diff options
Diffstat (limited to 'system/doc/extensions/bit_syntax.xml')
| -rw-r--r-- | system/doc/extensions/bit_syntax.xml | 403 |
1 files changed, 403 insertions, 0 deletions
diff --git a/system/doc/extensions/bit_syntax.xml b/system/doc/extensions/bit_syntax.xml new file mode 100644 index 0000000000..d86f73cd9a --- /dev/null +++ b/system/doc/extensions/bit_syntax.xml @@ -0,0 +1,403 @@ +<?xml version="1.0" encoding="latin1" ?> +<!DOCTYPE chapter SYSTEM "chapter.dtd"> + +<chapter> + <header> + <copyright> + <year>2000</year><year>2009</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>The Bit Syntax</title> + <prepared>Björn Gustavsson</prepared> + <responsible>Bjarne Däcker</responsible> + <docno>1</docno> + <approved>Bjarne DäKer</approved> + <checked></checked> + <date>00-06-21</date> + <rev>PA1</rev> + <file>bit_syntax.sgml</file> + </header> + <p>This section describes the "bit syntax" which was added to + the Erlang language in release 5.0 (R7). + Compared to the original bit syntax prototype + by Claes Wikström and Tony Rogvall (presented on the + Erlang User's Conference 1999), this implementation differs + primarily in the following respects, + </p> + <list type="ordered"> + <item> + <p>the character pairs '<<' and '>>' are used to delimit + a binary patterns and constructor (not '<' and '>' as in + the prototype), + </p> + </item> + <item> + <p>the tail syntax ('|Variable') has been eliminated, + </p> + </item> + <item> + <p>all size expressions must be bound, + </p> + </item> + <item> + <p>a type <c>unit:U</c> has been added, + </p> + </item> + <item> + <p>lists and tuples cannot be generated + </p> + </item> + <item> + <p>there are no paddings whatsoever. + </p> + </item> + </list> + + <section> + <title>Introduction</title> + <p>In Erlang a Bin is used for constructing binaries and + matching binary patterns. A Bin is written with the + following syntax:</p> + <code type="none"><![CDATA[ + <<E1, E2, ... En>> + ]]></code> + <p>A Bin is a low-level sequence of bytes. The purpose of a + Bin is to be able to, from a high level, + <em>construct</em> a binary, + </p> + <code type="none"><![CDATA[ + Bin = <<E1, E2, ... En>> + ]]></code> + <p>in which case all elements must be bound, or to + <em>match</em> a binary, + </p> + <code type="none"><![CDATA[ + <<E1, E2, ... En>> = Bin + ]]></code> + <p>where <c>Bin</c> is bound, and where the elements are bound or unbound, + as in any match. + </p> + <p>Each element specifies a certain <em>segment</em> of the binary. + A segment is is a set of contiguous bits of the binary (not + necessarily on a byte boundary). The first element specifies + the initial segment, the second element specifies the following + segment etc. + </p> + <p>The following examples illustrate how binaries are constructed + or matched, and how elements and tails are specified. + </p> + + <section> + <title>Examples</title> + <p><em>Example 1: </em>A binary can be constructed from a set of + constants or a string literal:</p> + <code type="none"><![CDATA[ + Bin11 = <<1, 17, 42>>, + Bin12 = <<"abc">> + ]]></code> + <p>yields binaries of size 3; <c>binary_to_list(Bin11)</c> + evaluates to <c>[1, 17, 42]</c>, and + <c>binary_to_list(Bin12)</c> evaluates to <c>[97, 98, 99]</c>. + </p> + <p><em>Example 2: </em>Similarly, a binary can be constructed + from a set of bound variables:</p> + <code type="none"><![CDATA[ + A = 1, B = 17, C = 42, + Bin2 = <<A, B, C:16>> + ]]></code> + <p>yields a binary of size 4, and <c>binary_to_list(Bin2)</c> + evaluates to <c>[1, 17, 00, 42]</c> too. Here we used a + <em>size expression</em> for the variable <c>C</c> in order to + specify a 16-bits segment of <c>Bin2</c>. + </p> + <p><em>Example 3: </em>A Bin can also be used for matching: if + <c>D</c>, <c>E</c>, and <c>F</c> are unbound variables, and + <c>Bin2</c> is bound as in the former example,</p> + <code type="none"><![CDATA[ + <<D:16, E, F/binary>> = Bin2 + ]]></code> + <p>yields <c>D = 273</c>, <c>E = 00</c>, and F binds to a binary + of size 1: <c>binary_to_list(F) = [42]</c>. + </p> + <p><em>Example 4: </em>The following is a more elaborate example + of matching, where <c>Dgram</c> is bound to the consecutive + bytes of an IP datagram of IP protocol version 4, and where we + want to extract the header and the data of the datagram:</p> + <code type="none"><![CDATA[ + -define(IP_VERSION, 4). + -define(IP_MIN_HDR_LEN, 5). + + DgramSize = byte_size(Dgram), + case Dgram of + <<?IP_VERSION:4, HLen:4, SrvcType:8, TotLen:16, + ID:16, Flgs:3, FragOff:13, + TTL:8, Proto:8, HdrChkSum:16, + SrcIP:32, + DestIP:32, RestDgram/binary>> when HLen >= 5, 4*HLen =< DgramSize -> + OptsLen = 4*(HLen - ?IP_MIN_HDR_LEN), + <<Opts:OptsLen/binary,Data/binary>> = RestDgram, + ... + end. + ]]></code> + <p>Here the segment corresponding to the <c>Opts</c> variable + has a <em>type modifier</em> specifying that <c>Opts</c> should + bind to a binary. All other variables have the default type + equal to unsigned integer. + </p> + <p>An IP datagram header is of variable length, and its length - + measured in the number of 32-bit words - is given in the segment + corresponding to <c>HLen</c>, the minimum value of which is + 5. It is the segment corresponding to <c>Opts</c> that is + variable: if <c>HLen</c> is equal to 5, <c>Opts</c> will be an + empty binary. + </p> + <p>The tail variables <c>RestDgram</c> and <c>Data</c> bind to + binaries, as all tail variables do. Both may bind to empty + binaries. + </p> + <p>If the first 4-bits segment of <c>Dgram</c> is not equal to + 4, or if <c>HLen</c> is less than 5, or if the size of + <c>Dgram</c> is less than <c>4*HLen</c>, the match of + <c>Dgram</c> fails. + </p> + </section> + </section> + + <section> + <title>A Lexical Note</title> + <p>Note that "<c><![CDATA[B=<<1>>]]></c>" will be interpreted as + "<c><![CDATA[B =< ;<1>>]]></c>", which is a syntax error. + The correct way to write the expression is "<c><![CDATA[B = <<1>>]]></c>".</p> + </section> + + <section> + <title>Segments</title> + <p>Each segment has the following general syntax:</p> + <p><c>Value:Size/TypeSpecifierList</c></p> + <p>Both the <c>Size</c> and the <c>TypeSpecifier</c> or both may be + omitted; thus the following variations are allowed: + </p> + <p><c>Value</c></p> + <p><c>Value:Size</c></p> + <p><c>Value/TypeSpecifierList</c></p> + <p>Default values will be used for missing specifications. The default + values are described in the section "Defaults" below. + </p> + <p>Used in binary construction, the <c>Value</c> part is any expression. + Used in binary matching, the <c>Value</c> part must be a literal or + variable. You can read more about the <c>Value</c> part in the + sections about constructing binaries and matching binaries. + </p> + <p>The <c>Size</c> part of the segment multiplied by the unit in the + <c>TypeSpecifierList</c> (described below) gives the number of bits + for the segment. In construction, <c>Size</c> is any expression that + evaluates to an integer. In matching, <c>Size</c> must be a constant + expression or a variable. + </p> + <p>The <c>TypeSpecifierList</c> is a list of type specifiers separated by + hyphens. + </p> + <taglist> + <tag>Type</tag> + <item>The type can be <c>integer</c>, <c>float</c>, or <c>binary</c>.</item> + <tag>Signedness</tag> + <item>The signedness specification can be either <c>signed</c> + or <c>unsigned</c>. Note that signedness only matters for matching.</item> + <tag>Endianness</tag> + <item>The endianness specification can be either <c>big</c>, + <c>little</c>, or <c>native</c>. Native-endian means that + the endian will be resolved at load time to be either + big-endian or little-endian, depending on what is "native" + for the CPU that the Erlang machine is run on.</item> + <tag>Unit</tag> + <item>The unit size is given as <c>unit:IntegerLiteral</c>. + The allowed range is 1-256. It will be multiplied by the <c>Size</c> + specifier to give the effective size of the segment.</item> + </taglist> + <p>Example: + </p> + <code type="none"> +X:4/little-signed-integer-unit:8 + </code> + <p>This element has a total size of 4*8 = 32 bits, and it contains a + signed integer in little-endian order.</p> + </section> + + <section> + <title>Defaults</title> + <p>The default type for a segment is <c>integer</c>. The default type + does <em>not</em> depend on the value, even if the value is a literal. + For instance, the default type in '<c><![CDATA[<<3.14>>]]></c>' is <c>integer</c>, + not <c>float</c>. + </p> + <p>The default <c>Size</c> depends on the type. + For <c>integer</c> it is 8. For <c>float</c> it is 64. + For <c>binary</c> it is all of the binary. In matching, this default + value is only valid for the very last element. All other binary elements + in matching must have a size specification. + </p> + <p>The default unit depends on the the type. + For <c>integer</c> and <c>float</c> it is 1. + For <c>binary</c> it is 8. + </p> + <p>The default signedness is <c>unsigned</c>. + </p> + <p>The default endianness is <c>big</c>.</p> + </section> + + <section> + <title>Constructing Binaries</title> + <p>This section describes the rules for constructing binaries using + the bit syntax. Unlike when constructing lists or tuples, the construction + of a binary can fail with a <c>badarg</c> exception. + </p> + <p>There can be zero or more segments in a binary to be constructed. + The expression '<c><![CDATA[<<>>]]></c>' constructs a zero length binary. + </p> + <p>Each segment in a binary can consist of zero or more bits. + There are no alignment rules for individual segments, but the total + number of bits in all segments must be evenly divisible by 8, + or in other words, the resulting binary must consist of a whole number + of bytes. An <c>badarg</c> exception will be thrown if the resulting + binary is not byte-aligned. Example: + </p> + <code type="none"><![CDATA[ +<<X:1,Y:6>> + ]]></code> + <p>The total number of bits is 7, which is not evenly divisible by 8; + thus, there will be <c>badarg</c> exception (and a compiler warning + as well). The following example + </p> + <code type="none"><![CDATA[ +<<X:1,Y:6,Z:1>> + ]]></code> + <p>will successfully construct a binary of 8 bits, or one byte. (Provided + that all of X, Y and Z are integers.) + </p> + <p>As noted earlier, segments have the following general syntax: + </p> + <p><c>Value:Size/TypeSpecifierList</c></p> + <p>When constructing binaries, <c>Value</c> and <c>Size</c> can be + any Erlang expression. However, for syntactical reasons, + both <c>Value</c> and <c>Size</c> must be enclosed in parenthesis + if the expression consists of anything more than a single literal + or variable. The following gives a compiler syntax error: + </p> + <code type="none"><![CDATA[ +<<X+1:8>> + ]]></code> + <p>This expression must be rewritten to + </p> + <code type="none"><![CDATA[ +<<(X+1):8>> + ]]></code> + <p>in order to be accepted by the compiler. + </p> + + <section> + <title>Including Literal Strings</title> + <p>As syntactic sugar, an literal string may be written instead of a + element.</p> + <code type="none"><![CDATA[ +<<"hello">> ]]></code> + <p>which is syntactic sugar for</p> + <code type="none"><![CDATA[ +<<$h,$e,$l,$l,$o>> ]]></code> + </section> + </section> + + <section> + <title>Matching Binaries</title> + <p>This section describes the rules for matching binaries using the + bit syntax. + </p> + <p>There can be zero or more segments in a binary binary pattern. + A binary pattern can occur in every place patterns are allowed, also + inside other patterns. Binary patterns cannot be nested. + </p> + <p>The pattern '<c><![CDATA[<<>>]]></c>' matches a zero length binary. + </p> + <p>Each segment in a binary can consist of zero or more bits. + </p> + <p>A segment of type <c>binary</c> must have a size evenly divisible by 8. + </p> + <p>This means that the following head will never match:</p> + <code type="none"><![CDATA[ +foo(<<X:7/binary,Y:1/binary>>) -> ]]></code> + <p>As noted earlier, segments have the following general syntax: + </p> + <p><c>Value:Size/TypeSpecifierList</c></p> + <p>When matching <c>Value</c> value must be either a variable or an integer + or floating point literal. Expressions are not allowed. + </p> + <p><c>Size</c> must be an integer literal, or a previously bound variable. + Note that the following is not allowed:</p> + <code type="none"><![CDATA[ +foo(N, <<X:N,T/binary>>) -> + {X,T}. ]]></code> + <p>The two occurrences of <c>N</c> are not related. The compiler + will complain that the <c>N</c> in the size field is unbound. + </p> + <p>The correct way to write this example is like this:</p> + <code type="none"><![CDATA[ +foo(N, Bin) -> + <<X:N,T/binary>> = Bin, + {X,T}. ]]></code> + + <section> + <title>Getting the Rest of the Binary</title> + <p>To match out the rest of binary, specify a binary field without size:</p> + <code type="none"><![CDATA[ +foo(<<A:8,Rest/binary>>) -> ]]></code> + <p>As always, the size of the tail must be evenly divisible by 8. + </p> + </section> + </section> + + <section> + <title>Traps and Pitfalls</title> + <p>Assume that we need a function that creates a binary out of a + list of triples of integers. A first (inefficient) version of such + a function could look like this:</p> + <code type="none"><![CDATA[ +triples_to_bin(T) -> + triples_to_bin(T, <<>>). + +triples_to_bin([{X,Y,Z} | T], Acc) -> + triples_to_bin(T, <<Acc/binary, X:32, Y:32, Z:32>>); % inefficient +triples_to_bin([], Acc) -> + Acc. ]]></code> + <p>The reason for the inefficiency of this function is that for + each triple, the binary constructed so far (<c>Acc</c>) is copied. + (Note: The original bit syntax prototype avoided the copy operation + by using segmented binaries, which are not implemented in R7.) + </p> + <p>The efficient way to write this function in R7 is:</p> + <code type="none"><![CDATA[ +triples_to_bin(T) -> + triples_to_bin(T, []). + +triples_to_bin([{X,Y,Z} | T], Acc) -> + triples_to_bin(T, [<<X:32, Y:32, Z:32>> | Acc]); +triples_to_bin([], Acc) -> + list_to_binary(lists:reverse(Acc)). ]]></code> + <p>Note that <c>list_to_binary/1</c> handles deep lists of binaries + and small integers. (This fact was previously undocumented.) + </p> + </section> +</chapter> + |