|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
/** |
|
4
|
|
|
* This file is part of the sweetrdf/InMemoryStoreSqlite package and licensed under |
|
5
|
|
|
* the terms of the GPL-2 license. |
|
6
|
|
|
* |
|
7
|
|
|
* (c) Konrad Abicht <[email protected]> |
|
8
|
|
|
* (c) Benjamin Nowack |
|
9
|
|
|
* |
|
10
|
|
|
* For the full copyright and license information, please view the LICENSE |
|
11
|
|
|
* file that was distributed with this source code. |
|
12
|
|
|
*/ |
|
13
|
|
|
|
|
14
|
|
|
namespace sweetrdf\InMemoryStoreSqlite; |
|
15
|
|
|
|
|
16
|
|
|
function calcURI(string $path, ?string $base = null): string |
|
17
|
|
|
{ |
|
18
|
|
|
/* quick check */ |
|
19
|
98 |
|
if (preg_match("/^[a-z0-9\_]+\:/i", $path)) {/* abs path or bnode */ |
|
20
|
98 |
|
return $path; |
|
21
|
|
|
} |
|
22
|
3 |
|
if (preg_match('/^\$\{.*\}/', $path)) {/* placeholder, assume abs URI */ |
|
23
|
|
|
return $path; |
|
24
|
|
|
} |
|
25
|
3 |
|
if (preg_match("/^\/\//", $path)) {/* net path, assume http */ |
|
26
|
|
|
return 'http:'.$path; |
|
27
|
|
|
} |
|
28
|
|
|
/* other URIs */ |
|
29
|
3 |
|
$base = $base ?: NamespaceHelper::BASE_NAMESPACE; |
|
30
|
3 |
|
$base = preg_replace('/\#.*$/', '', $base); |
|
31
|
3 |
|
if (true === $path) {/* empty (but valid) URIref via turtle parser: <> */ |
|
|
|
|
|
|
32
|
|
|
return $base; |
|
33
|
|
|
} |
|
34
|
3 |
|
$path = preg_replace("/^\.\//", '', $path); |
|
35
|
3 |
|
$root = preg_match('/(^[a-z0-9]+\:[\/]{1,3}[^\/]+)[\/|$]/i', $base, $m) ? $m[1] : $base; /* w/o trailing slash */ |
|
36
|
3 |
|
$base .= ($base == $root) ? '/' : ''; |
|
37
|
3 |
|
if (preg_match('/^\//', $path)) {/* leading slash */ |
|
38
|
|
|
return $root.$path; |
|
39
|
|
|
} |
|
40
|
3 |
|
if (!$path) { |
|
41
|
|
|
return $base; |
|
42
|
|
|
} |
|
43
|
3 |
|
if (preg_match('/^([\#\?])/', $path, $m)) { |
|
44
|
1 |
|
return preg_replace('/\\'.$m[1].'.*$/', '', $base).$path; |
|
45
|
|
|
} |
|
46
|
2 |
|
if (preg_match('/^(\&)(.*)$/', $path, $m)) {/* not perfect yet */ |
|
47
|
|
|
return preg_match('/\?/', $base) ? $base.$m[1].$m[2] : $base.'?'.$m[2]; |
|
48
|
|
|
} |
|
49
|
2 |
|
if (preg_match("/^[a-z0-9]+\:/i", $path)) {/* abs path */ |
|
50
|
|
|
return $path; |
|
51
|
|
|
} |
|
52
|
|
|
/* rel path: remove stuff after last slash */ |
|
53
|
2 |
|
$base = substr($base, 0, strrpos($base, '/') + 1); |
|
54
|
|
|
|
|
55
|
|
|
/* resolve ../ */ |
|
56
|
2 |
|
while (preg_match('/^(\.\.\/)(.*)$/', $path, $m)) { |
|
57
|
|
|
$path = $m[2]; |
|
58
|
|
|
$base = ($base == $root.'/') ? $base : preg_replace('/^(.*\/)[^\/]+\/$/', '\\1', $base); |
|
59
|
|
|
} |
|
60
|
|
|
|
|
61
|
2 |
|
return $base.$path; |
|
62
|
|
|
} |
|
63
|
|
|
|
|
64
|
|
|
function calcBase(string $path): string |
|
65
|
|
|
{ |
|
66
|
2 |
|
$r = $path; |
|
67
|
2 |
|
$r = preg_replace('/\#.*$/', '', $r); /* remove hash */ |
|
68
|
2 |
|
$r = preg_replace('/^\/\//', 'http://', $r); /* net path (//), assume http */ |
|
69
|
2 |
|
if (preg_match('/^[a-z0-9]+\:/', $r)) {/* scheme, abs path */ |
|
70
|
2 |
|
while (preg_match('/^(.+\/)(\.\.\/.*)$/U', $r, $m)) { |
|
71
|
|
|
$r = calcURI($m[1], $m[2]); |
|
72
|
|
|
} |
|
73
|
|
|
|
|
74
|
2 |
|
return $r; |
|
75
|
|
|
} |
|
76
|
|
|
|
|
77
|
|
|
return 'file://'.realpath($r); /* real path */ |
|
78
|
|
|
} |
|
79
|
|
|
|
|
80
|
|
|
/** |
|
81
|
|
|
* Normalize value for ORDER BY operations. |
|
82
|
|
|
*/ |
|
83
|
|
|
function getNormalizedValue(string $val): string |
|
84
|
|
|
{ |
|
85
|
|
|
/* try date (e.g. 21 August 2007) */ |
|
86
|
|
|
if ( |
|
87
|
95 |
|
preg_match('/^[0-9]{1,2}\s+[a-z]+\s+[0-9]{4}/i', $val) |
|
88
|
95 |
|
&& ($uts = strtotime($val)) |
|
89
|
95 |
|
&& (-1 !== $uts) |
|
90
|
|
|
) { |
|
91
|
1 |
|
return (string) date("Y-m-d\TH:i:s", $uts); |
|
92
|
|
|
} |
|
93
|
|
|
|
|
94
|
|
|
/* xsd date (e.g. 2009-05-28T18:03:38+09:00 2009-05-28T18:03:38GMT) */ |
|
95
|
95 |
|
if (true === (bool) strtotime($val)) { |
|
96
|
10 |
|
return (string) date('Y-m-d\TH:i:s\Z', strtotime($val)); |
|
97
|
|
|
} |
|
98
|
|
|
|
|
99
|
93 |
|
if (is_numeric($val)) { |
|
100
|
29 |
|
$val = sprintf('%f', $val); |
|
101
|
29 |
|
if (preg_match("/([\-\+])([0-9]*)\.([0-9]*)/", $val, $m)) { |
|
102
|
|
|
return $m[1].sprintf('%018s', $m[2]).'.'.sprintf('%-015s', $m[3]); |
|
103
|
|
|
} |
|
104
|
29 |
|
if (preg_match("/([0-9]*)\.([0-9]*)/", $val, $m)) { |
|
105
|
29 |
|
return '+'.sprintf('%018s', $m[1]).'.'.sprintf('%-015s', $m[2]); |
|
106
|
|
|
} |
|
107
|
|
|
|
|
108
|
|
|
return $val; |
|
109
|
|
|
} |
|
110
|
|
|
|
|
111
|
|
|
/* any other string: remove tags, linebreaks etc., but keep MB-chars */ |
|
112
|
|
|
// [\PL\s]+ ( = non-Letters) kills digits |
|
113
|
69 |
|
$re = '/[\PL\s]+/isu'; |
|
|
|
|
|
|
114
|
69 |
|
$re = '/[\s\'\"\´\`]+/is'; |
|
115
|
69 |
|
$val = trim(preg_replace($re, '-', strip_tags($val))); |
|
116
|
69 |
|
if (\strlen($val) > 35) { |
|
117
|
5 |
|
$fnc = \function_exists('mb_substr') ? 'mb_substr' : 'substr'; |
|
118
|
5 |
|
$val = $fnc($val, 0, 17).'-'.$fnc($val, -17); |
|
119
|
|
|
} |
|
120
|
|
|
|
|
121
|
69 |
|
return $val; |
|
122
|
|
|
} |
|
123
|
|
|
|
|
124
|
|
|
/** |
|
125
|
|
|
* @return array<string,string> |
|
126
|
|
|
*/ |
|
127
|
|
|
function splitURI($v): array |
|
128
|
|
|
{ |
|
129
|
|
|
/* |
|
130
|
|
|
* the following namespaces may lead to conflated URIs, |
|
131
|
|
|
* we have to set the split position manually |
|
132
|
|
|
*/ |
|
133
|
|
|
if (strpos($v, 'www.w3.org')) { |
|
134
|
|
|
/* |
|
135
|
|
|
* @todo port to NamespaceHelper |
|
136
|
|
|
*/ |
|
137
|
|
|
$specials = [ |
|
138
|
|
|
'http://www.w3.org/XML/1998/namespace', |
|
139
|
|
|
'http://www.w3.org/2005/Atom', |
|
140
|
|
|
'http://www.w3.org/1999/xhtml', |
|
141
|
|
|
]; |
|
142
|
|
|
foreach ($specials as $ns) { |
|
143
|
|
|
if (str_contains($v, $ns)) { |
|
144
|
|
|
$local_part = substr($v, \strlen($ns)); |
|
145
|
|
|
if (!preg_match('/^[\/\#]/', $local_part)) { |
|
146
|
|
|
return [$ns, $local_part]; |
|
147
|
|
|
} |
|
148
|
|
|
} |
|
149
|
|
|
} |
|
150
|
|
|
} |
|
151
|
|
|
/* auto-splitting on / or # */ |
|
152
|
|
|
//$re = '^(.*?)([A-Z_a-z][-A-Z_a-z0-9.]*)$'; |
|
153
|
|
|
if (preg_match('/^(.*[\/\#])([^\/\#]+)$/', $v, $m)) { |
|
154
|
|
|
return [$m[1], $m[2]]; |
|
155
|
|
|
} |
|
156
|
|
|
/* auto-splitting on last special char, e.g. urn:foo:bar */ |
|
157
|
|
|
if (preg_match('/^(.*[\:\/])([^\:\/]+)$/', $v, $m)) { |
|
158
|
|
|
return [$m[1], $m[2]]; |
|
159
|
|
|
} |
|
160
|
|
|
|
|
161
|
|
|
return [$v, '']; |
|
162
|
|
|
} |
|
163
|
|
|
|