mingw/html/lib/Encode/PerlIO.html

   1 <?xml version="1.0" ?>
   2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
   3 <html xmlns="http://www.w3.org/1999/xhtml">
   4 <head>
   5 <title>Encode::PerlIO -- a detailed document on Encode and PerlIO</title>
   6 <meta http-equiv="content-type" content="text/html; charset=utf-8" />
   7 <link rev="made" href="mailto:" />
   8 </head>
   9
  10 <body style="background-color: white">
  11 <table border="0" width="100%" cellspacing="0" cellpadding="3">
  12 <tr><td class="block" style="background-color: #cccccc" valign="middle">
  13 <big><strong><span class="block">&nbsp;Encode::PerlIO -- a detailed document on Encode and PerlIO</span></strong></big>
  14 </td></tr>
  15 </table>
  16
  17 <p><a name="__index__"></a></p>
  18 <!-- INDEX BEGIN -->
  19
  20 <ul>
  21
  22         <li><a href="#name">NAME</a></li>
  23         <li><a href="#overview">Overview</a></li>
  24         <li><a href="#how_does_it_work">How does it work?</a></li>
  25         <li><a href="#line_buffering">Line Buffering</a></li>
  26         <ul>
  27
  28                 <li><a href="#how_can_i_tell_whether_my_encoding_fully_supports_perlio_">How can I tell whether my encoding fully supports PerlIO ?</a></li>
  29         </ul>
  30
  31         <li><a href="#see_also">SEE ALSO</a></li>
  32 </ul>
  33 <!-- INDEX END -->
  34
  35 <hr />
  36 <p>
  37 </p>
  38 <h1><a name="name">NAME</a></h1>
  39 <p>Encode::PerlIO -- a detailed document on Encode and PerlIO</p>
  40 <p>
  41 </p>
  42 <hr />
  43 <h1><a name="overview">Overview</a></h1>
  44 <p>It is very common to want to do encoding transformations when
  45 reading or writing files, network connections, pipes etc.
  46 If Perl is configured to use the new 'perlio' IO system then
  47 <code>Encode</code> provides a ``layer'' (see <a href="file://C|\msysgit\mingw\html/lib/Encode/PerlIO.html">the PerlIO manpage</a>) which can transform
  48 data as it is read or written.</p>
  49 <p>Here is how the blind poet would modernise the encoding:</p>
  50 <pre>
  51     use Encode;
  52     open(my $iliad,'&lt;:encoding(iso-8859-7)','iliad.greek');
  53     open(my $utf8,'&gt;:utf8','iliad.utf8');
  54     my @epic = &lt;$iliad&gt;;
  55     print $utf8 @epic;
  56     close($utf8);
  57     close($illiad);</pre>
  58 <p>In addition, the new IO system can also be configured to read/write
  59 UTF-8 encoded characters (as noted above, this is efficient):</p>
  60 <pre>
  61     open(my $fh,'&gt;:utf8','anything');
  62     print $fh &quot;Any \x{0021} string \N{SMILEY FACE}\n&quot;;</pre>
  63 <p>Either of the above forms of ``layer'' specifications can be made the default
  64 for a lexical scope with the <code>use open ...</code> pragma. See <a href="file://C|\msysgit\mingw\html/lib/open.html">the open manpage</a>.</p>
  65 <p>Once a handle is open, its layers can be altered using <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_binmode"><code>binmode</code></a>.</p>
  66 <p>Without any such configuration, or if Perl itself is built using the
  67 system's own IO, then write operations assume that the file handle
  68 accepts only <em>bytes</em> and will <code>die</code> if a character larger than 255 is
  69 written to the handle. When reading, each octet from the handle becomes
  70 a byte-in-a-character. Note that this default is the same behaviour
  71 as bytes-only languages (including Perl before v5.6) would have,
  72 and is sufficient to handle native 8-bit encodings e.g. iso-8859-1,
  73 EBCDIC etc. and any legacy mechanisms for handling other encodings
  74 and binary data.</p>
  75 <p>In other cases, it is the program's responsibility to transform
  76 characters into bytes using the API above before doing writes, and to
  77 transform the bytes read from a handle into characters before doing
  78 ``character operations'' (e.g. <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_lc"><code>lc</code></a>, <code>/\W+/</code>, ...).</p>
  79 <p>You can also use PerlIO to convert larger amounts of data you don't
  80 want to bring into memory.  For example, to convert between ISO-8859-1
  81 (Latin 1) and UTF-8 (or UTF-EBCDIC in EBCDIC machines):</p>
  82 <pre>
  83     open(F, &quot;&lt;:encoding(iso-8859-1)&quot;, &quot;data.txt&quot;) or die $!;
  84     open(G, &quot;&gt;:utf8&quot;,                 &quot;data.utf&quot;) or die $!;
  85     while (&lt;F&gt;) { print G }</pre>
  86 <pre>
  87     # Could also do &quot;print G &lt;F&gt;&quot; but that would pull
  88     # the whole file into memory just to write it out again.</pre>
  89 <p>More examples:</p>
  90 <pre>
  91     open(my $f, &quot;&lt;:encoding(cp1252)&quot;)
  92     open(my $g, &quot;&gt;:encoding(iso-8859-2)&quot;)
  93     open(my $h, &quot;&gt;:encoding(latin9)&quot;)       # iso-8859-15</pre>
  94 <p>See also <a href="file://C|\msysgit\mingw\html/lib/encoding.html">the encoding manpage</a> for how to change the default encoding of the
  95 data in your script.</p>
  96 <p>
  97 </p>
  98 <hr />
  99 <h1><a name="how_does_it_work">How does it work?</a></h1>
 100 <p>Here is a crude diagram of how filehandle, PerlIO, and Encode
 101 interact.</p>
 102 <pre>
 103   filehandle &lt;-&gt; PerlIO        PerlIO &lt;-&gt; scalar (read/printed)
 104                        \      /
 105                         Encode</pre>
 106 <p>When PerlIO receives data from either direction, it fills a buffer
 107 (currently with 1024 bytes) and passes the buffer to Encode.
 108 Encode tries to convert the valid part and passes it back to PerlIO,
 109 leaving invalid parts (usually a partial character) in the buffer.
 110 PerlIO then appends more data to the buffer, calls Encode again,
 111 and so on until the data stream ends.</p>
 112 <p>To do so, PerlIO always calls (de|en)code methods with CHECK set to 1.
 113 This ensures that the method stops at the right place when it
 114 encounters partial character.  The following is what happens when
 115 PerlIO and Encode tries to encode (from utf8) more than 1024 bytes
 116 and the buffer boundary happens to be in the middle of a character.</p>
 117 <pre>
 118    A   B   C   ....   ~     \x{3000}    ....
 119   41  42  43   ....  7E   e3   80   80  ....
 120   &lt;- buffer ---------------&gt;
 121   &lt;&lt; encoded &gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;
 122                        &lt;- next buffer ------</pre>
 123 <p>Encode converts from the beginning to \x7E, leaving \xe3 in the buffer
 124 because it is invalid (partial character).</p>
 125 <p>Unfortunately, this scheme does not work well with escape-based
 126 encodings such as ISO-2022-JP.</p>
 127 <p>
 128 </p>
 129 <hr />
 130 <h1><a name="line_buffering">Line Buffering</a></h1>
 131 <p>Now let's see what happens when you try to decode from ISO-2022-JP and
 132 the buffer ends in the middle of a character.</p>
 133 <pre>
 134                           JIS208-ESC   \x{5f3e}
 135    A   B   C   ....   ~   \e   $   B  |DAN | ....
 136   41  42  43   ....  7E   1b  24  41  43  46 ....
 137   &lt;- buffer ---------------------------&gt;
 138   &lt;&lt; encoded &gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;&gt;</pre>
 139 <p>As you see, the next buffer begins with \x43.  But \x43 is 'C' in
 140 ASCII, which is wrong in this case because we are now in JISX 0208
 141 area so it has to convert \x43\x46, not \x43.  Unlike utf8 and EUC,
 142 in escape-based encodings you can't tell if a given octet is a whole
 143 character or just part of it.</p>
 144 <p>Fortunately PerlIO also supports line buffer if you tell PerlIO to use
 145 one instead of fixed buffer.  Since ISO-2022-JP is guaranteed to revert to ASCII at the end of the line, partial
 146 character will never happen when line buffer is used.</p>
 147 <p>To tell PerlIO to use line buffer, implement -&gt;needs_lines method
 148 for your encoding object.  See  <a href="file://C|\msysgit\mingw\html/lib/Encode/Encoding.html">the Encode::Encoding manpage</a> for details.</p>
 149 <p>Thanks to these efforts most encodings that come with Encode support
 150 PerlIO but that still leaves following encodings.</p>
 151 <pre>
 152   iso-2022-kr
 153   MIME-B
 154   MIME-Header
 155   MIME-Q</pre>
 156 <p>Fortunately iso-2022-kr is hardly used (according to Jungshik) and
 157 MIME-* are very unlikely to be fed to PerlIO because they are for mail
 158 headers.  See <a href="file://C|\msysgit\mingw\html/lib/Encode/MIME/Header.html">the Encode::MIME::Header manpage</a> for details.</p>
 159 <p>
 160 </p>
 161 <h2><a name="how_can_i_tell_whether_my_encoding_fully_supports_perlio_">How can I tell whether my encoding fully supports PerlIO ?</a></h2>
 162 <p>As of this writing, any encoding whose class belongs to Encode::XS and
 163 Encode::Unicode works.  The Encode module has a <code>perlio_ok</code> method
 164 which you can use before applying PerlIO encoding to the filehandle.
 165 Here is an example:</p>
 166 <pre>
 167   my $use_perlio = perlio_ok($enc);
 168   my $layer = $use_perlio ? &quot;&lt;:raw&quot; : &quot;&lt;:encoding($enc)&quot;;
 169   open my $fh, $layer, $file or die &quot;$file : $!&quot;;
 170   while(&lt;$fh&gt;){
 171     $_ = decode($enc, $_) unless $use_perlio;
 172     # ....
 173   }</pre>
 174 <p>
 175 </p>
 176 <hr />
 177 <h1><a name="see_also">SEE ALSO</a></h1>
 178 <p><a href="file://C|\msysgit\mingw\html/lib/Encode/Encoding.html">the Encode::Encoding manpage</a>,
 179 <a href="file://C|\msysgit\mingw\html/lib/Encode/Supported.html">the Encode::Supported manpage</a>,
 180 <a href="file://C|\msysgit\mingw\html/lib/Encode/PerlIO.html">the Encode::PerlIO manpage</a>,
 181 <a href="file://C|\msysgit\mingw\html/lib/encoding.html">the encoding manpage</a>,
 182 <a href="file://C|\msysgit\mingw\html/pod/perlebcdic.html">the perlebcdic manpage</a>,
 183 <a href="file://C|\msysgit\mingw\html/pod/perlfunc.html#item_open">open in the perlfunc manpage</a>,
 184 <a href="file://C|\msysgit\mingw\html/pod/perlunicode.html">the perlunicode manpage</a>,
 185 <a href="file://C|\msysgit\mingw\html/lib/utf8.html">the utf8 manpage</a>,
 186 the Perl Unicode Mailing List &lt;<a href="mailto:perl-unicode@perl.org">perl-unicode@perl.org</a>&gt;</p>
 187 <table border="0" width="100%" cellspacing="0" cellpadding="3">
 188 <tr><td class="block" style="background-color: #cccccc" valign="middle">
 189 <big><strong><span class="block">&nbsp;Encode::PerlIO -- a detailed document on Encode and PerlIO</span></strong></big>
 190 </td></tr>
 191 </table>
 192
 193 </body>
 194
 195 </html>