1 |
85ad3d82
|
Assos Assos
|
<?php
|
2 |
|
|
|
3 |
|
|
/**
|
4 |
|
|
* @file
|
5 |
|
|
* Contains CSV Parser.
|
6 |
|
|
*
|
7 |
|
|
* Functions in this file are independent of the Feeds specific implementation.
|
8 |
|
|
* Thanks to jpetso http://drupal.org/user/56020 for most of the code in this
|
9 |
|
|
* file.
|
10 |
|
|
*/
|
11 |
|
|
|
12 |
|
|
/**
|
13 |
|
|
* Text lines from file iterator.
|
14 |
|
|
*/
|
15 |
|
|
class ParserCSVIterator implements Iterator {
|
16 |
|
|
private $handle;
|
17 |
|
|
private $currentLine;
|
18 |
|
|
private $currentPos;
|
19 |
|
|
|
20 |
|
|
public function __construct($filepath) {
|
21 |
|
|
$this->handle = fopen($filepath, 'r');
|
22 |
|
|
$this->currentLine = NULL;
|
23 |
|
|
$this->currentPos = NULL;
|
24 |
|
|
}
|
25 |
|
|
|
26 |
|
|
function __destruct() {
|
27 |
|
|
if ($this->handle) {
|
28 |
|
|
fclose($this->handle);
|
29 |
|
|
}
|
30 |
|
|
}
|
31 |
|
|
|
32 |
|
|
public function rewind($pos = 0) {
|
33 |
|
|
if ($this->handle) {
|
34 |
|
|
fseek($this->handle, $pos);
|
35 |
|
|
$this->next();
|
36 |
|
|
}
|
37 |
|
|
}
|
38 |
|
|
|
39 |
|
|
public function next() {
|
40 |
|
|
if ($this->handle) {
|
41 |
|
|
$this->currentLine = feof($this->handle) ? NULL : fgets($this->handle);
|
42 |
|
|
$this->currentPos = ftell($this->handle);
|
43 |
|
|
return $this->currentLine;
|
44 |
|
|
}
|
45 |
|
|
}
|
46 |
|
|
|
47 |
|
|
public function valid() {
|
48 |
|
|
return isset($this->currentLine);
|
49 |
|
|
}
|
50 |
|
|
|
51 |
|
|
public function current() {
|
52 |
|
|
return $this->currentLine;
|
53 |
|
|
}
|
54 |
|
|
|
55 |
|
|
public function currentPos() {
|
56 |
|
|
return $this->currentPos;
|
57 |
|
|
}
|
58 |
|
|
|
59 |
|
|
public function key() {
|
60 |
|
|
return 'line';
|
61 |
|
|
}
|
62 |
|
|
}
|
63 |
|
|
|
64 |
|
|
/**
|
65 |
|
|
* Functionality to parse CSV files into a two dimensional array.
|
66 |
|
|
*/
|
67 |
|
|
class ParserCSV {
|
68 |
|
|
private $delimiter;
|
69 |
|
|
private $skipFirstLine;
|
70 |
|
|
private $columnNames;
|
71 |
|
|
private $timeout;
|
72 |
|
|
private $timeoutReached;
|
73 |
|
|
private $startByte;
|
74 |
|
|
private $lineLimit;
|
75 |
|
|
private $lastLinePos;
|
76 |
|
|
|
77 |
|
|
public function __construct() {
|
78 |
|
|
$this->delimiter = ',';
|
79 |
|
|
$this->skipFirstLine = FALSE;
|
80 |
|
|
$this->columnNames = FALSE;
|
81 |
|
|
$this->timeout = FALSE;
|
82 |
|
|
$this->timeoutReached = FALSE;
|
83 |
|
|
$this->startByte = 0;
|
84 |
|
|
$this->lineLimit = 0;
|
85 |
|
|
$this->lastLinePos = 0;
|
86 |
|
|
ini_set('auto_detect_line_endings', TRUE);
|
87 |
|
|
}
|
88 |
|
|
|
89 |
|
|
/**
|
90 |
|
|
* Set the column delimiter string.
|
91 |
|
|
* By default, the comma (',') is used as delimiter.
|
92 |
|
|
*/
|
93 |
|
|
public function setDelimiter($delimiter) {
|
94 |
|
|
$this->delimiter = $delimiter;
|
95 |
|
|
}
|
96 |
|
|
|
97 |
|
|
/**
|
98 |
|
|
* Set this to TRUE if the parser should skip the first line of the CSV text,
|
99 |
|
|
* which might be desired if the first line contains the column names.
|
100 |
|
|
* By default, this is set to FALSE and the first line is not skipped.
|
101 |
|
|
*/
|
102 |
|
|
public function setSkipFirstLine($skipFirstLine) {
|
103 |
|
|
$this->skipFirstLine = $skipFirstLine;
|
104 |
|
|
}
|
105 |
|
|
|
106 |
|
|
/**
|
107 |
|
|
* Specify an array of column names if you know them in advance, or FALSE
|
108 |
|
|
* (which is the default) to unset any prior column names. If no column names
|
109 |
|
|
* are set, the parser will put each row into a simple numerically indexed
|
110 |
|
|
* array. If column names are given, the parser will create arrays with
|
111 |
|
|
* these column names as array keys instead.
|
112 |
|
|
*/
|
113 |
|
|
public function setColumnNames($columnNames) {
|
114 |
|
|
$this->columnNames = $columnNames;
|
115 |
|
|
}
|
116 |
|
|
|
117 |
|
|
/**
|
118 |
|
|
* Define the time (in milliseconds) after which the parser stops parsing,
|
119 |
|
|
* even if it has not yet finished processing the CSV data. If the timeout
|
120 |
|
|
* has been reached before parsing is done, the parse() method will return
|
121 |
|
|
* an incomplete list of rows - a single row will never be cut off in the
|
122 |
|
|
* middle, though. By default, no timeout (@p $timeout == FALSE) is defined.
|
123 |
|
|
*
|
124 |
|
|
* You can check if the timeout has been reached by calling the
|
125 |
|
|
* timeoutReached() method after parse() has been called.
|
126 |
|
|
*/
|
127 |
|
|
public function setTimeout($timeout) {
|
128 |
|
|
$this->timeout = $timeout;
|
129 |
|
|
}
|
130 |
|
|
|
131 |
|
|
/**
|
132 |
|
|
* After calling the parse() method, determine if the timeout (set by the
|
133 |
|
|
* setTimeout() method) has been reached.
|
134 |
|
|
*
|
135 |
|
|
* @deprecated Use lastLinePos() instead to determine whether a file has
|
136 |
|
|
* finished parsing.
|
137 |
|
|
*/
|
138 |
|
|
public function timeoutReached() {
|
139 |
|
|
return $this->timeoutReached;
|
140 |
|
|
}
|
141 |
|
|
|
142 |
|
|
/**
|
143 |
|
|
* Define the number of lines to parse in one parsing operation.
|
144 |
|
|
*
|
145 |
|
|
* By default, all lines of a file are being parsed.
|
146 |
|
|
*/
|
147 |
|
|
public function setLineLimit($lines) {
|
148 |
|
|
$this->lineLimit = $lines;
|
149 |
|
|
}
|
150 |
|
|
|
151 |
|
|
/**
|
152 |
|
|
* Get the byte number where the parser left off after last parse() call.
|
153 |
|
|
*
|
154 |
|
|
* @return
|
155 |
|
|
* 0 if all lines or no line has been parsed, the byte position of where a
|
156 |
|
|
* timeout or the line limit has been reached otherwise. This position can be
|
157 |
|
|
* used to set the start byte for the next iteration after parse() has
|
158 |
|
|
* reached the timeout set with setTimeout() or the line limit set with
|
159 |
|
|
* setLineLimit().
|
160 |
|
|
*
|
161 |
|
|
* @see ParserCSV::setStartByte()
|
162 |
|
|
*/
|
163 |
|
|
public function lastLinePos() {
|
164 |
|
|
return $this->lastLinePos;
|
165 |
|
|
}
|
166 |
|
|
|
167 |
|
|
/**
|
168 |
|
|
* Set the byte where file should be started to read.
|
169 |
|
|
*
|
170 |
|
|
* Useful when parsing a file in batches.
|
171 |
|
|
*/
|
172 |
|
|
public function setStartByte($start) {
|
173 |
|
|
return $this->startByte = $start;
|
174 |
|
|
}
|
175 |
|
|
|
176 |
|
|
/**
|
177 |
|
|
* Parse CSV files into a two dimensional array.
|
178 |
|
|
*
|
179 |
|
|
* @param Iterator $lineIterator
|
180 |
|
|
* An Iterator object that yields line strings, e.g. ParserCSVIterator.
|
181 |
|
|
* @param $start
|
182 |
|
|
* The byte number from where to start parsing the file.
|
183 |
|
|
* @param $lines
|
184 |
|
|
* The number of lines to parse, 0 for all lines.
|
185 |
|
|
* @return
|
186 |
|
|
* Two dimensional array that contains the data in the CSV file.
|
187 |
|
|
*/
|
188 |
|
|
public function parse(Iterator $lineIterator) {
|
189 |
|
|
$skipLine = $this->skipFirstLine;
|
190 |
|
|
$rows = array();
|
191 |
|
|
|
192 |
|
|
$this->timeoutReached = FALSE;
|
193 |
|
|
$this->lastLinePos = 0;
|
194 |
|
|
$maxTime = empty($this->timeout) ? FALSE : (microtime() + $this->timeout);
|
195 |
|
|
$linesParsed = 0;
|
196 |
|
|
|
197 |
|
|
for ($lineIterator->rewind($this->startByte); $lineIterator->valid(); $lineIterator->next()) {
|
198 |
|
|
|
199 |
|
|
// Make really sure we've got lines without trailing newlines.
|
200 |
|
|
$line = trim($lineIterator->current(), "\r\n");
|
201 |
|
|
|
202 |
|
|
// Skip empty lines.
|
203 |
|
|
if (empty($line)) {
|
204 |
|
|
continue;
|
205 |
|
|
}
|
206 |
|
|
// If the first line contains column names, skip it.
|
207 |
|
|
if ($skipLine) {
|
208 |
|
|
$skipLine = FALSE;
|
209 |
|
|
continue;
|
210 |
|
|
}
|
211 |
|
|
|
212 |
|
|
// The actual parser. explode() is unfortunately not suitable because the
|
213 |
|
|
// delimiter might be located inside a quoted field, and that would break
|
214 |
|
|
// the field and/or require additional effort to re-join the fields.
|
215 |
|
|
$quoted = FALSE;
|
216 |
|
|
$currentIndex = 0;
|
217 |
|
|
$currentField = '';
|
218 |
|
|
$fields = array();
|
219 |
|
|
|
220 |
|
|
// We must use strlen() as we're parsing byte by byte using strpos(), so
|
221 |
|
|
// drupal_strlen() will not work properly.
|
222 |
|
|
while ($currentIndex <= strlen($line)) {
|
223 |
|
|
if ($quoted) {
|
224 |
|
|
$nextQuoteIndex = strpos($line, '"', $currentIndex);
|
225 |
|
|
|
226 |
|
|
if ($nextQuoteIndex === FALSE) {
|
227 |
|
|
// There's a line break before the quote is closed, so fetch the
|
228 |
|
|
// next line and start from there.
|
229 |
|
|
$currentField .= substr($line, $currentIndex);
|
230 |
|
|
$lineIterator->next();
|
231 |
|
|
|
232 |
|
|
if (!$lineIterator->valid()) {
|
233 |
|
|
// Whoa, an unclosed quote! Well whatever, let's just ignore
|
234 |
|
|
// that shortcoming and record it nevertheless.
|
235 |
|
|
$fields[] = $currentField;
|
236 |
|
|
break;
|
237 |
|
|
}
|
238 |
|
|
// Ok, so, on with fetching the next line, as mentioned above.
|
239 |
|
|
$currentField .= "\n";
|
240 |
|
|
$line = trim($lineIterator->current(), "\r\n");
|
241 |
|
|
$currentIndex = 0;
|
242 |
|
|
continue;
|
243 |
|
|
}
|
244 |
|
|
|
245 |
|
|
// There's actually another quote in this line...
|
246 |
|
|
// find out whether it's escaped or not.
|
247 |
|
|
$currentField .= substr($line, $currentIndex, $nextQuoteIndex - $currentIndex);
|
248 |
|
|
|
249 |
|
|
if (isset($line[$nextQuoteIndex + 1]) && $line[$nextQuoteIndex + 1] === '"') {
|
250 |
|
|
// Escaped quote, add a single one to the field and proceed quoted.
|
251 |
|
|
$currentField .= '"';
|
252 |
|
|
$currentIndex = $nextQuoteIndex + 2;
|
253 |
|
|
}
|
254 |
|
|
else {
|
255 |
|
|
// End of the quoted section, close the quote and let the
|
256 |
|
|
// $quoted == FALSE block finalize the field.
|
257 |
|
|
$quoted = FALSE;
|
258 |
|
|
$currentIndex = $nextQuoteIndex + 1;
|
259 |
|
|
}
|
260 |
|
|
}
|
261 |
|
|
else { // $quoted == FALSE
|
262 |
|
|
// First, let's find out where the next character of interest is.
|
263 |
|
|
$nextQuoteIndex = strpos($line, '"', $currentIndex);
|
264 |
|
|
$nextDelimiterIndex = strpos($line, $this->delimiter, $currentIndex);
|
265 |
|
|
|
266 |
|
|
if ($nextQuoteIndex === FALSE) {
|
267 |
|
|
$nextIndex = $nextDelimiterIndex;
|
268 |
|
|
}
|
269 |
|
|
elseif ($nextDelimiterIndex === FALSE) {
|
270 |
|
|
$nextIndex = $nextQuoteIndex;
|
271 |
|
|
}
|
272 |
|
|
else {
|
273 |
|
|
$nextIndex = min($nextQuoteIndex, $nextDelimiterIndex);
|
274 |
|
|
}
|
275 |
|
|
|
276 |
|
|
if ($nextIndex === FALSE) {
|
277 |
|
|
// This line is done, add the rest of it as last field.
|
278 |
|
|
$currentField .= substr($line, $currentIndex);
|
279 |
|
|
$fields[] = $currentField;
|
280 |
|
|
break;
|
281 |
|
|
}
|
282 |
|
|
elseif ($line[$nextIndex] === $this->delimiter[0]) {
|
283 |
|
|
$length = ($nextIndex + strlen($this->delimiter) - 1) - $currentIndex;
|
284 |
|
|
$currentField .= substr($line, $currentIndex, $length);
|
285 |
|
|
$fields[] = $currentField;
|
286 |
|
|
$currentField = '';
|
287 |
|
|
$currentIndex += $length + 1;
|
288 |
|
|
// Continue with the next field.
|
289 |
|
|
}
|
290 |
|
|
else { // $line[$nextIndex] == '"'
|
291 |
|
|
$quoted = TRUE;
|
292 |
|
|
$currentField .= substr($line, $currentIndex, $nextIndex - $currentIndex);
|
293 |
|
|
$currentIndex = $nextIndex + 1;
|
294 |
|
|
// Continue this field in the $quoted == TRUE block.
|
295 |
|
|
}
|
296 |
|
|
}
|
297 |
|
|
}
|
298 |
|
|
// End of CSV parser. We've now got all the fields of the line as strings
|
299 |
|
|
// in the $fields array.
|
300 |
|
|
|
301 |
|
|
if (empty($this->columnNames)) {
|
302 |
|
|
$row = $fields;
|
303 |
|
|
}
|
304 |
|
|
else {
|
305 |
|
|
$row = array();
|
306 |
|
|
foreach ($this->columnNames as $columnName) {
|
307 |
|
|
$field = array_shift($fields);
|
308 |
|
|
$row[$columnName] = isset($field) ? $field : '';
|
309 |
|
|
}
|
310 |
|
|
}
|
311 |
|
|
$rows[] = $row;
|
312 |
|
|
|
313 |
|
|
// Quit parsing if timeout has been reached or requested lines have been
|
314 |
|
|
// reached.
|
315 |
|
|
if (!empty($maxTime) && microtime() > $maxTime) {
|
316 |
|
|
$this->timeoutReached = TRUE;
|
317 |
|
|
$this->lastLinePos = $lineIterator->currentPos();
|
318 |
|
|
break;
|
319 |
|
|
}
|
320 |
|
|
$linesParsed++;
|
321 |
|
|
if ($this->lineLimit && $linesParsed >= $this->lineLimit) {
|
322 |
|
|
$this->lastLinePos = $lineIterator->currentPos();
|
323 |
|
|
break;
|
324 |
|
|
}
|
325 |
|
|
}
|
326 |
|
|
return $rows;
|
327 |
|
|
}
|
328 |
|
|
} |