Magic functions to detect encoding and separator of CSV file and parsing it

php-csv-parse-detect-encoding-separator.php

<?php

function parseCsv($file) {
    $content = file_get_contents($file);
    if (!$content) {
        throw Exception('Unable to read the CSV file');
    }
    
    $encoding = detectEncoding($content);
    $content = iconv($encoding, 'UTF-8//TRANSLIT', $content);
    
    echo $content."\n\n";
    
    $fileSample = substr($content, 0, 100);
    $separator = ';';
    if (strpos($fileSample, ',') !== false) {
        $separator = ',';
    }
    
    $csv = [];
    $fp = fopen("php://temp", 'r+');
    fputs($fp, $content);
    rewind($fp);
    
    // parse header
    $cols = [];
    if (($data = fgetcsv($fp, 10000, $separator)) !== false) {
        $cols = $data;
    } else {
        throw Exception('No content found in the CSV file');
    }
    
    // parse rows
    while (($data = fgetcsv($fp, 10000, $separator)) !== false) {
        $row = [];
        foreach ($data as $i => $value) {
            $row[$cols[$i]] = $value;
        }
        $csv[] = $row;
    }
    fclose($fp);
    return $csv;
}

function detectEncoding($content) {
    // try to detect BOM first
    $first2 = substr($content, 0, 2);
    $first3 = substr($content, 0, 3);
    $first4 = substr($content, 0, 3);
    
    $UTF8_BOM = chr(0xEF) . chr(0xBB) . chr(0xBF);
    $UTF32_BIG_ENDIAN_BOM = chr(0x00) . chr(0x00) . chr(0xFE) . chr(0xFF);
    $UTF32_LITTLE_ENDIAN_BOM = chr(0xFF) . chr(0xFE) . chr(0x00) . chr(0x00);
    $UTF16_BIG_ENDIAN_BOM = chr(0xFE) . chr(0xFF);
    $UTF16_LITTLE_ENDIAN_BOM = chr(0xFF) . chr(0xFE);

    if ($first3 == $UTF8_BOM) return 'UTF-8';
    elseif ($first4 == $UTF32_BIG_ENDIAN_BOM) return 'UTF-32BE';
    elseif ($first4 == $UTF32_LITTLE_ENDIAN_BOM) return 'UTF-32LE';
    elseif ($first2 == $UTF16_BIG_ENDIAN_BOM) return 'UTF-16BE';
    elseif ($first2 == $UTF16_LITTLE_ENDIAN_BOM) return 'UTF-16LE';
    
    // when no BOM available, use mb_detect_encoding
    $encoding = mb_detect_encoding($content, 'auto');
    if ($encoding == 'UTF-8') { // returned for any single-byte encoding
        if (preg_match('!!u', $content)) {
            return 'UTF-8';
        }
        // if not valid UTF-8, suppose Windows encoding
        return 'Windows-1250';
    }
    return $encoding;
}