/**
* @inheritdoc
*/
public function needsTransforming(ResourceInterface $resource)
{
$file = $resource->getFile()->getPathname();
$handle = fopen($file, 'r');
$result = (bool) preg_match($this->xmlDeclarationRegEx, fread($handle, 96));
fclose($handle);
return $result;
}
/**
* Creates a reader for a resource.
*
* @param ResourceInterface $resource
*/
protected function createReader(ResourceInterface $resource)
{
try {
$jsonFile = $resource->getFile()->getPathname();
$this->fileObject = new \SplFileObject($jsonFile);
} catch (EmptyResponseException $e) {
// getting an empty response is not a problem for jsonlines files, this
// just means that there are no records.
$this->fileObject = new \ArrayIterator();
}
}
/**
* @inheritdoc
*/
public function needsTransforming(ResourceInterface $resource)
{
$file = $resource->getFile();
while (!$file->eof()) {
$line = $file->fgets();
if (!mb_check_encoding($line, $this->toEncoding)) {
return true;
}
}
return false;
}
/**
* @inheritdoc
*/
public function transform(ResourceInterface $resource, ResourceCollection $collection)
{
$file = $resource->getFile()->getPathname();
$tmpFile = tempnam(sys_get_temp_dir(), $file);
// remove control characters
$old = fopen($file, 'r');
$new = fopen($tmpFile, 'w');
// list control characters, but leave out \t\r\n
$chars = array_map('chr', range(0, 31));
$chars[] = chr(127);
unset($chars[9], $chars[10], $chars[13]);
while (!feof($old)) {
fwrite($new, str_replace($chars, '', fread($old, $this->length)));
}
fclose($old);
fclose($new);
// atomic write
$this->rename($tmpFile, $file);
$transport = FileTransport::create($file);
if ($resource->getTransport()) {
$transport->setDestinationDir($resource->getTransport()->getDestinationDir());
}
return new FileResource($transport);
}
/**
* @inheritdoc
*/
protected function createReader(ResourceInterface $resource)
{
$this->fileObject = new SplFileObject($resource->getFile()->getPathname());
$this->fileObject->setFlags(SplFileObject::READ_CSV | SplFileObject::DROP_NEW_LINE | SplFileObject::SKIP_EMPTY);
$this->fileObject->setCsvControl($this->delimiter, $this->enclosure, $this->escape);
if ($this->useFirstRow) {
/** @var array $mapping */
$mapping = $this->fileObject->current();
$this->setFieldMapping($mapping);
$this->fileObject->next();
}
}
/**
* @inheritdoc
*/
protected function createReader(ResourceInterface $resource)
{
$this->reader = new \XmlReader();
$this->open($resource->getFile()->getPathname());
$this->key = -1;
$this->next();
}
/**
* @param ResourceInterface $resource
*
* @return array
*/
protected function breakup(ResourceInterface $resource)
{
$originalFile = $resource->getFile();
$baseFile = $originalFile->getPathname();
$this->reader->setResources(new ResourceCollection([$resource]));
$partCount = 0;
$started = false;
while ($this->reader->valid()) {
if ($this->reader->key() % $this->size === 0) {
if ($this->reader->key() > 0) {
$this->endPart();
}
$file = sprintf('%s.part%s', $baseFile, ++$partCount);
$this->startPart($file);
$started = true;
}
$this->writer->write($this->reader->current());
$this->writer->flush();
$this->reader->next();
}
if ($started) {
$this->endPart();
}
return $this->getPartFiles($resource);
}
/**
* @param ResourceInterface $resource
*
* @return string
*/
protected function getTargetDir(ResourceInterface $resource)
{
return $this->target ?: $resource->getFile()->getPath();
}
