regex - Extract from string latest substring with alpa character with or not other numbers on PHP -

i'm trying normalize data series , find problem in 1 of fields.

from data, supposed should normalized , return string of no more 14 characters, composed of 2 alpha characters , rest numbers

but sometime between 90 million items, there may have 1 or 2 additional items start undetermined or serializable alpha character, , number (or not)

normalized values (aa + 000000000000)

ep0123456789    es123456 fr1234567890123 

incorrect values (aa + 00000000 + a) or (aa + 00000000 + a0)

ep1025364758a  fr1920393874b1  ch172637488858a cn727363525252w2 

a -> alpha
0 -> number (positive)

for extract normalized values (aa00000000 code, a0 kindcode) use bit complish code. think there're best algorithm

$pat = 'fr1920393874b1'; if (preg_match("/[a-z]/i",substr($pat, -2))) {    $fail = substr($pat, -2);    if (preg_match('/[\a-za-z]+/', $fail, $match, preg_offset_capture)) {       $kind = substr($fail,$match[0][1]);         // b1        $pat  = str_replace($kind,'',$pat);     // fr1920393874    } } 

so, need 2 values out of input string:

1. the first 2 alpha chars , 1 or more digits after them
   
2. the rest of string
   

so, fr1920393874b1, want fr1920393874 , b1 separate values.

it turns out need split codes rest of string , 2 values in output.

use ^([a-za-z]{2}\d+)(.*) pattern:

$pat = 'fr1920393874b1'; if (preg_match('~^([a-za-z]{2}\d+)(.*)~', $pat, $m)) {     echo "val: " . $m[1] . "\nkind: " . $m[2]; } 

see php demo

details:

• ^ - start of string
• ([a-za-z]{2}\d+) - capturing group 1 ($m[1]): 2 ascii letters , 1+ digits
• (.*) - capturing group 2 ($m[2]): 0+ chars other line break chars, many possible (the rest of line)