1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
/** |
4
|
|
|
* League.Csv (https://csv.thephpleague.com). |
5
|
|
|
* |
6
|
|
|
* @author Ignace Nyamagana Butera <[email protected]> |
7
|
|
|
* @license https://github.com/thephpleague/csv/blob/master/LICENSE (MIT License) |
8
|
|
|
* @version 9.2.0 |
9
|
|
|
* @link https://github.com/thephpleague/csv |
10
|
|
|
* |
11
|
|
|
* For the full copyright and license information, please view the LICENSE |
12
|
|
|
* file that was distributed with this source code. |
13
|
|
|
*/ |
14
|
|
|
|
15
|
|
|
declare(strict_types=1); |
16
|
|
|
|
17
|
|
|
namespace League\Csv\Polyfill; |
18
|
|
|
|
19
|
|
|
use Generator; |
20
|
|
|
use League\Csv\Stream; |
21
|
|
|
use SplFileObject; |
22
|
|
|
use TypeError; |
23
|
|
|
use function explode; |
24
|
|
|
use function get_class; |
25
|
|
|
use function gettype; |
26
|
|
|
use function in_array; |
27
|
|
|
use function is_object; |
28
|
|
|
use function ltrim; |
29
|
|
|
use function rtrim; |
30
|
|
|
use function sprintf; |
31
|
|
|
use function str_replace; |
32
|
|
|
use function substr; |
33
|
|
|
|
34
|
|
|
/** |
35
|
|
|
* A Polyfill to PHP's SplFileObject behavior when reading a CSV document |
36
|
|
|
* with the SplFileObject::READ_CSV and SplFileObject::SKIP_EMPTY flags on |
37
|
|
|
* and the empty string as the escape parameter. |
38
|
|
|
* |
39
|
|
|
* <code> |
40
|
|
|
* $file = new SplFileObject('/path/to/file.csv', 'r'); |
41
|
|
|
* $file->setFlags(SplFileObject::READ_CSV | SplFileObject::READ_AHEAD | SplFileObject::SKIP_EMPTY); |
42
|
|
|
* $file->setCsvControl($delimiter, $enclosure, ''); //this does not currently in any PHP stable release |
43
|
|
|
* </code> |
44
|
|
|
* |
45
|
|
|
* instead you can do this |
46
|
|
|
* |
47
|
|
|
* <code> |
48
|
|
|
* $file = new SplFileObject('/path/to/file.csv', 'r'); |
49
|
|
|
* $file->setFlags(SplFileObject::READ_CSV | SplFileObject::READ_AHEAD | SplFileObject::SKIP_EMPTY); |
50
|
|
|
* $file->setCsvControl($delimiter, $enclosure, $escape); |
51
|
|
|
* EmptyEscapeParser::parse($file); //parsing will be done while ignoring the escape character value. |
52
|
|
|
* </code> |
53
|
|
|
* |
54
|
|
|
* @see https://php.net/manual/en/function.fgetcsv.php |
55
|
|
|
* @see https://php.net/manual/en/function.fgets.php |
56
|
|
|
* @see https://tools.ietf.org/html/rfc4180 |
57
|
|
|
* @see http://edoceo.com/utilitas/csv-file-format |
58
|
|
|
* |
59
|
|
|
* @internal used internally to parse a CSV document without using the escape character |
60
|
|
|
*/ |
61
|
|
|
final class EmptyEscapeParser |
62
|
|
|
{ |
63
|
|
|
/** |
64
|
|
|
* @internal |
65
|
|
|
*/ |
66
|
|
|
const FIELD_BREAKS = [false, '', "\r\n", "\n", "\r"]; |
67
|
|
|
|
68
|
|
|
/** |
69
|
|
|
* @var SplFileObject|Stream |
70
|
|
|
*/ |
71
|
|
|
private static $document; |
72
|
|
|
|
73
|
|
|
/** |
74
|
|
|
* @var string |
75
|
|
|
*/ |
76
|
|
|
private static $delimiter; |
77
|
|
|
|
78
|
|
|
/** |
79
|
|
|
* @var string |
80
|
|
|
*/ |
81
|
|
|
private static $enclosure; |
82
|
|
|
|
83
|
|
|
/** |
84
|
|
|
* @var string |
85
|
|
|
*/ |
86
|
|
|
private static $trim_mask; |
87
|
|
|
|
88
|
|
|
/** |
89
|
|
|
* @var string|bool |
90
|
|
|
*/ |
91
|
|
|
private static $line; |
92
|
|
|
|
93
|
|
|
/** |
94
|
|
|
* Converts the document into a CSV record iterator. |
95
|
|
|
* |
96
|
|
|
* Each record array contains strings elements. |
97
|
|
|
* |
98
|
|
|
* @param SplFileObject|Stream $document |
99
|
|
|
* |
100
|
|
|
* @return Generator|array[] |
101
|
|
|
*/ |
102
|
36 |
|
public static function parse($document): Generator |
103
|
|
|
{ |
104
|
36 |
|
self::$document = self::filterDocument($document); |
105
|
33 |
|
list(self::$delimiter, self::$enclosure, ) = self::$document->getCsvControl(); |
106
|
33 |
|
self::$trim_mask = str_replace([self::$delimiter, self::$enclosure], '', " \t\0\x0B"); |
107
|
33 |
|
self::$document->setFlags(0); |
108
|
33 |
|
self::$document->rewind(); |
109
|
33 |
|
while (self::$document->valid()) { |
110
|
33 |
|
$record = self::extractRecord(); |
111
|
33 |
|
if ([null] !== $record) { |
112
|
33 |
|
yield $record; |
113
|
|
|
} |
114
|
|
|
} |
115
|
33 |
|
} |
116
|
|
|
|
117
|
|
|
/** |
118
|
|
|
* Filters the submitted document. |
119
|
|
|
* |
120
|
|
|
* @param SplFileObject|Stream $document |
121
|
|
|
* |
122
|
|
|
* @return SplFileObject|Stream |
123
|
|
|
*/ |
124
|
9 |
|
private static function filterDocument($document) |
125
|
|
|
{ |
126
|
9 |
|
if ($document instanceof Stream || $document instanceof SplFileObject) { |
127
|
6 |
|
return $document; |
128
|
|
|
} |
129
|
|
|
|
130
|
3 |
|
throw new TypeError(sprintf( |
131
|
3 |
|
'Expected a %s or an SplFileObject object, %s given', |
132
|
3 |
|
Stream::class, |
133
|
3 |
|
is_object($document) ? get_class($document) : gettype($document) |
134
|
|
|
)); |
135
|
|
|
} |
136
|
|
|
|
137
|
|
|
/** |
138
|
|
|
* Extracts a record form the CSV document. |
139
|
|
|
*/ |
140
|
3 |
|
private static function extractRecord(): array |
141
|
|
|
{ |
142
|
3 |
|
$record = []; |
143
|
3 |
|
self::$line = self::$document->fgets(); |
144
|
|
|
do { |
145
|
3 |
|
$method = 'extractFieldContent'; |
146
|
3 |
|
$buffer = ltrim(self::$line, self::$trim_mask); |
147
|
3 |
|
if (($buffer[0] ?? '') === self::$enclosure) { |
148
|
3 |
|
$method = 'extractEnclosedFieldContent'; |
149
|
3 |
|
self::$line = $buffer; |
150
|
|
|
} |
151
|
|
|
|
152
|
3 |
|
$record[] = self::$method(); |
153
|
3 |
|
} while (false !== self::$line); |
154
|
|
|
|
155
|
3 |
|
return $record; |
156
|
|
|
} |
157
|
|
|
|
158
|
|
|
/** |
159
|
|
|
* Extracts the content from a field without enclosure. |
160
|
|
|
* |
161
|
|
|
* - Leading and trailing whitespaces must be removed. |
162
|
|
|
* - trailing line-breaks must be removed. |
163
|
|
|
* |
164
|
|
|
* @return null|string |
165
|
|
|
*/ |
166
|
33 |
|
private static function extractFieldContent() |
167
|
|
|
{ |
168
|
33 |
|
if (in_array(self::$line, self::FIELD_BREAKS, true)) { |
169
|
3 |
|
self::$line = false; |
170
|
|
|
|
171
|
3 |
|
return null; |
172
|
|
|
} |
173
|
|
|
|
174
|
30 |
|
list($content, self::$line) = explode(self::$delimiter, self::$line, 2) + [1 => false]; |
175
|
30 |
|
if (false === self::$line) { |
176
|
24 |
|
return rtrim($content, "\r\n"); |
177
|
|
|
} |
178
|
|
|
|
179
|
30 |
|
return $content; |
180
|
|
|
} |
181
|
|
|
|
182
|
|
|
/** |
183
|
|
|
* Extracts the content from a field with enclosure. |
184
|
|
|
* |
185
|
|
|
* - Field content can spread on multiple document lines. |
186
|
|
|
* - Content inside enclosure must be preserved. |
187
|
|
|
* - Double enclosure sequence must be replaced by single enclosure character. |
188
|
|
|
* - Trailing line break must be removed if they are not part of the field content. |
189
|
|
|
* - Invalid field content are treated as per fgetcsv behavior. |
190
|
|
|
*/ |
191
|
24 |
|
private static function extractEnclosedFieldContent(): string |
192
|
|
|
{ |
193
|
24 |
|
if ((self::$line[0] ?? '') === self::$enclosure) { |
194
|
24 |
|
self::$line = substr(self::$line, 1); |
195
|
|
|
} |
196
|
|
|
|
197
|
24 |
|
$content = ''; |
198
|
24 |
|
while (false !== self::$line) { |
199
|
24 |
|
list($buffer, $remainder) = explode(self::$enclosure, self::$line, 2) + [1 => false]; |
200
|
24 |
|
$content .= $buffer; |
201
|
24 |
|
if (false !== $remainder) { |
202
|
21 |
|
self::$line = $remainder; |
203
|
21 |
|
break; |
204
|
|
|
} |
205
|
9 |
|
self::$line = self::$document->fgets(); |
206
|
|
|
} |
207
|
|
|
|
208
|
24 |
|
if (in_array(self::$line, self::FIELD_BREAKS, true)) { |
209
|
12 |
|
self::$line = false; |
210
|
|
|
|
211
|
12 |
|
return rtrim($content, "\r\n"); |
212
|
|
|
} |
213
|
|
|
|
214
|
21 |
|
$char = self::$line[0] ?? ''; |
215
|
21 |
|
if (self::$delimiter === $char) { |
216
|
15 |
|
self::$line = substr(self::$line, 1); |
217
|
|
|
|
218
|
15 |
|
return $content; |
219
|
|
|
} |
220
|
|
|
|
221
|
18 |
|
if (self::$enclosure === $char) { |
222
|
9 |
|
return $content.self::$enclosure.self::extractEnclosedFieldContent(); |
223
|
|
|
} |
224
|
|
|
|
225
|
9 |
|
return $content.self::extractFieldContent(); |
226
|
|
|
} |
227
|
|
|
} |
228
|
|
|
|