how to find out if csv file fields are tab delimited or comma delimited

Question

how to find out if csv file fields are tab delimited or comma delimited. I need php validation for this. Can anyone plz help. Thanks in advance.

Answer 1

I'm just counting the occurrences of the different delimiters in the CSV file, the one with the most should probably be the correct delimiter:

//The delimiters array to look through
$delimiters = array(
    'semicolon' => ";",
    'tab'       => "\t",
    'comma'     => ",",
);

//Load the csv file into a string
$csv = file_get_contents($file);
foreach ($delimiters as $key => $delim) {
    $res[$key] = substr_count($csv, $delim);
}

//reverse sort the values, so the [0] element has the most occured delimiter
arsort($res);

reset($res);
$first_key = key($res);

return $delimiters[$first_key]; 

Answer 2

Thanks for all your inputs, I made mine using your tricks : preg_split, fgetcsv, loop, etc.

But I implemented something that was surprisingly not here, the use of fgets instead of reading the whole file, way better if the file is heavy!

Here's the code :

ini_set("auto_detect_line_endings", true);
function guessCsvDelimiter($filePath, $limitLines = 5) {
    if (!is_readable($filePath) || !is_file($filePath)) {
        return false;
    }

    $delimiters = array(
        'tab'       => "\t",
        'comma'     => ",",
        'semicolon' => ";"
    );

    $fp = fopen($filePath, 'r', false);
    $lineResults = array(
        'tab'       => array(),
        'comma'     => array(),
        'semicolon' => array()
    );

    $lineIndex = 0;
    while (!feof($fp)) {
        $line = fgets($fp);

        foreach ($delimiters as $key=>$delimiter) {
            $lineResults[$key][$lineIndex] = count (fgetcsv($fp, 1024, $delimiter)) - 1;
        }

        $lineIndex++;
        if ($lineIndex > $limitLines) break;
    }
    fclose($fp);

    // Calculating average
    foreach ($lineResults as $key=>$entry) {
        $lineResults[$key] = array_sum($entry)/count($entry);
    }

    arsort($lineResults);
    reset($lineResults);
    return ($lineResults[0] !== $lineResults[1]) ? $delimiters[key($lineResults)] : $delimiters['comma'];
}

Answer 3

Here's what I do.

1. Parse the first 5 lines of a CSV file
2. Count the number of delimiters [commas, tabs, semicolons and colons] in each line
3. Compare the number of delimiters in each line. If you have a properly formatted CSV, then one of the delimiter counts will match in each row.

This will not work 100% of the time, but it is a decent starting point. At minimum, it will reduce the number of possible delimiters (making it easier for your users to select the correct delimiter).

/* Rearrange this array to change the search priority of delimiters */
$delimiters = array('tab'       => "\t",
                'comma'     => ",",
                'semicolon' => ";"
                );

$handle = file( $file );    # Grabs the CSV file, loads into array

$line = array();            # Stores the count of delimiters in each row

$valid_delimiter = array(); # Stores Valid Delimiters

# Count the number of Delimiters in Each Row
for ( $i = 1; $i < 6; $i++ ){
foreach ( $delimiters as $key => $value ){
    $line[$key][$i] = count( explode( $value, $handle[$i] ) ) - 1;
}
}


# Compare the Count of Delimiters in Each line
foreach ( $line as $delimiter => $count ){

# Check that the first two values are not 0
if ( $count[1] > 0 and $count[2] > 0 ){
    $match = true;

    $prev_value = '';
    foreach ( $count as $value ){

        if ( $prev_value != '' )
            $match = ( $prev_value == $value and $match == true ) ? true : false;

        $prev_value = $value;
    }

} else { 
    $match = false;
}

if ( $match == true )    $valid_delimiter[] = $delimiter;

}//foreach

# Set Default delimiter to comma
$delimiter = ( $valid_delimiter[0] != '' ) ? $valid_delimiter[0] : "comma";


/*  !!!! This is good enough for my needs since I have the priority set to "tab"
!!!! but you will want to have to user select from the delimiters in $valid_delimiter
!!!! if multiple dilimiter counts match
*/

# The Delimiter for the CSV
echo $delimiters[$delimiter];