Data analysys, data diapasons/masks creating

I got a big data, which have columns with brand_name and brand_item_number. I need a script, which can analise Brands items array and give back masks of numbering (like diapasons, where this items numbers could be). Following steps is to get number of brand_items for all maked brand_items masks. Following step is creating new items for existing brands, if user trying to create items, with a brand, which not fit any mask of brand - reject it. This script needs, to prevent, if some one tryes to create new items for brand with item number not fit existing data masks. Full data of brands/items stored in PostgreSql, but it could be uploaded to the script by CSV. Script have to reed full data (more than 1500 brands, more than 1500000 items, and then give back list: masks, number of items, which fits each brand mask, each brand items list, which included to each mask). Also this script could be part of SQL script, or PHP based script, which well be cobined with existing program. Data example attached