Total Complexity | 68 |
Total Lines | 454 |
Duplicated Lines | 0 % |
Changes | 1 | ||
Bugs | 0 | Features | 0 |
Complex classes like CsvBulkLoader often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use CsvBulkLoader, and based on these observations, apply Extract Interface, too.
1 | <?php |
||
22 | class CsvBulkLoader extends BulkLoader |
||
23 | { |
||
24 | |||
25 | /** |
||
26 | * Delimiter character (Default: comma). |
||
27 | * |
||
28 | * @var string |
||
29 | */ |
||
30 | public $delimiter = ','; |
||
31 | |||
32 | /** |
||
33 | * Enclosure character (Default: doublequote) |
||
34 | * |
||
35 | * @var string |
||
36 | */ |
||
37 | public $enclosure = '"'; |
||
38 | |||
39 | /** |
||
40 | * Identifies if csv the has a header row. |
||
41 | * |
||
42 | * @var boolean |
||
43 | */ |
||
44 | public $hasHeaderRow = true; |
||
45 | |||
46 | /** |
||
47 | * Number of lines to split large CSV files into. |
||
48 | * |
||
49 | * @var int |
||
50 | * |
||
51 | * @config |
||
52 | */ |
||
53 | private static $lines = 1000; |
||
54 | |||
55 | /** |
||
56 | * @inheritDoc |
||
57 | */ |
||
58 | public function preview($filepath) |
||
61 | } |
||
62 | |||
63 | /** |
||
64 | * @param string $filepath |
||
65 | * @param boolean $preview |
||
66 | * |
||
67 | * @return null|BulkLoader_Result |
||
68 | */ |
||
69 | protected function processAll($filepath, $preview = false) |
||
70 | { |
||
71 | $this->extend('onBeforeProcessAll', $filepath, $preview); |
||
72 | |||
73 | $result = BulkLoader_Result::create(); |
||
74 | |||
75 | try { |
||
76 | $filepath = Director::getAbsFile($filepath); |
||
77 | $csvReader = Reader::createFromPath($filepath, 'r'); |
||
78 | $csvReader->setDelimiter($this->delimiter); |
||
79 | |||
80 | // league/csv 9 |
||
81 | if (method_exists($csvReader, 'skipInputBOM')) { |
||
82 | $csvReader->skipInputBOM(); |
||
83 | // league/csv 8 |
||
84 | } else { |
||
85 | $csvReader->stripBom(true); |
||
86 | } |
||
87 | |||
88 | $tabExtractor = function ($row, $rowOffset) { |
||
89 | foreach ($row as &$item) { |
||
90 | // [SS-2017-007] Ensure all cells with leading tab and then [@=+] have the tab removed on import |
||
91 | if (preg_match("/^\t[\-@=\+]+.*/", $item)) { |
||
92 | $item = ltrim($item, "\t"); |
||
93 | } |
||
94 | } |
||
95 | return $row; |
||
96 | }; |
||
97 | |||
98 | if ($this->columnMap) { |
||
99 | $headerMap = $this->getNormalisedColumnMap(); |
||
100 | |||
101 | $remapper = function ($row, $rowOffset) use ($headerMap, $tabExtractor) { |
||
102 | $row = $tabExtractor($row, $rowOffset); |
||
103 | foreach ($headerMap as $column => $renamedColumn) { |
||
104 | if ($column == $renamedColumn) { |
||
105 | continue; |
||
106 | } |
||
107 | if (array_key_exists($column, $row)) { |
||
108 | if (strpos($renamedColumn, '_ignore_') !== 0) { |
||
109 | $row[$renamedColumn] = $row[$column]; |
||
110 | } |
||
111 | unset($row[$column]); |
||
112 | } |
||
113 | } |
||
114 | return $row; |
||
115 | }; |
||
116 | } else { |
||
117 | $remapper = $tabExtractor; |
||
118 | } |
||
119 | |||
120 | if ($this->hasHeaderRow) { |
||
121 | if (method_exists($csvReader, 'fetchAssoc')) { |
||
122 | $rows = $csvReader->fetchAssoc(0, $remapper); |
||
123 | } else { |
||
124 | $csvReader->setHeaderOffset(0); |
||
125 | $rows = new MapIterator($csvReader->getRecords(), $remapper); |
||
126 | } |
||
127 | } elseif ($this->columnMap) { |
||
128 | if (method_exists($csvReader, 'fetchAssoc')) { |
||
129 | $rows = $csvReader->fetchAssoc($headerMap, $remapper); |
||
130 | } else { |
||
131 | $rows = new MapIterator($csvReader->getRecords($headerMap), $remapper); |
||
132 | } |
||
133 | } |
||
134 | |||
135 | foreach ($rows as $row) { |
||
136 | $this->processRecord($row, $this->columnMap, $result, $preview); |
||
137 | } |
||
138 | } catch (\Exception $e) { |
||
139 | $failedMessage = sprintf("Failed to parse %s", $filepath); |
||
140 | if (Director::isDev()) { |
||
141 | $failedMessage = sprintf($failedMessage . " because %s", $e->getMessage()); |
||
142 | } |
||
143 | print $failedMessage . PHP_EOL; |
||
144 | } |
||
145 | |||
146 | $this->extend('onAfterProcessAll', $result, $preview); |
||
147 | |||
148 | return $result; |
||
149 | } |
||
150 | |||
151 | protected function getNormalisedColumnMap() |
||
167 | } |
||
168 | |||
169 | /** |
||
170 | * Splits a large file up into many smaller files. |
||
171 | * |
||
172 | * @param string $path Path to large file to split |
||
173 | * @param int $lines Number of lines per file |
||
174 | * |
||
175 | * @return array List of file paths |
||
176 | */ |
||
177 | protected function splitFile($path, $lines = null) |
||
178 | { |
||
179 | Deprecation::notice('5.0', 'splitFile is deprecated, please process files using a stream'); |
||
180 | |||
181 | if (!is_int($lines)) { |
||
182 | $lines = $this->config()->get("lines"); |
||
183 | } |
||
184 | |||
185 | $new = $this->getNewSplitFileName(); |
||
186 | |||
187 | $to = fopen($new, 'w+'); |
||
188 | $from = fopen($path, 'r'); |
||
189 | |||
190 | $header = null; |
||
191 | |||
192 | if ($this->hasHeaderRow) { |
||
193 | $header = fgets($from); |
||
194 | fwrite($to, $header); |
||
195 | } |
||
196 | |||
197 | $files = []; |
||
198 | $files[] = $new; |
||
199 | |||
200 | $count = 0; |
||
201 | |||
202 | while (!feof($from)) { |
||
203 | fwrite($to, fgets($from)); |
||
204 | |||
205 | $count++; |
||
206 | |||
207 | if ($count >= $lines) { |
||
208 | fclose($to); |
||
209 | |||
210 | // get a new temporary file name, to write the next lines to |
||
211 | $new = $this->getNewSplitFileName(); |
||
212 | |||
213 | $to = fopen($new, 'w+'); |
||
214 | |||
215 | if ($this->hasHeaderRow) { |
||
216 | // add the headers to the new file |
||
217 | fwrite($to, $header); |
||
218 | } |
||
219 | |||
220 | $files[] = $new; |
||
221 | |||
222 | $count = 0; |
||
223 | } |
||
224 | } |
||
225 | fclose($to); |
||
226 | |||
227 | return $files; |
||
228 | } |
||
229 | |||
230 | /** |
||
231 | * @return string |
||
232 | */ |
||
233 | protected function getNewSplitFileName() |
||
234 | { |
||
235 | Deprecation::notice('5.0', 'getNewSplitFileName is deprecated, please name your files yourself'); |
||
236 | return TEMP_PATH . DIRECTORY_SEPARATOR . uniqid(str_replace('\\', '_', static::class), true) . '.csv'; |
||
237 | } |
||
238 | |||
239 | /** |
||
240 | * @param string $filepath |
||
241 | * @param boolean $preview |
||
242 | * |
||
243 | * @return BulkLoader_Result |
||
244 | */ |
||
245 | protected function processChunk($filepath, $preview = false) |
||
246 | { |
||
247 | Deprecation::notice('5.0', 'processChunk is deprecated, please process rows individually'); |
||
248 | $results = BulkLoader_Result::create(); |
||
249 | |||
250 | $csv = new CSVParser( |
||
251 | $filepath, |
||
252 | $this->delimiter, |
||
253 | $this->enclosure |
||
254 | ); |
||
255 | |||
256 | // ColumnMap has two uses, depending on whether hasHeaderRow is set |
||
257 | if ($this->columnMap) { |
||
258 | // if the map goes to a callback, use the same key value as the map |
||
259 | // value, rather than function name as multiple keys may use the |
||
260 | // same callback |
||
261 | $map = []; |
||
262 | foreach ($this->columnMap as $k => $v) { |
||
263 | if (strpos($v, "->") === 0) { |
||
264 | $map[$k] = $k; |
||
265 | } else { |
||
266 | $map[$k] = $v; |
||
267 | } |
||
268 | } |
||
269 | |||
270 | if ($this->hasHeaderRow) { |
||
271 | $csv->mapColumns($map); |
||
272 | } else { |
||
273 | $csv->provideHeaderRow($map); |
||
274 | } |
||
275 | } |
||
276 | |||
277 | foreach ($csv as $row) { |
||
278 | $this->processRecord($row, $this->columnMap, $results, $preview); |
||
279 | } |
||
280 | |||
281 | return $results; |
||
282 | } |
||
283 | |||
284 | /** |
||
285 | * @todo Better messages for relation checks and duplicate detection |
||
286 | * Note that columnMap isn't used. |
||
287 | * |
||
288 | * @param array $record |
||
289 | * @param array $columnMap |
||
290 | * @param BulkLoader_Result $results |
||
291 | * @param boolean $preview |
||
292 | * |
||
293 | * @return int |
||
294 | */ |
||
295 | protected function processRecord($record, $columnMap, &$results, $preview = false) |
||
296 | { |
||
297 | $class = $this->objectClass; |
||
298 | |||
299 | // find existing object, or create new one |
||
300 | $existingObj = $this->findExistingObject($record, $columnMap); |
||
301 | /** @var DataObject $obj */ |
||
302 | $obj = ($existingObj) ? $existingObj : new $class(); |
||
303 | $schema = DataObject::getSchema(); |
||
304 | |||
305 | // first run: find/create any relations and store them on the object |
||
306 | // we can't combine runs, as other columns might rely on the relation being present |
||
307 | foreach ($record as $fieldName => $val) { |
||
308 | // don't bother querying of value is not set |
||
309 | if ($this->isNullValue($val)) { |
||
310 | continue; |
||
311 | } |
||
312 | |||
313 | // checking for existing relations |
||
314 | if (isset($this->relationCallbacks[$fieldName])) { |
||
315 | // trigger custom search method for finding a relation based on the given value |
||
316 | // and write it back to the relation (or create a new object) |
||
317 | $relationName = $this->relationCallbacks[$fieldName]['relationname']; |
||
318 | /** @var DataObject $relationObj */ |
||
319 | $relationObj = null; |
||
320 | if ($this->hasMethod($this->relationCallbacks[$fieldName]['callback'])) { |
||
321 | $relationObj = $this->{$this->relationCallbacks[$fieldName]['callback']}($obj, $val, $record); |
||
322 | } elseif ($obj->hasMethod($this->relationCallbacks[$fieldName]['callback'])) { |
||
323 | $relationObj = $obj->{$this->relationCallbacks[$fieldName]['callback']}($val, $record); |
||
324 | } |
||
325 | if (!$relationObj || !$relationObj->exists()) { |
||
326 | $relationClass = $schema->hasOneComponent(get_class($obj), $relationName); |
||
327 | $relationObj = new $relationClass(); |
||
328 | //write if we aren't previewing |
||
329 | if (!$preview) { |
||
330 | $relationObj->write(); |
||
331 | } |
||
332 | } |
||
333 | $obj->{"{$relationName}ID"} = $relationObj->ID; |
||
334 | //write if we are not previewing |
||
335 | if (!$preview) { |
||
336 | $obj->write(); |
||
337 | $obj->flushCache(); // avoid relation caching confusion |
||
338 | } |
||
339 | } elseif (strpos($fieldName, '.') !== false) { |
||
340 | // we have a relation column with dot notation |
||
341 | [$relationName, $columnName] = explode('.', $fieldName); |
||
342 | // always gives us an component (either empty or existing) |
||
343 | $relationObj = $obj->getComponent($relationName); |
||
344 | if (!$preview) { |
||
345 | $relationObj->write(); |
||
346 | } |
||
347 | $obj->{"{$relationName}ID"} = $relationObj->ID; |
||
348 | |||
349 | //write if we are not previewing |
||
350 | if (!$preview) { |
||
351 | $obj->write(); |
||
352 | $obj->flushCache(); // avoid relation caching confusion |
||
353 | } |
||
354 | } |
||
355 | } |
||
356 | |||
357 | // second run: save data |
||
358 | |||
359 | foreach ($record as $fieldName => $val) { |
||
360 | // break out of the loop if we are previewing |
||
361 | if ($preview) { |
||
362 | break; |
||
363 | } |
||
364 | |||
365 | // look up the mapping to see if this needs to map to callback |
||
366 | $mapped = $this->columnMap && isset($this->columnMap[$fieldName]); |
||
367 | |||
368 | if ($mapped && strpos($this->columnMap[$fieldName], '->') === 0) { |
||
369 | $funcName = substr($this->columnMap[$fieldName], 2); |
||
370 | |||
371 | $this->$funcName($obj, $val, $record); |
||
372 | } elseif ($obj->hasMethod("import{$fieldName}")) { |
||
373 | $obj->{"import{$fieldName}"}($val, $record); |
||
374 | } else { |
||
375 | $obj->update([$fieldName => $val]); |
||
376 | } |
||
377 | } |
||
378 | |||
379 | $isChanged = $obj->isChanged(); |
||
380 | |||
381 | // write record |
||
382 | if (!$preview) { |
||
383 | $obj->write(); |
||
384 | } |
||
385 | |||
386 | // @todo better message support |
||
387 | $message = ''; |
||
388 | |||
389 | // save to results |
||
390 | if ($existingObj) { |
||
391 | // We mark as updated regardless of isChanged, since custom formatters and importers |
||
392 | // might have affected relationships and other records. |
||
393 | $results->addUpdated($obj, $message); |
||
394 | } else { |
||
395 | $results->addCreated($obj, $message); |
||
396 | } |
||
397 | |||
398 | $this->extend('onAfterProcessRecord', $obj, $preview, $isChanged); |
||
399 | |||
400 | $objID = $obj->ID; |
||
401 | |||
402 | $obj->destroy(); |
||
403 | |||
404 | // memory usage |
||
405 | unset($existingObj, $obj); |
||
406 | |||
407 | return $objID; |
||
408 | } |
||
409 | |||
410 | /** |
||
411 | * Find an existing objects based on one or more uniqueness columns |
||
412 | * specified via {@link self::$duplicateChecks}. |
||
413 | * |
||
414 | * @todo support $columnMap |
||
415 | * |
||
416 | * @param array $record CSV data column |
||
417 | * @param array $columnMap |
||
418 | * @return DataObject |
||
419 | */ |
||
420 | public function findExistingObject($record, $columnMap = []) |
||
465 | } |
||
466 | |||
467 | /** |
||
468 | * Determine whether any loaded files should be parsed with a |
||
469 | * header-row (otherwise we rely on {@link self::$columnMap}. |
||
470 | * |
||
471 | * @return boolean |
||
472 | */ |
||
473 | public function hasHeaderRow() |
||
476 | } |
||
477 | } |
||
478 |