<?xml version="1.0" encoding="latin1" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">
<erlref>
<header>
<copyright>
<year>1996</year><year>2011</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
The contents of this file are subject to the Erlang Public License,
Version 1.1, (the "License"); you may not use this file except in
compliance with the License. You should have received a copy of the
Erlang Public License along with this software. If not, it can be
retrieved online at http://www.erlang.org/.
Software distributed under the License is distributed on an "AS IS"
basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
the License for the specific language governing rights and limitations
under the License.
</legalnotice>
<title>ets</title>
<prepared></prepared>
<docno></docno>
<date></date>
<rev></rev>
</header>
<module>ets</module>
<modulesummary>Built-In Term Storage</modulesummary>
<description>
<p>This module is an interface to the Erlang built-in term storage
BIFs. These provide the ability to store very large quantities of
data in an Erlang runtime system, and to have constant access
time to the data. (In the case of <c>ordered_set</c>, see below,
access time is proportional to the logarithm of the number of
objects stored).</p>
<p>Data is organized as a set of dynamic tables, which can store
tuples. Each table is created by a process. When the process
terminates, the table is automatically destroyed. Every table has
access rights set at creation.</p>
<p>Tables are divided into four different types, <c>set</c>,
<c>ordered_set</c>, <c>bag</c> and <c>duplicate_bag</c>.
A <c>set</c> or <c>ordered_set</c> table can only have one object
associated with each key. A <c>bag</c> or <c>duplicate_bag</c> can
have many objects associated with each key.</p>
<p>The number of tables stored at one Erlang node is limited.
The current default limit is approximately 1400 tables. The upper
limit can be increased by setting the environment variable
<c>ERL_MAX_ETS_TABLES</c> before starting the Erlang runtime
system (i.e. with the <c>-env</c> option to
<c>erl</c>/<c>werl</c>). The actual limit may be slightly higher
than the one specified, but never lower.</p>
<p>Note that there is no automatic garbage collection for tables.
Even if there are no references to a table from any process, it
will not automatically be destroyed unless the owner process
terminates. It can be destroyed explicitly by using
<c>delete/1</c>. The default owner is the process that created the
table. Table ownership can be transferred at process termination
by using the <seealso marker="#heir">heir</seealso> option or explicitly
by calling <seealso marker="#give_away/3">give_away/3</seealso>.</p>
<p>Some implementation details:</p>
<list type="bulleted">
<item>In the current implementation, every object insert and
look-up operation results in a copy of the object.</item>
<item><c>'$end_of_table'</c> should not be used as a key since
this atom is used to mark the end of the table when using
<c>first</c>/<c>next</c>.</item>
</list>
<p>Also worth noting is the subtle difference between
<em>matching</em> and <em>comparing equal</em>, which is
demonstrated by the different table types <c>set</c> and
<c>ordered_set</c>. Two Erlang terms <c>match</c> if they are of
the same type and have the same value, so that <c>1</c> matches
<c>1</c>, but not <c>1.0</c> (as <c>1.0</c> is a <c>float()</c>
and not an <c>integer()</c>). Two Erlang terms <em>compare equal</em> if they either are of the same type and value, or if
both are numeric types and extend to the same value, so that
<c>1</c> compares equal to both <c>1</c> and <c>1.0</c>. The
<c>ordered_set</c> works on the <em>Erlang term order</em> and
there is no defined order between an <c>integer()</c> and a
<c>float()</c> that extends to the same value, hence the key
<c>1</c> and the key <c>1.0</c> are regarded as equal in an
<c>ordered_set</c> table.</p>
</description>
<section>
<title>Failure</title>
<p>In general, the functions below will exit with reason
<c>badarg</c> if any argument is of the wrong format, if the
table identifier is invalid or if the operation is denied due to
table access rights (<seealso marker="#protected">protected</seealso>
or <seealso marker="#private">private</seealso>).</p>
</section>
<section><marker id="concurrency"></marker>
<title>Concurrency</title>
<p>This module provides some limited support for concurrent access.
All updates to single objects are guaranteed to be both <em>atomic</em>
and <em>isolated</em>. This means that an updating operation towards
a single object will either succeed or fail completely without any
effect at all (atomicy).
Nor can any intermediate results of the update be seen by other
processes (isolation). Some functions that update several objects
state that they even guarantee atomicy and isolation for the entire
operation. In database terms the isolation level can be seen as
"serializable", as if all isolated operations were carried out serially,
one after the other in a strict order.</p>
<p>No other support is available within ETS that would guarantee
consistency between objects. However, the <c>safe_fixtable/2</c>
function can be used to guarantee that a sequence of
<c>first/1</c> and <c>next/2</c> calls will traverse the table
without errors and that each existing object in the table is visited
exactly once, even if another process (or the same process)
simultaneously deletes or inserts objects into the table.
Nothing more is guaranteed; in particular objects that are inserted
or deleted during such a traversal may be visited once or not at all.
Functions that internally traverse over a table, like <c>select</c>
and <c>match</c>, will give the same guarantee as <c>safe_fixtable</c>.</p>
</section>
<section>
<marker id="match_spec"></marker>
<title>Match Specifications</title>
<p>Some of the functions uses a <em>match specification</em>,
match_spec. A brief explanation is given in
<seealso marker="#select/2">select/2</seealso>. For a detailed
description, see the chapter "Match specifications in Erlang" in
<em>ERTS User's Guide</em>.</p>
</section>
<datatypes>
<datatype>
<name><marker id="type-continuation">continuation()</marker></name>
<desc>
<p>Opaque continuation used by <seealso marker="#select/1">
<c>select/1</c></seealso> and <seealso marker="#select/3">
<c>select/3</c></seealso>.</p>
</desc>
</datatype>
<datatype>
<name name="match_spec"/>
<desc><p>A match specification, see above.</p></desc>
</datatype>
<datatype>
<name name="match_pattern"/>
</datatype>
<datatype>
<name name="tab"/>
</datatype>
<datatype>
<name name="tid"/>
<desc><p>A table identifier, as returned by new/2.</p></desc>
</datatype>
</datatypes>
<funcs>
<func>
<name>all() -> [Tab]</name>
<fsummary>Return a list of all ETS tables.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
</type>
<desc>
<p>Returns a list of all tables at the node. Named tables are
given by their names, unnamed tables are given by their
table identifiers.</p>
</desc>
</func>
<func>
<name>delete(Tab) -> true</name>
<fsummary>Delete an entire ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
</type>
<desc>
<p>Deletes the entire table <c>Tab</c>.</p>
</desc>
</func>
<func>
<name>delete(Tab, Key) -> true</name>
<fsummary>Delete all objects with a given key from an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
</type>
<desc>
<p>Deletes all objects with the key <c>Key</c> from the table
<c>Tab</c>.</p>
</desc>
</func>
<func>
<name>delete_all_objects(Tab) -> true</name>
<fsummary>Delete all objects in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
</type>
<desc>
<p>Delete all objects in the ETS table <c>Tab</c>.
The operation is guaranteed to be
<seealso marker="#concurrency">atomic and isolated</seealso>.</p>
</desc>
</func>
<func>
<name>delete_object(Tab,Object) -> true</name>
<fsummary>Deletes a specific from an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Object = tuple()</v>
</type>
<desc>
<p>Delete the exact object <c>Object</c> from the ETS table,
leaving objects with the same key but other differences
(useful for type <c>bag</c>). In a <c>duplicate_bag</c>, all
instances of the object will be deleted.</p>
</desc>
</func>
<func>
<name name="file2tab" arity="1"/>
<fsummary>Read an ETS table from a file.</fsummary>
<desc>
<p>Reads a file produced by <seealso
marker="#tab2file/2">tab2file/2</seealso> or
<seealso marker="#tab2file/3">tab2file/3</seealso> and creates the
corresponding table <c><anno>Tab</anno></c>.</p>
<p>Equivalent to <c>file2tab(<anno>Filename</anno>, [])</c>.</p>
</desc>
</func>
<func>
<name name="file2tab" arity="2"/>
<fsummary>Read an ETS table from a file.</fsummary>
<desc>
<p>Reads a file produced by <seealso
marker="#tab2file/2">tab2file/2</seealso> or
<seealso marker="#tab2file/3">tab2file/3</seealso> and creates the
corresponding table <c><anno>Tab</anno></c>.</p>
<p>The currently only supported option is <c>{verify,boolean()}</c>. If
verification is turned on (by means of specifying
<c>{verify,true}</c>), the function utilizes whatever
information is present in the file to assert that the
information is not damaged. How this is done depends on which
<c>extended_info</c> was written using
<seealso marker="#tab2file/3">tab2file/3</seealso>.</p>
<p>If no <c>extended_info</c> is present in the file and
<c>{verify,true}</c> is specified, the number of objects
written is compared to the size of the original table when the
dump was started. This might make verification fail if the
table was
<c>public</c> and objects were added or removed while the
table was dumped to file. To avoid this type of problems,
either do not verify files dumped while updated simultaneously
or use the <c>{extended_info, [object_count]}</c> option to
<seealso marker="#tab2file/3">tab2file/3</seealso>, which
extends the information in the file with the number of objects
actually written.</p>
<p>If verification is turned on and the file was written with
the option <c>{extended_info, [md5sum]}</c>, reading the file
is slower and consumes radically more CPU time than
otherwise.</p>
<p><c>{verify,false}</c> is the default.</p>
</desc>
</func>
<func>
<name>first(Tab) -> Key | '$end_of_table'</name>
<fsummary>Return the first key in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
</type>
<desc>
<p>Returns the first key <c>Key</c> in the table <c>Tab</c>.
If the table is of the <c>ordered_set</c> type, the first key
in Erlang term order will be returned. If the table is of any
other type, the first key according to the table's internal
order will be returned. If the table is empty,
<c>'$end_of_table'</c> will be returned.</p>
<p>Use <c>next/2</c> to find subsequent keys in the table.</p>
</desc>
</func>
<func>
<name name="foldl" arity="3"/>
<fsummary>Fold a function over an ETS table</fsummary>
<desc>
<p><c><anno>Acc0</anno></c> is returned if the table is empty.
This function is similar to <c>lists:foldl/3</c>. The order in
which the elements of the table are traversed is unspecified,
except for tables of type <c>ordered_set</c>, for which they
are traversed first to last.</p>
<p>If <c><anno>Function</anno></c> inserts objects into the table, or another
process inserts objects into the table, those objects <em>may</em>
(depending on key ordering) be included in the traversal.</p>
</desc>
</func>
<func>
<name name="foldr" arity="3"/>
<fsummary>Fold a function over an ETS table</fsummary>
<desc>
<p><c><anno>Acc0</anno></c> is returned if the table is empty.
This function is similar to <c>lists:foldr/3</c>. The order in
which the elements of the table are traversed is unspecified,
except for tables of type <c>ordered_set</c>, for which they
are traversed last to first.</p>
<p>If <c><anno>Function</anno></c> inserts objects into the table, or another
process inserts objects into the table, those objects <em>may</em>
(depending on key ordering) be included in the traversal.</p>
</desc>
</func>
<func>
<name name="from_dets" arity="2"/>
<fsummary>Fill an ETS table with objects from a Dets table.</fsummary>
<desc>
<p>Fills an already created ETS table with the objects in the
already opened Dets table named <c><anno>DetsTab</anno></c>. The existing
objects of the ETS table are kept unless overwritten.</p>
<p>Throws a badarg error if any of the tables does not exist or the
dets table is not open.</p>
</desc>
</func>
<func>
<name name="fun2ms" arity="1"/>
<fsummary>Pseudo function that transforms fun syntax to a match_spec.</fsummary>
<desc>
<p>Pseudo function that by means of a <c>parse_transform</c>
translates <c><anno>LiteralFun</anno></c> typed as parameter in the
function call to a
<seealso marker="#match_spec">match_spec</seealso>. With
"literal" is meant that the fun needs to textually be written
as the parameter of the function, it cannot be held in a
variable which in turn is passed to the function).</p>
<p>The parse transform is implemented in the module
<c>ms_transform</c> and the source <em>must</em> include the
file <c>ms_transform.hrl</c> in STDLIB for this
pseudo function to work. Failing to include the hrl file in
the source will result in a runtime error, not a compile
time ditto. The include file is easiest included by adding
the line
<c>-include_lib("stdlib/include/ms_transform.hrl").</c> to
the source file.</p>
<p>The fun is very restricted, it can take only a single
parameter (the object to match): a sole variable or a
tuple. It needs to use the <c>is_</c>XXX guard tests.
Language constructs that have no representation
in a match_spec (like <c>if</c>, <c>case</c>, <c>receive</c>
etc) are not allowed.</p>
<p>The return value is the resulting match_spec.</p>
<p>Example:</p>
<pre>
1> <input>ets:fun2ms(fun({M,N}) when N > 3 -> M end).</input>
[{{'$1','$2'},[{'>','$2',3}],['$1']}]</pre>
<p>Variables from the environment can be imported, so that this
works:</p>
<pre>
2> <input>X=3.</input>
3
3> <input>ets:fun2ms(fun({M,N}) when N > X -> M end).</input>
[{{'$1','$2'},[{'>','$2',{const,3}}],['$1']}]</pre>
<p>The imported variables will be replaced by match_spec
<c>const</c> expressions, which is consistent with the
static scoping for Erlang funs. Local or global function
calls can not be in the guard or body of the fun however.
Calls to builtin match_spec functions of course is allowed:</p>
<pre>
4> <input>ets:fun2ms(fun({M,N}) when N > X, is_atomm(M) -> M end).</input>
Error: fun containing local Erlang function calls
('is_atomm' called in guard) cannot be translated into match_spec
{error,transform_error}
5> <input>ets:fun2ms(fun({M,N}) when N > X, is_atom(M) -> M end).</input>
[{{'$1','$2'},[{'>','$2',{const,3}},{is_atom,'$1'}],['$1']}]</pre>
<p>As can be seen by the example, the function can be called
from the shell too. The fun needs to be literally in the call
when used from the shell as well. Other means than the
parse_transform are used in the shell case, but more or less
the same restrictions apply (the exception being records,
as they are not handled by the shell).</p>
<warning>
<p>If the parse_transform is not applied to a module which
calls this pseudo function, the call will fail in runtime
(with a <c>badarg</c>). The module <c>ets</c> actually
exports a function with this name, but it should never
really be called except for when using the function in the
shell. If the <c>parse_transform</c> is properly applied by
including the <c>ms_transform.hrl</c> header file, compiled
code will never call the function, but the function call is
replaced by a literal match_spec.</p>
</warning>
<p>For more information, see
<seealso marker="ms_transform#top">ms_transform(3)</seealso>.</p>
</desc>
</func>
<func>
<name>give_away(Tab, Pid, GiftData) -> true</name>
<fsummary>Change owner of a table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Pid = pid()</v>
<v>GiftData = term()</v>
</type>
<desc>
<p>Make process <c>Pid</c> the new owner of table <c>Tab</c>.
If successful, the message
<c>{'ETS-TRANSFER',Tab,FromPid,GiftData}</c> will be sent
to the new owner.</p>
<p>The process <c>Pid</c> must be alive, local and not already the
owner of the table. The calling process must be the table owner.</p>
<p>Note that <c>give_away</c> does not at all affect the
<seealso marker="#heir">heir</seealso> option of the table. A table
owner can for example set the <c>heir</c> to itself, give the table
away and then get it back in case the receiver terminates.</p>
</desc>
</func>
<func>
<name name="i" arity="0"/>
<fsummary>Display information about all ETS tables on tty.</fsummary>
<desc>
<p>Displays information about all ETS tables on tty.</p>
</desc>
</func>
<func>
<name name="i" arity="1"/>
<fsummary>Browse an ETS table on tty.</fsummary>
<desc>
<p>Browses the table <c><anno>Tab</anno></c> on tty.</p>
</desc>
</func>
<func>
<name>info(Tab) -> [{Item, Value}] | undefined</name>
<fsummary>Return information about an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Item = atom(), see below</v>
<v>Value = term(), see below</v>
</type>
<desc>
<p>Returns information about the table <c>Tab</c> as a list of
<c>{Item, Value}</c> tuples. If <c>Tab</c> has the correct type
for a table identifier, but does not refer to an existing ETS
table, <c>undefined</c> is returned. If <c>Tab</c> is not of the
correct type, this function fails with reason <c>badarg</c>.</p>
<list type="bulleted">
<item><c>Item=memory, Value=integer()</c> <br></br>
The number of words allocated to the table.</item>
<item><c>Item=owner, Value=pid()</c> <br></br>
The pid of the owner of the table.</item>
<item><c>Item=heir, Value=pid()|none</c> <br></br>
The pid of the heir of the table, or <c>none</c> if no heir is set.</item>
<item><c>Item=name, Value=atom()</c> <br></br>
The name of the table.</item>
<item><c>Item=size, Value=integer()</c> <br></br>
The number of objects inserted in the table.</item>
<item><c>Item=node, Value=atom()</c> <br></br>
The node where the table is stored. This field is no longer
meaningful as tables cannot be accessed from other nodes.</item>
<item><c>Item=named_table, Value=true|false</c> <br></br>
Indicates if the table is named or not.</item>
<item><c>Item=type, Value=set|ordered_set|bag|duplicate_bag</c> <br></br>
The table type.</item>
<item><c>Item=keypos, Value=integer()</c> <br></br>
The key position.</item>
<item><c>Item=protection, Value=public|protected|private</c> <br></br>
The table access rights.</item>
<item><c>Item=compressed, Value=true|false</c> <br></br>
Indicates if the table is compressed or not.</item>
</list>
</desc>
</func>
<func>
<name>info(Tab, Item) -> Value | undefined</name>
<fsummary>Return the information associated with given item for an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Item, Value - see below</v>
</type>
<desc>
<p>Returns the information associated with <c>Item</c> for
the table <c>Tab</c>, or returns <c>undefined</c> if <c>Tab</c>
does not refer an existing ETS table.
If <c>Tab</c> is not of the correct type, or if <c>Item</c> is not
one of the allowed values, this function fails with reason <c>badarg</c>.</p>
<warning><p>In R11B and earlier, this function would not fail but return
<c>undefined</c> for invalid values for <c>Item</c>.</p>
</warning>
<p>In addition to the <c>{Item,Value}</c>
pairs defined for <c>info/1</c>, the following items are
allowed:</p>
<list type="bulleted">
<item><c>Item=fixed, Value=true|false</c> <br></br>
Indicates if the table is fixed by any process or not.</item>
<item>
<p><c>Item=safe_fixed, Value={FirstFixed,Info}|false</c> <br></br>
</p>
<p>If the table has been fixed using <c>safe_fixtable/2</c>,
the call returns a tuple where <c>FirstFixed</c> is the
time when the table was first fixed by a process, which
may or may not be one of the processes it is fixed by
right now.</p>
<p><c>Info</c> is a possibly empty lists of tuples
<c>{Pid,RefCount}</c>, one tuple for every process the
table is fixed by right now. <c>RefCount</c> is the value
of the reference counter, keeping track of how many times
the table has been fixed by the process.</p>
<p>If the table never has been fixed, the call returns
<c>false</c>.</p>
<item><c>Item=stats, Value=tuple()</c> <br></br>
Returns internal statistics about set, bag and duplicate_bag tables on an internal format used by OTP test suites.
Not for production use.</item>
</item>
</list>
</desc>
</func>
<func>
<name name="init_table" arity="2"/>
<fsummary>Replace all objects of an ETS table.</fsummary>
<desc>
<p>Replaces the existing objects of the table <c><anno>Tab</anno></c> with
objects created by calling the input function <c><anno>InitFun</anno></c>,
see below. This function is provided for compatibility with
the <c>dets</c> module, it is not more efficient than filling
a table by using <c>ets:insert/2</c>.
</p>
<p>When called with the argument <c>read</c> the function
<c><anno>InitFun</anno></c> is assumed to return <c>end_of_input</c> when
there is no more input, or <c>{<anno>Objects</anno>, Fun}</c>, where
<c><anno>Objects</anno></c> is a list of objects and <c>Fun</c> is a new
input function. Any other value Value is returned as an error
<c>{error, {init_fun, Value}}</c>. Each input function will be
called exactly once, and should an error occur, the last
function is called with the argument <c>close</c>, the reply
of which is ignored.</p>
<p>If the type of the table is <c>set</c> and there is more
than one object with a given key, one of the objects is
chosen. This is not necessarily the last object with the given
key in the sequence of objects returned by the input
functions. This holds also for duplicated
objects stored in tables of type <c>bag</c>.</p>
</desc>
</func>
<func>
<name>insert(Tab, ObjectOrObjects) -> true</name>
<fsummary>Insert an object into an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>ObjectOrObjects = tuple() | [tuple()]</v>
</type>
<desc>
<p>Inserts the object or all of the objects in the list
<c>ObjectOrObjects</c> into the table <c>Tab</c>.
If the table is a <c>set</c> and the key of the inserted
objects <em>matches</em> the key of any object in the table,
the old object will be replaced. If the table is an
<c>ordered_set</c> and the key of the inserted object
<em>compares equal</em> to the key of any object in the
table, the old object is also replaced. If the list contains
more than one object with <em>matching</em> keys and the table is a
<c>set</c>, one will be inserted, which one is
not defined. The same thing holds for <c>ordered_set</c>, but
will also happen if the keys <em>compare equal</em>.</p>
<p>The entire operation is guaranteed to be
<seealso marker="#concurrency">atomic and isolated</seealso>,
even when a list of objects is inserted.</p>
</desc>
</func>
<func>
<name>insert_new(Tab, ObjectOrObjects) -> boolean()</name>
<fsummary>Insert an object into an ETS table if the key is not already present.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>ObjectOrObjects = tuple() | [tuple()]</v>
</type>
<desc>
<p>This function works exactly like <c>insert/2</c>, with the
exception that instead of overwriting objects with the same
key (in the case of <c>set</c> or <c>ordered_set</c>) or
adding more objects with keys already existing in the table
(in the case of <c>bag</c> and <c>duplicate_bag</c>), it
simply returns <c>false</c>. If <c>ObjectOrObjects</c> is a
list, the function checks <em>every</em> key prior to
inserting anything. Nothing will be inserted if not
<em>all</em> keys present in the list are absent from the
table. Like <c>insert/2</c>, the entire operation is guaranteed to be
<seealso marker="#concurrency">atomic and isolated</seealso>.</p>
</desc>
</func>
<func>
<name>is_compiled_ms(Term) -> boolean()</name>
<fsummary>Checks if an Erlang term is the result of ets:match_spec_compile</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>This function is used to check if a term is a valid
compiled <seealso marker="#match_spec">match_spec</seealso>.
The compiled match_spec is an opaque datatype which can
<em>not</em> be sent between Erlang nodes nor be stored on
disk. Any attempt to create an external representation of a
compiled match_spec will result in an empty binary
(<c><![CDATA[<<>>]]></c>). As an example, the following
expression:</p>
<code type="none">
ets:is_compiled_ms(ets:match_spec_compile([{'_',[],[true]}])).</code>
<p>will yield <c>true</c>, while the following expressions:</p>
<code type="none">
MS = ets:match_spec_compile([{'_',[],[true]}]),
Broken = binary_to_term(term_to_binary(MS)),
ets:is_compiled_ms(Broken).</code>
<p>will yield false, as the variable <c>Broken</c> will contain
a compiled match_spec that has passed through external
representation.</p>
<note>
<p>The fact that compiled match_specs has no external
representation is for performance reasons. It may be subject
to change in future releases, while this interface will
still remain for backward compatibility reasons.</p>
</note>
</desc>
</func>
<func>
<name>last(Tab) -> Key | '$end_of_table'</name>
<fsummary>Return the last key in an ETS table of type<c>ordered_set</c>.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
</type>
<desc>
<p>Returns the last key <c>Key</c> according to Erlang term
order in the table <c>Tab</c> of the <c>ordered_set</c> type.
If the table is of any other type, the function is synonymous
to <c>first/2</c>. If the table is empty,
<c>'$end_of_table'</c> is returned.</p>
<p>Use <c>prev/2</c> to find preceding keys in the table.</p>
</desc>
</func>
<func>
<name>lookup(Tab, Key) -> [Object]</name>
<fsummary>Return all objects with a given key in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
<v>Object = tuple()</v>
</type>
<desc>
<p>Returns a list of all objects with the key <c>Key</c> in
the table <c>Tab</c>.</p>
<p>In the case of <c>set, bag and duplicate_bag</c>, an object
is returned only if the given key <em>matches</em> the key
of the object in the table. If the table is an
<c>ordered_set</c> however, an object is returned if the key
given <em>compares equal</em> to the key of an object in the
table. The difference being the same as between <c>=:=</c>
and <c>==</c>. As an example, one might insert an object
with the
<c>integer()</c><c>1</c> as a key in an <c>ordered_set</c>
and get the object returned as a result of doing a
<c>lookup/2</c> with the <c>float()</c><c>1.0</c> as the
key to search for.</p>
<p>If the table is of type <c>set</c> or <c>ordered_set</c>,
the function returns either the empty list or a list with one
element, as there cannot be more than one object with the same
key. If the table is of type <c>bag</c> or
<c>duplicate_bag</c>, the function returns a list of
arbitrary length.</p>
<p>Note that the time order of object insertions is preserved;
The first object inserted with the given key will be first
in the resulting list, and so on.</p>
<p>Insert and look-up times in tables of type <c>set</c>,
<c>bag</c> and <c>duplicate_bag</c> are constant, regardless
of the size of the table. For the <c>ordered_set</c>
data-type, time is proportional to the (binary) logarithm of
the number of objects.</p>
</desc>
</func>
<func>
<name>lookup_element(Tab, Key, Pos) -> Elem</name>
<fsummary>Return the <c>Pos</c>:th element of all objects with a given key in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
<v>Pos = integer()</v>
<v>Elem = term() | [term()]</v>
</type>
<desc>
<p>If the table <c>Tab</c> is of type <c>set</c> or
<c>ordered_set</c>, the function returns the <c>Pos</c>:th
element of the object with the key <c>Key</c>.</p>
<p>If the table is of type <c>bag</c> or <c>duplicate_bag</c>,
the functions returns a list with the <c>Pos</c>:th element of
every object with the key <c>Key</c>.</p>
<p>If no object with the key <c>Key</c> exists, the function
will exit with reason <c>badarg</c>.</p>
<p>The difference between <c>set</c>, <c>bag</c> and
<c>duplicate_bag</c> on one hand, and <c>ordered_set</c> on
the other, regarding the fact that <c>ordered_set</c>'s
view keys as equal when they <em>compare equal</em>
whereas the other table types only regard them equal when
they <em>match</em>, naturally holds for
<c>lookup_element</c> as well as for <c>lookup</c>.</p>
</desc>
</func>
<func>
<name>match(Tab, Pattern) -> [Match]</name>
<fsummary>Match the objects in an ETS table against a pattern.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Pattern = tuple()</v>
<v>Match = [term()]</v>
</type>
<desc>
<p>Matches the objects in the table <c>Tab</c> against the
pattern <c>Pattern</c>.</p>
<p>A pattern is a term that may contain:</p>
<list type="bulleted">
<item>bound parts (Erlang terms),</item>
<item><c>'_'</c> which matches any Erlang term, and</item>
<item>pattern variables: <c>'$N'</c> where
<c>N</c>=0,1,...</item>
</list>
<p>The function returns a list with one element for each
matching object, where each element is an ordered list of
pattern variable bindings. An example:</p>
<pre>
6> <input>ets:match(T, '$1').</input> % Matches every object in the table
[[{rufsen,dog,7}],[{brunte,horse,5}],[{ludde,dog,5}]]
7> <input>ets:match(T, {'_',dog,'$1'}).</input>
[[7],[5]]
8> <input>ets:match(T, {'_',cow,'$1'}).</input>
[]</pre>
<p>If the key is specified in the pattern, the match is very
efficient. If the key is not specified, i.e. if it is a
variable or an underscore, the entire table must be searched.
The search time can be substantial if the table is very large.</p>
<p>On tables of the <c>ordered_set</c> type, the result is in
the same order as in a <c>first/next</c> traversal.</p>
</desc>
</func>
<func>
<name>match(Tab, Pattern, Limit) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Match the objects in an ETS table against a pattern and returns part of the answers.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Pattern = tuple()</v>
<v>Match = [term()]</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Works like <c>ets:match/2</c> but only returns a limited
(<c>Limit</c>) number of matching objects. The
<c>Continuation</c> term can then be used in subsequent calls
to <c>ets:match/1</c> to get the next chunk of matching
objects. This is a space efficient way to work on objects in a
table which is still faster than traversing the table object
by object using <c>ets:first/1</c> and <c>ets:next/1</c>.</p>
<p><c>'$end_of_table'</c> is returned if the table is empty.</p>
</desc>
</func>
<func>
<name>match(Continuation) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Continues matching objects in an ETS table.</fsummary>
<type>
<v>Match = [term()]</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Continues a match started with <c>ets:match/3</c>. The next
chunk of the size given in the initial <c>ets:match/3</c>
call is returned together with a new <c>Continuation</c>
that can be used in subsequent calls to this function.</p>
<p><c>'$end_of_table'</c> is returned when there are no more
objects in the table.</p>
</desc>
</func>
<func>
<name name="match_delete" arity="2"/>
<fsummary>Delete all objects which match a given pattern from an ETS table.</fsummary>
<desc>
<p>Deletes all objects which match the pattern <c><anno>Pattern</anno></c>
from the table <c><anno>Tab</anno></c>. See <c>match/2</c> for a
description of patterns.</p>
</desc>
</func>
<func>
<name>match_object(Tab, Pattern) -> [Object]</name>
<fsummary>Match the objects in an ETS table against a pattern.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Pattern = Object = tuple()</v>
</type>
<desc>
<p>Matches the objects in the table <c>Tab</c> against the
pattern <c>Pattern</c>. See <c>match/2</c> for a description
of patterns. The function returns a list of all objects which
match the pattern.</p>
<p>If the key is specified in the pattern, the match is very
efficient. If the key is not specified, i.e. if it is a
variable or an underscore, the entire table must be searched.
The search time can be substantial if the table is very large.</p>
<p>On tables of the <c>ordered_set</c> type, the result is in
the same order as in a <c>first/next</c> traversal.</p>
</desc>
</func>
<func>
<name>match_object(Tab, Pattern, Limit) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Match the objects in an ETS table against a pattern and returns part of the answers.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Pattern = tuple()</v>
<v>Match = [term()]</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Works like <c>ets:match_object/2</c> but only returns a
limited (<c>Limit</c>) number of matching objects. The
<c>Continuation</c> term can then be used in subsequent calls
to <c>ets:match_object/1</c> to get the next chunk of matching
objects. This is a space efficient way to work on objects in a
table which is still faster than traversing the table object
by object using <c>ets:first/1</c> and <c>ets:next/1</c>.</p>
<p><c>'$end_of_table'</c> is returned if the table is empty.</p>
</desc>
</func>
<func>
<name>match_object(Continuation) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Continues matching objects in an ETS table.</fsummary>
<type>
<v>Match = [term()]</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Continues a match started with <c>ets:match_object/3</c>.
The next chunk of the size given in the initial
<c>ets:match_object/3</c> call is returned together with a
new <c>Continuation</c> that can be used in subsequent calls
to this function.</p>
<p><c>'$end_of_table'</c> is returned when there are no more
objects in the table.</p>
</desc>
</func>
<func>
<name>match_spec_compile(MatchSpec) -> CompiledMatchSpec</name>
<fsummary>Compiles a match specification into its internal representation</fsummary>
<type>
<v>MatchSpec = match_spec()</v>
<v>CompiledMatchSpec = comp_match_spec()</v>
</type>
<desc>
<p>This function transforms a
<seealso marker="#match_spec">match_spec</seealso> into an
internal representation that can be used in subsequent calls
to <c>ets:match_spec_run/2</c>. The internal representation is
opaque and can not be converted to external term format and
then back again without losing its properties (meaning it can
not be sent to a process on another node and still remain a
valid compiled match_spec, nor can it be stored on disk).
The validity of a compiled match_spec can be checked using
<c>ets:is_compiled_ms/1</c>.</p>
<p>If the term <c>MatchSpec</c> can not be compiled (does not
represent a valid match_spec), a <c>badarg</c> fault is
thrown.</p>
<note>
<p>This function has limited use in normal code, it is used by
Dets to perform the <c>dets:select</c> operations.</p>
</note>
</desc>
</func>
<func>
<name>match_spec_run(List,CompiledMatchSpec) -> list()</name>
<fsummary>Performs matching, using a compiled match_spec, on a list of tuples</fsummary>
<type>
<v>List = [ tuple() ]</v>
<v>CompiledMatchSpec = comp_match_spec()</v>
</type>
<desc>
<p>This function executes the matching specified in a
compiled <seealso marker="#match_spec">match_spec</seealso> on
a list of tuples. The <c>CompiledMatchSpec</c> term should be
the result of a call to <c>ets:match_spec_compile/1</c> and
is hence the internal representation of the match_spec one
wants to use.</p>
<p>The matching will be executed on each element in <c>List</c>
and the function returns a list containing all results. If an
element in <c>List</c> does not match, nothing is returned
for that element. The length of the result list is therefore
equal or less than the the length of the parameter
<c>List</c>. The two calls in the following example will give
the same result (but certainly not the same execution
time...):</p>
<code type="none">
Table = ets:new...
MatchSpec = ....
% The following call...
ets:match_spec_run(ets:tab2list(Table),
ets:match_spec_compile(MatchSpec)),
% ...will give the same result as the more common (and more efficient)
ets:select(Table,MatchSpec),</code>
<note>
<p>This function has limited use in normal code, it is used by
Dets to perform the <c>dets:select</c> operations and by
Mnesia during transactions.</p>
</note>
</desc>
</func>
<func>
<name>member(Tab, Key) -> true | false</name>
<fsummary>Tests for occurrence of a key in an ETS table</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
</type>
<desc>
<p>Works like <c>lookup/2</c>, but does not return the objects.
The function returns <c>true</c> if one or more elements in
the table has the key <c>Key</c>, <c>false</c> otherwise.</p>
</desc>
</func>
<func>
<name>new(Name, Options) -> tid() | atom()</name>
<fsummary>Create a new ETS table.</fsummary>
<type>
<v>Name = atom()</v>
<v>Options = [Option]</v>
<v> Option = Type | Access | named_table | {keypos,Pos} | {heir,pid(),HeirData} | {heir,none} | Tweaks</v>
<v> Type = set | ordered_set | bag | duplicate_bag</v>
<v> Access = public | protected | private</v>
<v> Tweaks = {write_concurrency,boolean()} | {read_concurrency,boolean()} | compressed</v>
<v> Pos = integer()</v>
<v> HeirData = term()</v>
</type>
<desc>
<p>Creates a new table and returns a table identifier which can
be used in subsequent operations. The table identifier can be
sent to other processes so that a table can be shared between
different processes within a node.</p>
<p>The parameter <c>Options</c> is a list of atoms which
specifies table type, access rights, key position and if the
table is named or not. If one or more options are left out,
the default values are used. This means that not specifying
any options (<c>[]</c>) is the same as specifying
<c>[set,protected,{keypos,1},{heir,none},{write_concurrency,false},{read_concurrency,false}]</c>.</p>
<list type="bulleted">
<item>
<p><c>set</c>
The table is a <c>set</c> table - one key, one object,
no order among objects. This is the default table type.</p>
</item>
<item>
<p><c>ordered_set</c>
The table is a <c>ordered_set</c> table - one key, one
object, ordered in Erlang term order, which is the order
implied by the < and > operators. Tables of this type
have a somewhat different behavior in some situations
than tables of the other types. Most notably the
<c>ordered_set</c> tables regard keys as equal when they
<em>compare equal</em>, not only when they match. This
means that to an <c>ordered_set</c>, the
<c>integer()</c><c>1</c> and the <c>float()</c><c>1.0</c> are regarded as equal. This also means that the
key used to lookup an element not necessarily
<em>matches</em> the key in the elements returned, if
<c>float()</c>'s and <c>integer()</c>'s are mixed in
keys of a table.</p>
</item>
<item>
<p><c>bag</c>
The table is a <c>bag</c> table which can have many
objects, but only one instance of each object, per key.</p>
</item>
<item>
<p><c>duplicate_bag</c>
The table is a <c>duplicate_bag</c> table which can have
many objects, including multiple copies of the same
object, per key.</p>
</item>
<item>
<p><c>public</c>
Any process may read or write to the table.</p>
</item>
<item>
<marker id="protected"></marker>
<p><c>protected</c>
The owner process can read and write to the table. Other
processes can only read the table. This is the default
setting for the access rights.</p>
</item>
<item>
<marker id="private"></marker>
<p><c>private</c>
Only the owner process can read or write to the table.</p>
</item>
<item>
<p><c>named_table</c>
If this option is present, the name <c>Name</c> is
associated with the table identifier. The name can then
be used instead of the table identifier in subsequent
operations.</p>
</item>
<item>
<p><c>{keypos,Pos}</c>
Specfies which element in the stored tuples should be
used as key. By default, it is the first element, i.e.
<c>Pos=1</c>. However, this is not always appropriate. In
particular, we do not want the first element to be the
key if we want to store Erlang records in a table.</p>
<p>Note that any tuple stored in the table must have at
least <c>Pos</c> number of elements.</p>
</item>
<item>
<marker id="heir"></marker>
<p><c>{heir,Pid,HeirData} | {heir,none}</c><br></br>
Set a process as heir. The heir will inherit the table if
the owner terminates. The message
<c>{'ETS-TRANSFER',tid(),FromPid,HeirData}</c> will be sent to
the heir when that happens. The heir must be a local process.
Default heir is <c>none</c>, which will destroy the table when
the owner terminates.</p>
</item>
<item>
<marker id="new_2_write_concurrency"></marker>
<p><c>{write_concurrency,boolean()}</c>
Performance tuning. Default is <c>false</c>, in which case an operation that
mutates (writes to) the table will obtain exclusive access,
blocking any concurrent access of the same table until finished.
If set to <c>true</c>, the table is optimized towards concurrent
write access. Different objects of the same table can be mutated
(and read) by concurrent processes. This is achieved to some degree
at the expense of sequential access and concurrent reader performance.
The <c>write_concurrency</c> option can be combined with the
<seealso marker="#new_2_read_concurrency">read_concurrency</seealso>
option. You typically want to combine these when large concurrent
read bursts and large concurrent write bursts are common (see the
documentation of the
<seealso marker="#new_2_read_concurrency">read_concurrency</seealso>
option for more information).
Note that this option does not change any guarantees about
<seealso marker="#concurrency">atomicy and isolation</seealso>.
Functions that makes such promises over several objects (like
<c>insert/2</c>) will gain less (or nothing) from this option.</p>
<p>Table type <c>ordered_set</c> is not affected by this option in current
implementation.</p>
</item>
<item>
<marker id="new_2_read_concurrency"></marker>
<p><c>{read_concurrency,boolean()}</c>
Performance tuning. Default is <c>false</c>. When set to
<c>true</c>, the table is optimized for concurrent read
operations. When this option is enabled on a runtime system with
SMP support, read operations become much cheaper; especially on
systems with multiple physical processors. However, switching
between read and write operations becomes more expensive. You
typically want to enable this option when concurrent read
operations are much more frequent than write operations, or when
concurrent reads and writes comes in large read and write
bursts (i.e., lots of reads not interrupted by writes, and lots
of writes not interrupted by reads). You typically do
<em>not</em> want to enable this option when the common access
pattern is a few read operations interleaved with a few write
operations repeatedly. In this case you will get a performance
degradation by enabling this option. The <c>read_concurrency</c>
option can be combined with the
<seealso marker="#new_2_write_concurrency">write_concurrency</seealso>
option. You typically want to combine these when large concurrent
read bursts and large concurrent write bursts are common.</p>
</item>
<item>
<marker id="new_2_compressed"></marker>
<p><c>compressed</c>
If this option is present, the table data will be stored in a more compact format to
consume less memory. The downside is that it will make table operations slower.
Especially operations that need to inspect entire objects,
such as <c>match</c> and <c>select</c>, will get much slower. The key element
is not compressed in current implementation.</p>
</item>
</list>
</desc>
</func>
<func>
<name>next(Tab, Key1) -> Key2 | '$end_of_table'</name>
<fsummary>Return the next key in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key1 = Key2 = term()</v>
</type>
<desc>
<p>Returns the next key <c>Key2</c>, following the key
<c>Key1</c> in the table <c>Tab</c>. If the table is of the
<c>ordered_set</c> type, the next key in Erlang term order is
returned. If the table is of any other type, the next key
according to the table's internal order is returned. If there
is no next key, <c>'$end_of_table'</c> is returned.</p>
<p>Use <c>first/1</c> to find the first key in the table.</p>
<p>Unless a table of type <c>set</c>, <c>bag</c> or
<c>duplicate_bag</c> is protected using
<c>safe_fixtable/2</c>, see below, a traversal may fail if
concurrent updates are made to the table. If the table is of
type <c>ordered_set</c>, the function returns the next key in
order, even if the object does no longer exist.</p>
</desc>
</func>
<func>
<name>prev(Tab, Key1) -> Key2 | '$end_of_table'</name>
<fsummary>Return the previous key in an ETS table of type<c>ordered_set</c>.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key1 = Key2 = term()</v>
</type>
<desc>
<p>Returns the previous key <c>Key2</c>, preceding the key
<c>Key1</c> according the Erlang term order in the table
<c>Tab</c> of the <c>ordered_set</c> type. If the table is of
any other type, the function is synonymous to <c>next/2</c>.
If there is no previous key, <c>'$end_of_table'</c> is
returned.</p>
<p>Use <c>last/1</c> to find the last key in the table.</p>
</desc>
</func>
<func>
<name>rename(Tab, Name) -> Name</name>
<fsummary>Rename a named ETS table.</fsummary>
<type>
<v>Tab = Name = atom()</v>
</type>
<desc>
<p>Renames the named table <c>Tab</c> to the new name
<c>Name</c>. Afterwards, the old name can not be used to
access the table. Renaming an unnamed table has no effect.</p>
</desc>
</func>
<func>
<name name="repair_continuation" arity="2"/>
<fsummary>Repair a continuation from ets:select/1 or ets:select/3 that has passed through external representation</fsummary>
<desc>
<p>This function can be used to restore an opaque continuation
returned by <c>ets:select/3</c> or <c>ets:select/1</c> if the
continuation has passed through external term format (been
sent between nodes or stored on disk).</p>
<p>The reason for this function is that continuation terms
contain compiled match_specs and therefore will be
invalidated if converted to external term format. Given that
the original match_spec is kept intact, the continuation can
be restored, meaning it can once again be used in subsequent
<c>ets:select/1</c> calls even though it has been stored on
disk or on another node.</p>
<p>As an example, the following sequence of calls will fail:</p>
<code type="none">
T=ets:new(x,[]),
...
{_,C} = ets:select(T,ets:fun2ms(fun({N,_}=A)
when (N rem 10) =:= 0 ->
A
end),10),
Broken = binary_to_term(term_to_binary(C)),
ets:select(Broken).</code>
<p>...while the following sequence will work:</p>
<code type="none">
T=ets:new(x,[]),
...
MS = ets:fun2ms(fun({N,_}=A)
when (N rem 10) =:= 0 ->
A
end),
{_,C} = ets:select(T,MS,10),
Broken = binary_to_term(term_to_binary(C)),
ets:select(ets:repair_continuation(Broken,MS)).</code>
<p>...as the call to <c>ets:repair_continuation/2</c> will
reestablish the (deliberately) invalidated continuation
<c>Broken</c>.</p>
<note>
<p>This function is very rarely needed in application code. It
is used by Mnesia to implement distributed <c>select/3</c>
and <c>select/1</c> sequences. A normal application would
either use Mnesia or keep the continuation from being
converted to external format.</p>
<p>The reason for not having an external representation of a
compiled match_spec is performance. It may be subject to
change in future releases, while this interface will remain
for backward compatibility.</p>
</note>
</desc>
</func>
<func>
<name>safe_fixtable(Tab, true|false) -> true</name>
<fsummary>Fix an ETS table for safe traversal.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
</type>
<desc>
<p>Fixes a table of the <c>set</c>, <c>bag</c> or
<c>duplicate_bag</c> table type for safe traversal.</p>
<p>A process fixes a table by calling
<c>safe_fixtable(Tab,true)</c>. The table remains fixed until
the process releases it by calling
<c>safe_fixtable(Tab,false)</c>, or until the process
terminates.</p>
<p>If several processes fix a table, the table will remain fixed
until all processes have released it (or terminated).
A reference counter is kept on a per process basis, and N
consecutive fixes requires N releases to actually release
the table.</p>
<p>When a table is fixed, a sequence of <c>first/1</c> and
<c>next/2</c> calls are guaranteed to succeed and each object in
the table will only be returned once, even if objects
are removed or inserted during the traversal.
The keys for new objects inserted during the traversal <em>may</em>
be returned by <seealso marker="#next/2">next/2</seealso>
(it depends on the internal ordering of the keys). An example:</p>
<code type="none">
clean_all_with_value(Tab,X) ->
safe_fixtable(Tab,true),
clean_all_with_value(Tab,X,ets:first(Tab)),
safe_fixtable(Tab,false).
clean_all_with_value(Tab,X,'$end_of_table') ->
true;
clean_all_with_value(Tab,X,Key) ->
case ets:lookup(Tab,Key) of
[{Key,X}] ->
ets:delete(Tab,Key);
_ ->
true
end,
clean_all_with_value(Tab,X,ets:next(Tab,Key)).</code>
<p>Note that no deleted objects are actually removed from a
fixed table until it has been released. If a process fixes a
table but never releases it, the memory used by the deleted
objects will never be freed. The performance of operations on
the table will also degrade significantly.</p>
<p>Use <c>info/2</c> to retrieve information about which
processes have fixed which tables. A system with a lot of
processes fixing tables may need a monitor which sends alarms
when tables have been fixed for too long.</p>
<p>Note that for tables of the <c>ordered_set</c> type,
<c>safe_fixtable/2</c> is not necessary as calls to
<c>first/1</c> and <c>next/2</c> will always succeed.</p>
</desc>
</func>
<func>
<name>select(Tab, MatchSpec) -> [Match]</name>
<fsummary>Match the objects in an ETS table against a match_spec.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Match = term()</v>
<v>MatchSpec = match_spec()</v>
</type>
<desc>
<p>Matches the objects in the table <c>Tab</c> using a
<seealso marker="#match_spec">match_spec</seealso>. This is a
more general call than the <c>ets:match/2</c> and
<c>ets:match_object/2</c> calls. In its simplest forms the
match_specs look like this:</p>
<list type="bulleted">
<item>MatchSpec = [MatchFunction]</item>
<item>MatchFunction = {MatchHead, [Guard], [Result]}</item>
<item>MatchHead = "Pattern as in ets:match"</item>
<item>Guard = {"Guardtest name", ...}</item>
<item>Result = "Term construct"</item>
</list>
<p>This means that the match_spec is always a list of one or
more tuples (of arity 3). The tuples first element should be
a pattern as described in the documentation of
<c>ets:match/2</c>. The second element of the tuple should
be a list of 0 or more guard tests (described below). The
third element of the tuple should be a list containing a
description of the value to actually return. In almost all
normal cases the list contains exactly one term which fully
describes the value to return for each object.</p>
<p>The return value is constructed using the "match variables"
bound in the MatchHead or using the special match variables
<c>'$_'</c> (the whole matching object) and <c>'$$'</c> (all
match variables in a list), so that the following
<c>ets:match/2</c> expression:</p>
<code type="none">
ets:match(Tab,{'$1','$2','$3'})</code>
<p>is exactly equivalent to:</p>
<code type="none">
ets:select(Tab,[{{'$1','$2','$3'},[],['$$']}])</code>
<p>- and the following <c>ets:match_object/2</c> call:</p>
<code type="none">
ets:match_object(Tab,{'$1','$2','$1'})</code>
<p>is exactly equivalent to</p>
<code type="none">
ets:select(Tab,[{{'$1','$2','$1'},[],['$_']}])</code>
<p>Composite terms can be constructed in the <c>Result</c> part
either by simply writing a list, so that this code:</p>
<code type="none">
ets:select(Tab,[{{'$1','$2','$3'},[],['$$']}])</code>
<p>gives the same output as:</p>
<code type="none">
ets:select(Tab,[{{'$1','$2','$3'},[],[['$1','$2','$3']]}])</code>
<p>i.e. all the bound variables in the match head as a list. If
tuples are to be constructed, one has to write a tuple of
arity 1 with the single element in the tuple being the tuple
one wants to construct (as an ordinary tuple could be mistaken
for a <c>Guard</c>). Therefore the following call:</p>
<code type="none">
ets:select(Tab,[{{'$1','$2','$1'},[],['$_']}])</code>
<p>gives the same output as:</p>
<code type="none">
ets:select(Tab,[{{'$1','$2','$1'},[],[{{'$1','$2','$3'}}]}])</code>
<p>- this syntax is equivalent to the syntax used in the trace
patterns (see
<seealso marker="runtime_tools:dbg">dbg(3)</seealso>).</p>
<p>The <c>Guard</c>s are constructed as tuples where the first
element is the name of the test and the rest of the elements
are the parameters of the test. To check for a specific type
(say a list) of the element bound to the match variable
<c>'$1'</c>, one would write the test as
<c>{is_list, '$1'}</c>. If the test fails, the object in the
table will not match and the next <c>MatchFunction</c> (if
any) will be tried. Most guard tests present in Erlang can be
used, but only the new versions prefixed <c>is_</c> are
allowed (like <c>is_float</c>, <c>is_atom</c> etc).</p>
<p>The <c>Guard</c> section can also contain logic and
arithmetic operations, which are written with the same syntax
as the guard tests (prefix notation), so that a guard test
written in Erlang looking like this:</p>
<code type="none"><![CDATA[
is_integer(X), is_integer(Y), X + Y < 4711]]></code>
<p>is expressed like this (X replaced with '$1' and Y with
'$2'):</p>
<code type="none"><![CDATA[
[{is_integer, '$1'}, {is_integer, '$2'}, {'<', {'+', '$1', '$2'}, 4711}]]]></code>
<p>On tables of the <c>ordered_set</c> type, objects are visited
in the same order as in a <c>first/next</c>
traversal. This means that the match specification will be
executed against objects with keys in the <c>first/next</c>
order and the corresponding result list will be in the order of that
execution.</p>
</desc>
</func>
<func>
<name>select(Tab, MatchSpec, Limit) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Match the objects in an ETS table against a match_spec and returns part of the answers.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Match = term()</v>
<v>MatchSpec = match_spec()</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Works like <c>ets:select/2</c> but only returns a limited
(<c>Limit</c>) number of matching objects. The
<c>Continuation</c> term can then be used in subsequent calls
to <c>ets:select/1</c> to get the next chunk of matching
objects. This is a space efficient way to work on objects in a
table which is still faster than traversing the table object
by object using <c>ets:first/1</c> and <c>ets:next/1</c>.</p>
<p><c>'$end_of_table'</c> is returned if the table is empty.</p>
</desc>
</func>
<func>
<name>select(Continuation) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Continue matching objects in an ETS table.</fsummary>
<type>
<v>Match = term()</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Continues a match started with
<c>ets:select/3</c>. The next
chunk of the size given in the initial <c>ets:select/3</c>
call is returned together with a new <c>Continuation</c>
that can be used in subsequent calls to this function.</p>
<p><c>'$end_of_table'</c> is returned when there are no more
objects in the table.</p>
</desc>
</func>
<func>
<name>select_count(Tab, MatchSpec) -> NumMatched</name>
<fsummary>Match the objects in an ETS table against a match_spec and returns the number of objects for which the match_spec returned 'true'</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Object = tuple()</v>
<v>MatchSpec = match_spec()</v>
<v>NumMatched = integer()</v>
</type>
<desc>
<p>Matches the objects in the table <c>Tab</c> using a
<seealso marker="#match_spec">match_spec</seealso>. If the
match_spec returns <c>true</c> for an object, that object
considered a match and is counted. For any other result from
the match_spec the object is not considered a match and is
therefore not counted.</p>
<p>The function could be described as a <c>match_delete/2</c>
that does not actually delete any elements, but only counts
them.</p>
<p>The function returns the number of objects matched.</p>
</desc>
</func>
<func>
<name>select_delete(Tab, MatchSpec) -> NumDeleted</name>
<fsummary>Match the objects in an ETS table against a match_spec and deletes objects where the match_spec returns 'true'</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Object = tuple()</v>
<v>MatchSpec = match_spec()</v>
<v>NumDeleted = integer()</v>
</type>
<desc>
<p>Matches the objects in the table <c>Tab</c> using a
<seealso marker="#match_spec">match_spec</seealso>. If the
match_spec returns <c>true</c> for an object, that object is
removed from the table. For any other result from the
match_spec the object is retained. This is a more general
call than the <c>ets:match_delete/2</c> call.</p>
<p>The function returns the number of objects actually
deleted from the table.</p>
<note>
<p>The <c>match_spec</c> has to return the atom <c>true</c> if
the object is to be deleted. No other return value will get the
object deleted, why one can not use the same match specification for
looking up elements as for deleting them.</p>
</note>
</desc>
</func>
<func>
<name>select_reverse(Tab, MatchSpec) -> [Match]</name>
<fsummary>Match the objects in an ETS table against a match_spec.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Match = term()</v>
<v>MatchSpec = match_spec()</v>
</type>
<desc>
<p>Works like <c>select/2</c>, but returns the list in reverse
order for the <c>ordered_set</c> table type. For all other table
types, the return value is identical to that of <c>select/2</c>.</p>
</desc>
</func>
<func>
<name>select_reverse(Tab, MatchSpec, Limit) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Match the objects in an ETS table against a match_spec and returns part of the answers.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Match = term()</v>
<v>MatchSpec = match_spec()</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Works like <c>select/3</c>, but for the <c>ordered_set</c>
table type, traversing is done starting at the last object in
Erlang term order and moves towards the first. For all other
table types, the return value is identical to that of
<c>select/3</c>.</p>
<p>Note that this is <em>not</em> equivalent to
reversing the result list of a <c>select/3</c> call, as the result list
is not only reversed, but also contains the last <c>Limit</c>
matching objects in the table, not the first.</p>
</desc>
</func>
<func>
<name>select_reverse(Continuation) -> {[Match],Continuation} | '$end_of_table'</name>
<fsummary>Continue matching objects in an ETS table.</fsummary>
<type>
<v>Match = term()</v>
<v>Continuation = term()</v>
</type>
<desc>
<p>Continues a match started with
<c>ets:select_reverse/3</c>. If the table is an
<c>ordered_set</c>, the traversal of the table will continue
towards objects with keys earlier in the Erlang term order. The
returned list will also contain objects with keys in reverse
order.</p>
<p>For all other table types, the behaviour is exatly that of <c>select/1</c>.</p>
<p>Example:</p>
<code>
1> T = ets:new(x,[ordered_set]).
2> [ ets:insert(T,{N}) || N <- lists:seq(1,10) ].
...
3> {R0,C0} = ets:select_reverse(T,[{'_',[],['$_']}],4).
...
4> R0.
[{10},{9},{8},{7}]
5> {R1,C1} = ets:select_reverse(C0).
...
6> R1.
[{6},{5},{4},{3}]
7> {R2,C2} = ets:select_reverse(C1).
...
8> R2.
[{2},{1}]
9> '$end_of_table' = ets:select_reverse(C2).
...
</code>
</desc>
</func>
<func>
<name>setopts(Tab, Opts) -> true</name>
<fsummary>Set table options.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Opts = Opt | [Opt]</v>
<v>Opt = {heir,pid(),HeirData} | {heir,none}</v>
<v>HeirData = term()</v>
</type>
<desc>
<p>Set table options. The only option that currently is allowed to be
set after the table has been created is
<seealso marker="#heir">heir</seealso>. The calling process must be
the table owner.</p>
</desc>
</func>
<func>
<name>slot(Tab, I) -> [Object] | '$end_of_table'</name>
<fsummary>Return all objects in a given slot of an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>I = integer()</v>
<v>Object = tuple()</v>
</type>
<desc>
<p>This function is mostly for debugging purposes, Normally
one should use <c>first/next</c> or <c>last/prev</c> instead.</p>
<p>Returns all objects in the <c>I</c>:th slot of the table
<c>Tab</c>. A table can be traversed by repeatedly calling
the function, starting with the first slot <c>I=0</c> and
ending when <c>'$end_of_table'</c> is returned.
The function will fail with reason <c>badarg</c> if the
<c>I</c> argument is out of range.</p>
<p>Unless a table of type <c>set</c>, <c>bag</c> or
<c>duplicate_bag</c> is protected using
<c>safe_fixtable/2</c>, see above, a traversal may fail if
concurrent updates are made to the table. If the table is of
type <c>ordered_set</c>, the function returns a list
containing the <c>I</c>:th object in Erlang term order.</p>
</desc>
</func>
<func>
<name name="tab2file" arity="2"/>
<fsummary>Dump an ETS table to a file.</fsummary>
<desc>
<p>Dumps the table <c><anno>Tab</anno></c> to the file <c><anno>Filename</anno></c>.</p>
<p>Equivalent to <c>tab2file(<anno>Tab</anno>, <anno>Filename</anno>,[])</c></p>
</desc>
</func>
<func>
<name name="tab2file" arity="3"/>
<fsummary>Dump an ETS table to a file.</fsummary>
<desc>
<p>Dumps the table <c><anno>Tab</anno></c> to the file <c><anno>Filename</anno></c>.</p>
<p>When dumping the table, certain information about the table
is dumped to a header at the beginning of the dump. This
information contains data about the table type,
name, protection, size, version and if it's a named table. It
also contains notes about what extended information is added
to the file, which can be a count of the objects in the file
or a MD5 sum of the header and records in the file.</p>
<p>The size field in the header might not correspond to the
actual number of records in the file if the table is public
and records are added or removed from the table during
dumping. Public tables updated during dump, and that one wants
to verify when reading, needs at least one field of extended
information for the read verification process to be reliable
later.</p>
<p>The <c>extended_info</c> option specifies what extra
information is written to the table dump:</p>
<taglist>
<tag><c>object_count</c></tag>
<item><p>The number of objects actually written to the file is
noted in the file footer, why verification of file truncation
is possible even if the file was updated during
dump.</p></item>
<tag><c>md5sum</c></tag>
<item><p>The header and objects in the file are checksummed using
the built in MD5 functions. The MD5 sum of all objects is
written in the file footer, so that verification while reading
will detect the slightest bitflip in the file data. Using this
costs a fair amount of CPU time.</p></item>
</taglist>
<p>Whenever the <c>extended_info</c> option is used, it
results in a file not readable by versions of ets prior to
that in stdlib-1.15.1</p>
</desc>
</func>
<func>
<name name="tab2list" arity="1"/>
<fsummary>Return a list of all objects in an ETS table.</fsummary>
<desc>
<p>Returns a list of all objects in the table <c><anno>Tab</anno></c>.</p>
</desc>
</func>
<func>
<name name="tabfile_info" arity="1"/>
<fsummary>Return a list of all objects in an ETS table.</fsummary>
<desc>
<p>Returns information about the table dumped to file by
<seealso marker="#tab2file/2">tab2file/2</seealso> or
<seealso marker="#tab2file/3">tab2file/3</seealso></p>
<p>The following items are returned:</p>
<taglist>
<tag>name</tag>
<item><p>The name of the dumped table. If the table was a
named table, a table with the same name cannot exist when the
table is loaded from file with
<seealso marker="#file2tab/2">file2tab/2</seealso>. If the table is
not saved as a named table, this field has no significance
at all when loading the table from file.</p></item>
<tag>type</tag>
<item>The ets type of the dumped table (i.e. <c>set</c>, <c>bag</c>,
<c>duplicate_bag</c> or <c>ordered_set</c>). This type will be used
when loading the table again.</item>
<tag>protection</tag>
<item>The protection of the dumped table (i.e. <c>private</c>,
<c>protected</c> or <c>public</c>). A table loaded from the file
will get the same protection.</item>
<tag>named_table</tag>
<item><c>true</c> if the table was a named table when dumped
to file, otherwise <c>false</c>. Note that when a named table
is loaded from a file, there cannot exist a table in the
system with the same name.</item>
<tag>keypos</tag>
<item>The <c>keypos</c> of the table dumped to file, which
will be used when loading the table again.</item>
<tag>size</tag>
<item>The number of objects in the table when the table dump
to file started, which in case of a <c>public</c> table need
not correspond to the number of objects actually saved to the
file, as objects might have been added or deleted by another
process during table dump.</item>
<tag>extended_info</tag>
<item>The extended information written in the file footer to
allow stronger verification during table loading from file, as
specified to <seealso
marker="#tab2file/3">tab2file/3</seealso>. Note that this
function only tells <em>which</em> information is present, not
the values in the file footer. The value is a list containing
one or more of the atoms <c>object_count</c> and
<c>md5sum</c>.</item>
<tag>version</tag>
<item>A tuple <c>{<anno>Major</anno>,<anno>Minor</anno>}</c> containing the major and
minor version of the file format for ets table dumps. This
version field was added beginning with stdlib-1.5.1, files
dumped with older versions will return <c>{0,0}</c> in this
field.</item>
</taglist>
<p>An error is returned if the file is inaccessible,
badly damaged or not an file produced with <seealso
marker="#tab2file/2">tab2file/2</seealso> or <seealso
marker="#tab2file/3">tab2file/3</seealso>.</p>
</desc>
</func>
<func>
<name name="table" arity="1"/>
<name name="table" arity="2"/>
<fsummary>Return a QLC query handle.</fsummary>
<desc>
<p><marker id="qlc_table"></marker>Returns a QLC (Query List
Comprehension) query handle. The module <c>qlc</c> implements
a query language aimed mainly at Mnesia but ETS tables, Dets
tables, and lists are also recognized by QLC as sources of
data. Calling <c>ets:table/1,2</c> is the means to make the
ETS table <c>Tab</c> usable to QLC.</p>
<p>When there are only simple restrictions on the key position
QLC uses <c>ets:lookup/2</c> to look up the keys, but when
that is not possible the whole table is traversed. The
option <c>traverse</c> determines how this is done:</p>
<list type="bulleted">
<item>
<p><c>first_next</c>. The table is traversed one key at
a time by calling <c>ets:first/1</c> and
<c>ets:next/2</c>.</p>
</item>
<item>
<p><c>last_prev</c>. The table is traversed one key at
a time by calling <c>ets:last/1</c> and
<c>ets:prev/2</c>.</p>
</item>
<item>
<p><c>select</c>. The table is traversed by calling
<c>ets:select/3</c> and <c>ets:select/1</c>. The option
<c>n_objects</c> determines the number of objects
returned (the third argument of <c>select/3</c>); the
default is to return <c>100</c> objects at a time. The
<seealso marker="#match_spec">match_spec</seealso> (the
second argument of <c>select/3</c>) is assembled by QLC:
simple filters are translated into equivalent match_specs
while more complicated filters have to be applied to all
objects returned by <c>select/3</c> given a match_spec
that matches all objects.</p>
</item>
<item>
<p><c>{select, <anno>MatchSpec</anno>}</c>. As for <c>select</c>
the table is traversed by calling <c>ets:select/3</c> and
<c>ets:select/1</c>. The difference is that the
match_spec is explicitly given. This is how to state
match_specs that cannot easily be expressed within the
syntax provided by QLC.</p>
</item>
</list>
<p>The following example uses an explicit match_spec to
traverse the table:</p>
<pre>
9> <input>true = ets:insert(Tab = ets:new(t, []), [{1,a},{2,b},{3,c},{4,d}]),</input>
<input>MS = ets:fun2ms(fun({X,Y}) when (X > 1) or (X < 5) -> {Y} end),</input>
<input>QH1 = ets:table(Tab, [{traverse, {select, MS}}]).</input></pre>
<p>An example with implicit match_spec:</p>
<pre>
10> <input>QH2 = qlc:q([{Y} || {X,Y} <- ets:table(Tab), (X > 1) or (X < 5)]).</input></pre>
<p>The latter example is in fact equivalent to the former which
can be verified using the function <c>qlc:info/1</c>:</p>
<pre>
11> <input>qlc:info(QH1) =:= qlc:info(QH2).</input>
true</pre>
<p><c>qlc:info/1</c> returns information about a query handle,
and in this case identical information is returned for the
two query handles.</p>
</desc>
</func>
<func>
<name name="test_ms" arity="2"/>
<fsummary>Test a match_spec for use in ets:select/2.</fsummary>
<desc>
<p>This function is a utility to test a
<seealso marker="#match_spec">match_spec</seealso> used in
calls to <c>ets:select/2</c>. The function both tests
<c><anno>MatchSpec</anno></c> for "syntactic" correctness and runs the
match_spec against the object <c><anno>Tuple</anno></c>. If the match_spec
contains errors, the tuple <c>{error, <anno>Errors</anno>}</c> is returned
where <c><anno>Errors</anno></c> is a list of natural language
descriptions of what was wrong with the match_spec. If the
match_spec is syntactically OK, the function returns
<c>{ok,<anno>Result</anno>}</c> where <c><anno>Result</anno></c> is what would have been
the result in a real <c>ets:select/2</c> call or <c>false</c>
if the match_spec does not match the object <c><anno>Tuple</anno></c>.</p>
<p>This is a useful debugging and test tool, especially when
writing complicated <c>ets:select/2</c> calls.</p>
</desc>
</func>
<func>
<name name="to_dets" arity="2"/>
<fsummary>Fill a Dets table with objects from an ETS table.</fsummary>
<desc>
<p>Fills an already created/opened Dets table with the objects
in the already opened ETS table named <c><anno>Tab</anno></c>. The Dets
table is emptied before the objects are inserted.</p>
</desc>
</func>
<func>
<name>update_counter(Tab, Key, UpdateOp) -> Result</name>
<name>update_counter(Tab, Key, [UpdateOp]) -> [Result]</name>
<name>update_counter(Tab, Key, Incr) -> Result</name>
<fsummary>Update a counter object in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = term()</v>
<v>UpdateOp = {Pos,Incr} | {Pos,Incr,Threshold,SetValue}</v>
<v>Pos = Incr = Threshold = SetValue = Result = integer()</v>
</type>
<desc>
<p>This function provides an efficient way to update one or more
counters, without the hassle of having to look up an object, update
the object by incrementing an element and insert the resulting object
into the table again. (The update is done atomically; i.e. no process
can access the ets table in the middle of the operation.)
</p>
<p>It will destructively update the object with key <c>Key</c>
in the table <c>Tab</c> by adding <c>Incr</c> to the element
at the <c>Pos</c>:th position. The new counter value is
returned. If no position is specified, the element directly
following the key (<c><![CDATA[<keypos>+1]]></c>) is updated.</p>
<p>If a <c>Threshold</c> is specified, the counter will be
reset to the value <c>SetValue</c> if the following
conditions occur:</p>
<list type="bulleted">
<item>The <c>Incr</c> is not negative (<c>>= 0</c>) and the
result would be greater than (<c>></c>) <c>Threshold</c></item>
<item>The <c>Incr</c> is negative (<c><![CDATA[< 0]]></c>) and the
result would be less than (<c><![CDATA[<]]></c>)
<c>Threshold</c></item>
</list>
<p>A list of <c>UpdateOp</c> can be supplied to do several update
operations within the object. The operations are carried out in the
order specified in the list. If the same counter position occurs
more than one time in the list, the corresponding counter will thus
be updated several times, each time based on the previous result.
The return value is a list of the new counter values from each
update operation in the same order as in the operation list. If an
empty list is specified, nothing is updated and an empty list is
returned. If the function should fail, no updates will be done at
all.
</p>
<p>The given Key is used to identify the object by either
<em>matching</em> the key of an object in a <c>set</c> table,
or <em>compare equal</em> to the key of an object in an
<c>ordered_set</c> table (see
<seealso marker="#lookup/2">lookup/2</seealso> and
<seealso marker="#new/2">new/2</seealso>
for details on the difference).</p>
<p>The function will fail with reason <c>badarg</c> if:</p>
<list type="bulleted">
<item>the table is not of type <c>set</c> or
<c>ordered_set</c>,</item>
<item>no object with the right key exists,</item>
<item>the object has the wrong arity,</item>
<item>the element to update is not an integer,</item>
<item>the element to update is also the key, or,</item>
<item>any of <c>Pos</c>, <c>Incr</c>, <c>Threshold</c> or
<c>SetValue</c> is not an integer</item>
</list>
</desc>
</func>
<func>
<name>update_element(Tab, Key, {Pos,Value}) -> true | false</name>
<name>update_element(Tab, Key, [{Pos,Value}]) -> true | false</name>
<fsummary>Updates the <c>Pos</c>:th element of the object with a given key in an ETS table.</fsummary>
<type>
<v>Tab = tid() | atom()</v>
<v>Key = Value = term()</v>
<v>Pos = integer()</v>
</type>
<desc>
<p>This function provides an efficient way to update one or more
elements within an object, without the hassle of having to look up,
update and write back the entire object.
</p>
<p>It will destructively update the object with key <c>Key</c>
in the table <c>Tab</c>. The element at the <c>Pos</c>:th position
will be given the value <c>Value</c>. </p>
<p>A list of <c>{Pos,Value}</c> can be supplied to update several
elements within the same object. If the same position occurs more
than one in the list, the last value in the list will be written. If
the list is empty or the function fails, no updates will be done at
all. The function is also atomic in the sense that other processes
can never see any intermediate results.
</p>
<p>The function returns <c>true</c> if an object with the key
<c>Key</c> was found, <c>false</c> otherwise.
</p>
<p>The given Key is used to identify the object by either
<em>matching</em> the key of an object in a <c>set</c> table,
or <em>compare equal</em> to the key of an object in an
<c>ordered_set</c> table (see
<seealso marker="#lookup/2">lookup/2</seealso> and
<seealso marker="#new/2">new/2</seealso>
for details on the difference).</p>
<p>The function will fail with reason <c>badarg</c> if:</p>
<list type="bulleted">
<item>the table is not of type <c>set</c> or
<c>ordered_set</c>,</item>
<item><c>Pos</c> is less than 1 or greater than the object
arity, or,</item>
<item>the element to update is also the key</item>
</list>
</desc>
</func>
</funcs>
</erlref>
