <CODE>"lt"</CODE> event.
<P>
<DT><B>``concat'':</B><DD><A NAME="concatenation event"></A>
called when a concatenation is applied to non string operands.
<PRE>
function concat_event (op1, op2)
if (type(op1) == "string" or type(op1) == "number") and
(type(op2) == "string" or type(op2) == "number") then
return op1..op2 -- primitive string concatenation
else
local tm = getbinmethod(op1, op2, "concat")
if tm then
return tm(op1, op2, "concat")
else
error("unexpected type for concatenation")
end
end
end
</PRE>
<P>
<DT><B>``index'':</B><DD><A NAME="index event"></A>
called when Lua tries to retrieve the value of an index
not present in a table.
See event <CODE>"gettable"</CODE> for its semantics.
<P>
<DT><B>``getglobal'':</B><DD><A NAME="getglobal event"></A>
called whenever Lua needs the value of a global variable.
This method can only be set for <B>nil</B> and for tags
created by <CODE>newtag</CODE>.
<PRE>
function getglobal (varname)
local value = rawgetglobal(varname)
local tm = gettagmethod(tag(value), "getglobal")
if not tm then
return value
else
return tm(varname, value)
end
end
</PRE>
The function <CODE>getglobal</CODE> is pre-defined in Lua (see Section&nbsp;<A HREF="#predefined">6.1</A>).
<P>
<DT><B>``setglobal'':</B><DD><A NAME="setglobal event"></A>
called whenever Lua assigns to a global variable.
This method cannot be set for numbers, strings, and tables and
userdata with default tags.
<PRE>
function setglobal (varname, newvalue)
local oldvalue = rawgetglobal(varname)
local tm = gettagmethod(tag(oldvalue), "setglobal")
if not tm then
return rawsetglobal(varname, newvalue)
else
return tm(varname, oldvalue, newvalue)
end
end
</PRE>
Notice: the function <CODE>setglobal</CODE> is pre-defined in Lua (see Section&nbsp;<A HREF="#predefined">6.1</A>).
<P>
<DT><B>``gettable'':</B><DD><A NAME="gettable event"></A>
called whenever Lua accesses an indexed variable.
This method cannot be set for tables with default tag.
<PRE>
function gettable_event (table, index)
local tm = gettagmethod(tag(table), "gettable")
if tm then
return tm(table, index)
elseif type(table) ~= "table" then
error("indexed expression not a table");
else
local v = rawgettable(table, index)
tm = gettagmethod(tag(table), "index")
if v == nil and tm then
return tm(table, index)
else
return v
end
end
end
</PRE>
<P>
<DT><B>``settable'':</B><DD><A NAME="settable event"></A>
called when Lua assigns to an indexed variable.
This method cannot be set for tables with default tag.
<PRE>
function settable_event (table, index, value)
local tm = gettagmethod(tag(table), "settable")
if tm then
tm(table, index, value)
elseif type(table) ~= "table" then
error("indexed expression not a table")
else
rawsettable(table, index, value)
end
end
</PRE>
<P>
<DT><B>``function'':</B><DD><A NAME="function event"></A>
called when Lua tries to call a non function value.
<PRE>
function function_event (func, ...)
if type(func) == "function" then
return call(func, arg)
else
local tm = gettagmethod(tag(func), "function")
if tm then
local i = arg.n
while i &gt; 0 do
arg[i+1] = arg[i]
i = i-1
end
arg.n = arg.n+1
arg[1] = func
return call(tm, arg)
else
error("call expression not a function")
end
end
end
</PRE>
<P>
<DT><B>``gc'':</B><DD><A NAME="gc event"></A>
called when Lua is garbage collecting an object.
This method cannot be set for strings, numbers, functions,
and userdata with default tag.
For each object to be collected,
Lua does the equivalent of the following function:
<PRE>
function gc_event (obj)
local tm = gettagmethod(tag(obj), "gc")
if tm then
tm(obj)
end
end
</PRE>
Moreover, at the end of a garbage collection cycle,
Lua does the equivalent of the call <CODE>gc_event(nil)</CODE>.
<P>
</DL>
<P>
<P>
<P>
<A NAME="error"></A>
<A NAME="4.9"></A>
<H2>4.9 - Error Handling</H2>
<P>
Because Lua is an extension language,
all Lua actions start from C code in the host program
calling a function from the Lua library.
Whenever an error occurs during Lua compilation or execution,
the <A NAME="error method"><EM>error method</EM></A> is called,
and then the corresponding function from the library
(<CODE>lua_dofile</CODE>, <CODE>lua_dostring</CODE>,
<CODE>lua_dobuffer</CODE>, or <CODE>lua_callfunction</CODE>)
is terminated, returning an error condition.
<P>
The only argument to the error method is a string
describing the error.
The default method prints this message to <CODE>stderr</CODE>.
If needed, it is possible to change the error method with the
function <CODE>seterrormethod</CODE>,
which gets the new error handler as its only parameter
(see Section&nbsp;<A HREF="#pdf-seterrormethod">6.1</A>).
The standard I/O library uses this facility to redefine the error method,
using the debug facilities (see Section&nbsp;<A HREF="#debugI">7</A>),
in order to print some extra information,
such as the call stack.
<P>
To provide more information about errors,
Lua programs should include the compilation pragma <CODE>$debug</CODE>.
<A NAME="pragma"></A>
<A NAME="debug pragma"></A>
When an error occurs in a program compiled with this option,
the I/O error routine is able to print the number of the
lines where the calls (and the error) were made.
<P>
Lua code can explicitly generate an error by calling the built-in
function <CODE>error</CODE> (see Section&nbsp;<A HREF="#pdf-error">6.1</A>).
Lua code can ``catch'' an error using the built-in function
<CODE>call</CODE> (see Section&nbsp;<A HREF="#pdf-call">6.1</A>).
<P>
<P>
<P>
<!-- ====================================================================== -->
<HR>
<A NAME="5."></A>
<H1>5 - The Application Program Interface</H1>
<P>
This section describes the API for Lua, that is,
the set of C functions available to the host program to communicate
with the Lua library.
The API functions can be classified in the following categories:
<OL>
<LI>managing states;
<LI>exchanging values between C and Lua;
<LI>executing Lua code;
<LI>manipulating (reading and writing) Lua objects;
<LI>calling Lua functions;
<LI>C functions to be called by Lua;
<LI>manipulating references to Lua Objects.
</OL>
All API functions and related types and constants
are declared in the header file <CODE>lua.h</CODE>.
<P>
<A NAME="mangstate"></A>
<A NAME="5.1"></A>
<H2>5.1 - Managing States</H2>
The whole state of the Lua interpreter
(global variables, stack, tag methods, etc)
is stored in a dynamic structure pointed by<A NAME="lua_state"></A>
<PRE>
typedef struct lua_State lua_State;
extern lua_State *lua_state;
</PRE>
<P>
Before calling any API function,
this state must be initialized.
This is done by calling<A NAME="lua_open"></A>
<PRE>
void lua_open (void);
</PRE>
This function allocates and initializes some internal structures,
and defines all pre-defined functions of Lua.
If <CODE>lua_state</CODE> is already different from <CODE>NULL</CODE>,
<CODE>lua_open</CODE> has no effect;
therefore, it is safe to call this function multiple times.
All standard libraries call <CODE>lua_open</CODE> when they are opened.
<P>
Function <CODE>lua_setstate</CODE> is used to change the current state
of Lua:<A NAME="lua_setstate"></A>
<PRE>
lua_State *lua_setstate (lua_State *st);
</PRE>
It sets <CODE>lua_state</CODE> to <CODE>st</CODE> and returns the old state.
<P>
Multiple, independent states may be created.
For that, you must set <CODE>lua_state</CODE> back to <CODE>NULL</CODE> before
calling <CODE>lua_open</CODE>.
An easy way to do that is defining an auxiliary function:
<PRE>
lua_State *lua_newstate (void) {
lua_State *old = lua_setstate(NULL);
lua_open();
return lua_setstate(old);
}
</PRE>
This function creates a new state without changing the current state
of the interpreter.
Note that any new state is built with all predefined functions,
but any additional library (such as the standard libraries) must be
explicitly open in the new state, if needed.
<P>
If necessary, a state may be released:<A NAME="lua_close"></A>
<PRE>
void lua_close (void);
</PRE>
This function destroys all objects in the current Lua environment
(calling the correspondent garbage collector tag methods),
frees all dynamic memory used by the state,
and then sets <CODE>lua_state</CODE> to <CODE>NULL</CODE>.
Usually, there is no need to call this function,
since these resources are naturally released when the program ends.
If <CODE>lua_state</CODE> is already <CODE>NULL</CODE>,
<CODE>lua_close</CODE> has no effect.
<P>
If you are using multiple states,
you may find useful the following function,
which releases a given state:
<PRE>
void lua_freestate (lua_State *st) {
lua_State *old = lua_setstate(st);
lua_close();
if (old != st) lua_setstate(old);
}
</PRE>
<P>
<A NAME="valuesCLua"></A>
<A NAME="5.2"></A>
<H2>5.2 - Exchanging Values between C and Lua</H2>
Because Lua has no static type system,
all values passed between Lua and C have type
<CODE>lua_Object</CODE><A NAME="lua_Object"></A>,
which works like an abstract type in C that can hold any Lua value.
Values of type <CODE>lua_Object</CODE> have no meaning outside Lua;
for instance,
the comparison of two <CODE>lua_Object's</CODE> is undefined.
<P>
To check the type of a <CODE>lua_Object</CODE>,
the following functions are available:
<A NAME="lua_isnil"></A><A NAME="lua_isnumber"></A><A NAME="lua_isstring"></A>
<A NAME="lua_istable"></A><A NAME="lua_iscfunction"></A><A NAME="lua_isuserdata"></A>
<A NAME="lua_isfunction"></A>
<PRE>
int lua_isnil (lua_Object object);
int lua_isnumber (lua_Object object);
int lua_isstring (lua_Object object);
int lua_istable (lua_Object object);
int lua_isfunction (lua_Object object);
int lua_iscfunction (lua_Object object);
int lua_isuserdata (lua_Object object);
</PRE>
All macros return 1 if the object is compatible with the given type,
and 0 otherwise.
The function <CODE>lua_isnumber</CODE> accepts numbers and numerical strings,
whereas
<CODE>lua_isstring</CODE> accepts strings and numbers (see Section&nbsp;<A HREF="#coercion">4.3</A>),
and <CODE>lua_isfunction</CODE> accepts Lua functions and C functions.
<P>
To get the tag of a <CODE>lua_Object</CODE>,
the following function is available:
<A NAME="lua_tag"></A>
<PRE>
int lua_tag (lua_Object object);
</PRE>
<P>
To translate a value from type <CODE>lua_Object</CODE> to a specific C type,
the programmer can use:
<A NAME="lua_getnumber"></A><A NAME="lua_getstring"></A><A NAME="lua_strlen"></A>
<A NAME="lua_getcfunction"></A><A NAME="lua_getuserdata"></A>
<PRE>
double lua_getnumber (lua_Object object);
char *lua_getstring (lua_Object object);
long lua_strlen (lua_Object object);
lua_CFunction lua_getcfunction (lua_Object object);
void *lua_getuserdata (lua_Object object);
</PRE>
<P>
<CODE>lua_getnumber</CODE> converts a <CODE>lua_Object</CODE> to a floating-point number.
This <CODE>lua_Object</CODE> must be a number or a string convertible to number
(see Section&nbsp;<A HREF="#coercion">4.3</A>); otherwise, <CODE>lua_getnumber</CODE> returns&nbsp;0.
<P>
<CODE>lua_getstring</CODE> converts a <CODE>lua_Object</CODE> to a string (<CODE>char*</CODE>).
This <CODE>lua_Object</CODE> must be a string or a number;
otherwise, the function returns&nbsp;0 (the <CODE>NULL</CODE> pointer).
This function does not create a new string,
but returns a pointer to a string inside the Lua environment.
Those strings always have a 0 after their last character (like in C),
but may contain other zeros in their body.
If you do not know whether a string may contain zeros,
you can use <CODE>lua_strlen</CODE> to get the actual length.
Because Lua has garbage collection,
there is no guarantee that such pointer will be valid after the block ends
(see Section&nbsp;<A HREF="#GC">5.3</A>).
<P>
<CODE>lua_getcfunction</CODE> converts a <CODE>lua_Object</CODE> to a C function.
This <CODE>lua_Object</CODE> must have type <EM>CFunction</EM>;
otherwise, <CODE>lua_getcfunction</CODE> returns 0 (the <CODE>NULL</CODE> pointer).
The type <CODE>lua_CFunction</CODE> is explained in Section&nbsp;<A HREF="#LuacallC">5.7</A>.
<P>
<CODE>lua_getuserdata</CODE> converts a <CODE>lua_Object</CODE> to <CODE>void*</CODE>.
This <CODE>lua_Object</CODE> must have type <EM>userdata</EM>;
otherwise, <CODE>lua_getuserdata</CODE> returns 0 (the <CODE>NULL</CODE> pointer).
<P>
<A NAME="GC"></A>
<A NAME="5.3"></A>
<H2>5.3 - Garbage Collection</H2>
Because Lua has automatic memory management and garbage collection,
a <CODE>lua_Object</CODE> has a limited scope,
and is only valid inside the <EM>block</EM> where it has been created.
A C function called from Lua is a block,
and its parameters are valid only until its end.
It is good programming practice to convert Lua objects to C values
as soon as they are available,
and never to store <CODE>lua_Object</CODE>s in C global variables.
<P>
A garbage collection cycle can be forced by:
<A NAME="lua_collectgarbage"></A>
<PRE>
long lua_collectgarbage (long limit);
</PRE>
This function returns the number of objects collected.
The argument <CODE>limit</CODE> makes the next cycle occur only
after that number of new objects have been created.
If <CODE>limit</CODE>=0, then Lua uses an adaptive heuristics to set this limit.
<P>
<P>
All communication between Lua and C is done through two
abstract data types, called <A NAME="lua2C"><EM>lua2C</EM></A> and <A NAME="C2lua"><EM>C2lua</EM></A>.
The first one, as the name implies, is used to pass values
from Lua to C:
parameters when Lua calls C and results when C calls Lua.
The structure C2lua is used in the reverse direction:
parameters when C calls Lua and results when Lua calls C.
<P>
The structure lua2C is an abstract array,
which can be indexed with the function:
<A NAME="lua_lua2C"></A>
<PRE>
lua_Object lua_lua2C (int number);
</PRE>
where <CODE>number</CODE> starts with 1.
When called with a number larger than the array size,
this function returns <CODE>LUA_NOOBJECT</CODE><A NAME="LUA_NOOBJECT"></A>.
In this way, it is possible to write C functions that receive
a variable number of parameters,
and to call Lua functions that return a variable number of results.
Note that the structure lua2C cannot be directly modified by C code.
<P>
The second structure, C2lua, is an abstract stack.
Pushing elements into this stack
is done with the following functions and macros:
<A NAME="lua_pushnumber"></A><A NAME="lua_pushlstring"></A><A NAME="lua_pushstring"></A>
<A NAME="lua_pushcfunction"></A><A NAME="lua_pushusertag"></A>
<A NAME="lua_pushnil"></A><A NAME="lua_pushobject"></A>
<A NAME="pushing"></A>
<A NAME="lua_pushuserdata"></A>
<PRE>
void lua_pushnumber (double n);
void lua_pushlstring (char *s, long len);
void lua_pushstring (char *s);
void lua_pushusertag (void *u, int tag);
void lua_pushnil (void);
void lua_pushobject (lua_Object object);
void lua_pushcfunction (lua_CFunction f); /* macro */
</PRE>
All of them receive a C value,
convert it to a corresponding <CODE>lua_Object</CODE>,
and leave the result on the top of C2lua.
Particularly, functions <CODE>lua_pushlstring</CODE> and <CODE>lua_pushstring</CODE>
make an internal copy of the given string.
Function <CODE>lua_pushstring</CODE> can only be used to push proper C strings
(that is, strings that do not contain zeros and end with a zero);
otherwise you should use the more generic <CODE>lua_pushlstring</CODE>.
The function
<A NAME="lua_pop"></A>
<PRE>
lua_Object lua_pop (void);
</PRE>
returns a reference to the object at the top of the C2lua stack,
and pops it.
<P>
As a general rule, all API functions pop from the stack
all elements they use.
<P>
Because userdata are objects,
the function <CODE>lua_pushusertag</CODE> may create a new userdata.
If Lua has a userdata with the given value (<CODE>void*</CODE>) and tag,
that userdata is pushed.
Otherwise, a new userdata is created, with the given value and tag.
If this function is called with
<CODE>tag</CODE> equal to <CODE>LUA_ANYTAG</CODE><A NAME="LUA_ANYTAG"></A>,
then Lua will try to find any userdata with the given value,
regardless of its tag.
If there is no userdata with that value, then a new one is created,
with tag equal to 0.
<P>
Userdata can have different tags,
whose semantics are only known to the host program.
Tags are created with the function:
<A NAME="lua_newtag"></A>
<PRE>
int lua_newtag (void);
</PRE>
The function <CODE>lua_settag</CODE> changes the tag of
the object on the top of C2lua (and pops it);
the object must be a userdata or a table.
<A NAME="lua_settag"></A>
<PRE>
void lua_settag (int tag);
</PRE>
<CODE>tag</CODE> must be a value created with <CODE>lua_newtag</CODE>.
<P>
When C code calls Lua repeatedly, as in a loop,
objects returned by these calls can accumulate,
and may cause a stack overflow.
To avoid this,
nested blocks can be defined with the functions:
<PRE>
void lua_beginblock (void);
void lua_endblock (void);
</PRE>
After the end of the block,
all <CODE>lua_Object</CODE>'s created inside it are released.
The use of explicit nested blocks is good programming practice
and is strongly encouraged.
<P>
<A NAME="5.4"></A>
<H2>5.4 - Executing Lua Code</H2>
A host program can execute Lua chunks written in a file or in a string
using the following functions:
<A NAME="lua_dofile"></A><A NAME="lua_dostring"></A><A NAME="lua_dobuffer"></A>
<PRE>
int lua_dofile (char *filename);
int lua_dostring (char *string);
int lua_dobuffer (char *buff, int size, char *name);
</PRE>
All these functions return an error code:
0, in case of success; non zero, in case of errors.
More specifically, <CODE>lua_dofile</CODE> returns 2 if for any reason
it could not open the file.
When called with argument <CODE>NULL</CODE>,
<CODE>lua_dofile</CODE> executes the <CODE>stdin</CODE> stream.
Functions <CODE>lua_dofile</CODE> and <CODE>lua_dobuffer</CODE>
are both able to execute pre-compiled chunks.
They automatically detect whether the chunk is text or binary,
and load it accordingly (see program <A NAME="luac"><TT><A HREF="luac.html">luac</A></TT></A>).
Function <CODE>lua_dostring</CODE> executes only source code.
<P>
The third parameter to <CODE>lua_dobuffer</CODE> (<CODE>name</CODE>)
is the ``name of the chunk'',
used in error messages and debug information.
If <CODE>name</CODE> is <CODE>NULL</CODE>,
Lua gives a default name to the chunk.
In files this name is the file name,
and <CODE>lua_dostring</CODE> uses a small prefix
of the string as the chunk name.
<P>
These functions return, in structure lua2C,
any values eventually returned by the chunks.
They also empty the stack C2lua.
<P>
<P>
<A NAME="5.5"></A>
<H2>5.5 - Manipulating Lua Objects</H2>
To read the value of any global Lua variable,
one uses the function:
<A NAME="lua_getglobal"></A>
<PRE>
lua_Object lua_getglobal (char *varname);
</PRE>
As in Lua, this function may trigger a tag method.
To read the real value of any global variable,
without invoking any tag method,
use the <EM>raw</EM> version:
<A NAME="lua_rawgetglobal"></A>
<PRE>
lua_Object lua_rawgetglobal (char *varname);
</PRE>
<P>
To store a value previously pushed onto C2lua in a global variable,
there is the function:
<A NAME="lua_setglobal"></A>
<PRE>
void lua_setglobal (char *varname);
</PRE>
As in Lua, this function may trigger a tag method.
To set the real value of any global variable,
without invoking any tag method,
use the <EM>raw</EM> version:
<A NAME="lua_rawgetglobal"></A>
<PRE>
void lua_rawsetglobal (char *varname);
</PRE>
<P>
Tables can also be manipulated via the API.
The function
<A NAME="lua_gettable"></A>
<PRE>
lua_Object lua_gettable (void);
</PRE>
pops from the stack C2lua a table and an index,
and returns the contents of the table at that index.
As in Lua, this operation may trigger a tag method.
To get the real value of any table index,
without invoking any tag method,
use the <EM>raw</EM> version:
<A NAME="lua_rawgetglobal"></A>
<PRE>
lua_Object lua_rawgettable (void);
</PRE>
<P>
To store a value in an index,
the program must push the table, the index,
and the value onto C2lua,
and then call the function:
<A NAME="lua_settable"></A>
<PRE>
void lua_settable (void);
</PRE>
Again, the tag method for ``settable'' may be called.
To set the real value of any table index,
without invoking any tag method,
use the <EM>raw</EM> version:
<A NAME="lua_rawsettable"></A>
<PRE>
void lua_rawsettable (void);
</PRE>
<P>
Finally, the function
<A NAME="lua_createtable"></A>
<PRE>
lua_Object lua_createtable (void);
</PRE>
creates and returns a new, empty table.
<P>
<P>
<A NAME="5.6"></A>
<H2>5.6 - Calling Lua Functions</H2>
Functions defined in Lua by a chunk
can be called from the host program.
This is done using the following protocol:
first, the arguments to the function are pushed onto C2lua
(see Section&nbsp;<A HREF="#pushing">5.3</A>), in direct order, i.e., the first argument is pushed first.
Then, the function is called using
<A NAME="lua_callfunction"></A>
<PRE>
int lua_callfunction (lua_Object function);
</PRE>
This function returns an error code:
0, in case of success; non zero, in case of errors.
Finally, the results (a Lua function may return many values)
are returned in structure lua2C,
and can be retrieved with the macro <CODE>lua_getresult</CODE>,
<A NAME="lua_getresult"></A>
which is just another name to function <CODE>lua_lua2C</CODE>.
Note that function <CODE>lua_callfunction</CODE>
pops all elements from the C2lua stack.
<P>
The following example shows how a C program may do the
equivalent to the Lua code:
<PRE>
a,b = f("how", t.x, 4)
</PRE>
<PRE>
lua_pushstring("how"); /* 1st argument */
lua_pushobject(lua_getglobal("t")); /* push value of global 't' */
lua_pushstring("x"); /* push the string 'x' */
lua_pushobject(lua_gettable()); /* push result of t.x (2nd arg) */
lua_pushnumber(4); /* 3rd argument */
lua_callfunction(lua_getglobal("f")); /* call Lua function */
lua_pushobject(lua_getresult(1)); /* push first result of the call */
lua_setglobal("a"); /* sets global variable 'a' */
lua_pushobject(lua_getresult(2)); /* push second result of the call */
lua_setglobal("b"); /* sets global variable 'b' */
</PRE>
<P>
Some special Lua functions have exclusive interfaces.
A C function can generate a Lua error calling the function
<A NAME="lua_error"></A>
<PRE>
void lua_error (char *message);
</PRE>
This function never returns.
If the C function has been called from Lua,
then the corresponding Lua execution terminates,
as if an error had occurred inside Lua code.
Otherwise, the whole program terminates with a call to <CODE>exit(1)</CODE>.
The <CODE>message</CODE> is passed to the error handler method.
If <CODE>message</CODE> is <CODE>NULL</CODE>,
the error handler method is not called.
<P>
The error handler method (see Section&nbsp;<A HREF="#error">4.9</A>) can be
changed with: <A NAME="lua_seterrormethod"></A>
<PRE>
lua_Object lua_seterrormethod (void);
</PRE>
This function sets the object at the top of C2lua
as the new error method,
and returns the old error method value.
<P>
Tag methods can be changed with: <A NAME="lua_settagmethod"></A>
<PRE>
lua_Object lua_settagmethod (int tag, char *event);
</PRE>
The first parameter is the tag,
and the second is the event name (see Section&nbsp;<A HREF="#tag-method">4.8</A>);
the new method is pushed from C2lua.
This function returns a <CODE>lua_Object</CODE>,
which is the old tag method value.
To get just the current value of a tag method,
use the function <A NAME="lua_gettagmethod"></A>
<PRE>
lua_Object lua_gettagmethod (int tag, char *event);
</PRE>
<P>
It is also possible to copy all tag methods from one tag
to another: <A NAME="lua_copytagmethods"></A>
<PRE>
int lua_copytagmethods (int tagto, int tagfrom);
</PRE>
This function returns <CODE>tagto</CODE>.
<P>
<P>
<A NAME="LuacallC"></A>
<A NAME="5.7"></A>
<H2>5.7 - C Functions</H2>
To register a C function to Lua,
there is the following macro:
<A NAME="lua_register"></A>
<PRE>
#define lua_register(n,f) (lua_pushcfunction(f), lua_setglobal(n))
/* char *n; */
/* lua_CFunction f; */
</PRE>
which receives the name the function will have in Lua,
and a pointer to the function.
This pointer must have type <CODE>lua_CFunction</CODE>,
which is defined as
<A NAME="lua_CFunction"></A>
<PRE>
typedef void (*lua_CFunction) (void);
</PRE>
that is, a pointer to a function with no parameters and no results.
<P>
In order to communicate properly with Lua,
a C function must follow a protocol,
which defines the way parameters and results are passed.
<P>
A C function receives its arguments in structure lua2C;
to access them, it uses the macro <CODE>lua_getparam</CODE>, <A NAME="lua_getparam"></A>
again just another name for <CODE>lua_lua2C</CODE>.
To return values, a C function just pushes them onto the stack C2lua,
in direct order (see Section&nbsp;<A HREF="#valuesCLua">5.2</A>).
Like a Lua function, a C function called by Lua can also return
many results.
<P>
When a C function is created,
it is possible to associate some <EM>upvalues</EM> to it;
then these values are passed to the function whenever it is called,
as common arguments.
To associate upvalues to a function,
first these values must be pushed on C2lua.
Then the function:
<A NAME="lua_pushcclosure"></A>
<PRE>
void lua_pushcclosure (lua_CFunction fn, int n);
</PRE>
is used to put the C function on C2lua,
with the argument <CODE>n</CODE> telling how many upvalues must be
associated with the function;
in fact, the macro <CODE>lua_pushcfunction</CODE> is defined as
<CODE>lua_pushcclosure</CODE> with <CODE>n</CODE> set to 0.
Then, any time the function is called,
these upvalues are inserted as the first arguments to the function,
before the actual arguments provided in the call.
<P>
For some examples of C functions, see files <CODE>lstrlib.c</CODE>,
<CODE>liolib.c</CODE> and <CODE>lmathlib.c</CODE> in the official Lua distribution.
<P>
<A NAME="5.8"></A>
<H2>5.8 - References to Lua Objects</H2>
<P>
As noted in Section&nbsp;<A HREF="#GC">5.3</A>, <CODE>lua_Object</CODE>s are volatile.
If the C code needs to keep a <CODE>lua_Object</CODE>
outside block boundaries,
then it must create a <A NAME="reference"><EM>reference</EM></A> to the object.
The routines to manipulate references are the following:
<A NAME="lua_ref"></A><A NAME="lua_getref"></A>
<A NAME="lua_unref"></A>
<PRE>
int lua_ref (int lock);
lua_Object lua_getref (int ref);
void lua_unref (int ref);
</PRE>
The function <CODE>lua_ref</CODE> creates a reference
to the object that is on the top of the stack,
and returns this reference.
If <CODE>lock</CODE> is true, the object is <EM>locked</EM>:
this means the object will not be garbage collected.
Note that an unlocked reference may be garbage collected.
Whenever the referenced object is needed,
a call to <CODE>lua_getref</CODE>
returns a handle to it;
if the object has been collected,
<CODE>lua_getref</CODE> returns <CODE>LUA_NOOBJECT</CODE>.
<P>
When a reference is no longer needed,
it can be released with a call to <CODE>lua_unref</CODE>.
<P>
<P>
<P>
<!-- ====================================================================== -->
<HR>
<A NAME="6."></A>
<H1>6 - Predefined Functions and Libraries</H1>
<P>
The set of <A NAME="predefined functions">predefined functions</A> in Lua is small but powerful.
Most of them provide features that allow some degree of
<A NAME="reflexivity">reflexivity</A> in the language.
Some of these features cannot be simulated with the rest of the
language nor with the standard Lua API.
Others are just convenient interfaces to common API functions.
<P>
The libraries, on the other hand, provide useful routines
that are implemented directly through the standard API.
Therefore, they are not necessary to the language,
and are provided as separate C modules.
Currently there are three standard libraries:
<UL>
<LI>string manipulation;
<LI>mathematical functions (sin, log, etc);
<LI>input and output (plus some system facilities).
</UL>
To have access to these libraries,
the C host program must call the functions
<CODE>lua_strlibopen</CODE>, <CODE>lua_mathlibopen</CODE>,
and <CODE>lua_iolibopen</CODE>, declared in <CODE>lualib.h</CODE>.
<A NAME="lua_strlibopen"></A><A NAME="lua_mathlibopen"></A><A NAME="lua_iolibopen"></A>
<P>
<P>
<A NAME="predefined"></A>
<A NAME="6.1"></A>
<H2>6.1 - Predefined Functions</H2>
<P>
<h3> <TT>call (func, arg [, mode [, errmethod]])</TT></h3><A NAME="call"></A>
<A NAME="pdf-call"></A>
This function calls function <CODE>func</CODE> with
the arguments given by the table <CODE>arg</CODE>.
The call is equivalent to
<PRE>
func(arg[1], arg[2], ..., arg[arg.n])
</PRE>
If <CODE>arg.n</CODE> is not defined,
then Lua stops getting arguments at the first <B>nil</B> value.
<P>
By default,
all results from <CODE>func</CODE> are just returned by the call.
If the string <CODE>mode</CODE> contains <CODE>"p"</CODE>,
the results are <EM>packed</EM> in a single table.<A NAME="packed results"></A>
That is, <CODE>call</CODE> returns just one table;
at index <CODE>n</CODE>, the table has the total number of results
from the call;
the first result is at index 1, etc.
For instance, the following calls produce the following results:
<PRE>
a = call(sin, {5}) --&gt; a = 0.0871557 = sin(5)
a = call(max, {1,4,5; n=2}) --&gt; a = 4 (only 1 and 4 are arguments)
t = {x=1}
a = call(next, {t,nil;n=2}, "p") --&gt; a={"x", 1; n=2}
</PRE>
<P>
By default,
if an error occurs during the function call,
the error is propagated.
If the string <CODE>mode</CODE> contains <CODE>"x"</CODE>,
then the call is <EM>protected</EM>.<A NAME="protected calls"></A>
In this mode, function <CODE>call</CODE> does not generate an error,
whatever happens during the call.
Instead, it returns <B>nil</B> to signal the error
(besides calling the appropriated error method).
<P>
If provided, <CODE>errmethod</CODE> is temporarily set as the error method,
while <CODE>func</CODE> runs.
As a particular case, if <CODE>errmethod</CODE> is <B>nil</B>,
no error messages will be issued during the execution of the called function.
<P>
<h3> <TT>collectgarbage ([limit])</TT></h3><A NAME="collectgarbage"></A>
Forces a garbage collection cycle.
Returns the number of objects collected.
An optional argument, <CODE>limit</CODE>, is a number that
makes the next cycle occur only after that number of new
objects have been created.
If absent, Lua uses an adaptive algorithm to set
this limit.
<CODE>collectgarbage</CODE> is equivalent to
the API function <CODE>lua_collectgarbage</CODE>.
<P>
<h3> <TT>dofile (filename)</TT></h3><A NAME="dofile"></A>
This function receives a file name,
opens the file, and executes the file contents as a Lua chunk,
or as pre-compiled chunks.
When called without arguments,
<CODE>dofile</CODE> executes the contents of the standard input (<CODE>stdin</CODE>).
If there is any error executing the file,
then <CODE>dofile</CODE> returns <B>nil</B>.
Otherwise, it returns the values returned by the chunk,
or a non <B>nil</B> value if the chunk returns no values.
It issues an error when called with a non string argument.
<CODE>dofile</CODE> is equivalent to the API function <CODE>lua_dofile</CODE>.
<P>
<h3> <TT>dostring (string [, chunkname])</TT></h3><A NAME="dostring"></A>
This function executes a given string as a Lua chunk.
If there is any error executing the string,
<CODE>dostring</CODE> returns <B>nil</B>.
Otherwise, it returns the values returned by the chunk,
or a non <B>nil</B> value if the chunk returns no values.
An optional second parameter (<CODE>chunkname</CODE>)
is the ``name of the chunk'',
used in error messages and debug information.
<CODE>dostring</CODE> is equivalent to the API function <CODE>lua_dostring</CODE>.
<P>
<A NAME="pdf-newtag"></A>
<h3> <TT>newtag ()</TT></h3><A NAME="newtag"></A>
Returns a new tag.
<CODE>newtag</CODE> is equivalent to the API function <CODE>lua_newtag</CODE>.
<P>
<h3> <TT>next (table, index)</TT></h3><A NAME="next"></A>
This function allows a program to traverse all fields of a table.
Its first argument is a table and its second argument
is an index in this table.
It returns the next index of the table and the
value associated with the index.
When called with <B>nil</B> as its second argument,
the function returns the first index
of the table (and its associated value).
When called with the last index,
or with <B>nil</B> in an empty table,
it returns <B>nil</B>.
<P>
Lua has no declaration of fields;
semantically, there is no difference between a
field not present in a table or a field with value <B>nil</B>.
Therefore, the function only considers fields with non <B>nil</B> values.
The order in which the indices are enumerated is not specified,
<EM>even for numeric indices</EM>
(to traverse a table in numeric order, use a counter).
If the table is modified in any way during a traversal,
the semantics of <CODE>next</CODE> is undefined.
<P>
This function cannot be written with the standard API.
<P>
<h3> <TT>nextvar (name)</TT></h3><A NAME="nextvar"></A>
This function is similar to the function <CODE>next</CODE>,
but iterates instead over the global variables.
Its single argument is the name of a global variable,
or <B>nil</B> to get a first name.
Similarly to <CODE>next</CODE>, it returns the name of another variable
and its value,
or <B>nil</B> if there are no more variables.
There can be no assignments to global variables during the traversal;
otherwise the semantics of <CODE>nextvar</CODE> is undefined.
<P>
This function cannot be written with the standard API.
<P>
<h3> <TT>foreach (table, function)</TT></h3><A NAME="foreach"></A>
Executes the given <CODE>function</CODE> over all elements of <CODE>table</CODE>.
For each element, the function is called with the index and
respective value as arguments.
If the function returns any non-<B>nil</B> value,
the loop is broken, and the value is returned
as the final value of <CODE>foreach</CODE>.
<P>
This function could be defined in Lua:
<PRE>
function foreach (t, f)
local i, v = next(t, nil)
while i do
local res = f(i, v)
if res then return res end
i, v = next(t, i)
end
end
</PRE>
<P>
<h3> <TT>foreachvar (function)</TT></h3><A NAME="foreachvar"></A>
Executes <CODE>function</CODE> over all global variables.
For each variable,
the function is called with its name and its value as arguments.
If the function returns any non-nil value,
the loop is broken, and the value is returned
as the final value of <CODE>foreachvar</CODE>.
<P>
This function could be defined in Lua:
<PRE>
function foreachvar (f)
local n, v = nextvar(nil)
while n do
local res = f(n, v)
if res then return res end
n, v = nextvar(n)
end
end
</PRE>
<P>
<h3> <TT>tostring (e)</TT></h3><A NAME="tostring"></A>
This function receives an argument of any type and
converts it to a string in a reasonable format.
For complete control on how numbers are converted,
use function <CODE>format</CODE>.
<P>
<h3> <TT>print (e1, e2, ...)</TT></h3><A NAME="print"></A>
This function receives any number of arguments,
and prints their values using the strings returned by <CODE>tostring</CODE>.
This function is not intended for formatted output,
but only as a quick way to show a value,
for instance for error messages or debugging.
See Section&nbsp;<A HREF="#libio">6.4</A> for functions for formatted output.
<P>
<h3> <TT>tonumber (e [, base])</TT></h3><A NAME="tonumber"></A>
This function receives one argument,
and tries to convert it to a number.
If the argument is already a number or a string convertible
to a number, then <CODE>tonumber</CODE> returns that number;
otherwise, it returns <B>nil</B>.
<P>
An optional argument specifies the base to interpret the numeral.
The base may be any integer between 2 and 36 inclusive.
In bases above 10, the letter `A' (either upper or lower case)
represents 10, `B' represents 11, and so forth, with `Z' representing 35.
<P>
In base 10 (the default), the number may have a decimal part,
as well as an optional exponent part (see Section&nbsp;<A HREF="#coercion">4.3</A>).
In other bases, only integers are accepted.
<P>
<A NAME="pdf-type"></A>
<h3> <TT>type (v)</TT></h3><A NAME="type"></A>
This function allows Lua to test the type of a value.
It receives one argument, and returns its type, coded as a string.
The possible results of this function are
<CODE>"nil"</CODE> (a string, not the value <B>nil</B>),
<CODE>"number"</CODE>,
<CODE>"string"</CODE>,
<CODE>"table"</CODE>,
<CODE>"function"</CODE>,
and <CODE>"userdata"</CODE>.
<P>
<h3> <TT>tag (v)</TT></h3><A NAME="tag"></A>
This function allows Lua to test the tag of a value (see Section&nbsp;<A HREF="#TypesSec">3</A>).
It receives one argument, and returns its tag (a number).
<CODE>tag</CODE> is equivalent to the API function <CODE>lua_tag</CODE>.
<P>
<h3> <TT>settag (t, tag)</TT></h3><A NAME="settag"></A>
This function sets the tag of a given table (see Section&nbsp;<A HREF="#TypesSec">3</A>).
<CODE>tag</CODE> must be a value created with <CODE>newtag</CODE>
(see Section&nbsp;<A HREF="#pdf-newtag">6.1</A>).
It returns the value of its first argument (the table).
For security reasons,
it is impossible to change the tag of a userdata from Lua.
<P>
<P>
<h3> <TT>assert (v [, message])</TT></h3><A NAME="assert"></A>
This function issues an <EM>``assertion failed!''</EM> error
when its argument is <B>nil</B>.
This function is equivalent to the following Lua function:
<PRE>
function assert (v, m)
if not v then
m = m or ""
error("assertion failed! " .. m)
end
end
</PRE>
<P>
<A NAME="pdf-error"></A>
<h3> <TT>error (message)</TT></h3><A NAME="error"></A>
This function calls the error handler and then terminates
the last protected function called
(in&nbsp;C: <CODE>lua_dofile</CODE>, <CODE>lua_dostring</CODE>,
<CODE>lua_dobuffer</CODE>, or <CODE>lua_callfunction</CODE>;
in Lua: <CODE>dofile</CODE>, <CODE>dostring</CODE>, or <CODE>call</CODE> in protected mode).
If <CODE>message</CODE> is <B>nil</B>, the error handler is not called.
Function <CODE>error</CODE> never returns.
<CODE>error</CODE> is equivalent to the API function <CODE>lua_error</CODE>.
<P>
<h3> <TT>rawgettable (table, index)</TT></h3><A NAME="rawgettable"></A>
Gets the real value of <CODE>table[index]</CODE>,
without invoking any tag method.
<CODE>table</CODE> must be a table,
and <CODE>index</CODE> is any value different from <B>nil</B>.
<P>
<h3> <TT>rawsettable (table, index, value)</TT></h3><A NAME="rawsettable"></A>
Sets the real value of <CODE>table[index]</CODE> to <CODE>value</CODE>,
without invoking any tag method.
<CODE>table</CODE> must be a table,
<CODE>index</CODE> is any value different from <B>nil</B>,
and <CODE>value</CODE> is any Lua value.
<P>
<h3> <TT>rawsetglobal (name, value)</TT></h3><A NAME="rawsetglobal"></A>
This function assigns the given value to a global variable.
The string <CODE>name</CODE> does not need to be a
syntactically valid variable name.
Therefore,
this function can set global variables with strange names like
<CODE>"m v 1"</CODE> or <CODE>34</CODE>.
Function <CODE>rawsetglobal</CODE> returns the value of its second argument.
<P>
<h3> <TT>setglobal (name, value)</TT></h3><A NAME="setglobal"></A>
This function assigns the given value to a global variable,
or calls a tag method.
Its full semantics is explained in Section&nbsp;<A HREF="#tag-method">4.8</A>.
The string <CODE>name</CODE> does not need to be a
syntactically valid variable name.
Function <CODE>setglobal</CODE> returns the value of its second argument.
<P>
<h3> <TT>rawgetglobal (name)</TT></h3><A NAME="rawgetglobal"></A>
This function retrieves the value of a global variable.
The string <CODE>name</CODE> does not need to be a
syntactically valid variable name.
<P>
<h3> <TT>getglobal (name)</TT></h3><A NAME="getglobal"></A>
This function retrieves the value of a global variable,
or calls a tag method.
Its full semantics is explained in Section&nbsp;<A HREF="#tag-method">4.8</A>.
The string <CODE>name</CODE> does not need to be a
syntactically valid variable name.
<P>
<h3> <TT>seterrormethod (newmethod)</TT></h3>
<A NAME="pdf-seterrormethod"></A>
Sets the error handler (see Section&nbsp;<A HREF="#error">4.9</A>).
<CODE>newmethod</CODE> must be a function or <B>nil</B>,
in which case the error handler does nothing.
Returns the old error handler.
<P>
<h3> <TT>settagmethod (tag, event, newmethod)</TT></h3>
<A NAME="settagmethod"></A>
This function sets a new tag method to the given pair <EM>(tag, event)</EM>.
It returns the old method.
If <CODE>newmethod</CODE> is <B>nil</B>,
<CODE>settagmethod</CODE> restores the default behavior for the given event.
<P>
<h3> <TT>gettagmethod (tag, event)</TT></h3>
<A NAME="gettagmethod"></A>
This function returns the current tag method
for a given pair <EM>(tag, event)</EM>.
<P>
<h3> <TT>copytagmethods (tagto, tagfrom)</TT></h3>
<A NAME="copytagmethods"></A>
This function copies all tag methods from one tag to another;
it returns <CODE>tagto</CODE>.
<P>
<P>
<A NAME="6.2"></A>
<H2>6.2 - String Manipulation</H2>
This library provides generic functions for string manipulation,
such as finding and extracting substrings and pattern matching.
When indexing a string, the first character is at position&nbsp;1
(not at&nbsp;0, as in C).
<P>
<h3> <TT>strfind (str, pattern [, init [, plain]])</TT></h3>
<A NAME="strfind"></A>
This function looks for the first <EM>match</EM> of
<CODE>pattern</CODE> in <CODE>str</CODE>.
If it finds one, then it returns the indices on <CODE>str</CODE>
where this occurrence starts and ends;
otherwise, it returns <B>nil</B>.
If the pattern specifies captures,
the captured strings are returned as extra results.
A third optional numerical argument specifies where to start the search;
its default value is 1.
If <CODE>init</CODE> is negative,
it is replaced by the length of the string minus its
absolute value plus 1.
Therefore, <I>-1</I> points to the last character of <CODE>str</CODE>.
A value of 1 as a fourth optional argument
turns off the pattern matching facilities,
so the function does a plain ``find substring'' operation,
with no characters in <CODE>pattern</CODE> being considered ``magic''.
<P>
<h3> <TT>strlen (s)</TT></h3><A NAME="strlen"></A>
Receives a string and returns its length.
<P>
<h3> <TT>strsub (s, i [, j])</TT></h3><A NAME="strsub"></A>
Returns another string, which is a substring of <CODE>s</CODE>,
starting at <CODE>i</CODE> and running until <CODE>j</CODE>.
If <CODE>i</CODE> or <CODE>j</CODE> are negative,
they are replaced by the length of the string minus their
absolute value plus 1.
Therefore, <I>-1</I> points to the last character of <CODE>s</CODE>
and <I>-2</I> to the previous one.
If <CODE>j</CODE> is absent, it is assumed to be equal to <I>-1</I>
(which is the same as the string length).
In particular,
the call <CODE>strsub(s,1,j)</CODE> returns a prefix of <CODE>s</CODE>
with length <CODE>j</CODE>,
and the call <CODE>strsub(s, -i)</CODE> returns a suffix of <CODE>s</CODE>
with length <CODE>i</CODE>.
<P>
<h3> <TT>strlower (s)</TT></h3><A NAME="strlower"></A>
Receives a string and returns a copy of that string with all
upper case letters changed to lower case.
All other characters are left unchanged.
The definition of what is an upper case
letter depends on the current locale.
<P>
<h3> <TT>strupper (s)</TT></h3><A NAME="strupper"></A>
Receives a string and returns a copy of that string with all
lower case letters changed to upper case.
All other characters are left unchanged.
The definition of what is a lower case
letter depends on the current locale.
<P>
<h3> <TT>strrep (s, n)</TT></h3><A NAME="strrep"></A>
Returns a string that is the concatenation of <CODE>n</CODE> copies of
the string <CODE>s</CODE>.
<P>
<h3> <TT>strbyte (s [, i])</TT></h3><A NAME="strbyte"></A>
Returns the internal numerical code of the character <CODE>s[i]</CODE>.
If <CODE>i</CODE> is absent, then it is assumed to be 1.
If <CODE>i</CODE> is negative,
it is replaced by the length of the string minus its
absolute value plus 1.
Therefore, <I>-1</I> points to the last character of <CODE>s</CODE>.
<P>
Note that numerical codes are not necessarily portable across platforms.
<P>
<h3> <TT>strchar (i1, i2, ...)</TT></h3><A NAME="strchar"></A>
Receives 0 or more integers.
Returns a string with length equal to the number of arguments,
wherein each character has the internal numerical code equal
to its correspondent argument.
<P>
Note that numerical codes are not necessarily portable across platforms.
<P>
<h3> <TT>format (formatstring, e1, e2, ...)</TT></h3><A NAME="format"></A>
<A NAME="format"></A>
This function returns a formatted version of its variable number of arguments
following the description given in its first argument (which must be a string).
The format string follows the same rules as the <CODE>printf</CODE> family of
standard C functions.
The only differences are that the options/modifiers
<CODE>*</CODE>, <CODE>l</CODE>, <CODE>L</CODE>, <CODE>n</CODE>, <CODE>p</CODE>,
and <CODE>h</CODE> are not supported,
and there is an extra option, <CODE>q</CODE>.
This option formats a string in a form suitable to be safely read
back by the Lua interpreter;
that is,
the string is written between double quotes,
and all double quotes, returns and backslashes in the string
are correctly escaped when written.
For instance, the call
<PRE>
format('%q', 'a string with "quotes" and \n new line')
</PRE>
will produce the string:
<PRE>
"a string with \"quotes\" and \
new line"
</PRE>
<P>
Conversions can be applied to the n-th argument in the argument list,
rather than the next unused argument.
In this case, the conversion character <CODE>%</CODE> is replaced
by the sequence <CODE>%d$</CODE>, where <CODE>d</CODE> is a
decimal digit in the range [1,9],
giving the position of the argument in the argument list.
For instance, the call <CODE>format("%2$d -&gt; %1$03d", 1, 34)</CODE> will
result in <CODE>"34 -&gt; 001"</CODE>.
The same argument can be used in more than one conversion.
<P>
The options <CODE>c</CODE>, <CODE>d</CODE>, <CODE>E</CODE>, <CODE>e</CODE>, <CODE>f</CODE>,
<CODE>g</CODE>, <CODE>G</CODE>, <CODE>i</CODE>, <CODE>o</CODE>, <CODE>u</CODE>, <CODE>X</CODE>, and <CODE>x</CODE> all
expect a number as argument,
whereas <CODE>q</CODE> and <CODE>s</CODE> expect a string.
Note that the <CODE>*</CODE> modifier can be simulated by building
the appropriate format string.
For example, <CODE>"%*g"</CODE> can be simulated with
<CODE>"%"..width.."g"</CODE>.
<P>
<EM>Note: function <TT>format</TT> can only be used with strings that do not contain zeros.</EM>
<P>
<h3> <TT>gsub (s, pat, repl [, n])</TT></h3>
<A NAME="gsub"></A>
Returns a copy of <CODE>s</CODE>,
where all occurrences of the pattern <CODE>pat</CODE> have been
replaced by a replacement string specified by <CODE>repl</CODE>.
This function also returns, as a second value,
the total number of substitutions made.
<P>
If <CODE>repl</CODE> is a string, then its value is used for replacement.
Any sequence in <CODE>repl</CODE> of the form <CODE>%n</CODE>
with <CODE>n</CODE> between 1 and 9
stands for the value of the n-th captured substring.
<P>
If <CODE>repl</CODE> is a function, then this function is called every time a
match occurs, with all captured substrings passed as arguments,
in order (see below).
If the value returned by this function is a string,
then it is used as the replacement string;
otherwise, the replacement string is the empty string.
<P>
A last optional parameter <CODE>n</CODE> limits
the maximum number of substitutions to occur.
For instance, when <CODE>n</CODE> is 1 only the first occurrence of
<CODE>pat</CODE> is replaced.
<P>
See some examples below:
<PRE>
x = gsub("hello world", "(%w%w*)", "%1 %1")
--&gt; x="hello hello world world"
<P>
x = gsub("hello world", "(%w%w*)", "%1 %1", 1)
--&gt; x="hello hello world"
<P>
x = gsub("hello world from Lua", "(%w%w*)%s*(%w%w*)", "%2 %1")
--&gt; x="world hello Lua from"
<P>
x = gsub("home = $HOME, user = $USER", "$(%w%w*)", getenv)
--&gt; x="home = /home/roberto, user = roberto" (for instance)
<P>
x = gsub("4+5 = $return 4+5$", "$(.-)%$", dostring)
--&gt; x="4+5 = 9"
<P>
local t = {name="lua", version="3.1"}
x = gsub("$name - $version", "$(%w%w*)", function (v) return %t[v] end)
--&gt; x="lua - 3.1"
<P>
t = {n=0}
gsub("first second word", "(%w%w*)",
function (w) %t.n = %t.n+1; %t[%t.n] = w end)
--&gt; t={"first", "second", "word"; n=3}
</PRE>
<P>
<P>
<A NAME="pm"></A>
<h3>Patterns</h3>
<P>
<H4>Character Class:</H4>
a <A NAME="character class"><EM>character class</EM></A> is used to represent a set of characters.
The following combinations are allowed in describing a character class:
<DL>
<DT><B><EM>x</EM></B><DD> (where <EM>x</EM> is any character not in the list <CODE>()%.[*-?</CODE>)
- represents the character <EM>x</EM> itself.
<DT><B><TT>.</TT></B><DD> - (a dot) represents all characters.
<DT><B><TT>%a</TT></B><DD> - represents all letters.
<DT><B><TT>%A</TT></B><DD> - represents all non letter characters.
<DT><B><TT>%d</TT></B><DD> - represents all digits.
<DT><B><TT>%D</TT></B><DD> - represents all non digits.
<DT><B><TT>%l</TT></B><DD> - represents all lower case letters.
<DT><B><TT>%L</TT></B><DD> - represents all non lower case letter characters.
<DT><B><TT>%s</TT></B><DD> - represents all space characters.
<DT><B><TT>%S</TT></B><DD> - represents all non space characters.
<DT><B><TT>%u</TT></B><DD> - represents all upper case letters.
<DT><B><TT>%U</TT></B><DD> - represents all non upper case letter characters.
<DT><B><TT>%w</TT></B><DD> - represents all alphanumeric characters.
<DT><B><TT>%W</TT></B><DD> - represents all non alphanumeric characters.
<DT><B><TT>%</TT><EM>x</EM></B><DD> (where <EM>x</EM> is any non alphanumeric character) -
represents the character <EM>x</EM>.
This is the standard way to escape the magic characters <CODE>()%.[*-?</CODE>.
<DT><B><TT>[char-set</TT></B>]<DD> -
Represents the class which is the union of all
characters in char-set.
To include a <CODE>]</CODE> in char-set, it must be the first character.
A range of characters may be specified by
separating the end characters of the range with a <CODE>-</CODE>.
If <CODE>-</CODE> appears as the first or last character of char-set,
then it represents itself.
All classes <CODE>%</CODE><EM>x</EM> described above can also be used as
components in a char-set.
All other characters in char-set represent themselves.
E.g., assuming an <EM>ascii</EM> character set,
<CODE>[%dA-Fa-f]</CODE> specifies the hexa-decimal digits.
<DT><B><TT>[^char-set</TT></B>]<DD> -
represents the complement of char-set,
where char-set is interpreted as above.
</DL>
<P>
The definitions of letter, space, etc. depend on the current locale.
In particular, the class <CODE>[a-z]</CODE> may not be equivalent to <CODE>%l</CODE>.
The second form should be preferred for more portable programs.
<P>
<H4>Pattern Item:</H4>
a <A NAME="pattern item"><EM>pattern item</EM></A> may be:
<UL>
<LI>
a single character class,
which matches any single character in the class;
<LI>
a single character class followed by <CODE>*</CODE>,
which matches 0 or more repetitions of characters in the class.
These repetition items will always match the longest possible sequence.
<LI>
a single character class followed by <CODE>-</CODE>,
which also matches 0 or more repetitions of characters in the class.
Unlike <CODE>*</CODE>,
these repetition items will always match the shortest possible sequence.
<LI>
a single character class followed by <CODE>?</CODE>,
which matches 0 or 1 occurrence of a character in the class;
<LI>
<TT>%<EM>n</TT></EM>, for <EM>n</EM> between 1 and 9;
such item matches a sub-string equal to the n-th captured string
(see below);
<LI>
<TT>%b<EM>xy</TT></EM>, where <EM>x</EM> and <EM>y</EM> are two distinct characters;
such item matches strings that start with <EM>x</EM>, end with <EM>y</EM>,
and where the <EM>x</EM> and <EM>y</EM> are <EM>balanced</EM>.
That means that, if one reads the string from left to write,
counting plus 1 for an <EM>x</EM> and minus 1 for a <EM>y</EM>,
the ending <EM>y</EM> is the first where the count reaches 0.
For instance, the item <CODE>%b()</CODE> matches expressions with
balanced parentheses.
</UL>
<P>
<H4>Pattern:</H4>
a <A NAME="pattern"><EM>pattern</EM></A> is a sequence of pattern items.
A <CODE>^</CODE> at the beginning of a pattern anchors the match at the
beginning of the subject string.
A <CODE>$</CODE> at the end of a pattern anchors the match at the
end of the subject string.
<P>
<H4>Captures:</H4>
a pattern may contain sub-patterns enclosed in parentheses,
that describe <A NAME="captures"><EM>captures</EM></A>.
When a match succeeds, the sub-strings of the subject string
that match captures are stored (<EM>captured</EM>) for future use.
Captures are numbered according to their left parentheses.
For instance, in the pattern <CODE>"(a*(.)%w(%s*))"</CODE>,
the part of the string matching <CODE>"a*(.)%w(%s*)"</CODE> is
stored as the first capture (and therefore has number&nbsp;1);
the character matching <CODE>.</CODE> is captured with number&nbsp;2,
and the part matching <CODE>%s*</CODE> has number&nbsp;3.
<P>
<P>
<A NAME="mathlib"></A>
<A NAME="6.3"></A>
<H2>6.3 - Mathematical Functions</H2>
<P>
This library is an interface to some functions of the standard C math library.
In addition, it registers a tag method for the binary operator <CODE>^</CODE> that
returns <I>x^y</I> when applied to numbers <CODE>x^y</CODE>.
<P>
The library provides the following functions:
<A NAME="abs"></A><A NAME="acos"></A><A NAME="asin"></A><A NAME="atan"></A>
<A NAME="atan2"></A><A NAME="ceil"></A><A NAME="cos"></A><A NAME="floor"></A>
<A NAME="log"></A><A NAME="log10"></A><A NAME="max"></A><A NAME="min"></A>
<A NAME="mod"></A><A NAME="sin"></A><A NAME="sqrt"></A><A NAME="tan"></A>
<A NAME="frexp"></A><A NAME="ldexp"></A>
<A NAME="random"></A><A NAME="randomseed"></A>
<PRE>
abs acos asin atan atan2 ceil cos deg floor log log10
max min mod rad sin sqrt tan frexp ldexp
random randomseed
</PRE>
plus a global variable <A NAME="PI"><TT>PI</TT></A>.
Most of them
are only interfaces to the homonymous functions in the C library,
except that, for the trigonometric functions,
all angles are expressed in <EM>degrees</EM>, not radians.
Functions <A NAME="deg"><TT>deg</TT></A> and <A NAME="rad"><TT>rad</TT></A> can be used to convert
between radians and degrees.
<P>
The function <CODE>max</CODE> returns the maximum
value of its numeric arguments.
Similarly, <CODE>min</CODE> computes the minimum.
Both can be used with 1, 2 or more arguments.
<P>
The functions <CODE>random</CODE> and <CODE>randomseed</CODE> are interfaces to
the simple random generator functions <CODE>rand</CODE> and <CODE>srand</CODE>,
provided by ANSI C.
The function <CODE>random</CODE>, when called without arguments,
returns a pseudo-random real number in the range <I>[0,1)</I>.
When called with a number <I>n</I>,
<CODE>random</CODE> returns a pseudo-random integer in the range <I>[1,n]</I>.
<P>
<P>
<A NAME="libio"></A>
<A NAME="6.4"></A>
<H2>6.4 - I/O Facilities</H2>
<P>
All input and output operations in Lua are done over two
<A NAME="file handles"><EM>file handles</EM></A>, one for reading and one for writing.
These handles are stored in two Lua global variables,
called <CODE>_INPUT</CODE> and <CODE>_OUTPUT</CODE>.
The global variables
<CODE>_STDIN</CODE>, <CODE>_STDOUT</CODE> and <CODE>_STDERR</CODE>
are initialized with file descriptors for
<CODE>stdin</CODE>, <CODE>stdout</CODE> and <CODE>stderr</CODE>.
Initially, <CODE>_INPUT=_STDIN</CODE> and <CODE>_OUTPUT=_STDOUT</CODE>.
<A NAME="_INPUT"></A><A NAME="_OUTPUT"></A>
<A NAME="_STDIN"></A><A NAME="_STDOUT"></A><A NAME="_STDERR"></A>
<P>
A file handle is a userdata containing the file stream <CODE>FILE*</CODE>,
and with a distinctive tag created by the I/O library.
<P>
<P>
Unless otherwise stated,
all I/O functions return <B>nil</B> on failure and
some value different from <B>nil</B> on success.
<P>
<h3> <TT>readfrom (filename)</TT></h3><A NAME="readfrom"></A>
<P>
This function may be called in two ways.
When called with a file name, it opens the named file,
sets its handle as the value of <CODE>_INPUT</CODE>,
and returns this value.
It does not close the current input file.
When called without parameters,
it closes the <CODE>_INPUT</CODE> file,
and restores <CODE>stdin</CODE> as the value of <CODE>_INPUT</CODE>.
<P>
If this function fails, it returns <B>nil</B>,
plus a string describing the error.
<P>
<CITE>
<EM>System dependent</EM>: if <CODE>filename</CODE> starts with a <CODE>|</CODE>,
then a <A NAME="piped input">piped input</A> is opened, via function <A NAME="popen"><TT>popen</TT></A>.
Not all systems implement pipes.
Moreover,
the number of files that can be open at the same time is
usually limited and depends on the system.
</CITE>
<P>
<h3> <TT>writeto (filename)</TT></h3><A NAME="writeto"></A>
<P>
This function may be called in two ways.
When called with a file name,
it opens the named file,
sets its handle as the value of <CODE>_OUTPUT</CODE>,
and returns this value.
It does not close the current output file.
Note that, if the file already exists,
then it will be <EM>completely erased</EM> with this operation.
When called without parameters,
this function closes the <CODE>_OUTPUT</CODE> file,
and restores <CODE>stdout</CODE> as the value of <CODE>_OUTPUT</CODE>.
<A NAME="closing a file"></A>
<P>
If this function fails, it returns <B>nil</B>,
plus a string describing the error.
<P>
<CITE>
<EM>System dependent</EM>: if <CODE>filename</CODE> starts with a <CODE>|</CODE>,
then a <A NAME="piped output">piped output</A> is opened, via function <A NAME="popen"><TT>popen</TT></A>.
Not all systems implement pipes.
Moreover,
the number of files that can be open at the same time is
usually limited and depends on the system.
</CITE>
<P>
<h3> <TT>appendto (filename)</TT></h3><A NAME="appendto"></A>
<P>
This function opens a file named <CODE>filename</CODE> and sets it as the
value of <CODE>_OUTPUT</CODE>.
Unlike the <CODE>writeto</CODE> operation,
this function does not erase any previous content of the file.
If this function fails, it returns <B>nil</B>,
plus a string describing the error.
<P>
Note that function <CODE>writeto</CODE> is
available to close an output file opened by <CODE>appendto</CODE>.
<P>
<h3> <TT>remove (filename)</TT></h3><A NAME="remove"></A>
<P>
This function deletes the file with the given name.
If this function fails, it returns <B>nil</B>,
plus a string describing the error.
<P>
<h3> <TT>rename (name1, name2)</TT></h3><A NAME="rename"></A>
<P>
This function renames file named <CODE>name1</CODE> to <CODE>name2</CODE>.
If this function fails, it returns <B>nil</B>,
plus a string describing the error.
<P>
<h3> <TT>tmpname ()</TT></h3><A NAME="tmpname"></A>
<P>
This function returns a string with a file name that can safely
be used for a temporary file.
The file must be explicitly removed when no longer needed.
<P>
<h3> <TT>read ([filehandle] [readpattern])</TT></h3><A NAME="read"></A>
<P>
This function reads the file <CODE>_INPUT</CODE>,
or from <CODE>filehandle</CODE> if this argument is given,
according to a read pattern, which specifies how much to read;
characters are read from the input file until
the read pattern fails or ends.
The function <CODE>read</CODE> returns a string with the characters read,
even if the pattern succeeds only partially,
or <B>nil</B> if the read pattern fails <EM>and</EM>
the result string would be empty.
When called without parameters,
it uses a default pattern that reads the next line
(see below).
<P>
A <A NAME="read pattern"><EM>read pattern</EM></A> is a sequence of read pattern items.
An item may be a single character class
or a character class followed by <CODE>?</CODE> or by <CODE>*</CODE>.
A single character class reads the next character from the input
if it belongs to the class, otherwise it fails.
A character class followed by <CODE>?</CODE> reads the next character
from the input if it belongs to the class;
it never fails.
A character class followed by <CODE>*</CODE> reads until a character that
does not belong to the class, or end of file;
since it can match a sequence of zero characters, it never fails.
Note that the behavior of read patterns is slightly different from
the regular pattern matching behavior,
where a <CODE>*</CODE> expands to the maximum length <EM>such that</EM>
the rest of the pattern does not fail.
With the read pattern behavior
there is no need for backtracking the reading.
<P>
A pattern item may contain sub-patterns enclosed in curly brackets,
that describe <A NAME="skips"><EM>skips</EM></A>.
Characters matching a skip are read,
but are not included in the resulting string.
<P>
Following are some examples of read patterns and their meanings:
<UL>
<LI><CODE>"."</CODE> returns the next character, or <B>nil</B> on end of file.
<LI><CODE>".*"</CODE> reads the whole file.
<LI><CODE>"[^\n]*{\n}"</CODE> returns the next line
(skipping the end of line), or <B>nil</B> on end of file.
This is the default pattern.
<LI><CODE>"{%s*}%S%S*"</CODE> returns the next word
(maximal sequence of non white-space characters),
skipping spaces if necessary,
or <B>nil</B> on end of file.
<LI><CODE>"{%s*}[+-]?%d%d*"</CODE> returns the next integer
or <B>nil</B> if the next characters do not conform to an integer format.
</UL>
<P>
<h3> <TT>write ([filehandle, ] value1, ...)</TT></h3><A NAME="write"></A>
<P>
This function writes the value of each of its arguments to the
file <CODE>_OUTPUT</CODE>,
or to <CODE>filehandle</CODE> if this argument is given,
The arguments must be strings or numbers.
To write other values,
use <CODE>tostring</CODE> or <CODE>format</CODE> before <CODE>write</CODE>.
If this function fails, it returns <B>nil</B>,
plus a string describing the error.
<P>
<h3> <TT>date ([format])</TT></h3><A NAME="date"></A>
<P>
This function returns a string containing date and time
formatted according to the given string <CODE>format</CODE>,
following the same rules of the ANSI C function <CODE>strftime</CODE>.
When called without arguments,
it returns a reasonable date and time representation that depends on
the host system and the locale.
<P>
<h3> <TT>clock ()</TT></h3><A NAME="clock"></A>
<P>
This function returns an approximation of the amount of CPU time
used by the program, in seconds.
<P>
<h3> <TT>exit ([code])</TT></h3><A NAME="exit"></A>
<P>
This function calls the C function <CODE>exit</CODE>,
with an optional <CODE>code</CODE>,
to terminate the program.
The default value for <CODE>code</CODE> is 1.
<P>
<h3> <TT>getenv (varname)</TT></h3><A NAME="getenv"></A>
<P>
Returns the value of the process environment variable <CODE>varname</CODE>,
or <B>nil</B> if the variable is not defined.
<P>
<h3> <TT>execute (command)</TT></h3><A NAME="execute"></A>
<P>
This function is equivalent to the C function <CODE>system</CODE>.
It passes <CODE>command</CODE> to be executed by an operating system shell.
It returns an error code, which is system-dependent.
<P>
<h3> <TT>setlocale (locale [, category])</TT></h3><A NAME="setlocale"></A>
<P>
This function is an interface to the ANSI C function <CODE>setlocale</CODE>.
<CODE>locale</CODE> is a string specifying a locale;
<CODE>category</CODE> is an optional string describing which category to change:
<CODE>"all"</CODE>, <CODE>"collate"</CODE>, <CODE>"ctype"</CODE>,
<CODE>"monetary"</CODE>, <CODE>"numeric"</CODE>, or <CODE>"time"</CODE>;
the default category is <CODE>"all"</CODE>.
The function returns the name of the new locale,
or <B>nil</B> if the request cannot be honored.
<P>
<P>
<A NAME="debugI"></A>
<!-- ====================================================================== -->
<HR>
<A NAME="7."></A>
<H1>7 - The Debugger Interface</H1>
<P>
Lua has no built-in debugging facilities.
Instead, it offers a special interface,
by means of functions and <EM>hooks</EM>,
which allows the construction of different
kinds of debuggers, profilers, and other tools
that need ``inside information'' from the interpreter.
This interface is declared in the header file <CODE>luadebug.h</CODE>.
<P>
<A NAME="7.1"></A>
<H2>7.1 - Stack and Function Information</H2>
<P>
The main function to get information about the interpreter stack
is
<PRE>
lua_Function lua_stackedfunction (int level);
</PRE>
It returns a handle (<CODE>lua_Function</CODE>) to the <EM>activation record</EM>
of the function executing at a given level.
Level&nbsp;0 is the current running function,
while level <I>n+1</I> is the function that has called level <I>n</I>.
When called with a level greater than the stack depth,
<CODE>lua_stackedfunction</CODE> returns <CODE>LUA_NOOBJECT</CODE>.
<P>
The type <CODE>lua_Function</CODE> is just another name
to <CODE>lua_Object</CODE>.
Although, in this library,
a <CODE>lua_Function</CODE> can be used wherever a <CODE>lua_Object</CODE> is required,
when a parameter has type <CODE>lua_Function</CODE>
it accepts only a handle returned by
<CODE>lua_stackedfunction</CODE>.
<P>
Three other functions produce extra information about a function:
<PRE>
void lua_funcinfo (lua_Object func, char **filename, int *linedefined);
int lua_currentline (lua_Function func);
char *lua_getobjname (lua_Object o, char **name);
</PRE>
<CODE>lua_funcinfo</CODE> gives the file name and the line where the
given function has been defined.
If the ``function'' is in fact the main code of a chunk,
then <CODE>linedefined</CODE> is 0.
If the function is a C function,
then <CODE>linedefined</CODE> is <I>-1</I>, and <CODE>filename</CODE> is <CODE>"(C)"</CODE>.
<P>
The function <CODE>lua_currentline</CODE> gives the current line where
a given function is executing.
It only works if the function has been compiled with debug
information (see Section&nbsp;<A HREF="#pragma">4.9</A>).
When no line information is available,
<CODE>lua_currentline</CODE> returns <I>-1</I>.
<P>
Function <CODE>lua_getobjname</CODE> tries to find a reasonable name for
a given function.
Because functions in Lua are first class values,
they do not have a fixed name:
Some functions may be the value of many global variables,
while others may be stored only in a table field.
Function <CODE>lua_getobjname</CODE> first checks whether the given
function is a tag method.
If so, it returns the string <CODE>"tag-method"</CODE>,
and <CODE>name</CODE> is set to point to the event name.
Otherwise, if the given function is the value of a global variable,
then <CODE>lua_getobjname</CODE> returns the string <CODE>"global"</CODE>,
and <CODE>name</CODE> points to the variable name.
If the given function is neither a tag method nor a global variable,
then <CODE>lua_getobjname</CODE> returns the empty string,
and <CODE>name</CODE> is set to <CODE>NULL</CODE>.
<P>
<A NAME="7.2"></A>
<H2>7.2 - Manipulating Local Variables</H2>
<P>
The following functions allow the manipulation of the
local variables of a given activation record.
They only work if the function has been compiled with debug
information (see Section&nbsp;<A HREF="#pragma">4.9</A>).
<PRE>
lua_Object lua_getlocal (lua_Function func, int local_number, char **name);
int lua_setlocal (lua_Function func, int local_number);
</PRE>
<CODE>lua_getlocal</CODE> returns the value of a local variable,
and sets <CODE>name</CODE> to point to the variable name.
<CODE>local_number</CODE> is an index for local variables.
The first parameter has index 1, and so on, until the
last active local variable.
When called with a <CODE>local_number</CODE> greater than the
number of active local variables,
or if the activation record has no debug information,
<CODE>lua_getlocal</CODE> returns <CODE>LUA_NOOBJECT</CODE>.
Formal parameters are the first local variables.
<P>
The function <CODE>lua_setlocal</CODE> sets the local variable
<CODE>local_number</CODE> to the value previously pushed on the stack
(see Section&nbsp;<A HREF="#valuesCLua">5.2</A>).
If the function succeeds, then it returns 1.
If <CODE>local_number</CODE> is greater than the number
of active local variables,
or if the activation record has no debug information,
then this function fails and returns 0.
<P>
<A NAME="7.3"></A>
<H2>7.3 - Hooks</H2>
<P>
The Lua interpreter offers two hooks for debugging purposes:
<PRE>
typedef void (*lua_CHFunction) (lua_Function func, char *file, int line);
extern lua_CHFunction lua_callhook;
<P>
typedef void (*lua_LHFunction) (int line);
extern lua_LHFunction lua_linehook;
</PRE>
The first one is called whenever the interpreter enters or leaves a
function.
When entering a function,
its parameters are a handle to the function activation record,
plus the file and the line where the function is defined (the same
information which is provided by <CODE>lua_funcinfo</CODE>);
when leaving a function, <CODE>func</CODE> is <CODE>LUA_NOOBJECT</CODE>,
<CODE>file</CODE> is <CODE>"(return)"</CODE>, and <CODE>line</CODE> is 0.
<P>
The other hook is called every time the interpreter changes
the line of code it is executing.
Its only parameter is the line number
(the same information which is provided by the call
<CODE>lua_currentline(lua_stackedfunction(0))</CODE>).
This second hook is only called if the active function
has been compiled with debug information (see Section&nbsp;<A HREF="#pragma">4.9</A>).
<P>
A hook is disabled when its value is <CODE>NULL</CODE>,
which is the initial value of both hooks.
<P>
<P>
<P>
<A NAME="lua-sa"></A>
<!-- ====================================================================== -->
<HR>
<A NAME="8."></A>
<H1>8 - Lua Stand-alone</H1>
<P>
Although Lua has been designed as an extension language,
the language can also be used as a stand-alone interpreter.
An implementation of such an interpreter,
called simply <CODE>lua</CODE>,
is provided with the standard distribution.
This program can be called with any sequence of the following arguments:
<DL>
<DT><B><TT>-v</TT></B><DD> prints version information.
<DT><B><TT>-d</TT></B><DD> turns on debug information.
<DT><B><TT>-e stat</TT></B><DD> executes <CODE>stat</CODE> as a Lua chunk.
<DT><B><TT>-i</TT></B><DD> runs interactively,
accepting commands from standard input until an <CODE>EOF</CODE>.
Each line entered is immediately executed.
<DT><B><TT>-q</TT></B><DD> same as <TT>-i</TT>, but without a prompt (quiet mode).
<DT><B><TT>-</TT></B><DD> executes <CODE>stdin</CODE> as a file.
<DT><B><TT>var=value</TT></B><DD> sets global <CODE>var</CODE> with string <CODE>"value"</CODE>.
<DT><B><TT>filename</TT></B><DD> executes file <CODE>filename</CODE> as a Lua chunk.
</DL>
When called without arguments,
Lua behaves as <CODE>lua -v -i</CODE> when <CODE>stdin</CODE> is a terminal,
and as <CODE>lua -</CODE> otherwise.
<P>
All arguments are handled in order.
For instance, an invocation like
<PRE>
$ lua -i a=test prog.lua
</PRE>
will first interact with the user until an <CODE>EOF</CODE>,
then will set <CODE>a</CODE> to <CODE>"test"</CODE>,
and finally will run the file <CODE>prog.lua</CODE>.
<P>
When in interactive mode,
a multi-line statement can be written finishing intermediate
lines with a backslash (<CODE>\</CODE>).
The prompt presented is the value of the global variable <CODE>_PROMPT</CODE>.
Therefore, the prompt can be changed like below:
<PRE>
$ lua _PROMPT='myprompt&gt; ' -i
</PRE>
<P>
In Unix systems, Lua scripts can be made into executable programs
by using the <CODE>#!</CODE> form,
as in <CODE>#!/usr/local/bin/lua</CODE>.
<P>
<HR>
<A NAME="Acknowledgments"></A>
<h1>Acknowledgments</h1>
<P>
The authors would like to thank CENPES/PETROBRAS which,
jointly with TeCGraf, used extensively early versions of
this system and gave valuable comments.
The authors would also like to thank Carlos Henrique Levy,
who found the name of the game.
Lua means <EM>moon</EM> in Portuguese.
<P>
<P>
<P>
<P>
<HR>
<A NAME="Incompatibilities"></A>
<h1>Incompatibilities with Previous Versions</h1>
<P>
Although great care has been taken to avoid incompatibilities with
the previous public versions of Lua,
some differences had to be introduced.
Here is a list of all these incompatibilities.
<P>
<h2>Incompatibilities with <A NAME="version 3.0</h2>">version 3.0</h2></A>
<UL>
<P>
<LI>To support multiple contexts,
Lua 3.1 must be explicitly opened before used,
with function <CODE>lua_open</CODE>.
However, all standard libraries check whether Lua is already opened,
so any existing program that opens at least one standard
library before calling Lua does not need to be modified.
<P>
<LI>Function <CODE>dostring</CODE> no longer accepts an optional second argument,
with a temporary error method.
This facility is now provided by function <CODE>call</CODE>.
<P>
<LI>Function <CODE>gsub</CODE> no longer accepts an optional fourth argument
(a callback data, a table).
Closures replace this feature with advantage.
<P>
<LI>The syntax for function declaration is now more restricted;
for instance, the old syntax <CODE>function f[exp] (x) ... end</CODE> is not
accepted in Lua 3.1.
In these cases,
programs should use an explicit assignment instead, such as
<CODE>f[exp] = function (x) ... end</CODE>.
<P>
<LI>Old pre-compiled code is obsolete, and must be re-compiled.
<P>
<LI>The option <CODE>a=b</CODE> in Lua stand-alone now sets <CODE>a</CODE> to the
<EM>string</EM> <CODE>b</CODE>, and not to the value of <CODE>b</CODE>.
<P>
</UL>
<P>
<P>
<HR>
Last update:
Fri Jul 10 15:10:14 EST 1998
by <A HREF="http://www.tecgraf.puc-rio.br/~lhf/">lhf</A>.
</BODY>
</HTML>
Reference in a new issue View git blame Copy permalink