Question

We are trying to import a CSV file via PHP into MySQL. The character encoding is utf-8 (checked, normalised and verified). When importing using a MySQL client interface tool all of the rows are present. The code is as follows:

 $handle = fopen($this->file_name, 'r+') or die ('cant open a '.$this->file_name);
 $new_filename = $upload_directory.'/'.date('Y_m_d_H_i_s').'_data_import.csv';
 $for_import = fopen($new_filename, 'x+') or die ('cant open a '.$new_filename.'_for_import');

 $data = array();
 $row_counter = 0;

 while (($row = fgetss($handle)) !== false)
 {
    $row = fgetss($handle);
    if (($row_counter > $this->rows_to_ignore))
{
    $enc = mb_detect_encoding($row);
    $row = iconv($enc, 'UTF-8', $row);
    $row = str_replace(", ", '\\, ', $row);
    $row = trim($row);
    $row = str_getcsv($row, ',', '"', '\\');
    fputcsv($for_import, $row, ',', '"');
}

$row_counter++;

 }
 fclose($for_import);
 fclose($handle);

The row_counter variable returns approx 50% of all rows. Any ideas?

Thanks

Answer 1

You’re calling $row = fgetts($handle) twice. The first call is in your while() condition and consumes a row, the second is inside the loop and consumes another row. The rows from the first call are being lost.