Bash script Linux CentOS

The script needs: 1 - Check a folder and verify if the SOURCE file is newer 2 - If the SOURCE file is the same, just send an email. If it is newer, the flow will be: 3 - read the file and extract the numbers from the file. These numbers will build a MYSQL REPLACE command into a pinsets table. The source file has these informations: 0123=000111=BLABLABALABLA=A 0987=233901=TTTTTTTAAA=A 8765=234567=YYYYYAKAA=A I will need a MYSQL STRING like this: REPLACE into pinsets VALUES ('1','000111\n233901\n234567','pinsetname-2','1',''); So, the script needs to analyse the SOURCE FILE, create a MYSQL REPLACE (and run it!), or, if the source file has not change... just send an email. We will discard all other informations about the SOURCE file. The first number sequence, and the text after the second number sequence. I really need the SECOND number sequence. And, in MYSQL REPLACE, the numbers need to have \n between all of them because they will be in the same cell. I want to run this script for all time with one hour intervals (cron), ok!?