<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head><meta http-equiv="content-type" content="text/html; charset=utf-8" /><style type="text/css"><!--
#msg dl { border: 1px #006 solid; background: #369; padding: 6px; color: #fff; }
#msg dt { float: left; width: 6em; font-weight: bold; }
#msg dt:after { content:':';}
#msg dl, #msg dt, #msg ul, #msg li, #header, #footer { font-family: verdana,arial,helvetica,sans-serif; font-size: 10pt; }
#msg dl a { font-weight: bold}
#msg dl a:link { color:#fc3; }
#msg dl a:active { color:#ff0; }
#msg dl a:visited { color:#cc6; }
h3 { font-family: verdana,arial,helvetica,sans-serif; font-size: 10pt; font-weight: bold; }
#msg pre { overflow: auto; background: #ffc; border: 1px #fc0 solid; padding: 6px; }
#msg ul, pre { overflow: auto; }
#header, #footer { color: #fff; background: #636; border: 1px #300 solid; padding: 6px; }
#patch { width: 100%; }
#patch h4 {font-family: verdana,arial,helvetica,sans-serif;font-size:10pt;padding:8px;background:#369;color:#fff;margin:0;}
#patch .propset h4, #patch .binary h4 {margin:0;}
#patch pre {padding:0;line-height:1.2em;margin:0;}
#patch .diff {width:100%;background:#eee;padding: 0 0 10px 0;overflow:auto;}
#patch .propset .diff, #patch .binary .diff {padding:10px 0;}
#patch span {display:block;padding:0 10px;}
#patch .modfile, #patch .addfile, #patch .delfile, #patch .propset, #patch .binary, #patch .copfile {border:1px solid #ccc;margin:10px 0;}
#patch ins {background:#dfd;text-decoration:none;display:block;padding:0 10px;}
#patch del {background:#fdd;text-decoration:none;display:block;padding:0 10px;}
#patch .lines, .info {color:#888;background:#fff;}
--></style>
<title>[12174] trunk/wp-includes/pomo: pomo performance improvements.</title>
</head>
<body>
<div id="msg">
<dl>
<dt>Revision</dt> <dd><a href="http://trac.wordpress.org/changeset/12174">12174</a></dd>
<dt>Author</dt> <dd>ryan</dd>
<dt>Date</dt> <dd>2009-11-12 16:05:43 +0000 (Thu, 12 Nov 2009)</dd>
</dl>
<h3>Log Message</h3>
<pre>pomo performance improvements. Props nbachiyski. fixes <a href="http://trac.wordpress.org/ticket/10165">#10165</a></pre>
<h3>Modified Paths</h3>
<ul>
<li><a href="#trunkwpincludespomomophp">trunk/wp-includes/pomo/mo.php</a></li>
<li><a href="#trunkwpincludespomostreamsphp">trunk/wp-includes/pomo/streams.php</a></li>
</ul>
</div>
<div id="patch">
<h3>Diff</h3>
<a id="trunkwpincludespomomophp"></a>
<div class="modfile"><h4>Modified: trunk/wp-includes/pomo/mo.php (12173 => 12174)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/wp-includes/pomo/mo.php 2009-11-12 14:43:15 UTC (rev 12173)
+++ trunk/wp-includes/pomo/mo.php 2009-11-12 16:05:43 UTC (rev 12174)
</span><span class="lines">@@ -2,7 +2,7 @@
</span><span class="cx"> /**
</span><span class="cx"> * Class for working with MO files
</span><span class="cx"> *
</span><del>- * @version $Id: mo.php 221 2009-09-07 21:08:21Z nbachiyski $
</del><ins>+ * @version $Id: mo.php 293 2009-11-12 15:43:50Z nbachiyski $
</ins><span class="cx"> * @package pomo
</span><span class="cx"> * @subpackage mo
</span><span class="cx"> */
</span><span class="lines">@@ -21,10 +21,9 @@
</span><span class="cx"> * @param string $filename MO file to load
</span><span class="cx"> */
</span><span class="cx"> function import_from_file($filename) {
</span><del>- $reader = new POMO_CachedIntFileReader($filename);
- if (isset($reader->error)) {
</del><ins>+ $reader = new POMO_FileReader($filename);
+ if (!$reader->is_resource())
</ins><span class="cx"> return false;
</span><del>- }
</del><span class="cx"> return $this->import_from_reader($reader);
</span><span class="cx"> }
</span><span class="cx">
</span><span class="lines">@@ -113,61 +112,111 @@
</span><span class="cx"> }
</span><span class="cx">
</span><span class="cx"> function import_from_reader($reader) {
</span><del>- $reader->setEndian('little');
- $endian = MO::get_byteorder($reader->readint32());
- if (false === $endian) {
</del><ins>+ $endian_string = MO::get_byteorder($reader->readint32());
+ if (false === $endian_string) {
</ins><span class="cx"> return false;
</span><span class="cx"> }
</span><del>- $reader->setEndian($endian);
</del><ins>+ $reader->setEndian($endian_string);
</ins><span class="cx">
</span><del>- $revision = $reader->readint32();
- $total = $reader->readint32();
- // get addresses of array of lenghts and offsets for original string and translations
- $originals_lenghts_addr = $reader->readint32();
- $translations_lenghts_addr = $reader->readint32();
</del><ins>+ $endian = ('big' == $endian_string)? 'N' : 'V';
</ins><span class="cx">
</span><ins>+ $header = $reader->read(24);
+ if ($reader->strlen($header) != 24)
+ return false;
+
+ // parse header
+ $header = unpack("{$endian}revision/{$endian}total/{$endian}originals_lenghts_addr/{$endian}translations_lenghts_addr/{$endian}hash_length/{$endian}hash_addr", $header);
+ if (!is_array($header))
+ return false;
+
+ extract( $header );
+
+ // support revision 0 of MO format specs, only
+ if ($revision != 0)
+ return false;
+
+ // seek to data blocks
</ins><span class="cx"> $reader->seekto($originals_lenghts_addr);
</span><del>- $originals_lenghts = $reader->readint32array($total * 2); // each of
- $reader->seekto($translations_lenghts_addr);
- $translations_lenghts = $reader->readint32array($total * 2);
</del><span class="cx">
</span><del>- $length = create_function('$i', 'return $i * 2 + 1;');
- $offset = create_function('$i', 'return $i * 2 + 2;');
</del><ins>+ // read originals' indices
+ $originals_lengths_length = $translations_lenghts_addr - $originals_lenghts_addr;
+ if ( $originals_lengths_length != $total * 8 )
+ return false;
</ins><span class="cx">
</span><del>- for ($i = 0; $i < $total; ++$i) {
- $reader->seekto($originals_lenghts[$offset($i)]);
- $original = $reader->read($originals_lenghts[$length($i)]);
- $reader->seekto($translations_lenghts[$offset($i)]);
- $translation = $reader->read($translations_lenghts[$length($i)]);
- if ('' == $original) {
</del><ins>+ $originals = $reader->read($originals_lengths_length);
+ if ( $reader->strlen( $originals ) != $originals_lengths_length )
+ return false;
+
+ // read translations' indices
+ $translations_lenghts_length = $hash_addr - $translations_lenghts_addr;
+ if ( $translations_lenghts_length != $total * 8 )
+ return false;
+
+ $translations = $reader->read($translations_lenghts_length);
+ if ( $reader->strlen( $translations ) != $translations_lenghts_length )
+ return false;
+
+ // transform raw data into set of indices
+ $originals = $reader->str_split( $originals, 8 );
+ $translations = $reader->str_split( $translations, 8 );
+
+ // skip hash table
+ $strings_addr = $hash_addr + $hash_length * 4;
+
+ $reader->seekto($strings_addr);
+
+ $strings = $reader->read_all();
+ $reader->close();
+
+ for ( $i = 0; $i < $total; $i++ ) {
+ $o = unpack( "{$endian}length/{$endian}pos", $originals[$i] );
+ $t = unpack( "{$endian}length/{$endian}pos", $translations[$i] );
+ if ( !$o || !$t ) return false;
+
+ // adjust offset due to reading strings to separate space before
+ $o['pos'] -= $strings_addr;
+ $t['pos'] -= $strings_addr;
+
+ $original = $reader->substr( $strings, $o['pos'], $o['length'] );
+ $translation = $reader->substr( $strings, $t['pos'], $t['length'] );
+
+ if ('' === $original) {
</ins><span class="cx"> $this->set_headers($this->make_headers($translation));
</span><span class="cx"> } else {
</span><del>- $this->add_entry($this->make_entry($original, $translation));
</del><ins>+ $entry = &$this->make_entry($original, $translation);
+ $this->entries[$entry->key()] = &$entry;
</ins><span class="cx"> }
</span><span class="cx"> }
</span><span class="cx"> return true;
</span><span class="cx"> }
</span><span class="cx">
</span><span class="cx"> /**
</span><ins>+ * Build a Translation_Entry from original string and translation strings,
+ * found in a MO file
+ *
</ins><span class="cx"> * @static
</span><ins>+ * @param string $original original string to translate from MO file. Might contain
+ * 0x04 as context separator or 0x00 as singular/plural separator
+ * @param string $translation translation string from MO file. Might contain
+ * 0x00 as a plural translations separator
</ins><span class="cx"> */
</span><span class="cx"> function &make_entry($original, $translation) {
</span><del>- $args = array();
</del><ins>+ $entry = & new Translation_Entry();
</ins><span class="cx"> // look for context
</span><span class="cx"> $parts = explode(chr(4), $original);
</span><span class="cx"> if (isset($parts[1])) {
</span><span class="cx"> $original = $parts[1];
</span><del>- $args['context'] = $parts[0];
</del><ins>+ $entry->context = $parts[0];
</ins><span class="cx"> }
</span><span class="cx"> // look for plural original
</span><span class="cx"> $parts = explode(chr(0), $original);
</span><del>- $args['singular'] = $parts[0];
</del><ins>+ $entry->singular = $parts[0];
</ins><span class="cx"> if (isset($parts[1])) {
</span><del>- $args['plural'] = $parts[1];
</del><ins>+ $entry->is_plural = true;
+ $entry->plural = $parts[1];
</ins><span class="cx"> }
</span><span class="cx"> // plural translations are also separated by \0
</span><del>- $args['translations'] = explode(chr(0), $translation);
- $entry = & new Translation_Entry($args);
</del><ins>+ $entry->translations = explode(chr(0), $translation);
</ins><span class="cx"> return $entry;
</span><span class="cx"> }
</span><span class="cx">
</span><span class="lines">@@ -178,7 +227,5 @@
</span><span class="cx"> function get_plural_forms_count() {
</span><span class="cx"> return $this->_nplurals;
</span><span class="cx"> }
</span><del>-
-
</del><span class="cx"> }
</span><span class="cx"> endif;
</span><span class="cx">\ No newline at end of file
</span></span></pre></div>
<a id="trunkwpincludespomostreamsphp"></a>
<div class="modfile"><h4>Modified: trunk/wp-includes/pomo/streams.php (12173 => 12174)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/wp-includes/pomo/streams.php 2009-11-12 14:43:15 UTC (rev 12173)
+++ trunk/wp-includes/pomo/streams.php 2009-11-12 16:05:43 UTC (rev 12174)
</span><span class="lines">@@ -3,64 +3,182 @@
</span><span class="cx"> * Classes, which help reading streams of data from files.
</span><span class="cx"> * Based on the classes from Danilo Segan <danilo@kvota.net>
</span><span class="cx"> *
</span><del>- * @version $Id: streams.php 223 2009-09-07 21:20:13Z nbachiyski $
</del><ins>+ * @version $Id: streams.php 293 2009-11-12 15:43:50Z nbachiyski $
</ins><span class="cx"> * @package pomo
</span><span class="cx"> * @subpackage streams
</span><span class="cx"> */
</span><span class="cx">
</span><ins>+if ( !class_exists( 'POMO_Reader' ) ):
+class POMO_Reader {
+
+ var $endian = 'little';
+ var $_post = '';
+
+ function POMO_Reader() {
+ $this->is_overloaded = ((ini_get("mbstring.func_overload") & 2) != 0) && function_exists('mb_substr');
+ $this->_pos = 0;
+ }
+
+ /**
+ * Sets the endianness of the file.
+ *
+ * @param $endian string 'big' or 'little'
+ */
+ function setEndian($endian) {
+ $this->endian = $endian;
+ }
</ins><span class="cx">
</span><del>-if ( !class_exists( 'POMO_StringReader' ) ):
-/**
- * Provides file-like methods for manipulating a string instead
- * of a physical file.
- */
-class POMO_StringReader {
- var $_pos;
- var $_str;
</del><ins>+ /**
+ * Reads a 32bit Integer from the Stream
+ *
+ * @return mixed The integer, corresponding to the next 32 bits from
+ * the stream of false if there are not enough bytes or on error
+ */
+ function readint32() {
+ $bytes = $this->read(4);
+ if (4 != $this->strlen($bytes))
+ return false;
+ $endian_letter = ('big' == $this->endian)? 'N' : 'V';
+ $int = unpack($endian_letter, $bytes);
+ return array_shift($int);
+ }
</ins><span class="cx">
</span><del>- function POMO_StringReader($str = '') {
- $this->_str = $str;
- $this->_pos = 0;
- $this->is_overloaded = ((ini_get("mbstring.func_overload") & 2) != 0) && function_exists('mb_substr');
</del><ins>+ /**
+ * Reads an array of 32-bit Integers from the Stream
+ *
+ * @param integer count How many elements should be read
+ * @return mixed Array of integers or false if there isn't
+ * enough data or on error
+ */
+ function readint32array($count) {
+ $bytes = $this->read(4 * $count);
+ if (4*$count != $this->strlen($bytes))
+ return false;
+ $endian_letter = ('big' == $this->endian)? 'N' : 'V';
+ return unpack($endian_letter.$count, $bytes);
</ins><span class="cx"> }
</span><del>-
- function _substr($string, $start, $length) {
</del><ins>+
+
+ function substr($string, $start, $length) {
</ins><span class="cx"> if ($this->is_overloaded) {
</span><del>- return mb_substr($string,$start,$length,'ascii');
</del><ins>+ return mb_substr($string, $start, $length, 'ascii');
</ins><span class="cx"> } else {
</span><del>- return substr($string,$start,$length);
</del><ins>+ return substr($string, $start, $length);
</ins><span class="cx"> }
</span><span class="cx"> }
</span><span class="cx">
</span><del>- function _strlen($string) {
</del><ins>+ function strlen($string) {
</ins><span class="cx"> if ($this->is_overloaded) {
</span><del>- return mb_strlen($string,'ascii');
</del><ins>+ return mb_strlen($string, 'ascii');
</ins><span class="cx"> } else {
</span><span class="cx"> return strlen($string);
</span><span class="cx"> }
</span><span class="cx"> }
</span><ins>+
+ function str_split($string, $chunk_size) {
+ if (!function_exists('str_split')) {
+ $length = $this->strlen($string);
+ $out = array();
+ for ($i = 0; $i < $length; $i += $chunk_size)
+ $out[] = $this->substr($string, $i, $chunk_size);
+ return $out;
+ } else {
+ return str_split( $string, $chunk_size );
+ }
+ }
+
+
+ function pos() {
+ return $this->_pos;
+ }
</ins><span class="cx">
</span><ins>+ function is_resource() {
+ return true;
+ }
+
+ function close() {
+ return true;
+ }
+}
+endif;
+
+if ( !class_exists( 'POMO_FileReader' ) ):
+class POMO_FileReader extends POMO_Reader {
+ function POMO_FileReader($filename) {
+ parent::POMO_Reader();
+ $this->_f = fopen($filename, 'r');
+ }
+
</ins><span class="cx"> function read($bytes) {
</span><del>- $data = $this->_substr($this->_str, $this->_pos, $bytes);
</del><ins>+ return fread($this->_f, $bytes);
+ }
+
+ function seekto($pos) {
+ if ( -1 == fseek($this->_f, $pos, SEEK_SET)) {
+ return false;
+ }
+ $this->_pos = $pos;
+ return true;
+ }
+
+ function is_resource() {
+ return is_resource($this->_f);
+ }
+
+ function feof() {
+ return feof($this->_f);
+ }
+
+ function close() {
+ return fclose($this->_f);
+ }
+
+ function read_all() {
+ $all = '';
+ while ( !$this->feof() )
+ $all .= $this->read(4096);
+ return $all;
+ }
+}
+endif;
+
+if ( !class_exists( 'POMO_StringReader' ) ):
+/**
+ * Provides file-like methods for manipulating a string instead
+ * of a physical file.
+ */
+class POMO_StringReader extends POMO_Reader {
+
+ var $_str = '';
+
+ function POMO_StringReader($str = '') {
+ parent::POMO_Reader();
+ $this->_str = $str;
+ $this->_pos = 0;
+ }
+
+
+ function read($bytes) {
+ $data = $this->substr($this->_str, $this->_pos, $bytes);
</ins><span class="cx"> $this->_pos += $bytes;
</span><del>- if ($this->_strlen($this->_str) < $this->_pos) $this->_pos = $this->_strlen($this->_str);
</del><ins>+ if ($this->strlen($this->_str) < $this->_pos) $this->_pos = $this->strlen($this->_str);
</ins><span class="cx"> return $data;
</span><span class="cx"> }
</span><span class="cx">
</span><span class="cx"> function seekto($pos) {
</span><span class="cx"> $this->_pos = $pos;
</span><del>- if ($this->_strlen($this->_str) < $this->_pos) $this->_pos = $this->_strlen($this->_str);
</del><ins>+ if ($this->strlen($this->_str) < $this->_pos) $this->_pos = $this->strlen($this->_str);
</ins><span class="cx"> return $this->_pos;
</span><span class="cx"> }
</span><span class="cx">
</span><del>- function pos() {
- return $this->_pos;
</del><ins>+ function length() {
+ return $this->strlen($this->_str);
</ins><span class="cx"> }
</span><span class="cx">
</span><del>- function length() {
- return $this->_strlen($this->_str);
</del><ins>+ function read_all() {
+ return $this->substr($this->_str, $this->_pos, $this->strlen($this->_str));
</ins><span class="cx"> }
</span><del>-
</del><ins>+
</ins><span class="cx"> }
</span><span class="cx"> endif;
</span><span class="cx">
</span><span class="lines">@@ -81,61 +199,11 @@
</span><span class="cx">
</span><span class="cx"> if ( !class_exists( 'POMO_CachedIntFileReader' ) ):
</span><span class="cx"> /**
</span><del>- * Allows reading integers from a file.
</del><ins>+ * Reads the contents of the file in the beginning.
</ins><span class="cx"> */
</span><span class="cx"> class POMO_CachedIntFileReader extends POMO_CachedFileReader {
</span><del>-
- var $endian = 'little';
-
- /**
- * Opens a file and caches it.
- *
- * @param $filename string name of the file to be opened
- * @param $endian string endianness of the words in the file, allowed
- * values are 'little' or 'big'. Default value is 'little'
- */
- function POMO_CachedIntFileReader($filename, $endian = 'little') {
- $this->endian = $endian;
</del><ins>+ function POMO_CachedIntFileReader($filename) {
</ins><span class="cx"> parent::POMO_CachedFileReader($filename);
</span><span class="cx"> }
</span><del>-
- /**
- * Sets the endianness of the file.
- *
- * @param $endian string 'big' or 'little'
- */
- function setEndian($endian) {
- $this->endian = $endian;
- }
-
- /**
- * Reads a 32bit Integer from the Stream
- *
- * @return mixed The integer, corresponding to the next 32 bits from
- * the stream of false if there are not enough bytes or on error
- */
- function readint32() {
- $bytes = $this->read(4);
- if (4 != $this->_strlen($bytes))
- return false;
- $endian_letter = ('big' == $this->endian)? 'N' : 'V';
- $int = unpack($endian_letter, $bytes);
- return array_shift($int);
- }
-
- /**
- * Reads an array of 32-bit Integers from the Stream
- *
- * @param integer count How many elements should be read
- * @return mixed Array of integers or false if there isn't
- * enough data or on error
- */
- function readint32array($count) {
- $bytes = $this->read(4 * $count);
- if (4*$count != $this->_strlen($bytes))
- return false;
- $endian_letter = ('big' == $this->endian)? 'N' : 'V';
- return unpack($endian_letter.$count, $bytes);
- }
</del><span class="cx"> }
</span><span class="cx"> endif;
</span><span class="cx">\ No newline at end of file
</span></span></pre>
</div>
</div>
</body>
</html>