<?xml version="1.0" encoding="latin1" ?>
<!DOCTYPE chapter SYSTEM "chapter.dtd">
<chapter>
<header>
<copyright>
<year>1999</year><year>2010</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>Match specifications in Erlang</title>
<prepared>Patrik Nyblom</prepared>
<responsible></responsible>
<docno></docno>
<approved></approved>
<checked></checked>
<date>1999-06-01</date>
<rev>PA1</rev>
<file>match_spec.xml</file>
</header>
<p>A "match specification" (match_spec) is an Erlang term describing a
small "program" that will try to match something (either the
parameters to a function as used in the <c><![CDATA[erlang:trace_pattern/2]]></c>
BIF, or the objects in an ETS table.).
The match_spec in many ways works like a small function in Erlang, but is
interpreted/compiled by the Erlang runtime system to something much more
efficient than calling an Erlang function. The match_spec is also
very limited compared to the expressiveness of real Erlang functions.</p>
<p>Match specifications are given to the BIF <c><![CDATA[erlang:trace_pattern/2]]></c> to
execute matching of function arguments as well as to define some actions
to be taken when the match succeeds (the <c><![CDATA[MatchBody]]></c> part). Match
specifications can also be used in ETS, to specify objects to be
returned from an <c><![CDATA[ets:select/2]]></c> call (or other select
calls). The semantics and restrictions differ slightly when using
match specifications for tracing and in ETS, the differences are
defined in a separate paragraph below.</p>
<p>The most notable difference between a match_spec and an Erlang fun is
of course the syntax. Match specifications are Erlang terms, not
Erlang code. A match_spec also has a somewhat strange concept of
exceptions. An exception (e.g., <c><![CDATA[badarg]]></c>) in the <c><![CDATA[MatchCondition]]></c>
part,
which resembles an Erlang guard, will generate immediate failure,
while an exception in the <c><![CDATA[MatchBody]]></c> part, which resembles the body of an
Erlang function, is implicitly caught and results in the single atom
<c><![CDATA['EXIT']]></c>.
</p>
<section>
<title>Grammar</title>
<p>A match_spec can be described in this <em>informal</em> grammar:</p>
<list type="bulleted">
<item>MatchExpression ::= [ MatchFunction, ... ]
</item>
<item>MatchFunction ::= { MatchHead, MatchConditions, MatchBody }
</item>
<item>MatchHead ::= MatchVariable | <c><![CDATA['_']]></c> | [ MatchHeadPart, ... ]
</item>
<item>MatchHeadPart ::= term() | MatchVariable | <c><![CDATA['_']]></c></item>
<item>MatchVariable ::= '$<number>'
</item>
<item>MatchConditions ::= [ MatchCondition, ...] | <c><![CDATA[[]]]></c></item>
<item>MatchCondition ::= { GuardFunction } |
{ GuardFunction, ConditionExpression, ... }
</item>
<item>BoolFunction ::= <c><![CDATA[is_atom]]></c> | <c><![CDATA[is_constant]]></c> |
<c><![CDATA[is_float]]></c> | <c><![CDATA[is_integer]]></c> | <c><![CDATA[is_list]]></c> |
<c><![CDATA[is_number]]></c> | <c><![CDATA[is_pid]]></c> | <c><![CDATA[is_port]]></c> |
<c><![CDATA[is_reference]]></c> | <c><![CDATA[is_tuple]]></c> | <c><![CDATA[is_binary]]></c> |
<c><![CDATA[is_function]]></c> | <c><![CDATA[is_record]]></c> | <c><![CDATA[is_seq_trace]]></c> |
<c><![CDATA['and']]></c> | <c><![CDATA['or']]></c> | <c><![CDATA['not']]></c> | <c><![CDATA['xor']]></c> |
<c><![CDATA[andalso]]></c> | <c><![CDATA[orelse]]></c></item>
<item>ConditionExpression ::= ExprMatchVariable | { GuardFunction } |
{ GuardFunction, ConditionExpression, ... } | TermConstruct
</item>
<item>ExprMatchVariable ::= MatchVariable (bound in the MatchHead) |
<c><![CDATA['$_']]></c> | <c><![CDATA['$$']]></c></item>
<item>TermConstruct = {{}} | {{ ConditionExpression, ... }} |
<c><![CDATA[[]]]></c> | [ConditionExpression, ...] | NonCompositeTerm | Constant
</item>
<item>NonCompositeTerm ::= term() (not list or tuple)
</item>
<item>Constant ::= {<c><![CDATA[const]]></c>, term()}
</item>
<item>GuardFunction ::= BoolFunction | <c><![CDATA[abs]]></c> |
<c><![CDATA[element]]></c> | <c><![CDATA[hd]]></c> | <c><![CDATA[length]]></c> | <c><![CDATA[node]]></c> |
<c><![CDATA[round]]></c> | <c><![CDATA[size]]></c> | <c><![CDATA[tl]]></c> | <c><![CDATA[trunc]]></c> |
<c><![CDATA['+']]></c> | <c><![CDATA['-']]></c> | <c><![CDATA['*']]></c> | <c><![CDATA['div']]></c> |
<c><![CDATA['rem']]></c> | <c><![CDATA['band']]></c> | <c><![CDATA['bor']]></c> | <c><![CDATA['bxor']]></c> |
<c><![CDATA['bnot']]></c> | <c><![CDATA['bsl']]></c> | <c><![CDATA['bsr']]></c> | <c><![CDATA['>']]></c> |
<c><![CDATA['>=']]></c> | <c><![CDATA['<']]></c> | <c><![CDATA['=<']]></c> | <c><![CDATA['=:=']]></c> |
<c><![CDATA['==']]></c> | <c><![CDATA['=/=']]></c> | <c><![CDATA['/=']]></c> | <c><![CDATA[self]]></c> |
<c><![CDATA[get_tcw]]></c></item>
<item>MatchBody ::= [ ActionTerm ]
</item>
<item>ActionTerm ::= ConditionExpression | ActionCall
</item>
<item>ActionCall ::= {ActionFunction} |
{ActionFunction, ActionTerm, ...}
</item>
<item>ActionFunction ::= <c><![CDATA[set_seq_token]]></c> |
<c><![CDATA[get_seq_token]]></c> | <c><![CDATA[message]]></c> |
<c><![CDATA[return_trace]]></c> | <c><![CDATA[exception_trace]]></c> | <c><![CDATA[process_dump]]></c> |
<c><![CDATA[enable_trace]]></c> | <c><![CDATA[disable_trace]]></c> | <c><![CDATA[trace]]></c> |
<c><![CDATA[display]]></c> | <c><![CDATA[caller]]></c> | <c><![CDATA[set_tcw]]></c> |
<c><![CDATA[silent]]></c></item>
</list>
</section>
<section>
<title>Function descriptions</title>
<section>
<title>Functions allowed in all types of match specifications</title>
<p>The different functions allowed in <c><![CDATA[match_spec]]></c> work like this:
</p>
<p><em>is_atom, is_constant, is_float, is_integer, is_list, is_number, is_pid, is_port, is_reference, is_tuple, is_binary, is_function: </em> Like the corresponding guard tests in
Erlang, return <c><![CDATA[true]]></c> or <c><![CDATA[false]]></c>.
</p>
<p><em>is_record: </em>Takes an additional parameter, which SHALL
be the result of <c><![CDATA[record_info(size, <record_type>)]]></c>,
like in <c><![CDATA[{is_record, '$1', rectype, record_info(size, rectype)}]]></c>.
</p>
<p><em>'not': </em>Negates its single argument (anything other
than <c><![CDATA[false]]></c> gives <c><![CDATA[false]]></c>).
</p>
<p><em>'and': </em>Returns <c><![CDATA[true]]></c> if all its arguments
(variable length argument list) evaluate to <c><![CDATA[true]]></c>, else
<c><![CDATA[false]]></c>. Evaluation order is undefined.
</p>
<p><em>'or': </em>Returns <c><![CDATA[true]]></c> if any of its arguments
evaluates to <c><![CDATA[true]]></c>. Variable length argument
list. Evaluation order is undefined.
</p>
<p><em>andalso: </em>Like <c><![CDATA['and']]></c>, but quits evaluating its
arguments as soon as one argument evaluates to something else
than true. Arguments are evaluated left to right.
</p>
<p><em>orelse: </em>Like <c><![CDATA['or']]></c>, but quits evaluating as soon
as one of its arguments evaluates to <c><![CDATA[true]]></c>. Arguments are
evaluated left to right.
</p>
<p><em>'xor': </em>Only two arguments, of which one has to be true
and the other false to return <c><![CDATA[true]]></c>; otherwise
<c><![CDATA['xor']]></c> returns false.
</p>
<p><em>abs, element, hd, length, node, round, size, tl, trunc, '+', '-', '*', 'div', 'rem', 'band', 'bor', 'bxor', 'bnot', 'bsl', 'bsr', '>', '>=', '<', '=<', '=:=', '==', '=/=', '/=', self: </em>Work as the corresponding Erlang bif's (or
operators). In case of bad arguments, the result depends on
the context. In the <c><![CDATA[MatchConditions]]></c> part of the
expression, the test fails immediately (like in an Erlang
guard), but in the <c><![CDATA[MatchBody]]></c>, exceptions are implicitly
caught and the call results in the atom <c><![CDATA['EXIT']]></c>.</p>
</section>
<section>
<title>Functions allowed only for tracing</title>
<p><em>is_seq_trace: </em>Returns <c><![CDATA[true]]></c> if a sequential
trace token is set for the current process, otherwise <c><![CDATA[false]]></c>.
</p>
<p><em>set_seq_token:</em> Works like
<c><![CDATA[seq_trace:set_token/2]]></c>, but returns <c><![CDATA[true]]></c> on success
and <c><![CDATA['EXIT']]></c> on error or bad argument. Only allowed in the
<c><![CDATA[MatchBody]]></c> part and only allowed when tracing.
</p>
<p><em>get_seq_token:</em> Works just like
<c><![CDATA[seq_trace:get_token/0]]></c>, and is only allowed in the
<c><![CDATA[MatchBody]]></c> part when tracing.
</p>
<p><em>message:</em> Sets an additional message appended to the
trace message sent. One can only set one additional message in
the body; subsequent calls will replace the appended message. As
a special case, <c><![CDATA[{message, false}]]></c> disables sending of
trace messages ('call' and 'return_to')
for this function call, just like if the match_spec had not matched,
which can be useful if only the side effects of
the <c><![CDATA[MatchBody]]></c> are desired.
Another special case is <c><![CDATA[{message, true}]]></c> which
sets the default behavior, as if the function had no match_spec,
trace message is sent with no extra
information (if no other calls to <c><![CDATA[message]]></c> are placed
before <c><![CDATA[{message, true}]]></c>, it is in fact a "noop").
</p>
<p>Takes one argument, the message. Returns <c><![CDATA[true]]></c> and can
only be used in the <c><![CDATA[MatchBody]]></c> part and when tracing.
</p>
<p><em>return_trace:</em> Causes a <c><![CDATA[return_from]]></c> trace
message to be sent upon return from the current function.
Takes no arguments, returns <c><![CDATA[true]]></c> and can only be used
in the <c><![CDATA[MatchBody]]></c> part when tracing.
If the process trace flag <c><![CDATA[silent]]></c>
is active the <c><![CDATA[return_from]]></c> trace message is inhibited.
</p>
<p>NOTE! If the traced function is tail recursive, this match
spec function destroys that property.
Hence, if a match spec executing this function is used on a
perpetual server process, it may only be active for a limited
time, or the emulator will eventually use all memory in the host
machine and crash. If this match_spec function is inhibited
using the <c><![CDATA[silent]]></c> process trace flag
tail recursiveness still remains.
</p>
<p><em>exception_trace:</em> Same as <em>return_trace</em>,
plus; if the traced function exits due to an exception,
an <c><![CDATA[exception_from]]></c> trace message is generated,
whether the exception is caught or not.
</p>
<p><em>process_dump:</em> Returns some textual information about
the current process as a binary. Takes no arguments and is only
allowed in the <c><![CDATA[MatchBody]]></c> part when tracing.
</p>
<p><em>enable_trace:</em> With one parameter this function turns
on tracing like the Erlang call <c><![CDATA[erlang:trace(self(), true, [P2])]]></c>, where <c><![CDATA[P2]]></c> is the parameter to
<c><![CDATA[enable_trace]]></c>. With two parameters, the first parameter
should be either a process identifier or the registered name of
a process. In this case tracing is turned on for the designated
process in the same way as in the Erlang call <c><![CDATA[erlang:trace(P1, true, [P2])]]></c>, where P1 is the first and P2 is the second
argument. The process <c><![CDATA[P1]]></c> gets its trace messages sent to the same
tracer as the process executing the statement uses. <c><![CDATA[P1]]></c>
can <em>not</em> be one of the atoms <c><![CDATA[all]]></c>, <c><![CDATA[new]]></c> or
<c><![CDATA[existing]]></c> (unless, of course, they are registered names).
<c><![CDATA[P2]]></c> can <em>not</em> be <c><![CDATA[cpu_timestamp]]></c> nor
<c><![CDATA[{tracer,_}]]></c>.
Returns <c><![CDATA[true]]></c> and may only be used in
the <c><![CDATA[MatchBody]]></c> part when tracing.
</p>
<p><em>disable_trace:</em> With one parameter this function
disables tracing like the Erlang call <c><![CDATA[erlang:trace(self(), false, [P2])]]></c>, where <c><![CDATA[P2]]></c> is the parameter to
<c><![CDATA[disable_trace]]></c>. With two parameters it works like the
Erlang call <c><![CDATA[erlang:trace(P1, false, [P2])]]></c>, where P1 can
be either a process identifier or a registered name and is given
as the first argument to the match_spec function.
<c><![CDATA[P2]]></c> can <em>not</em> be <c><![CDATA[cpu_timestamp]]></c> nor
<c><![CDATA[{tracer,_}]]></c>. Returns
<c><![CDATA[true]]></c> and may only be used in the <c><![CDATA[MatchBody]]></c> part
when tracing.
</p>
<p><em>trace:</em> With two parameters this function takes a list
of trace flags to disable as first parameter and a list
of trace flags to enable as second parameter. Logically, the
disable list is applied first, but effectively all changes
are applied atomically. The trace flags
are the same as for <c><![CDATA[erlang:trace/3]]></c> not including
<c><![CDATA[cpu_timestamp]]></c> but including <c><![CDATA[{tracer,_}]]></c>. If a
tracer is specified in both lists, the tracer in the
enable list takes precedence. If no tracer is specified the
same tracer as the process executing the match spec is
used. With three parameters to this function the first is
either a process identifier or the registered name of a
process to set trace flags on, the second is the disable
list, and the third is the enable list. Returns
<c><![CDATA[true]]></c> if any trace property was changed for the
trace target process or <c><![CDATA[false]]></c> if not. It may only
be used in the <c><![CDATA[MatchBody]]></c> part when tracing.
</p>
<p><em>caller:</em>
Returns the calling function as a tuple {Module,
Function, Arity} or the atom <c><![CDATA[undefined]]></c> if the calling
function cannot be determined. May only be used in the
<c><![CDATA[MatchBody]]></c> part when tracing.
</p>
<p>Note that if a "technically built in function" (i.e. a
function not written in Erlang) is traced, the <c><![CDATA[caller]]></c>
function will sometimes return the atom <c><![CDATA[undefined]]></c>. The calling
Erlang function is not available during such calls.
</p>
<p><em>display:</em> For debugging purposes only; displays the
single argument as an Erlang term on stdout, which is seldom
what is wanted. Returns <c><![CDATA[true]]></c> and may only be used in the
<c><![CDATA[MatchBody]]></c> part when tracing.
</p>
<p> <marker id="get_tcw"></marker>
<em>get_tcw:</em>
Takes no argument and returns the value of the node's trace
control word. The same is done by
<c><![CDATA[erlang:system_info(trace_control_word)]]></c>.
</p>
<p>The trace control word is a 32-bit unsigned integer intended for
generic trace control. The trace control word can be tested and
set both from within trace match specifications and with BIFs.
This call is only allowed when tracing.
</p>
<p> <marker id="set_tcw"></marker>
<em>set_tcw:</em>
Takes one unsigned integer argument, sets the value of
the node's trace control word to the value of the argument
and returns the previous value. The same is done by
<c><![CDATA[erlang:system_flag(trace_control_word, Value)]]></c>. It is only
allowed to use <c><![CDATA[set_tcw]]></c> in the <c><![CDATA[MatchBody]]></c> part
when tracing.
</p>
<p><em>silent:</em>
Takes one argument. If the argument is <c><![CDATA[true]]></c>, the call
trace message mode for the current process is set to silent
for this call and all subsequent, i.e call trace messages
are inhibited even if <c><![CDATA[{message, true}]]></c> is called in the
<c><![CDATA[MatchBody]]></c> part for a traced function.
</p>
<p>This mode can also be activated with the <c><![CDATA[silent]]></c> flag
to <c><![CDATA[erlang:trace/3]]></c>.
</p>
<p>If the argument is <c><![CDATA[false]]></c>, the call trace message mode
for the current process is set to normal (non-silent) for
this call and all subsequent.
</p>
<p>If the argument is neither <c><![CDATA[true]]></c> nor <c><![CDATA[false]]></c>,
the call trace message mode is unaffected.</p>
</section>
<p><em>Note</em> that all "function calls" have to be tuples,
even if they take no arguments. The value of <c><![CDATA[self]]></c> is
the atom() <c><![CDATA[self]]></c>, but the value of <c><![CDATA[{self}]]></c> is
the pid() of the current process.</p>
</section>
<section>
<title>Variables and literals</title>
<p>Variables take the form <c><![CDATA['$<number>']]></c> where
<c><![CDATA[<number>]]></c> is an integer between 0 (zero) and
100000000 (1e+8), the behavior if the number is outside these
limits is <em>undefined</em>. In the <c><![CDATA[MatchHead]]></c> part, the special
variable <c><![CDATA['_']]></c> matches anything, and never gets bound (like
<c><![CDATA[_]]></c> in Erlang). In the <c><![CDATA[MatchCondition/MatchBody]]></c>
parts, no unbound variables are allowed, why <c><![CDATA['_']]></c> is
interpreted as itself (an atom). Variables can only be bound in
the <c><![CDATA[MatchHead]]></c> part. In the <c><![CDATA[MatchBody]]></c> and
<c><![CDATA[MatchCondition]]></c> parts, only variables bound previously may
be used. As a special case, in the
<c><![CDATA[MatchCondition/MatchBody]]></c> parts, the variable <c><![CDATA['$_']]></c>
expands to the whole expression which matched the
<c><![CDATA[MatchHead]]></c> (i.e., the whole parameter list to the possibly
traced function or the whole matching object in the ets table)
and the variable <c><![CDATA['$$']]></c> expands to a list
of the values of all bound variables in order
(i.e. <c><![CDATA[['$1','$2', ...]]]></c>).
</p>
<p>In the <c><![CDATA[MatchHead]]></c> part, all literals (except the variables
noted above) are interpreted as is. In the
<c><![CDATA[MatchCondition/MatchBody]]></c> parts, however, the
interpretation is in some ways different. Literals in the
<c><![CDATA[MatchCondition/MatchBody]]></c> can either be written as is,
which works for all literals except tuples, or by using the
special form <c><![CDATA[{const, T}]]></c>, where <c><![CDATA[T]]></c> is any Erlang
term. For tuple literals in the match_spec, one can also use
double tuple parentheses, i.e., construct them as a tuple of
arity one containing a single tuple, which is the one to be
constructed. The "double tuple parenthesis" syntax is useful to
construct tuples from already bound variables, like in
<c><![CDATA[{{'$1', [a,b,'$2']}}]]></c>. Some examples may be needed:
</p>
<table>
<row>
<cell align="left" valign="middle">Expression </cell>
<cell align="left" valign="middle">Variable bindings </cell>
<cell align="left" valign="middle">Result </cell>
</row>
<row>
<cell align="left" valign="middle">{{'$1','$2'}} </cell>
<cell align="left" valign="middle">'$1' = a, '$2' = b</cell>
<cell align="left" valign="middle">{a,b}</cell>
</row>
<row>
<cell align="left" valign="middle">{const, {'$1', '$2'}} </cell>
<cell align="left" valign="middle">doesn't matter</cell>
<cell align="left" valign="middle">{'$1', '$2'}</cell>
</row>
<row>
<cell align="left" valign="middle">a </cell>
<cell align="left" valign="middle">doesn't matter </cell>
<cell align="left" valign="middle">a</cell>
</row>
<row>
<cell align="left" valign="middle">'$1' </cell>
<cell align="left" valign="middle">'$1' = [] </cell>
<cell align="left" valign="middle">[]</cell>
</row>
<row>
<cell align="left" valign="middle">['$1'] </cell>
<cell align="left" valign="middle">'$1' = [] </cell>
<cell align="left" valign="middle">[[]]</cell>
</row>
<row>
<cell align="left" valign="middle">[{{a}}] </cell>
<cell align="left" valign="middle">doesn't matter</cell>
<cell align="left" valign="middle">[{a}]</cell>
</row>
<row>
<cell align="left" valign="middle">42 </cell>
<cell align="left" valign="middle">doesn't matter</cell>
<cell align="left" valign="middle">42</cell>
</row>
<row>
<cell align="left" valign="middle">"hello" </cell>
<cell align="left" valign="middle">doesn't matter</cell>
<cell align="left" valign="middle">"hello"</cell>
</row>
<row>
<cell align="left" valign="middle">$1 </cell>
<cell align="left" valign="middle">doesn't matter</cell>
<cell align="left" valign="middle">49 (the ASCII value for the character '1')</cell>
</row>
<tcaption>Literals in the MatchCondition/MatchBody parts of a match_spec</tcaption>
</table>
</section>
<section>
<title>Execution of the match</title>
<p>The execution of the match expression, when the runtime system
decides whether a trace message should be sent, goes as follows:
</p>
<p>For each tuple in the <c><![CDATA[MatchExpression]]></c> list and while no
match has succeeded:</p>
<list type="bulleted">
<item>Match the <c><![CDATA[MatchHead]]></c> part against the arguments to the
function,
binding the <c><![CDATA['$<number>']]></c> variables (much like in
<c><![CDATA[ets:match/2]]></c>).
If the <c><![CDATA[MatchHead]]></c> cannot match the arguments, the match fails.
</item>
<item>Evaluate each <c><![CDATA[MatchCondition]]></c> (where only
<c><![CDATA['$<number>']]></c> variables previously bound in the
<c><![CDATA[MatchHead]]></c> can occur) and expect it to return the atom
<c><![CDATA[true]]></c>. As soon as a condition does not evaluate to
<c><![CDATA[true]]></c>, the match fails. If any BIF call generates an
exception, also fail.
</item>
<item>
<list type="bulleted">
<item><em>If the match_spec is executing when tracing:</em><br></br>
Evaluate each <c><![CDATA[ActionTerm]]></c> in the same way as the
<c><![CDATA[MatchConditions]]></c>, but completely ignore the return
values. Regardless of what happens in this part, the match has
succeeded.</item>
<item><em>If the match_spec is executed when selecting objects from an ETS table:</em><br></br>
Evaluate the expressions in order and return the value of
the last expression (typically there is only one expression
in this context)</item>
</list>
</item>
</list>
</section>
<section>
<title>Differences between match specifications in ETS and tracing</title>
<p>ETS match specifications are there to produce a return
value. Usually the expression contains one single
<c><![CDATA[ActionTerm]]></c> which defines the return value without having
any side effects. Calls with side effects are not allowed in the
ETS context.</p>
<p>When tracing there is no return value to produce, the
match specification either matches or doesn't. The effect when the
expression matches is a trace message rather then a returned
term. The <c><![CDATA[ActionTerm]]></c>'s are executed as in an imperative
language, i.e. for their side effects. Functions with side effects
are also allowed when tracing.</p>
<p>In ETS the match head is a <c><![CDATA[tuple()]]></c> (or a single match
variable) while it is a list (or a single match variable) when
tracing.</p>
</section>
<section>
<title>Examples</title>
<p>Match an argument list of three where the first and third arguments
are equal:</p>
<code type="none"><![CDATA[
[{['$1', '_', '$1'],
[],
[]}]
]]></code>
<p>Match an argument list of three where the second argument is
a number greater than three:</p>
<code type="none"><![CDATA[
[{['_', '$1', '_'],
[{ '>', '$1', 3}],
[]}]
]]></code>
<p>Match an argument list of three, where the third argument
is a tuple containing argument one and two <em>or</em> a list
beginning with argument one and two (i. e. <c><![CDATA[[a,b,[a,b,c]]]]></c> or
<c><![CDATA[[a,b,{a,b}]]]></c>):
</p>
<code type="none"><![CDATA[
[{['$1', '$2', '$3'],
[{orelse,
{'=:=', '$3', {{'$1','$2'}}},
{'and',
{'=:=', '$1', {hd, '$3'}},
{'=:=', '$2', {hd, {tl, '$3'}}}}}],
[]}]
]]></code>
<p>The above problem may also be solved like this:</p>
<code type="none"><![CDATA[
[{['$1', '$2', {'$1', '$2}], [], []},
{['$1', '$2', ['$1', '$2' | '_']], [], []}]
]]></code>
<p>Match two arguments where the first is a tuple beginning with
a list which in turn begins with the second argument times
two (i. e. [{[4,x],y},2] or [{[8], y, z},4])</p>
<code type="none"><![CDATA[
[{['$1', '$2'],[{'=:=', {'*', 2, '$2'}, {hd, {element, 1, '$1'}}}],
[]}]
]]></code>
<p>Match three arguments. When all three are equal and are
numbers, append the process dump to the trace message, else
let the trace message be as is, but set the sequential trace
token label to 4711.</p>
<code type="none"><![CDATA[
[{['$1', '$1', '$1'],
[{is_number, '$1'}],
[{message, {process_dump}}]},
{'_', [], [{set_seq_token, label, 4711}]}]
]]></code>
<p>As can be noted above, the parameter list can be matched
against a single <c><![CDATA[MatchVariable]]></c> or an <c><![CDATA['_']]></c>. To replace the
whole
parameter list with a single variable is a special case. In all
other cases the <c><![CDATA[MatchHead]]></c> has to be a <em>proper</em> list.
</p>
<p>Match all objects in an ets table where the first element is
the atom 'strider' and the tuple arity is 3 and return the whole
object.</p>
<code type="none"><![CDATA[
[{{strider,'_'.'_'},
[],
['$_']}]
]]></code>
<p>Match all objects in an ets table with arity > 1 and the first
element is 'gandalf', return element 2.</p>
<code type="none"><![CDATA[
[{'$1',
[{'==', gandalf, {element, 1, '$1'}},{'>=',{size, '$1'},2}],
[{element,2,'$1'}]}]
]]></code>
<p>In the above example, if the first element had been the key,
it's much more efficient to match that key in the <c><![CDATA[MatchHead]]></c>
part than in the <c><![CDATA[MatchConditions]]></c> part. The search space of
the tables is restricted with regards to the <c><![CDATA[MatchHead]]></c> so
that only objects with the matching key are searched.
</p>
<p>Match tuples of 3 elements where the second element is either
'merry' or 'pippin', return the whole objects.</p>
<code type="none"><![CDATA[
[{{'_',merry,'_'},
[],
['$_']},
{{'_',pippin,'_'},
[],
['$_']}]
]]></code>
<p>The function <c><![CDATA[ets:test_ms/2]]></c> can be useful for testing
complicated ets matches.</p>
</section>
</chapter>
