PHP strlen() nor mb_strlen() returning unexpected results

Question

With PHP, I'm Trying to determine the length (number of characters) in strings such as these:

1
1.1
1.1.1
1.1.2
1.1.3
1.1.3.1
1.1.3.2
1.1.4
1.1.5
1.1.6
1.1.7

etc.

When the length of these strings are measured with mb_strlen() or strlen(), the results are

------------------------------
value   | mb_strlen() | strlen()
------------------------------
1       | 1           | 1
------------------------------
1.1     | 5           | 5
------------------------------
1.1.1   | 9           | 9
------------------------------
1.1.1.1 | 13          | 13
------------------------------
1.1.1.2 | 13          | 13
------------------------------
1.1.1.3 | 13          | 13
------------------------------

It appears that it's counting "." as 3 characters? I'm wondering about just doing a small function to compensate for the predictable "miscount", but am wondering why it's counting the "." as 3 characters to begin with.

I have already looked through several places including this SO article and read the article mentioned, adding the suggested conversions to the page:

mb_language('uni');
mb_internal_encoding('UTF-8');
$str = mb_convert_encoding($str, 'UTF-8', 'UTF-8');

What gives?

EDIT: The strings are imported as part of a csv import.

Here is code:

<?
    $f = fopen("s2db.csv", "r");
    while (($line = fgetcsv($f)) !== false) {

            $colcount = 0;
            foreach ($line as $cell) {
                //lets get the lines into variables first
                //there only five, so just count
                switch ($colcount) {
                    case '0':
                        $item = $cell;
                        break;
                    case '1':
                        $itemtitle = htmlspecialchars($cell);
                        break;
                    case '2':
                        $itemsubject = htmlspecialchars($cell);
                        break;
                    case '3':
                        $itemnumber = htmlspecialchars($cell);
                        break;
                    case '4':
                        $itemqty = htmlspecialchars($cell);
                        break;
                    case '5':
                        $itemfilename = htmlspecialchars($cell);
                        break;                    
                }
                $colcount++;
            }
            $itemlen = strlen($item);
            echo "Value = " . $item . " | strlen() Length = " . $itemlen .  "|  mb_strlen() = " . mb_strlen($item) . "</br>";
    }
?>

Here are results

Value = 1 | strlen() Length = 3| mb_strlen() = 3
Value = 1.1 | strlen() Length = 7| mb_strlen() = 7
Value = 1.1.1 | strlen() Length = 11| mb_strlen() = 11
Value = 1.1.1.1 | strlen() Length = 15| mb_strlen() = 15
Value = 1.1.1.2 | strlen() Length = 15| mb_strlen() = 15
Value = 1.1.1.3 | strlen() Length = 15| mb_strlen() = 15
Value = 1.1.1.3.1 | strlen() Length = 19| mb_strlen() = 19
Value = 1.1.1.3.2 | strlen() Length = 19| mb_strlen() = 19
Value = 1.1.1.3.3 | strlen() Length = 19| mb_strlen() = 19
Value = 1.1.1.4 | strlen() Length = 15| mb_strlen() = 15

SOLUTION:

I gave @hek2mgl the vote because his hexdump helped me determine that I wasn't crazy and it really was counting the "." as 3, as shown here.

Nothing I can do about the import format, so I'm just going to add code to compensate:

Thanks everyone for the help!

Answer 1

I know it's not an answer, but for code formatting reasons.

The following, saved in a UTF-8 file, on my setup...

<?php

echo 'mbstring.internal_encoding: '    . ini_get( 'mbstring.internal_encoding' ) . "\r\n";
echo 'mbstring.func_overload: '        . ini_get( 'mbstring.func_overload' ) . "\r\n";
echo 'mbstring.language: '             . ini_get( 'mbstring.language' ) . "\r\n";
echo 'mbstring.strict_detection: '     . ini_get( 'mbstring.strict_detection' ) . "\r\n";
echo 'mbstring.substitute_character: ' . ini_get( 'mbstring.substitute_character' ) . "\r\n";
echo 'mbstring.detect_order: '         . ini_get( 'mbstring.detect_order' ) . "\r\n";
echo 'mbstring.encoding_translation: ' . ini_get( 'mbstring.encoding_translation' ) . "\r\n";
echo "\r\n";

function outputLengths( $sString )  {
    echo( "mb_strlen('$sString', 'utf-8') = " . mb_strlen($sString, 'utf-8')  ."\r\n" );
    echo( "strlen('$sString') = " . strlen($sString)  ."\r\n\r\n" );
}

outputLengths( '1' );
outputLengths( '1.1' );
outputLengths( '1.1.1' );
outputLengths( '1.1.3.1' );

Outputs:

mbstring.internal_encoding: UTF-8
mbstring.func_overload: 0
mbstring.language: neutral
mbstring.strict_detection: 0
mbstring.substitute_character:
mbstring.detect_order:
mbstring.encoding_translation: 0

mb_strlen('1', 'utf-8') = 1
strlen('1') = 1

mb_strlen('1.1', 'utf-8') = 3
strlen('1.1') = 3

mb_strlen('1.1.1', 'utf-8') = 5
strlen('1.1.1') = 5

mb_strlen('1.1.3.1', 'utf-8') = 7
strlen('1.1.3.1') = 7

What do you get?

Answer 2

Where does you variables come from? Could you please show us the real code (instead of pseudocode)?

I tried to reproduce the described behavior and couldn't. Here's some test I conducted:

$strArray = array(
    '.',
    '1',
    '1.1',
    '1.1.1',
    1,
    1.1,
);

for ($i = 0; $i<count($strArray); ++$i) {
    print "{$strArray[$i]} -> strlen: ".strlen($strArray[$i])." <br/>";
    print "{$strArray[$i]} -> mb_strlen: ".mb_strlen($strArray[$i])." <br/>";
    print '<br>';  
}

This outputs:

. -> strlen: 1 
. -> mb_strlen: 1 

1 -> strlen: 1 
1 -> mb_strlen: 1 

1.1 -> strlen: 3 
1.1 -> mb_strlen: 3 

1.1.1 -> strlen: 5 
1.1.1 -> mb_strlen: 5 

1 -> strlen: 1 
1 -> mb_strlen: 1 

1.1 -> strlen: 3 
1.1 -> mb_strlen: 3

as expected

Answer 3

I got:

<?php

$str = '1.1.1';
var_dump(mb_strlen($str, 'utf-8'));  // 5
var_dump(strlen($str));              // 5

as expected. Seems the . in your case isn't the regular dot but a special unicode char. Please show a hexdump of your input data. You can use Hexdump (I wrote the package for such cases):

Installation:

sudo pear channel-discover www.metashock.de/pear
sudo pear install metashock/Hexdump

Usage:

<?php

require_once 'Hexdump.php';
hexdump('1.1.1');

Would be interesting to see what are the real characters behind the scenes.