/* lightbenc.php
Dear Bram Cohen,
You are an arse
WHAT were you smoking ?
This implementation should use one order of magnitude less memory then the TBdev version.
The bdecoding speed is similar to TBdev, bencoding is faster, and much faster then bdecoding.
Call the bdecode() function with the bencoded string:
$str="d7:oneListl8:a stringe10:oneIntegeri34ee";
var_dump(bdecode($str));
array(3) {
["oneList"]=>
array(1) {
[0]=>
string(8) "a string"
}
["oneInteger"]=>
int(34)
["isDct"]=>
bool(true)
}
The returned value is a nested data type with the following type of elements:
- ints (test type with is_integer($x))
- strings (test type with is_string($x))
- lists (test type with is_array($x) && !isset($x[isDct])
- dicts (test type with is_array($x) && isset($x[isDct])
All elements have the native PHP type, except for the dictionary which is an array with an "isDct" key.
This is necessary since PHP makes no distinction between flat and associative arrays. Note that the isDct
key is allways set as a bool, so that even if the dictionary contains an actual "isDct" value, the
functions behave transparently, i.e. they don't strip out or overwrite actual "isDct" keys.
As such, this implementation is not a drop-in replacement of the TBDev code, hence the new function names
For all practical purposes, it's just as flexible, and very easy to use. For example:
// decode the torrent file
$dict= bdecode_file($torrentfilename);
// change announce url
$dict['announce']='http://inferno.demonoid.com';
// add private tracker flag
$dict['info']['private']=1;
// compute infohash
$infohash = pack("H*", sha1(bencode($dict["info"])));
// recreate the torrent file
$torrentfile=bencode($dict);
After calling bencode(), the passed nested array will have all it's dictionaries sorted by key.
The bencoded data generated by bencode() will have sorted dictionaries, but bdecode() does not require
this in the input stream, and will keep the order unchanged.
This implementation is hereby released under the GFYPL, version 1.00.
The Go Fuck Yourself Public License, version 1.00
Article 1
You can go fuck yourself.
END OF ALL TERMS AND CONDITIONS
*/
class Lightbenc{
public static function bdecode($s, &$pos = 0) {
if ($pos >= strlen($s)) {
return null;
}
switch ($s[$pos]) {
case 'd' :
$pos++;
$retval = array();
while ($s[$pos] != 'e') {
$key = Lightbenc::bdecode($s, $pos);
$val = Lightbenc::bdecode($s, $pos);
if ($key === null || $val === null)
break;
$retval[$key] = $val;
}
$retval["isDct"] = true;
$pos++;
return $retval;
case 'l' :
$pos++;
$retval = array();
while ($s[$pos] != 'e'){
$val = Lightbenc::bdecode($s, $pos);
if ($val === null)
break;
$retval[] = $val;
}
$pos++;
return $retval;
case 'i' :
$pos++;
$digits = strpos($s, 'e', $pos) - $pos;
$val = (int)substr($s, $pos, $digits);
$pos += $digits + 1;
return $val;
//case "0": case "1": case "2": case "3": case "4":
//case "5": case "6": case "7": case "8": case "9":
default :
$digits = strpos($s, ':', $pos) - $pos;
if ($digits < 0 || $digits > 20)
return null;
$len = (int)substr($s, $pos, $digits);
$pos += $digits + 1;
$str = substr($s, $pos, $len);
$pos += $len;
//echo "pos: $pos str: [$str] len: $len digits: $digits\n
";
return (string)$str;
}
return null;
}
public static function bencode(&$d) {
if (is_array($d)) {
$ret = "l";
if (isset($d["isDct"])) { // todo:$d["isDct"]
$isDict = 1;
$ret = "d";
// this is required by the specs, and BitTornado actualy chokes on unsorted dictionaries
ksort($d, SORT_STRING);
}
foreach ($d as $key=>$value) {
if (isset($isDict)) { // todo:$isDict
// skip the isDct element, only if it's set by us
if ($key == "isDct" && is_bool($value))
continue;
$ret .= strlen($key). ":" .$key;
}
if (is_string($value)) {
$ret .= strlen($value). ":" .$value;
} else if (is_int($value)) {
$ret .= "i${value}e";
} else {
$ret .= Lightbenc::bencode ($value);
}
}
return $ret."e";
} else if (is_string($d)) {// fallback if we're given a single bencoded string or int
return strlen($d). ":" .$d;
} else if (is_int($d))
return "i${d}e";
else
return null;
}
public static function bdecode_file($filename) {
$f = file_get_contents($filename, FILE_BINARY);
return Lightbenc::bdecode($f);
}
public static function bdecode_getinfo($filename) {
$t = Lightbenc::bdecode(file_get_contents($filename, FILE_BINARY));
$t['info_hash'] = sha1(Lightbenc::bencode($t['info']));
if (is_array($t['info']['files'])) { //multifile
$t['info']['size'] = 0;
$t['info']['filecount'] = 0;
foreach ($t['info']['files'] as $file) {
$t['info']['filecount']++;
$t['info']['size'] += $file['length'];
}
} else {
$t['info']['size'] = $t['info']['length'];
$t['info']["filecount"] = 1;
$t['info']['files'][0]['path'] = $t['info']['name'];
$t['info']['files'][0]['length'] = $t['info']['length'];
}
return $t;
}
}
?>
一键复制
编辑
Web IDE
原始数据
按行查看
历史