Pattern Matching for Sequentinal numbers like 1234

sriram_r15 · January 18th, 2005, 10:09 PM

Hi,

Iam trying to match strings which contain 123456, 234567, 7890, 3456; i.e. numbers which are sequential. Number like 11234556 is not included and of course not 112339474 and 11111. So what could be the pattern matching syntax?

Cheers,
Sriram

BasilFawlty · January 20th, 2005, 03:21 PM

There may be a better way to do this, and with programming there are millions of possibilities, but I would do this.

Code:

<?php
// $numString is your string to test against

// This assumes that you wont be looking to test strings that are 
// longer than 10 characters long. If not, your logic in your 
// loop needs to be a little more complex.

$numString = trim($numString); // Remove any whitespace

$strCheck = ""; // Initialize string to test against

for ($x = 0; $x < strlen($numString); $x++)
{
    if ($x + $numString{0} > 9)
        $strCheck .= $numString{0} + $x - 10;
    else
        $strCheck .= $numString{0} + $x;
}

if ($strCheck == $numString)
     echo "True";
else
     echo "False";
?>

If you are going to do it multiple times, you probably want to put it into a function.

Good Luck
Jeff

nikolai · January 20th, 2005, 05:45 PM

I have a couple things to add. For increased efficiency, you should move the strlen() call to outside of the loop. The way you've written it, strlen() gets called each iteration of the loop, even though you know you're not modifying the length of the string.

Also, I don't understand what it is your loop is actually doing. What does the loop iteration variable, $x, have anything to do with finding matches in the original string?

It seems that this function can return any number of possibilities:
  1) the longest substring of incrementing numbers,
  2) the first substring of incrementing numbers,
  3) a structure or array containing ALL substrings of incrementing numbers.

In each case, you can save yourself the overhead of copying the substring into new strings by returning the start position and length of the substring.

For example:

// This function finds the first substring, if any, where at least
// two characters in succession are incrementing digits.

function find_incrementing_substring(&$str) // pass by reference
{
    $len = strlen($str)
    if ($len < 2)
    {
        return false; // can't match empty or 1-char strings
    }

    $last = (int)$str[0];
    $substr_start = 0;
    $substr_length = 1;

    for ($i = 1; $i < $len; ++$i) // start loop on 2nd char in str
    {
        if ((int)$str[$i] == ($last + 1))
        {
            ++$substr_length;
        }
        else if ($substr_length > 1) // already matched SOMEthing
        {
            break;
        }
        else // no match yet, move start position to current char.
        {
            $substr_start = $i;
        }

        $last = (int)$str[$i];
    }

    return $substr_length? array($substr_start, $substr_length) : false;
}

Do you follow the logic? Basically, we keep tabs of if, where a sequence starts, and how long it is.

If we STOP matching an existing sequence, we break from the loop and return the start/length of the found sequence.

Otherwise, we assume we haven't started matching anything yet, and advance our start pointer.

It would be very easy to modify this function to return an array of ALL such embedded sequential substrings instead of just the first one.

Take care,

Nik
http://www.bigaction.org/

nikolai · January 20th, 2005, 05:46 PM

Also, keep in mind that I haven't tested the above code -- I just typed it all into the msg window. It's your responsibility to verify that it works, if you need it to.

Take care,

Nik
http://www.bigaction.org/